sflS-:;

§m#«

MARY GUNN LIBRARY

0000017006

South Pfrican National

Biodiversity Institute

Digitized by the Internet Archive

in 2016

https://archive.org/details/bothaliavolume1515unse

BOTHALIA

Volume 15

Published by the

Botanical Research Institute, Pretoria

Department of Agriculture and Water Supply

Gepubliseer deur die

Navorsingsinstituut vir Plantkunde, Pretoria

Departement van Landbou en Watervoorsiening

Edited by — Onder redaksie van

D. J. B. KILLICK

Index prepared by — Indeks voorberei deur

B. A. MOMBERG

■2 5^7/

ns

^CONTENTS— INHOUD

Page/Bladsy Date/Datum

No. 1 & 2 1—344 August 1984

No. 3 & 4 345-779 August 1985

Page

Bladsy

ANTHONY, NICOLA C. New combinations in Herschelianthe (Orchidaceae) 554

ANTHONY, NICOLA C. and SCHELPE, E.A.C.L.E. A checklist of the Pteridophytes of the Flora of

Southern Africa region 541

Two new taxa and a new combination in southern African Pteridophyta (Pteridophyta) 554

x Pleopodium — a putative intergeneric fern hybrid from Africa (Pteridophyta) 555

BARKER, N.P. Evidence of a volatile attractant in Ficus ingens (Moraceae) 607

BRUSSE, F. New species and combinations in Parmelia (Lichenes) from southern Africa 315

Corynecystis, a new lichen genus from the Karoo, South Africa (Heppiaceae, Lichenes) 552

CODD, L.E. The genus Tetradenia Benth. (Lamiaceae). II. Malagasy Republic 1

The genus Isodon (Schrad. ex Benth.) Spach in Africa and a new genus Rabdosiella Codd (Lamiaceae) . 7

A new species of Plectranthus (Lamiaceae) 142

CODD, L.E. and GUNN, MARY. Additional biographical notes on plant collectors in southern Africa ... 631

COWLING, R.M. A syntaxonomic and synecological study in the Humansdorp region of the Fynbos Biome 175

DOUGLAS, K.H. The identity of Nerine flexuosa (Amaryllidaceae) 545

The relationship between Nerine flexuosa and N. humilis (Amaryllidaceae) 545

ELLIS, R.P. Leaf anatomy of the South African Danthonieae (Poaceae). IX. Asthenatherum glaucum . . 153

Leaf anatomy of the South African Danthonieae (Poaceae). X. Pseudopentameris 561

Leaf anatomy of the South African Danthonieae (Poaceae). XL Pentameris longiglumis and Pentameris

sp. nov 567

Leaf anatomy of the South African Danthonieae (Poaceae). XII. Pentameris thuarii 573

Leaf anatomy of the South African Danthonieae (Poaceae). XIII. Pentameris macrocalycina and P. ob-

tusifolia 579

FADEN, R.B. New taxa of Aneilema R. Br. (Commelinaceae) from southern and tropical East Africa ... 89

FRANKLIN HENNESSY, E.F. The genus Scleria in southern Africa 505

GIBBS RUSSELL, G.E. A new species of Ehrharta (Poaceae) 145

Notes on species of Ehrharta with a short first sterile lemma (Poaceae) 149

Analysis of the size and composition of the southern African flora 613

GIBBS RUSSELL, G.E. and GLEN, H.F. Register of names and types: a comparison between Mesembry-

anthemaceae and Poaceae 125

GIBBS RUSSELL, G.E., RETIEF, E. and SMOOK, L. Intensity of plant collecting in southern Africa ... 131

HENDERSON, L. Survey of exotic woody plant invaders of the Transvaal. Addendum 749

HENDERSON, L. and MUSIL, K.J. Exotic woody plant invaders of the Transvaal 297

IMMELMAN, K.L. Flowering in Kirkia wilmsii Engl. (Simaroubaceae) 151

LINDER, H.P. A phylogenetic classification of the genera of the African Restionaceae 11

Conspectus of the African species of Restionaceae 387

A new name for Disa patens (Orchidaceae) 553

MUSIL, C.F. and BREEN, C.M. The development from kinetic coefficients of a predictive model for the

growth of Eichhornia crassipes in the field. I. Generating kinetic coefficients for the model in green-

house culture 689

The development from kinetic coefficients of a predictive model for the growth of Eichhornia crassipes

in the field. II. Testing and refining the model under field conditions 705

The development from kinetic coefficients of a predictive model for the growth of Eichhornia crassipes

in the field. III. Testing a model for predicting growth rates from plant nutrient concentrations . . . 725

The development from kinetic coefficients of a predictive model for the growth of Eichhornia crassipes

in the field. IV. Application of the model to the Vernon Hooper Dam — a eutrophied South African

impoundment 733

OBERMEYER, A. A. Revision of the genus Myrsiphyllum Willd

The genus Protasparagus in southern Africa (Asparagaceae) 548

OLIVER, E.G.H. A new species of Philippia from the Drakensberg (Ericaceae) 550

PANARELLO, H.O. and SANCHEZ, E. The Kranz syndrome in the Eragrostoideae (Chloridoideae,

Poaceae) as indicated by carbon isotopic ratios 587

Page

Bladsy

REID, C. and ARNOLD, T.H. A new species of Carpha from the Natal Drakensberg, South Africa

(Cyperaceae) 139

RET1EF, E. and REYNEKE, W.F. The genus Thunbergia in southern Africa 107

RUTHERFORD, M.C. and WESTFALL, R.H. Sectors of the Transvaal province of South Africa 294

SCHRIRE, B.D. A new combination in Tephrosia (Fabaceae) 551

SMOOK, L. and GIBBS RUSSELL, G.E. Name changes and additional species of southern African Poaceae

(Poaceae) 147

SPIES, J.J. Embryo sac development in some South African Lantana species (Verbenaceae) 161

SPIES, J.J. and DU PLESSIS, H. The genus Rubus in South Africa. I. Chromosome numbers and geo-

graphical distribution of species 591

SPIES, J.J., DU PLESSIS, H. and LIEBENBERG, H. The genus Rubus in South Africa. II. Meiotic chromo-

some behaviour 597

STAFF OF THE NATIONAL HERBARIUM. New taxa, new records and name changes for southern

African plants 751

STIRTON, C.H. Notes on the genus Rubus in southern Africa 101

TANSLEY, S.A. and SCHELPE, E.A.C.L.E. Two new combinations in Melianthus (Melianthaceae) 143

An excluded species in Melianthus (Melianthaceae) 145

TAYLOR, H.C. A vegetation survey of the Cape of Good Hope Nature Reserve. I. The use of association-

analysis and Braun-Blanquet methods 245

A vegetation survey of the Cape of Good Hope Nature Reserve. II. Descriptive account 259

UBBINK, B. and BREDENKAMP, G.J. A new record of Acorns calamus in South Africa (Araceae) 547

VAN DAALEN, J.C. Distinguishing features of forest species on nutrient and poor soils in the Southern

Cape 229

VAN DER WALT, J.J.A. A taxonomic revision of the type section of Pelargonium L’Herit. (Geraniaceae) 345

VOLK, O.H. and PEROLD, S.M. Studies in the liverwort genus Riccia (Marchantiales) from the south-west

Cape 117

Studies in the genus Riccia (Marchantiales) from southern Africa. 1 . Two new species of the section

Pilifer-. R. duthieae and R. alatospora 531

VORSTER, T.B. and LIEBENBERG, H. Classification of embryo sacs in the Eragrostis curvula Complex . 167

WESTFALL, R.H. Phytocap. A field-data capture programme for the phytotab program package 749

WESTFALL, R.H. and DREWES, R. Grass root pattern in an Orange Free State floodplain 293

WESTFALL, R.H. and PANAGOS, M.D. A cover meter for canopy and basal cover estimations 241

WESTFALL, R.H., VAN ROOYEN, N. and THERON, G.K. The plant ecology of the farm Groothoek,

Thabazimbi District. II. Classification 655

Bothalia 15, 1 & 2 : 1-6 (1984)

The genus Tetradenia Benth. (Lamiaceae). II. Malagasy Republic

L.E. CODD*

Keywords: Lamiaceae, Malagasy Republic, taxonomy, Tetradenia

ABSTRACT

Three species of Tetradenia are recognized in the Malagasy Republic: T. fruticosa Benth., T. goudotii Briq.

(= T. hildebrandtii Briq.) and the newly described T. nervosa Codd.

INTRODUCTION

Since the African species of Tetradenia were dealt

with in Bothalia 14: 177-183(1983), some Malagasy

material, including type specimens, has been

received on loan from the herbaria of the Royal

Botanic Gardens, Kew, and the Conservatoire et

Jardin Botaniques, Geneva, whose kind assistance is

gratefully acknowledged. This permits an opinion to

be expressed regarding the Malagasy species, though

a good deal more material is desired, especially

living plants, for a reliable assessment to be made.

The male and female specimens collected by Mr

D.S. Hardy and Prof. W. Rauh in the south-east

extremity of the island, mentioned in the earlier

article, are unmatched in the loaned material and

are described below as T. nervosa. This brings the

species recognized in the Malagasy Republic to

three.

KEY TO SPECIES

Under-surface of leaf obscured by a fine or coarse, dense

whitish to grey tomentum:

Leaf margin finely and often obscurely crenulate;

under-surface finely white-felted, upper-surface

usually dark grey 1 . T. fruticosa

Leaf margin crenate; both leaf surfaces densely greyish

lanate-tomentose 2. T. goudotii

Under-surface of leaf stipitate-glandular, markedly

reticulate-veined 3. T. nervosa

1. Tetradenia fruticosa Benth. in Bot. Reg. sub

t.1300 (1830); Lab. 164 (1833); in DC., Prodr. 12:

159 (1848); in Hooker’s Icon. PI. 13: t. 1282 (1879);

Briq. in Bull. Herb. Boissier 2: 131 (1894); in

Natiirl. Pfl.Fam. 4, 3a: 331 (1897). Type: Malagasy

Republic, near Tananarive, Hilsenberg & Bojer s.n.

(K, holo.!).

Shrub about 1—1,5 m tall; stems terete, densely

felted with fine branched hairs. Leaves petiolate;

blade ovate-oblong, 50—60 x 25—30 mm, densely

felted with fine branched hairs on both surfaces,

upper-surface darker than lower, apex obtuse, base

obtuse to truncate, margin finely crenulate; petiole

3-10 mm long. Inflorescence a terminal panicle

150-180 x 70—100 mm; male and female flower-

spikes dense, 10-25 mm long; bracts broadly ovate,

1.5 x 2 mm. Calyx 1 mm long, divided to the base

below; upper lobe ovate, lateral lobes bifid. Corolla

2.5 mm long, 4-lobed, the uppermost lobe deeply

bifid giving the corolla a 5-lobed appearance; lobes

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

oblong, the lowest lobe the longest. Disc with 2

lobes exceeding the ovary in length. Figs 1 & 2.

Recorded from central, northern and north-

western districts.

Fig. 1. — Tetradenia fruticosa , holotype in K (left-hand specimen);

the right-hand specimen is T. goudotii.

Malagasy Republic. — Without locality: Goudot s.n. (G);

Lyall 278 (K); Tregellis Fox s.n. (K). Near Tananarive,

Hilsenberg & Bojer s.n. (K). Fianarantsoa Province, 5 km S of

Antanifotsy, Mabberley 1028 (K).

The species is characterized by the small to

medium-sized leaves with a finely crenulate margin,

and the short dense dendroid tomentum on leaves,

stems, rhachis and calyx; stipitate glands are absent,

but all parts of the plant are freely gland-dotted.

2

THE GENUS TETRADENIA BENTH. (LAMIACEAE). II. MALAGASY REPUBLIC

n, izsi

Fig. 2.—Tetradenia fruticosa in Hooker’s Icon. PI. 13: 1. 1282

(1879).

2. Tetradenia goudotii Briq. in Bull. Herb.

Boissier 2: 132 (1894). Syntypes: Malagasy Repub-

lic: near Tananarive, Goudot s.n. (G); Betsileo,

Hildebrandt 3971 (sic) (G). A sheet of Goudot

s.n. in G, annotated by Briquet, is selected as the

lectotype (see note below).

T. hildebrandtii Briq. in Bull. Herb. Boissier 2: 131 (1894).

Type (only specimen cited in protologue): Malagasy Republic,

near Tananarive, Hildebrandt 3471 (sic). This sheet in G is

annotated by Briquet as T. goudotii and No. 3971 (above) is

annotated as T. hildebrandtii.

Shrub up to 2 m tall; stems stout, quadrangular

above, terete below, densely clothed with multicel-

lular hairs and short glandular hairs. Leaves

petiolate; blade ovate to ovate-oblong, up to 90 x 60

mm, densely felted with a greyish-tawny tomentum

of multicellular hairs and short glandular hairs

obscuring both surfaces, matted on the under-

surface, appressed on the upper-surface, apex

obtuse, base obtuse to truncate, margin crenate;

petiole 3-12 mm long. Inflorescence a terminal

panicle up to 250 x 120 mm; male and female

flower-spikes dense, 15—35 mm long; bracts broadly

ovate, obtuse, 1 x 2 mm. Calyx 1 mm long at

flowering, enlarging to 2,5 mm long in fruit,

glandular-hispid; upper lobe ovate, lateral lobes

oblong, shortly bifid. Corolla 2 mm long, 4-lobed,

the uppermost lobe deeply bifid giving the corolla a

5-lobed appearance; lowermost lobe the longest,

Fig. 3 .-—Tetradenia goudotii, lectotype (Goudot s.n. ) in G.

lateral lobes the shortest. Disc with 2 lobes

exceeding the ovary in length. Nutlets oblong, 1 x

0,25 mm (Fig. 3).

Recorded from central, north-central and north-

western districts.

Malagasy Republic.— Without locality, Scott Elliot 3057 (K).

Near Tananarive, Goudot s.n. (G); Malassi, near Tananarive,

Hildebrandt 3471 (G); Betsileo, Hildebrandt 3971(G). ‘Chiefly

from north-west Madagascar,’ Baron 5198 (K);

Distinguished from the former by the more

coarsely crenate leaf-margin (though T. hildebrand-

tii is somewhat intermediate in this respect) and the

dense, coarse tomentum of multicellular hairs and

short glandular hairs of the stems, leaves and

rhachis; all parts of the plant are freely gland-dotted.

The shape of the leaf, the leaf-margin and the mixed

hairs resemble some forms of the African species, T.

riparia, and the two are closely related. However, in

T. goudotii the tomentum is coarser and denser, and

the disc is 2-lobed whereas in T. riparia it is 1-lobed.

There is a contradiction between Briquet’s

citation of Hildebrandt specimens in Bull. Herb.

Boissier 2: 131, 132 (1894) and the names which are

on the specimens in G, in Briquet’s handwriting. For

example, the herbarium sheet of Hildebrandt 3971

from Betsileo bears the annotation T. hildebrandtii

in Briquet’s handwriting, but is cited under T.

goudotii in Bull. Herb. Boissier. On the other hand,

L. E. CODD

3

THE GENUS TETRADENIA BENTH. (LAMIACEAE). II. MALAGASY REPUBLIC

National Herbarium. Pretoria

PRE Neg. No. 81.91.., , ,

iMiMrin i rrn i

Fig. 5. — Tetradenia nervosa, female plant: a, habit, much reduced; b, part of inflorescence, x0,75; c, immature leaves, x 0,75; d,

flower-spikes, x 2,5 (Hardy & Rauh 2910).

L. E. CODD

5

Fig. 6. — Tetradenia nervosa, a-f, male plant: a, bract, x 10; b, calyx, x 7; c, flower, x 10; d, male flower, x 10; e, abnormal

flower with 5 stamens, X 10; f, abortive ovary and style, x 20 ( Hardy & Rauh 2870). g-j, female plant: g, bract, x 10; h,

flower, x 10; i, corolla, opened, x 10; j, fertile ovary and style, x 20 ( Hardy & Rauh 2910).

Hildebrandt 3471 from near Tananarive bears the

annotation T. goudotii by Briquet, and has been

provided with a type label, but is cited as the only

specimen (and hence the type) of T. hildebrandtii.

A comparison of the specimens with the descriptions

indicates that the specimens as annotated by Briquet

compare better with the descriptions than the

specimens as cited in the publication. However, T.

goudotii and T. hildebrandtii are now regarded as

conspecific, so the problem of their typification will

be of only academic interest if the present treatment

is followed: i.e. if T. hildebrandtii is placed as a

synonym of T. goudotii and if a Goudot sheet is

selected as the lectotype of T. goudotii.

3. Tetradenia nervosa Codd, sp. nov., T.

fruticosae Benth. affinis, sed foliis lanceolatis vel

oblongo-lanceolatis, utrinque glandulis stipitatis

obsitis, subtus valde reticulatis differt.

Frutex 1— 2 m altus, ramosus; caules subsucculen-

ti, crassiusculi, teretes, tomentulosi. Folia breviter

petiolata; lamina lanceolata vel oblongo-lanceolata,

50-120 x 20—35 mm, supra glanduloso-puberula,

bullata, subtus valde reticulata glandulis stipitatis

obsita, apice rotundato, basi truncata, margine

crenulato; petiolus 3—10 mm longus. Inflorescentia

terminalis, paniculata, usque ad 400 x 300 mm;

spicae densae, 10-18 mm longae; rhachis dense

glanduloso-tomentulosa; bracteae late ovatae,

1-1,25 x 2—2,25 mm; verticillastri plerumque

6-flori; flores dioeci, subsessiles. Calyx glandulosus,

1 mm longus; lobus posticus ovatus; lobi laterales

breviter bifidi. Corolla dilute malvina, 1,75—2,25

mm longa; tubus anguste infundibularis; limbus plus

minusve asymmetricus, 4-lobatus, lobo postico

bifido. Stamina 4, erecto-divergentia, breviter

exserta. Discus purpureus, 2-lobatus, lobis ovarium

superantibus. Stylus florum masculorum 1,5 mm

longus, florum femineorum 2-2,5 mm longus;

stigma bifidum.

Type. — Malagasy Republic, Fort Dauphin area,

Pinanihy Beach, Hardy & Rauh 2870 (PRE, holo.).

Soft shrub 1— 2 m tall, freely branched; stems

semisucculent, rather stout, terete, tomentulose.

Leaves shortly petiolate; blade lanceolate to

oblong-lanceolate, 50—120 x 20-35 mm, drying

brown, upper-surface glandular-puberulous, bullate,

under-surface stipitate-glandular, strongly

reticulate-veined, apex rounded, base truncate,

margin regularly crenulate; petiole 3 — 10 mm long.

Inflorescence terminal, paniculate, up to 400 x 300

mm in male specimens, smaller in female; inflore-

scence branches subtended by leaves which are

smaller towards the apex; flower-spikes dense,

10—18 mm long; rhachis densely glandular-

tomentose; bracts broadly ovate, nearly twice as

broad as dong, 1 — 1,25 x 2-2,25 mm; verticils

usually 6-flowered; flowers dioecious, subsessile.

Calyx glandular, 1 mm long; posticous lobe ovate,

lateral lobes oblong, shortly bifid. Corolla pale

mauve, 1,75-2,25 mm long; tube narrowly funnel-

shaped; limb somewhat asymmetric, 4-lobed, posti-

cous lobe bifid. Stamens 4, erect-spreading, shortly

exserted. Disc purple, 2-lobed; lobes exceeding the

ovary. Style of male flowers 1,5 mm long, of female

6

THE GENUS TETRADENIA BENTH. (LAMIACEAE). II. MALAGASY REPUBLIC

flowers 2-2,5 mm long; stigma bifid. Figs 4,5 & 6.

Recorded from the Fort Dauphin area; on rocky

outcrops.

Malagasy Republic. — Fort Dauphin area: Pic St Louis,

Decary 10055; Pinanihy Beach, Hardy & Rauh 2870 (male);

Italy Bay, Hardy & Rauh 2910 (female).

Distinguished from the other Malagasy species by

the lanceolate to oblong-lanceolate leaves which are

glandular-puberulous above and stipitate-glandular

below with the reticulate-veined surface visible, not

densely felted as in T. fruticosa and T. goudotii. The

leaf shape and regularly crenulate margin distinguish

it from any of the African species.

UITTREKSEL

Drie Tetradenia spesies word in die Malgassiese

Republiek erken: T. fruticosa Benth., T. goudotii

Briq. (= T. hildebrandtii Briq.) en die nuutomskre-

we T. nervosa Codd.

Bothalia 15,1 & 2: 7-10 (1984)

The genus Isodon (Schrad. ex Benth.) Spach in Africa and a new

genus Rabdosiella Codd (Lamiaceae)

L. E. CODD*

Keywords: Africa, Isodon, Lamiaceae, Rabdosiella, taxonomy

ABSTRACT

The typification of the genus Isodon (Schrad. ex Benth.) Spach and its occurrence in Africa are discussed; an

allied genus Rabdosiella Codd is described and the combinations R. calycina (Benth.) Codd and R. ternifolia (D.

Don) Codd (the latter an Indian species) are effected.

INTRODUCTION

In Bentham’s treatment of the genus Plectranthus

in his Labiatarum Genera et Species (1832) and in

DC., Prodr. 12: 55—61 (1848) he divided the genus

into seven sections: sect. Germanea (Lam.) Benth.,

sect. Coleoides Benth., sect. Heterocylix Benth.,

sect. Melissoides Benth., sect. Isodon Schrad. ex

Benth. (based on Isodon Schrad. in herb.), sect.

Pyramidium Benth. and sect. Amethystoides Benth.

In Benth. & Hook, f., Gen. PI. 2,2: 1175 (1876),

the arrangement of sections was revised and two

primary sections were recognized: sect. Germanea,

which contained the true Plectranthus spp. in which

the calyx is 2-lipped with the upper lip consisting of a

single tooth usually distinctly larger than the

remaining 4 somewhat subulate teeth; and sect.

Isodon in which the calyx is equally 5-toothed or

obscurely 2-lipped with the upper lip composed of 3

teeth and the lower of 2 teeth, all 5 teeth being more

or less equal in size and triangular in shape. Sect.

Isodon included Isodon, Pyramidium, Amethy-

stoides and Melissoides as subsections. Briquet in

Nattirl. PflFam. 4,3a: 352—357 (1897) treated

Germanea and Isodon as subgenera.

With the exception of certain species which have

been retained in Plectranthus, and sect. Pyramidi-

um which is now raised to generic rank, the

majority of species of subgen. Isodon have been

regarded, by most modern workers in Lamiaceae, as

being worthy of separate generic status. In Taiwan,

Kudo in Mem. Fac. Sci. Agric. Taihoku imp. Univ.

2: 118-141 (1929) took up the name Isodon for this

group, attributing it to Schrader and making 37

combinations of species names in the genus. Several

more have been added since. In Japan, Nakai in Bot.

Mag., Tokyo 48: 785 (1934) described the genus

Amethystanthus Nakai, based on the sect. Amethy-

stoides Benth., and this genus was reduced to a

synonym of Isodon by Hara, Enum. Spermat. Jap.

Vol. 1 (1949). In Africa the genus Homalocheilos J.

K. Morton in J. Linn. Soc., Bot. 58: 268 (1962) was

erected to accommodate African representatives.

On the other hand, Keng in FI. Malesiana 1,8: 382

(1978), an area where only a few members of this

group occur, retained Plectranthus in a very broad

sense.

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

Isodon, Amethystanthus and Homalocheilos are

considered to be congeneric (Codd in Taxon 17:

239, 1968; Blake in Contr. Queensland Herb. 9: 4,

1971; Hara in J. Jap. Bot. 47: 193, 1972). In

addition, Blake pointed out that the generic name

Rabdosia (Bl.) Hassk. in Flora 25, Beibl. 2: 25

(1842) is applicable to this group. This resulted in

Hara (l.c.) recombining over 90 Asiatic species in

Rabdosia and, in South Africa, the generic name

was taken up (wrongly it now appears) for Rabdosia

calycina (Benth.) Codd in Bothalia 11:426 (1975).

Recently the compilers of the Index Genericorum

(Vol. 2, 1979) have revealed the still earlier

publication of Isodon (Schrad. ex Benth.) Spach,

Hist. Nat. Veg. Phan. 9: 162 (1840), so that Isodon

must be reinstated as the correct name for this

group.

TYPIFICATION OF ISODON

In lectotypifying Plectranthus sect. Isodon

Schrad. ex Benth. (and hence the genus Isodon )

one must consider the 13 species which Bentham

originally included in the section. Here only one

binomial is attributed to Schrader, namely, Isodon

plectranthoides Schrad., a name only, which

Bentham listed in synonymy under Plectranthus

rugosus Wall. This species agrees with the

description of the section and it seems a reasonable

choice. It was put forward as the lectotype in Taxon

17: 239 (1968), where the combination Isodon

rugosus (Wall.) Codd was effected.

A point which appears to have resulted in some

confused inferences is that Wallich, after his

description of P. rugosus (PI. As. Rar. 2: 17, 1831),

added: ‘ Ocimum densiflorum Roth, Nov. PI. Sp.

275 (?)’. The question mark indicates that he was not

sure of the identity of Roth’s species and so P.

rugosus Wall, should not be regarded as a

superfluous name. Bentham was equally uncertain

of the identity of Roth’s species.

Kudo, op. cit. p. 120, erroneously took up the

name Isodon plectranthoides, attributing it to

‘Schrad. apud Benth., Lab. Gen. et Spec. p. 43, pro

syn.’ and placed P. rugosus Wall, as a synonym. The

Index Genericorum goes a stage further when

they give the type of Isodon as: 7. plectranthoides

Schrad. ex Kudo, nom. illeg. (Ocimum densiflorum

Roth)’. In my opinion, this conclusion is a wrong

8 THE GENUS ISODON (SCHRAD. EX BENTH.) SPACH IN AFRICA AND A NEW GENUS RABDOSIELLA

CODD (LAMIACEAE)

interpretation of Wallich’s and Bentham’s presenta-

tions. The actual specimen which presumably was

annotated as 4 Isodon plectranthoides Schrad.’, and

which was seen by Bentham, has not been traced

and whether it is correctly identified as belonging to

P. rugosus has not been verified. However, neither

the identity of this specimen nor the identity of

Ocimum densiflorum Roth need be considered in

lectotypifying the genus Isodon.

The genus is concentrated mainly in Asia, with

outliers extending to Malesia and tropical Africa.

ISODON IN AFRICA

Isodon (Schrad. ex Benth.) Spach, Hist. Nat.

Veg. Phan. 9: 162 (1840); Kudo in Mem. Fac. Sci.

Agric. Taihoku imp. Univ. 2: 118 (1929); Hara,

Enum. Spermat. Jap. 1: 204 (1949); Codd in Taxon

17: 239 (1968); in Mitt. bot. StSamml., Miinch. 10:

250 (1971); Farr, Leussnik & Stafleu, Index

Genericorum 2: 880 (1979).

Plectranthus sect. Isodon Schrad. ex Benth.,

Lab. Gen. Sp. 40 (1832).

Elsholtzia sect. Rabdosia Blume, Bijdr. 825

(1826).

Rabdosia (Blume) Hasskarl in Flora 25: Beibl. 2:

25 (1842); Blake in Contr. Queensland Herb. 9: 4

(1971); Hara in J. Jap. Bot. 47: 193 (1972); Codd in

Bothalia 11: 436 (1975); Tseng-Chieng Huang &

Wu-Tsang Cheng in FI. Taiwan 4: 504 (1978).

Amethystanthus Nakai in Bot. Mag., Tokyo 48:

785 (1934).

Homalocheilos J. K. Morton in J. Linn. Soc.,

Bot. 58: 249, 268 (1962); in FI. W. Trop. Afr. edn 2,

2: 460 (1963).

The genus is characterized by the flowers being

borne in dichasia, in dichotomously branched

axillary as well as terminal panicles in which the

bracts are not sharply differentiated from the leaves,

but become progressively smaller towards the' apex

of the inflorescence; the calyx is subequally

5-toothed, sometimes obscurely separated into a

3-toothed upper lip and a 2-toothed lower lip; the

calyx tube is usually ventricose and somewhat

circinnate; the corolla tube is relatively straight and

not markedly ventricose or saccate at the base; the

lower lip of the corolla is small and almost flat; and

the stamens are attached at the mouth of the corolla

tube with filaments all free to the base.

Isodon ramosissimus (Hook. f. ) Codd in Taxon

17: 239 (1968). Type: Fernando Po, Mann 624 (K,

holo.!).

Plectranthus ramosissimus Hook. f. in J. Linn. Soc., Bot. 6: 17

(1861); Bak. in FI. Trop. Afr. 5; 418 (1900). Homalocheilos

ramosissimum (Hook, f.) J. K. Morton in J. Linn. Soc., Bot. 58:

268 (1962).

P. schimperi Va'tke in Linnaea 37: 317 (1871); Bak., l.c. 418

(1900). Syntypes: Ethiopia, Schimper 1174; 1179 (PRE!).

P. hoslundioides Bak., l.c. 418 (1900). Type: Tanzania,

Thomson s.n. (K, holo.!).

P. whytei Bak., l.c. 419 (1900). Syntypes: Malawi, Tanganyika

Plateau, Whyte s.n. (K!); between Mpata and Tanganyika

Fig. 1. — Isodon ramosissimus, after Morton in J. Linn. Soc.,

Bot. 58: 249 (1962).

Plateau, Whyte s.n. (K!); Masuka Plateau, Whyte s.n. (K!) ;

north Nyasa, Whyte s.n. (K!).

P. paniculatus Bak., l.c. 419 (1900). Syntypes: Malawi, Nyika

Plateau, Whyte 200 (K!); Blantyre, Buchanan 105 (K!).

P. bullatus Robyns & Lebrun in Rev. Zool. Bot. Afr. 16: 355

(1928), nom. illegit. Type: Zaire, Robyns 2191 (BR, holo.!).

Although there is variation in size and texture of

leaves, this may be due to habitat differences in its

wide range of distribution, at fairly high altitudes, in

West Tropical Africa and in Ethiopia southwards to

Zimbabwe. The species now placed in synonymy

were separated by Baker, l.c., mainly on flower

size, but this is a character of doubtful diagnostic

value and the impression gained is of one variable

species. Figs 1-4.

An examination of type material of the above

species, kindly sent on loan by the Director, Royal

Botanic Gardens, Kew, has emphasized the diffe-

rences which exist between the genus Isodon and

the South African species previously combined as

Rabdosia calycina (Benth.) Codd. The latter is now

placed in a new genus, Rabdosiella Codd.

L. E. CODD

9

Fig. 2. — Holotype of Plectranthus hoslundioides in K: Tanzania,

Thomson s.n.

Fig. 3. — Syntypes of Plectranthus whytei in K: Malawi,

Tanganyika Plateau, Whyte s.n. (left), between Mpata and

Tanganyika Plateau, Whyte s.n. (right).

Fig. 4. — Syntype of Plectranthus paniculatus in K: Malawi,

Blantyre, Buchanan 105.

RABDOSIELLA, GEN. NOV.

Rabdosiella Codd, gen. nov., a Isodonte

(Schrad. ex Benth.) Spach inflorescentia terminali,

dense paniculati, calyce fructifero erecto, tubuloso,

dentibus calycis anguste deltoideis, corolla basi

gibbosa vel saccata, declinata differt.

Rabdosia sensu Codd in Bothalia 11: 436 (1975).

Type species. — R. calycina (Benth.) Codd

( Plectranthus calycinus Benth.)

Herba perennis vel suffrutex; caules erecti, striati.

Folia ternata vel opposita, rugosa. Inflorescentia

non nisi terminalis, dense paniculata. Calyx fructifer

erectus, aequaliter 5-dentatus; tubus tubulosus,

10-nervosus; dentes anguste deltoidei. Corolla

bilabiata; tubus basi gibbosus vel saccatus, declina-

tus; labium supernum breve, obscure 4-lobatum;

labium infernum cymbiforme. Stamina 4, filamentis

liberis. Stigma breviter bilobatum.

The genus is allied to lsodon and Plectranthus. It

differs from lsodon in the stouter, erect stems which

are distinctly ribbed; the dense terminal panicles in

which the branches are ascending, not dichoto-

mously spreading as in lsodon; the tubular, erect

fruiting calyx which is distinctly 10-nerved, with the

teeth tending to close the mouth of the tube; and the

corolla with its saccate-based, declinate tube and the

distinctly concave lower Up. Fig. 5.

10 THE GENUS ISODON (SCHRAD. EX BENTH.) SPACH IN AFRICA AND A NEW GENUS RABDOSIELLA

CODD (LAMIACEAE)

From Plectranthus it differs mainly in the bracts

being leaf-like and becoming progressively smaller

towards the apex of the inflorescence. Also, the

plants tend to be more woody and the leaves are

usually ternately arranged (especially near the base),

a character which is rare in Plectranthus. There are

a few species of Plectranthus in which the calyx is

subequally 5-toothed (subgen. Burnatastrum ), but

these represent a different line of evolution from the

Isodon-Rabdosiella complex.

I

If

Fig. 5. — Rabdosiella calycina, Umzimkulu, Ward 6279.

Two species of Rabdosiella are recognized, both

of which were included in Plectranthus sect.

Pyramidium Benth. in DC., Prodr. 12: 61 (1848):

R. calycina (Benth.) Codd from South Africa and

R. ternifolia (D. Don) Codd from northern India.

Despite their wide geographical separation, they are

remarkably similar in appearance. The main

differences are that R. calycina tends to have leaves

broader in relation to their length and a longer calyx

tube with teeth more narrowly deltoid than R.

ternifolia.

Rabdosiella calycina (Benth.) Codd, comb.

nov.

Plectranthus calycinus Benth. in E. Mey., Comm. 230 (1837);

Drege, Zwei Pfl. Doc. 148, 152 (1843); Benth. in DC., Prodr. 12:

61 (1848); Briq. in Natiirl. PflFam. 4,3a: 352 (1897); Cooke in FI.

Cap. 5,1: 270 (1910); Trauseld, Wild Flow. Drakensberg 160

(1969); Compton, FI. Swazild 502 (1976). Rabdosia calycina

(Benth.) Codd in Bothalia 11: 117 (1973); ibid. IT. 426 (1975);

Codd ex Ross, FI. Natal 305 (1972), non rite publ. Lectotype:

Cape, between St Johns and Umsikaba Rivers, Drege 3584 (K,

lecto. ! ; = Drege b in G ! ; MO!; P!; S!).

P. pyramidatus Giirke in Bull. Herb. Boissier 6: 552 (1898).

Type: Transvaal, Houtbosch, Rehmann 6179 (Z, holo.!).

P. pachystachyus Briq. in Bull. Herb. Boissier ser. 2, 3: 1003

(1903). P. calycinus var. pachystachyus (Briq.) T. Cooke in FI.

Cap. 5,1: 270 (1910). Type: Natal, Umkomaas, Medley Wood

4621 (K!).

Its distribution extends from the Blouberg and

Soutpansberg in northern Transvaal along the

eastern escarpment to Swaziland, Natal, eastern

Orange Free State, Transkei and eastern Cape

Province to around Stutterheim. Over most of its

range, it grows in dense grassland, usually forming

several rigid, erect, unoranched stems arising

annually from a woody rootstock, probably in

response to cold and/or periodic burning. However,

several specimens from the Blouberg and Soutpans-

berg indicate that it may also form a branched

perennial shrub, presumably in the absence of fire or

frost. Fig. 5.

Rabdosiella ternifolia (D. Don) Codd, comb.

nov.

Plectranthus ternifolius D. Don, Prodr. FI. Nepal. 117 (1825);

Benth., Lab. Gen. Sp. 44 (1832); in DC., Prodr. 12: 61 (1848);

Hook, f., FI. Brit. India 4: 621 (1885). Isodon ternifolius (D.

Don) Kudo in Mem. Fac. Sci. Agric. Taihoku imp. Univ. 2: 140

(1929). Rabdosia ternifolia (D. Don) Hara in J. Jap. Bot. 47: 201

(1972). Type: India, Nepal, Hamilton s.n.

From the description it appears to be an erect

shrub 1 — 1,5 m tall. According to Hara, l.c., it occurs

in the mountainous parts of north-eastern India,

Burma, Thailand, Indo-China and south-western

China.

UITTREKSEL

Die tipifikasie van die genus Isodon (Schrad. ex

Benth.) Spach word bespreek; ’n verwante genus

Rabdosiella Codd word beskryf en die kombinasies

R. calycina (Benth.) Codd en R. ternifolia (D.

Don) Codd (die laasgenoemde ’n spesie uit Indie)

word gemaak.

Bothalia 15, 1 & 2: 11-76 (1984)

A phylogenetic classification of the genera of the African Restionaceae

H. P. LINDER*

Keywords: African, anatomy, classification, palynology, Restionaceae, seed coat

ABSTRACT

The generic limits of the African genera of the Restionaceae are revised. Three new genera (Nevillea

Esterhuysen & Linder, Platycaulos Linder and Hydrophilos Linder) are described, the limits of Restio Rottb.,

Chondropetalum Rottb., Mastersiella Gilg-Benedict, Willdenowia Thunb. and Anthochortus Nees are drastically

changed, and Ischyrolepis Steud. , Dovea Kunth, Askidiosperma Steud. , Rhodocoma Nees and Ceratocaryum Nees

are taken into use again. This revision is based on a cladistic analysis of macro-morphological, seed coat

morphological, anatomical, phytochemical and palynological data.

Three new combinations are made, namely Platycaulos compressus (Rottb.) Linder, Nevillea obtusissimus

(Steud.) Linder and Hydrophilos rattrayi (Pillans) Linder.

CONTENTS

Introduction 11

Methods 12

Phytochemistry 13

Cytology 13

Macro-morphology 13

Seed wall structure and morphology 19

Culm anatomy 30

Palynology 46

Phylogenetic analyses 60

Synopsis of the genera 63

Conclusion 67

Acknowledgements 68

References 68

Appendixes

1 Vouchers for the SEM study of seed sur-

face micro-morphology 70

2 Vouchers for the LM study of seed coats ... 72

3 Vouchers for the study of culm anatomy .... 72

4 Vouchers for the palynological study 75

INTRODUCTION

The Restionaceae is a family of evergreen,

rush-like xerophytes with erect, photosynthetic

culms, and with the leaves generally reduced to

sheaths. The flowers are small, aggregated into

spikelets and wind-pollinated. With few exceptions

the flowers are dioecious and there can often be

striking sexual dimorphism in the inflorescences.

They are related to the Poaceae (Dahlgren &

Clifford, 1982) and to a group of small families

(Anarthriaceae, Ecdeiocoleaceae, Flagellariaceae,

Centro lepidaceae and Hanguanaceae). The rela-

tionships between the Restionaceae and these small

families are still very obscure.

The family is almost restricted to the southern

hemisphere. There are approximately 320 species in

Africa, about 100 in Australia, three in New

Zealand (Moore & Edgar, 1970), one in Malaysia

and South-east Asia (Larsen, 1972; Keng, 1978) and

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria, 0001. Present address: c/o The

Herbarium, Royal Botanic Gardens, Kew, Richmond, TW9

3AE, Surrey, United Kingdom.

one in Chili (Pizarro, 1959). This distribution

pattern has led the family to be regarded as a

Gondwanaland relic (Cutler, 1972; Goldblatt, 1978).

In Africa, about 300 species are endemic to the Cape

Floristic Region (Goldblatt, 1978) in the Cape

Province of South Africa and a few species occur

further north in South Africa, with one species

reaching Zaire and one occurring in Madagascar.

The vast majority of the species occur in the

Mediterranean type ecosystems of south-western

South Africa and Western Australia. In these

climatic regions the plants occur on nutrient-poor

soils (Specht, 1981). They can often be the dominant

plants over large areas. In the tropical areas the

plants generally occur in swamps or boggy areas

(Keng, 1978: 356). Within the Mediterranean

systems of South Africa and to a lesser extent

Australia, these plants can be ecologically very

important.

The taxonomy of the Restionaceae has always

been difficult, due to a shortage of convenient

macro-morphological characters. Even at the indivi-

dual level, it can be difficult to match male and

female plants of the same species. At the

supra-specific level, the generic classification has

been very unstable and almost every taxonomist,

who has worked on the group since Linnaeus (1767)

and Bergius (1767) described the first species and

genus, has emended the generic delimitations. The

most prominent papers are Thunberg (1788, 1811),

Nees von Esenbeck (1830, 1836), Kunth (1841),

Steudel (1855), Masters (1878, 1897), Bentham &

Hooker (1883), Pillans (1928) and Gilg-Benedict

(1930). All classifications, with the exception of

Gilg-Benedict, used only macro-morphological data.

Gilg-Benedict used anatomical data as well. For the

African genera, Pillans’s (1928) classification is still

followed in herbaria, and in literature (i.e. Dyer,

1976). However, data from palynology (Chanda,

1966), anatomy (Gilg, 1891, Cutler, 1969) and

flavonoid chemistry (Harborne, 1979) are not

consistent with it.

The supra-generic classification has generally

been more stable. The first groupings were

suggested by Masters (1878). He recognized two

12

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

tribes: the Restioideae with 1— 3-locular capsules

and the Willdenovieae with a 1-locular nutlet. This

classification was used in Flora Capensis (Masters,

1897) and by Pillans (1928). Bentham & Hooker

(1883) developed a more resolved classification. The

primary division was based on the number of cells

per anther, with the Diplantherae having two cells

and the Haplantherae having one. Further subdivi-

sion of the Haplantherae was based on the nature of

the fruit, the number of styles and the nature of the

male spikelet. Gilg-Benedict (1930) also recognized

the Diplantherae and Haplantherae, but did not

subdivide the Haplantherae any further. Johnson &

Briggs (1981) published a phylogenetic tree for the

Restionaceae, which still basically recognizes the

Haplantherae, but is more articulated (Fig. 1), and

which indicates that the African genera form a

monophyletic group.

descriptions of the methodology). The advantages of

this approach include a logically sound and explicit

framework for the assessment of character homo-

logy, a clear phylogeny on which to base classifica-

tion, and a patently predictive system.

As the data sets are, at the descriptive and

comparative level, independent, they are here

presented independently. The specific technical

methods used are discussed with each data set. For

all the sets, the non- African genera are taken to

constitute the outgroup. This is based on the

phylogenetic tree of Johnson & Briggs (1981). In

each data set, character homology is determined first

by the structural and positional criteria (Wiley,

1981) and secondly, where possible, by outgroup

criteria (Eldredge & Cracraft, 1980; Watrous &

Wheeler, 1981; Wiley, 1981). By these criteria,

Fig. 1. — The phylogeny of the

Restionaceae, adapted from

Johnson & Briggs, 1981.

This paper is aimed at providing a preliminary

generic framework for the more detailed study and

preparation of the African species and genera for the

Flora of Southern Africa. Data from as wide a range

of sources as possible have been used: morphology,

anatomy, palynology, seed coat morphology and

structure and flavonoid chemistry. However, some

potentially important data sets are still missing, such

as cytology and non-flavonoid chemistry. Other

data-sets still have large gaps, such as detailed seed

coat structure and some aspects of morphology and

anatomy. However, further detailed work is beyond

the scope of this study, as my brief is to write a flora

account.

METHODS

The basic approach I have followed is that of

cladistic or phylogenetic analysis (see Hennig, 1966;

Bremer & Wanntorp, 1978; Wiley, 1981 and

Eldredge & Cracraft, 1980 for more detailed

characters suitable for use in the general cladogram

are determined.

Extrapolating the data from the specimens from

which they were obtained to species, species groups

or genera presented problems because the existing

classification cannot be trusted. Consequently, all

data are, where possible, related to the voucher

specimens from which they were originally obtained

(see appendices). These specimens were then placed

in species, named according to Pillans’s (1928 et

seq.) classification. Where no published names are

available, manuscript names proposed by Miss E.

Esterhuysen are used. These are grouped into the

genera according to Pillans (1928). They should be

published shortly (Esterhuysen & Linder, in prep.).

For each data set the species are then grouped into

phenetic groups based on that particular data set to

facilitate discussion of the data. In some cases,

where the data set is sufficiently complex, clado-

H. P. LINDER

13

grams are presented to show to what extent the

phenetic groups are monophyletic.

Finally, all the data are synthesized in one

cladogram (Fig. 30). This allows further assessment

of homologies, as well as the use of functional

outgroups (Watrous & Wheeler, 1981) (conditional

synapomorphies of Vrba, 1979) to further resolve

the cladogram. Based on the cladogram, a generic

classification is proposed. Formal supra-generic

groups will only be proposed when Johnson and

Briggs have completed their study of the non-

African genera of the Restionaceae.

The present study is predominantly a pattern

analysis and hypotheses on function, functional

relationships, evolutionary modes, biogeography

and evolutionary tendencies are avoided.

PHYTOCHEMISTRY

The flavonoid chemistry of the Restionaceae is

being studied by Harbome (Harbome & Clifford,

1969; Harborne, 1979; Harbome, in prep.). Several

taxonomically interesting flavonoids have been

found and their distribution and significance will be

reported on in detail in a later publication.

Myricetin derivatives (myricetin 3-galactoside,

larycetrin and syringetin) are very rare in the

monocotyledons. In the Restionaceae they occur

only in Elegia and about half the species of

Chondropetalum. As they do not occur in the

outgroup, they are clearly apomorphous.

Gossypetins are also rare in the monocotyledons.

They occur fairly widespread in the non-African

Restionaceae, and in all the species of Staberoha

that have been investigated. However, as yet

unconfirmed reports indicate that gossypetins may

also occur in one species of Thamnochortus and one

of Willdenowia. The phylogenetic status of gossype-

tins is still not clear, and it is possible that they are

plesiomorphous in the African Restionaceae. Al-

though outgroup comparison does not support

gossypetins as a character for Staberoha, it is still a

useful flavonoid marker for the genus in Africa.

Although there are numerous other flavones and

flavonoids present in the culms of the Restionaceae,

no pattern has been detected to date in their

distribution.

CYTOLOGY

The only published cytological data for African

species are three counts by Krupko (1962, 1966) viz

Elegia racemosa, 2n — 40; Staberoha cernua, n = 16

and Hypodiscus aristatus, n = 16. Numerous counts

are available for the Australian species (Briggs,

1963, 1966) and these have proved to be taxonomic-

ally valuable (see Johnson & Cutler, 1973).

MACRO-MORPHOLOGY

Introduction

The macro-morphology is probably the best

known aspect of the Restionaceae, as it has been

studied extensively and repeatedly by generations of

taxonomists as a data-base for their classifications.

The results of their studies can be found in their

descriptions of the species and genera (see, e.g.

Masters, 1897; Pillans, 1928; Gilg-Benedict, 1930).

Despite this extensive knowledge, there is only one

detailed morphological study, and it deals with the

sexual dimorphism of the inflorescences and the

flowers (Ueberfeld, 1925). The empirical base of this

study was, unfortunately, too narrow, consequently

the study lost most of its potential value. Detailed

morphological analyses of the branching patterns,

rhizome organization, the nature of the sterile

shoots, the floral morphology and the construction

of the inflorescences are unfortunately still out-

standing.

Methods

The available herbarium material at BOL and K

was investigated by hand lens and by dissecting

microscope, and details recorded by drawing by

camera lucida. Where necessary, such as in floral

morphology, material was dissected, either unpre-

pared, or after treating three to five minutes in

boiling water. As many species as possible were also

observed and described from fresh material col-

lected in the natural habitat.

For a few species the gynoecium was studied

microscopically. Here, material collected from

herbarium sheets was boiled, fixed on a freezing

stage, and sectioned at 10-50 pm by sledge

microtome. The sections were stained with Alcian

Blue and Safranin and mounted in euparal or

glycerine jelly.

Results

1. Rhizome

Although there is extensive variation in the

appearance of the rhizome, it is probably one of the

least known structures, as it is rarely collected in

large enough sections to reveal its organization.

Initial analysis indicates the following types of

organization:

(a) culm-bases aggregating, with the basal nodes

producing roots. New culms are formed as shoots

from one of the basal nodes;

(b) some culms spreading underground, acting as

stolons, occasionally rooting at the nodes and

producing clusters of aerial culms;

(c) creeping rhizomes produced, that are very

different in appearance from the aerial culms. These

produce a regular sequence of aerial culms from the

top of the rhizome and roots from below (Van

Greuning & Van der Schijff, 1975).

Numerous intermediate forms and variants exist,

making the above classification rather arbitrary. In

some taxa the rhizomes are densely covered in

golden scale-leaves and are usually ascending, in

others, they can be almost bare. It is probable that

detailed analyses of rhizome organization, in

conjunction with studies of comparative rhizome

anatomy, may yield useful data, but at present the

empirical data base is too weak and a theoretical

understanding of rhizome organization too un-

developed, to allow use of this character.

14

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

2. Culms

The size and branching patterns of the culms

dictate the overall aspect of the plants. In the culms

themselves four characters are found:

(a) the surface appearance is determined by the

presence of tubercles or hairs, or grooves or ridges,

or compression of the culms. These features have

been studied in more detail in the anatomical

section, but the data base is broadened by

morphological observation. Masters (1897) paid

particular attention to these features in order to

place individuals of different sexes into their correct

species;

(b) three types of branching patterns are found in

culms : unbranched (simple), simply branched and

verticillately branched. An analysis of the outgroup

shows that branching is a variable character in most

genera, but this cannot be read to mean that it will

also be a variable character within the African

genera;

(c) the size of the culms is a continuous character,

not exhibiting any discrete states. Consequently it

cannot be adequately treated. It is partially linked to

the branching pattern and the overall habit — very

tangled plants have slender, frequently branching

culms;

(d) ‘sterile’ culms probably occur in all African

taxa, but with the exception of Restio tetraphyllus I

have not seen any in non- African taxa. They are

profusely branched culm systems, in which the

sheaths are modified, so that the awn becomes long

and culm-like. The whole forms a tangled cluster of

material. In the majority of the African genera

sterile culms are found only in young plants and in

plants regenerating after disturbance, such as fire.

As such they are rarely recorded in herbarium

material. In some taxa they are regularly produced

and may even sprout from the nodes on the

flowering culms, sometimes only in the year after the

culm has flowered.

3. Leaves

The leaves in all African species are reduced to

sheaths, split to the base, usually with a small awn or

mucro, and often with a hyaline apex that has been

regarded as a ligule (Dahlgren & Clifford, 1982).

On sterile culms and often on the new shoots of

seedlings, the sheaths have much expanded awns,

that may be much larger than the sheaths and

indistinguishable from the culms. In these situations

the sheaths are dimorphic.

The sheaths are quite variable and are valuable in

matching different sexes of the same species. The

following variants are found:

(a) the sheaths may be caducous, with a distinct

abscission zone. This feature occurs in some genera

from both Africa and Australia.

(b) they may be either tightly or loosely

convoluted around the culms. In the majority of taxa

they are tightly convoluted, but in a few genera,

both in the study group and the outgroup, they can

be quite free.

(c) the sheath margin may be variously modified,

e.g. entire, lacerated, acute, obtuse, fimbriated, etc.

4. Inflorescences

The inflorescences are rather difficult to describe

and as yet, no detailed analyses have been

published. Masters (1897) considered them ‘simple,

or generally cymose, much or little branched’. In his

generic diagnoses he recognized spicate, panicled,

spicately cymose, paniculate-cymose, cymose and

thyrsoid inflorescences. Pillans (1928) described the

inflorescences as ‘spikelike or branched’ and

avoided terminology in his generic key.

Each flower is borne in the axil of a bract, and

may be pedicellate or sessile. This results in the

terminal inflorescence units being technically spicate

or thyrsoid (Weberling, 1981). However, as there is

a continuous range between pedicellate and sessile

flowers, the distinctions between spicate and

thyrsoid inflorescences is probably trivial. Pillans

(1928) and Johnson & Briggs (1981) recognize

flowers organized in spikelets or not in spikelets (i.e.

racemose, etc.). This distinction appears to be rather

a function of the spacing of the flowers and the size

of the bracts.

The terminal units may be directly sessile on the

main axis (Fig. 2e), or may be part of a larger

branching system (i.e. spicate or panicled) (Fig. 2c).

However, several branches may arise from the same

node, all subtended by a single bract (termed a

spathe by Pillans, 1928) (Fig. 2a, c, f, h). This

curious system is probably what Masters (1897)

described as ‘spicately cymose’ or ‘paniculate-

cymose’, although the inflorescences are indetermi-

nate, and not strictly speaking cymose.

Sexual dimorphism varies from absent to extreme.

It was studied in a small number of species by

Ueberfeld (1925). The basic tendency is for the

female inflorescence to develop fewer, larger, and

fewer-flowered spikelets than the male inflores-

cences (see Fig. 2). It is obvious that the male and

female inflorescences have to be analysed separ-

ately.

Generally, the bracts are more or less coriaceous

to osseous, at least as tall as the perianth and more

or less obscuring the flowers. This is also the

situation in the outgroup. In the African taxa there

are several deviations from this situation:

(a) the bracts may be much shorter than the

perianth (usually 1/3 to \ as long as the perianth (Fig.

2d, e, f, g);

(b) the bracts may be taller and wider than the

perianth, but may be more or less hyaline and

lacerated, so that they do not obscure the perianth;

(c) the bracts may be taller than the perianth, but

may be linear and hyaline and usually indistinguish-

able from the perianth segments (Fig. 2h).

The reduction in the number of flowers per

terminal unit (usually the spikelet) can be accom-

plished in two different ways:

H. P. LINDER

15

Restio ocreatus

Chondropetalum macrocarpum

Elegia grandis

Willdenowia luceana

Fig. 2. — Inflorescence types. Flowers are indicated by black spots, culms and spathes by heavy black lines, bracts and inflorescence

branches by light lines. Caducous spathes are indicated by T. The stylized inflorescence drawings show the various parts in

approximately the correct size ratios. Restio ocreatus, A, B, flowers in spikelets; Chondropetalum macrocarpum, C, D, E. In C

the male spikelets are not drawn in, they are shown in detail in D; Elegia grandis, G, F, flowers in thyrses; Willdenowia

luceana, H, I, male flowers in thyrses, female flowers in spikelets.

16

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

(a) the loss of flowers, but the retention of the

now sterile bracts, leading to a single terminal flower

surrounded by several to many bracts (Fig. 2). This

type of reduction is commonly found in the

outgroup, and presumably proceeds as reduction

from a spikelet;

(b) the occurrence of single flowers, subtended by

a single bract (Fig. 2f, g). This could either result

from reduction from a spikelet by loss of both

flowers and bracts or, more likely, from reduction

from a panicle or a thyrse, in which the flowers are

not aggregated into spikelets. This situation has not

been observed in the outgroup.

The flowers on the terminal inflorescence units

may be organized into compact spikelets (i.e.

flowers close together, if not adjacent, and more or

less sessile) (Fig. 2a, b), or more loosely into thyrses

(Fig. 2f, g, h). The distinction between spikelets and

thyrses is not absolute and the rule of thumb used is

that spikelets form discrete clusters of flowers (e.g.

Fig. 2d). In the outgroup all flowers are organized

into spikelets.

The distinction between spicate, paniculate and

cymose inflorescences which Masters (1897) made,

is not upheld here, because depauperate inflores-

cences will often appear spicate, whereas robust

individuals have paniculate or cymose inflorescen-

ces.

5. Perianth

The perianth consists of two whorls of three

segments each. As in the inflorescences, there can

ber strongly developed sexual dimorphism, conse-

quently the male and female perianth structures are

discussed separately.

B

G

H

Fig 3. Major perianth types, all x 6. A— D female, E— H male. A, outer lateral segments winged, Thamnochortus platypteris; B,

outer lateral segments conduplicate, carinate, villous, Restio bolusii; C, segments flat, cartilaginous, often nitid,

( hondropetalum hookerianum , D, segments flat, hyaline, Willdenowia luceana; E, segments valvate, hyaline, Willdenowia

argentea; F , outer lateral segments conduplicate, carinate, Restio bolusii; G, segments flat, cartilaginous, the outer whorl much

shorter than the inner whorl, Chondropetalum ebracteatum; H, segments linear, free, hyaline, Willdenowia luceana.

H. P. LINDER

17

In female individuals, the following types can be

recognized:

(a) segments concave, nitid, glabrous, occasion-

ally with the outer segments with a slightly raised

midrib. This is quite a distinctive type, often

associated with reduced bracts (Fig. 3c). It grades

into the next type;

(b) outer lateral segments conduplicate, often

slightly carinate, often more or less; woolly along the

carina. The woolliness can be highly variable and

may even be present only on one side of the flower

(Fig. 3b). Loss of woolliness and loss of the carina

results in this type grading into the previous type;

(c) outer lateral segments conduplicate, car-

inate, with the carina developed into a wide keel, so

that the perianth as a whole is winged (Fig. 3a).

Although Gilg-Benedict (1930) used this character

in her key to the genera, the distinction between the

carina and a keel is rather arbitrary. This perianth

type is unique to Africa. It is difficult to use though,

as it grades into type (b);

(d) perianth segments all more or less membra-

nous or chartaceous, flat and without particularly

developed characteristics (Fig. 3d). The size of the

segments relative to the nut is variable, but they are

rarely taller than the nut.

c

In male individuals the following perianth types

are recognized:

(a) segments concave, nitid, glabrous, the outer

whorl often much shorter than the inner whorl (Fig.

3g). This is rather similar to the female type (a);

(b) segments conduplicate, carinate, often woolly

along the carina. The inner whorl of segments are

usually hyaline, whereas the outer whorl is

cartilaginous (Fig. 3f). This type is similar to the

female type (b);

(c) segments equal or subequal, more or less

lanceolate, valvate, chartaceous, rarely the outer

whorl much shorter than the inner whorl [which

links these forms to type (a) ] (Fig. 3e);

(d) segments all equal or subequal, linear,

hyaline, the anthers dominating the appearance of

the flowers (Fig. 3f). This structure is unique to the

African Restionaceae, and has been used by Pillans

(1928) to partially delimit the genus Willdenowia.

6. Androecium

All African Restionaceae have three one-celled

anthers. The only variation is in the position of the

anther at dehiscence. In the majority of the species,

the filaments elongate dramatically just before

dehiscences, to push the anther out beyond the

perianth and the bract. This is also the general

situation in the outgroup. In one group of species the

filament does not elongate and the anther dehisces

inside the perianth. The perianth segments move

slightly apart to allow the pollen to escape.

7. Gynoecium

In the plesiomorphic condition, the gynoecium

consists of three dehiscent locules (Fig. 5g), each

with a pendulous ovule and three free styles. From

this situation, reduction to one locule and one style

has occurred, both in the study group and in the

outgroup.

The number of styles varies almost independently

from the number of locules and almost any kind of

combination may be found. Often, when there are

two styles, they are fused below to form a

stylopodium (Fig. 4b).

Reduction in the number of locules may or may

not leave traces of the lost locules (e.g. Fig. 5h, with

no trace of the lost locule). The fruits of the

two-or-three-locular ovaries are generally dehiscent

(Fig. 4h, g). Unilocular ovaries are often indehiscent

and several different kinds of nuts are found:

(a) nuts flattened or slightly compressed, rarely

lignified, sometimes with the walls compressed (Fig.

5a, b). Usually the perianth is persistent around the

nut;

c

Fig. 4. — Type of gynoecia, all x 8. A, unilocular, indehiscent

nutlet with 3 styles, Elegia juncea; B, bilocular, dehiscent

capsule with two partially fused styles, Restio ocreatus; C,

unilocular, indehiscent nutlet with a single style, Thamnoc-

hortus platypteris; D, trilocular, dehiscent capsule with three

free styles, Restio triticeus.

18

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Fig 5. — Fruit types. A, parenchymatous nut with three vascular traces. Dark cells below epidermis are filled with tannin,

Thamnochortus fruticosus, x 60; B, parenchymatous nut (note seed), Staberoha cernua, x 60; C, lignified nut, compressed on

one side, Cannomois virgata, x 16; D, detail of wall of C, note stone cells and compressed parenchymatous layer at foot of

wall, x 60; E, partially lignified nut, note large parenchymatous region, Elegia muirii, x 40; F, parenchymatous nut with a

suberized epidermis and the three vascular traces forming ribs, E. racemosa, x 60; G, three-locular dehiscent capsule, note

dehiscence lines and seed, Restio giganteus, x 20; H, two-locular dehiscent capsule without trace of the third locule, R.

filiformis, x 50.

H. P. LINDER

19

(b) nuts round in cross-section, very heavily

lignified, often with stone-cells in the walls (Fig. 5d).

The outer wall of the nut may be smooth or pitted.

These nuts tend to be large, about 5 mm long, and

are not found in the outgroup. Often they are

stipitate. The stipe is fleshy and often lobed (Fig. 3d)

and has been interpreted as an elaiosome (Slingsby

& Bond, 1981);

(c) nuts as above, but flattened on one side (Fig.

5c);

(d) nuts round or triangular in cross-section,

occasionally slightly lignified, especially near the

apex (Fig. 5e, f), usually dropping free from the

perianth.

Distribution of characters

It is not practical to summarize the distribution of

all these characters for all species in one table. Nor

can the variation be summarized by the genera

delimited by Pillans (1928), as the many new species

found since his revision have destroyed the

morphological homogeneity of his genera. So the

data are summarized in what appear to be

morphologically homogeneous groups (Table 1). In

the process, the data have been somewhat simpli-

fied.

The macro-morphological character distribution

in the Willdenowia-Cannomois-Hypolaena-

Hypodiscus group is explored in detail in Table 2.

Cutler (1969) suggested a redelimitation of the

genera in the group, so in order to avoid biassing the

morphological data by simplifying them into

unnatural groups, the data are listed species by

species.

In Restio and Leptocarpus several morphological

groups can be recognized. A group of about 50

species is separated by the possession of two more or

less fused styles and by the complete absence of any

staminodes. This group has been informally recog-

nized since Kunth (1841) separated it as ‘true’

Restio. A group of eight species can be recognized

by the possession of compressed culms. A further

three species have the aspect of Thamnochortus, i.e.

pendulous male spikelets, simple culms with clusters

of sterile culms at the nodes. Traditionally,

Leptocarpus has been separated from Restio by the

possession of an indehiscent, unilocular nutlet.

However, the distinction between a dehiscent and an

indehiscent unilocular fruit may be operationally

difficult and is probably evolutionarily trivial.

Further variation in the two genera, affecting

detailed perianth and bract shape, the number of

fertile bracts per spikelet and the number of

spikelets per inflorescence appears to be continuous,

making the recognition of further distinct groups

difficult.

The morphological distinction between Elegia and

the Elegia neesii group is not as simple as indicated

in Table 1, as style number is not correlated with

either inflorescence structure or branching pattern

and sheath persistence.

Within Chondropetalum, C. microcarpum is very

distinct by having branching culms, two styles, a

two-locular indehiscent ovary and persistent

sheaths. The overall aspect is very like Chondrope-

talum and this species is regarded as displaying

extensive anagenesis. Chondropetalum rectum and

C. nudum also have persistent sheaths.

Morphological data provide a wealth of informa-

tion. However, few characters stand up to rigorous

outgroup testing. Many of the characters that also

occur in the outgroup, such as caducous sheaths,

branching culms and female perianth type can still

be established by functional outgroup comparison.

Other morphological characters, although not being

useful in demonstrating monophyly, assist in placing

taxa of which little else is known (and there are

many such species) into the right groups.

SEED WALL STRUCTURE AND MORPHOLOGY

Introduction

The literature on the seeds and seed coats of the

Restionaceae is very sparse. Pillans (1945, 1952)

incorporated comments on seed surface appearance

in his descriptions of new species. Netolitzky (1926)

summarizes what was known to date about the seed

wall construction. Borwein et al. (1949) and Krupko

(1964) make some comments on the ontogeny of the

seed wall, incidental to the main theme of their

research on the embryology of the Restionaceae. Of

the African Restionaceae, only the members of the

genera Restio and Chondropetalum have dehiscent

fruit and have seeds with elaborate coats. Conse-

quently only these taxa were studied in detail.

Materials and methods

Seed was taken from herbarium specimens. As

many samples as available in the Bolus Herbarium,

University of Cape Town, were studied under a Wild

dissecting microscope. One or two seeds from one

collection of each species (where seed was available)

was selected for study by scanning electron

miscroscope. The untreated seeds were mounted on

double-sided sticky tape, coated with gold in an

‘Eiko Model IB-2’ ion coater to about 300—400 A,

and photographed in Model MSM-4 Hitachi Abashi

mini scanning electron microscope.

A further small sample was studied by light

microscopy at the Jodrell Laboratory, Royal Botanic

Gardens, Kew. The untreated seed was mounted

upright in ice on a freezing stage (i.e. micropyle

upwards), and sectioned at 10-20 pm with a sledge

microtome. The sections were stained with Alcian

Blue and Safranin, dehydrated in the alcohol series

and mounted in Euparal. A few sections were

mounted in glycerine jelly, and some sections were

stained with Sudan IV for cuticles and fats. Sections

were studied under a Wild microscope with normal

and polarized light and drawn by camera lucida.

A listing of the 120 species studied by SEM and

the 17 species studied by light microscope and their

vouchers, is given in Appendix 2.

Results

The seed is 2-4 mm long, varying from spheroidal

to ellipsoid to truncate and may be circular,

triangular or variously flattened in cross-section.

TABLE 1. — Distribution of macro-morphological characters in the African Restionaceae

I

20 A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

juasard auiosotEjg

punor XpooM jou jnjq

jejj XpooM jou jnjsi

# * * * * + *

+ + + + + +

+

m <N

+ + + +“ *

d - - -nnHmmmfsiNiNNMNriNNNN

X ro m cn w

m

+ + + + +

* + + + + +

+ *

* + + + + + +

* ++ + + + + +

+ +

+ + *

+ + ****+ + + *+ + + +

*•**+* + + + + +

+ + + i

+ * * + +

+ +

+

+ + + +

+■ +" * +"

+

+ + + +

+

+ ++ +++ + +++ ++

+" '+ +~

+ +~+ +" + + + + ? + + + +

s £ «

1 &S

■S 8-f

^ S.§

9 9 ft

£ £

P. 2

£ .

532

on co

S 8.

S 50

o "

ft » S'

O m sr<

s -a

? o

D <u <u

S- 9- 9-ts

g g g o

§ § § § ■£ 'c C S -c 3 g

,o .0,0 o .0 o ,o

§ Co '£* Cu ^

r 3 ^ 2 -S

P £ t 2 a

£« K a so*:

3 -S ft, ^ .§>

“ I|f1

-o g- g £

- cv ^

1 ^ rC rS*

i k) kj 6 6 S K

co Co Co El Co !

^ :

^ O' <1 ,

k ,5

' os ^ ^ ^

Hypodiscus

Willdenowia rugosus group

Willdenowia

H. P. LINDER

21

TABLE 2. — Distribution of macro-morphological characters in the Willdenowia-Hypodiscus-Hypolaena-Cannomois group

+ character present

* character rare or not well-developed

22

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Four groups of characters are recognized:

(a) seed shape

(b) seed surface

(c) cellular organization, and

(d) chemical nature of the walls.

(a) Seed shape

Two groups of seed shape can be postulated: seeds

that are more or less circular in cross-section and

seeds that are triangular in cross-section. The seed of

Anarthria scabra and Lyginia barbata are ellipsoid

in both longitudinal and cross-sections, which

suggests that the triangular seeds may be the more

derived form. Although there is variation in the

longitudinal shape of the seeds, the various shapes

intergrade and are difficult to quantify and are not

used here.

(b) Seed surface

There is a great range in the surface micro-

morphology of the seeds. This is partially visible by

hand-lens or dissecting microscope, but is best

studied by scanning electron microscope. It is here

described in the terminology suggested by Stearn

(1973). The surface features are best understood in

terms of the epidermal layer. The following types

can be recognized:

(i) smooth seeds, in which the epidermal cells

are square and the lumina are completely occluded

(Fig. 6A, B, C);

(ii) smooth seeds in which the epidermal cells are

elongate, with small elongate lumina (Fig. 6D— J);

(iii) smooth seeds in which the epidermal cells are

on flat plates (Fig. 10A, B) — although these at

higher magnifications appear strongly reticulated;

(iv) smooth seeds, in which the epidermal cells

are very elongated (Fig. 11). These surfaces usually

show deep cracks through the epidermal cells;

(v) scrobiculate surfaces, in which the epidermal

cells have concavely depressed outer walls (Fig. 7).

If the outer wall is only slightly depressed, the

surface appears alveolate (Fig. 7E). If the walls on

one side are higher than on the other, the surfaces

appear ridged;

(vi) tuberculate or verrucate surfaces, in which

the outer walls of the epidermal cells are much

thickened (Fig. 10G— K). The shape of the tubercles

appears to be quite variable;

(vii) colliculate surfaces, formed by the massive

thickening of the epidermal cell walls, leaving a

small lumen (Fig. 8A— G);

(viii) pseudo-colliculate surfaces, in which each

‘hill’ is surrounded by a flat surface (i.e. shaped like

a bowler hat) (Fig. 81, J). There are intermediate

forms between (vii) and (viii);

(ix) pitted rugose seeds, with the epidermal cells

thin-walled, tall, with concave outer walls with pits

in them (Fig. 9F, G);

(x) pitted colliculate seeds, with the epidermal

cells tall and thin-walled, but with the outer surface

convex (Fig. 9A, B, D);

(xi) surfaces ridged and scrobiculate, with no

cellular margins or boundaries evident in the

epidermal layer (Fig. 9H, I).

Secondary surface features may also be present.

These may be the nature of the cell margins

becoming visible on the surface, which may be

straight (i.e. Fig. 8B) or wavy and interdigitating

(i.e. Fig. 8G). Further fine ridges (i.e. Fig. 8B, G),

pits or tubercles may be present. These features are

as a rule only visible at higher magnifications.

(c) Cellular organization

The cellular organization of the seed coats is

difficult to assess, due to the high degree of

specialization of the cells, the technical difficulties of

making good sections and the lack of numerous

ontogenetic series. The ovule is bitegminous, with

each integument with two layers of cells (Borwein et

al., 1949; Krupko, 1964). These four layers are

manifested as follows in the mature seed wall:

(i) the inner epidermis of the inner integument

forms a tanniferous layer (as noticed by Borwein et

al., 1949), with the cells enlarging and the walls

often becoming suberized. The final actual cell

shape is variable, but usually obscured by the thick

tannin deposits;

(ii) the outer epidermis of the inner intergument

is sometimes present as a suberized layer, but is

more usually very squashed or totally lost. I call it

the ‘supra-tanniniferous layer’;

(iii) the inner epidermis of the outer integument

is highly variable in size and shape. It generally

stains pale blue with Alcian Blue and shows

birefringence and rarely stains lightly with Sudan IV.

In most instances the layer is very compressed, and

often no cellular structure can be detected in it (i.e.

Fig. 8F, J), while only in one case are the cells

enlarged (Fig. 6F). I call this layer the ‘subepidermal

layer’;

(iv) the outer epidermis of the outer integument

forms the surface morphology of the seed and I call

it the epidermal layer. This generally stains pale blue

with Alcian Blue, or often does not take up any stain

at all. It occasionally takes up very little Sudan IV.

Only in one species does it stain with Safranin.

Generally the layer shows birefringence. The shape

of the epidermal cells is highly diverse, as discussed

above.

(d) Chemical constituents of the cell walls

The chemical structure of the cell walls is inferred

from their staining properties and reaction to

polarized light. Although this is not a very

satisfactory basis to work on, the following

characters may be recognized:

(i) in some taxa the outer epidermal walls stain

dark blue with Alcian and are not birefringent —

presumably this is from a heavy thickening of

cellulose;

(ii) in several taxa the epidermis does not take up

any stain, but is still birefringent;

(iii) most epidermal cells are both birefringent

and stain pale blue — presumably a mixture of

hemicellulose and cellulose;

(iv) suberin and cutin, as detected from Sudan

IV, generally occurs in the tannin and supra-

tanniniferous layers, rarely in the epidermal layers.

H. P. LINDER

23

Fig. 6. — Seed surfaces smooth. Restio subverticillatus: A, x 200; B, x 240; C, x 30. R. egregius: D, x 18; F, x 200; G, x 600. R.

dispar: E, X 18; H, x 840. R. singulars: I, x 1200; J, x 36.

24

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

■■■!*

Fig 7— Seed surfaces scrobiculate: Restio virgeus. A, x 200; B, x 240; C, x 30. R. sieberi: D, x 30; F, x 200; G, x 600. Restio

cincinnatus, surface alveolate, E, x 600. R. nanus: H, X 200; I, x 600.

H. P. LINDER

25

Based on these characters, the following phenetic

groups can be recognized:

Group 1. Seeds more or less ellipsoid with the

ridge faintly developed. Surface scrobiculate, vary-

ing to reticulate-foveate, alveolate or ribbed with

faint cross-ribs (Fig. 7). In the majority the outline

of the cell margins is visible on the surface, varying

from straight to wavy. In transverse section the

subepidermal layer is squashed, with the cellular

organization sometimes visible, and staining lightly

with Alcian Blue and Sudan IV. The epidermal cells

have concave outer walls, and stain only with Alcian

Blue, whereas the outer margins of the scrobiculi

stain deep blue.

Restio virgeus, R. wittebergensis, R.

sieberi var. sieberi, R. sieberi var.

venustulus, R. gaudichaudianus, R. cin-

cinnatus, R. karooica, R. feminens, R.

tenuissimus, R. fraternus, R. coactilis, R.

sporadicus, R. nanus, R. gossypinus, R.

laniger, R. affinis, R. unispicatus, R.

intermedius, R. monanthus, R. macer, R.

curvibracteatus and R. nubigena.

Group 2. Seeds obliquely ellipsoid, ridge fairly

well-developed. Surface tuberculate to verrucate,

the exact shape of the verrucae highly variable.

Restio rottboellioides, with pustinulate outgrowths,

may be included and also R. paludosus with low,

elliptical oblong colliculae. In transverse section ( R .

marlothii. Fig. 10J) the surface features are formed

by the outer wall of the epidermis, which is greatly

thickened, stains deep blue with Alcian and does not

show any birefringence. In contrast, the subepider-

mal layer is quite flattened.

Restio marlothii, R. rivulus, R. papillo-

sus, R. dura, R. longiaristatus, R.

paludosus, R. rottboellioides.

Group 3. Seeds obliquely ellipsoid, with the

ridge fairly well-developed. Seed surface smooth,

with a fine patterning probably reflecting the cell

margins (Fig. 6A, B, C). In transverse section the

sub-epidermal layer appears very squashed, while

the epidermis consists of small, solid, square cells,

that hardly stain at all.

Restio gossypinus, R. subverticellatus, R.

setiger, R. triflorus, R. aridus and R.

hystrix.

Group 4. Seeds irregularly knobbly, more or less

ovoid with a prominent ridge. Surface with narrowly

elliptical colliculae, with pits either along the cell

margins or scattered (Fig. 9 A, B, D). In transverse

section the subepidermal layer is scarcely visible.

The epidermal layer consists of tall hollow cells with

convex upper margins, in which pits can be seen in

oblique sections. These cells stain pale blue.

Restio pygmaeus, R. saxatilis, R. praten-

sis, R. curviramis and R. cristatus.

Group 5. Seeds ellipsoid to oblong, with the

ridge faint. Surface smooth, faintly ribbed or

colliculate and cell margins visible (Fig. 10A, B, C).

In transverse section it is evident that the epidermis

consists of flat plates, which are either smooth or

carry sculpture, giving the ribbed or colliculate

appearance. Although no lumina are visible in these

plates, they are probably formed from the epidermal

cells. They take up no stain, but show birefringence.

The subepidermal layer is also well-developed, but

again no cellular organization is visible. The tannin

layer is relatively thick.

Chondropetalum esterhuyseniae, C. lon-

giflorum, C. andreaeanum, C. insigne, C.

nitidum, C. paniculatum and C. char-

taceum.

Group 6. Seeds usually oblong, rarely ellipsoid,

with a very prominent ridge. Surface very roughly

rugose, at higher magnifications densely foveolate or

pitted, sometimes with secondary ridges at right

angles to the main ridges (Fig. 9E, F, G). In

transverse section a relatively thick tannin layer is

evident. The subepidermal layer is not distinct. The

epidermal layer consists of tall hollow cells with

somewhat irregular column-like walls, and with the

outer wall pitted. This layer stains lightly with Alcian

Blue and shows birefringence.

Chondropetalum ebracteatum, C.

hookerianum, C. aggregatum, C. mucro-

natum, C. decipiens and C. marlothii.

Group 7. Seeds spheroidal or ellipsoid, with the

ridge weakly developed. Under a hand lens the

surface appears brilliant white. At higher magnifica-

tions the surface is ridged, and reticulate-foveate

between the ridges, rarely pitted (Fig. 9C, H, I). In

transverse section a well-developed tannin layer is

evident. The epidermis appears structureless, and

does not stain at all, but shows birefringence. It

appears to have complex shapes, but is very brittle,

and no really good sections were obtained.

Restio major, R. anceps, R. cascadensis,

R. subcompressus, R. callistachyus, R.

compressus. (It is possible that R.

echinatus should also be included here.)

Group 8. Seeds ellipsoid, spheroidal or oblong,

rarely truncate, with the ridge usually not deve-

loped. They are all colliculate. There is extensive

variation in detail, which allows the recognition of

the following groups:

(a) Colliculae large, conical, with fine ridges

more or less radiating from the summit to the base;

cell margins straight. In transverse section the tannin

layer is narrow and the subepidermal cells com-

pressed. The epidermal cells are large, pale blue

staining cells, with heavily thickened walls and fine

surface ridges (Fig. 8 A, B, D).

Restio similis, R. micans and R. debilis.

(b) Colliculae large, rounded, often with faint

ridges running lengthways across the colliculae. In

26

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Fig s. — Seed surfaces colliculatc: Restio similis: A, x 200; B, x 420; D, x 18. R. quinquefarius: C, x 30; E, x 240; F, x 200. R.

confusus: G, x36; I, x 420. R. filiformis: H, x 30; J, x 400; K, x 600.

H. P. LINDER

27

Fig. 9. — Seed surface pitted and colliculate: Restio pygmaeus: A, x 420; B, x 200; D, x 30. Seed surface pitted and rugose:

Chondropetalum ebracteatum: E, x 30; F, x 600; G, x 200. Seed surfaces ridged and scrobiculate: Restio anceps: C, x 30; H.

x 240. R. compressus: I, x 200.

28

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTION ACE AE

Fig. 10. — Seed surface of flat plates: Chondropetalum paniculatum: A, x 200; B, x 18; D, x 240. Restio giganteus: C, X 18- E x

1200; F, x 200. Seed surfaces tuberculate or verrucate: R. marlothii. G, X 30; J, x 400; K, x 600. R. dura: H, x 36; I, X 120o!

H. P. LINDER

29

several taxa the seeds are more or less triangular in

cross section. The cell margins are straight. In

transverse section the wall structure is similar to R.

similis.

Restio aureolus, R. peculiaris, R. triticeus,

R. inconspicuus, R. quinquefarius, R.

verrucosus, R. stereocaulis, R. sp. aff.

decipiens, R. feminens and R. bifarius.

(c) Colliculae as in the previous group, but with

the cell margins interdigitating extensively and the

bases of the colliculae developed into spurs and

valleys (Fig. 8G, I).

Restio pedicellatus, R. miser, R. confusus

and R. strictus.

Note that R. strictus and R. feminens are

somewhat intermediate between (b) and (c).

Group 9. Seeds more or less oblong, truncate at

both ends, triangular in cross section, with the ridge

along a raised zone. The surface is colliculate, with

regular cell margins. In most taxa the colliculae rise

only from the central part of the cell, with the

margins along a flat area (Fig. 8H, J, K). In

transverse section it is clear that the tannin layer is

relatively wide. No cellular structure could be

detected in the epidermis and subepidermal layer.

The colliculae are solid sausages of tissue that do not

take up any stain, but show birefringence. Presum-

ably they are formed from the epidermal layer. The

layer below stains lightly with Sudan IV, and is

presumably the subepidermal layer.

Restio filiformis, R. obscurus, R. rupico-

la, R. strobilifer, R. brunneus, R.

praeacutus, R. insignis, R. burchellii, R.

pachystachyus and R. bolusii.

This group is closely related to Group 8 and may

not be distinct from it.

Group 10. Seed ellipsoid to spheroidal, rather

irregularly lumpy. Surface either densely or finely

ribbed, or finely reticulate (Fig. 10 D, E, F). In

transverse section the tannin layer can be seen to be

thick. No subepidermal layer was observed. The

epidermis consists of flat plates, carrying the ridges.

Note the superficial resemblance to Group 5.

Restio fruticosus, R. foliosus and R.

rhodocoma.

Group 11. Seed oblong to ellipsoid, rarely

truncate, usually terete in cross section, occasionally

triangular. The surfaces are smooth, although at

higher magnification a range of surface detail may be

observed:

(a) Cell margins more or less distinct, outer

surfaces with slight longitudinal ridges or depres-

sions, seeds round in cross section with a raised

ridge.

Fig. 11. — Seed surface smooth with corky appearance: Chondropetalum macrocarpum: A, x 18; B, x 200; C, x 120.

30

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Restio aff. filiformis, R. fusiformis, R.

decipiens agg. R. dispar, R. arcuatus and

R. leptostachyus.

(b) Surface finely undulate or glebulate or

ruminate.

Restio perseverans, R. pulvinatus, R.

singularis, R. acockii, R. nodosus, R.

communis, R. purpurascens, R. ingens,

R. corneolus, R. scaberulus, R. implicatus

and R. sejunctus.

(c) Surfaces almost perfectly smooth.

Restio bifurcus, R. fragilis, R. egregius,

R. sp. aff. communis and R. involutis.

(d) Surfaces with cell margins visible.

Restio occultus, R. brachiatus, R.

perplexus, R. rarus, R. patens, R.

quadratus and R. sp .(EE 31762)

The walls in transverse section ( Restio dispar, R.

egregius, R. bifurcus), (Fig. 6D— J), show a

relatively thin tannin layer. The subepidermal layer

consists of large, very thickened cells with distinct

lumina. They stain lightly with Alcian and show

birefringence. The epidermal cells are much smaller,

erect, with the lumina almost lost and the cell walls

staining lightly with Alcian. However, not enough

species from this group have been sampled to

determine that this rather peculiar wall construction

is found in the entire group.

Group 12. Seeds oblong, circular in transverse

section, the ends obtuse (Fig. 11). The surface is

smooth, usually with deep cracks, giving the

appearance of bark. In transverse section a very

peculiar cell organization is visible. The epidermal

cells are thin-walled, very elongate and stain brown

with Safranin. The next layer of big cells stains

lightly with Alcian blue and is very thick-walled. It is

not clear whether this is the subepidermal layer or

the supra-tannin layer. The tannin layer consists of

very distinct, variously shaped cells, so that the layer

varies in thickness.

Chondropetalum macrocarpum.

Group 13. Seeds oblong, truncate, in transverse

section, triangular, the surface verrucate. No section

through the wall is available.

Restio papyraceus.

PHYLOGENETIC ASSESSMENT OF THE SEED-COAT

DATA

No published data are available on the seed-coat

morphology and anatomy of the non-African

Restionaceae. A light microscope study of seed of

Anarthria scabra and Lyginia barbata showed that

the same basic organization is found in the outgroup

of the African species, but that the detailed structure

differs. It is therefore not possible to establish any

polarity for the various characters found in the

African taxa.

In the absence of any polarity determination, each

‘phenetic’ group recognized is taken to represent a

monophyletic group, equivalent to the terminal units

on a cladogram. In some groups this hypothesis is

probably correct. Group 12 has an epidermal layer

not found in any other group, or in the outgroup

taxa investigated. On the other hand, Groups 13, 9

and 8 may form a single group. The minimum

hypothesis would be to leave the thirteen groups as

terminal taxa on an unresolved polychotomy.

CULM ANATOMY

Introduction

The culm anatomy of the Restionaceae is unusual

and highly specialized, presumably to cope with

seasonally arid environments and perennial, photo-

synthesizing culms. That they may be of great

systematic importance was first realized by Gilg

(1891) and supported and elaborated by Cutler

(1969). Gilg commented (1891: 594) ‘Nach der

anatomischen Methode dagegen ist das Feststellen

der einzelnen Gattungen meist ungemein leicht und

in sehr vielen Fallen lasst sich die Bestimmung auch

bis auf die Art durchfiihren’.

Masters (1865) studied the root anatomy of Restio

triflorus and the culm anatomy of Restio ferruginosus

(= R. gaudichaudianus). His anatomical observa-

tions are poor. Pfitzer (1869-1870) investigated the

culm anatomy of several Australian and African

species and observed the peculiar anatomical

features, which he described and related to

xeromorphism. This stimulated Gilg (1891) to do a

much more detailed study. Gilg related the great

diversity of structure to adaptation to desiccation

and produced excellent anatomical descriptions of

the groups. He also, in the mould of Engler,

interpreted the systematic implications of the

anatomical data. He suggested that no African

genera occurred in Australia and also that the

Hypodiscus - Willdenowia groups needed reorgani-

zation — comments repeated by Cutler (1969). He

was highly critical of Masters’s (1878) taxonomic

treatment of the family. Gilg’s recommendations

were largely implemented by his wife, Gilg-

Benedict, in 1930, but were totally ignored by Pillans

(1928).

Cutler (1969) produced a very detailed anatomical

survey of the family. The study is largely descriptive,

with very little adaptationist interpretation, but with

numerous excellent comments on the affinities of the

groups and the correct interpretation of structures.

Jordaan (1946) and Noel (1959) described the

anatomy of a few species. Under H.P. van der

Schijff, the anatomy of Elegia, Willdenowia and

Hypodiscus was further studied, resulting in papers

by Van Greuning & Van der Schijff (1973) on

Willdenowia and Hypodiscus, Botha on stem

anatomy of Elegia stipularis (1971), Botha , Van der

Schijff & Van Tonder (1972) on the ontogeny of

Elegia stomata; and Botha & Van der Schijff (1976)

on the ontogenic differentiation of the parenchyma

of Elegia.

H. P. LINDER

31

Cheadle (1955) and Cheadle & Kosakai (1975)

described and discussed the degree of specialization

of the vessels of the Restionaceae and the

Centrolepidaceae. However, their data are difficult

to assess, as average values are given for each genus,

and this paper is directed at a critical assessment of

the generic limits.

Materials and methods

This study is largely based on the anatomy slide

collection at the Jodrell Laboratory, Royal Botanic

Gardens, Kew, on which Cutler also based his work

(1965, 1966, 1969). Where adequate descriptions or

pertinent data were available in the literature, these

were also used. In addition, critical taxa, for which

no data were available, were sectioned, and the

slides are housed in the Jodrell Laboratory. Sections

of 186 species were studied, many of which are

represented by several different collections. The

material used, with vouchers, is listed in Appendix

3.

As adequate data over a sufficiently wide range of

taxa were available only for the culm anatomy, only

this aspect is reported on here. Descriptions of root,

rhizome and leaf anatomy may be found in Gilg

(1891), Noel (1959), Jordaan (1946) and Cutler

(1969). As Cutler (1969) gives excellent descriptions

of the anatomy of many of the species used here, no

full descriptions are given here.

Variation in culm anatomy

The culms of the Restionaceae are the major

photosynthetic organs. This feature, associated with

the xeromorphic structure and the needs for

mechanical strength dominate the organization of

the culms.

(a) Epidermis (Figs 12, 13)

The epidermis is usually a single layer of cells,

except in two genera, where there are two distinct

layers (Botha & Van der Schijff, 1976). The shape of

the cells is quite variable, ranging from square to

much longer than wide. The side walls vary from

being straight to flexuose and may be much

thickened. This results in the range of epidermal

types shown in Fig. 12.

Various kinds of outgrowths may occur on the

epidermal cells (Fig. 13). Tubercles are formed

when some cells are taller than others. Often

stomata are found on top of the tubercles. Peculiar

‘pimples’ are formed on some cells — these are

unicellular outgrowths, which may have several

forms, i.e. the ‘hairs’ of Thamnochortus argenteus

described by Gilg (1891).

(b) Stomata (Fig. 14)

The stomata are all paracytic. Cutler (1965)

discusses variation in the type of margin found in the

guard cells, but this was not investigated. In the

majority of the species the guard cells are flush with

the outer wall of the epidermis, whereas in a few

species they occur at the inner epidermal wall, or

occasionally in an intermediate position.

(c) Chlorenchyma (Fig. 15)

The chlorenchyma consists of 1—3 layers of cells,

which are more or less palisade-like. Generally there

are two rows of cells. A single row occurs

occasionally. When there are more than two rows

their organization tends to be variable.

The shape of the cells varies from more than 10 x

longer than wide to less than 3x longer than wide.

The short types often lack pegs and may even have

irregular shapes. The inner and outer chlorenchyma

layers may be dissimilar, usually with the inner layer

having wider and shorter cells. This may be

| manifested in its extreme form where the inner layer

,in a transverse section appears dumb-bell-shaped.

In a few species peculiar changes have occurred.

The inner or outer chlorenchyma layer may become

lignified. In one case the outer layer is lignified, and

the cells much enlarged relative to the inner layer

(Leptocarpus rattrayi Pillans).

(d) Protective cells (Fig. 16)

The substomatal cavity is, in the African

Restionaceae, lined with lignified or partially

lignified cells. Pfitzer (1869) called them ‘Schutzzel-

len’, whereas Cutler (1969) coined the term

‘protective cells.’ Van Greuning and Van der Schijff

(1973) follow Botha (1971) in calling them ‘lining

cells.’ There protective cells are developed from the

chlorenchyma cells. They are variable in number,

size and shape, but basically three types of

organization may be recognized:

1. The protective cells occur in two rows, roughly

equivalent to the two rows of chlorenchyma cells.

The inner tends to be less well developed than the

outer, but usually reaches the parenchyma layer

(Type A).

2. Only one layer of protective cells is developed,

the cells of which are usually somewhat longer than

the chlorenchyma cells, and bone-shaped, forming a

basket not reaching the parenchyma (Type B).

3. Only one layer of protective cells is developed,

the cells of which are short and squat and in the

extreme case, only lignified on the outer surface.

These cells tend to be identical to the chlorenchyma

cells (Type C).

It is not always possible to distinguish the various

types from one another with absolute certainty, as

they intergrade to some extent.

(e) Parenchyma layer

The parenchyma layer varies from one cell wide to

up to five cells wide. If only one cell wide, the cells

are usually remarkably uniform. If the sclerenchyma

ring forms ridges, these usually interrupt the

parenchyma layer, or alternatively result in a

reduction in the size of the parenchyma cells.

In the case of two species, the inner and side walls

of the parenchyma cells in some areas have become

much thickened, so that it appears as though the

parenchyma is being included in the sclerenchyma.

(f) Sclerenchyma cylinder (Fig. 17)

This structure varies extensively in thickness, but

32

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Pig 12. — Variation in the epidermis. A, a single layer of square cells, Restio ambiguus, x 100; B, a single layer of oblong cells,

Restio obtusissimus, x 100; C, a double layer of cells, Chondropetalum aggregatum, x 100; D, a double layer of cells, with the

upper layer partially lost, Elegia neesii, x 100; E, Thamnochortus-type cells, with the anticlinal walls wavy and thickened,

except at the base, Restio setiger, x 100; F, pseudo- Thamnochortus-type cells, Thamnochortus comptonii, X 100.

the variation may be more related to the age of the

stem than to intrinsic genetic differences. More

significant is the development of ridges and girders.

Generally this occurs opposite the vascular bundles,

presumably as the result of a thickening of the

sclerenchyma cap on the vascular bundle. However,

in most African taxa the ridges and girders alternate

with the vascular bundles. The ridges vary from

slight swellings, to humps, to ridges that penetrate

into the chlorenchyma layers. In some taxa the

ridges reach the epidermis, whereas in others they

stop short and the intervening chlorenchymatous

cells are lignified. These cells are analogous to the

pillar cells recorded from Australia, but as the

Australian pillar cells are developed from the

parenchymatous sheath, they cannot be considered

to be homologous (Cutler, 1969). I term them ‘false

pillar cells.’ However, it cannot be ruled out that

they are ontogenetically derived from parenchyma

cells, that would otherwise lignify to form the ridge.

(g) Central ground tissue (Fig. 18)

The central ground tissue (cgt) is quite variable,

mainly in the thickness of the cells’ walls, the

H. P. LINDER

33

Fig. 13. — Epidermal processes: tubercles. A, Restio zwartbergensis, x 100; B, Hypolaena diffusa, x 250; C, Restio marlothii, x 100;

D, Staberoha ornata, x 100; E, Restio madagascariensis, x 250; hairs, F, Thamnochortus argenteus, x 100.

34

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Fig. 14. — Stomata! types. A, guard cells superficial, Willdenowia

stokoei, x 250; B, guard cells near base of epidermal cells,

Cannomois dregei x 100; C, guard cells at base of epidermal

cells, Restio compressus, x 250.

distribution of cavities in it, whether the cavities are

caused by mechanical tearing of normal cells, or

whether they are the result of special thin-walled

cells breaking down, and in the distribution of the

vascular bundles through the tissue.

Generally, if the vascular bundles are evenly

scattered through the cgt, no central cavity is

formed, or if it is formed, torn vascular bundles can

often be seen at the edge of the cavity. In a small

group of taxa, small clusters of thin-walled cells

occur scattered between the vascular bundles in the

cgt, resulting in a mosaic of vascular bundles and

small cavities.

In another group of taxa the vascular bundles are

arranged approximately in a ring round the

periphery of the cgt, and in the centre of the cgt is

usually a large cavity without torn vascular bundles

at its edge. These central cavities are then usually

the result of thin- walled cells breaking down.

(h) Silica (Fig. 19)

The presence, nature and distribution in the

various tissues of silica is quite variable. Silica may

occur as sand and/or bodies. The bodies do not have

the rich diversity of shapes associated, for example,

with the silica bodies in the Poaceae. The silica

bodies are generally found in cells on the outer edge

of the sclerenchymatous cylinder (often on ridges) or

in the parenchymatous sheath, while the granular

silica is found either in the central ground tissue, or

more rarely in the chlorophyllous tissue or the

protective cells. Cutler (1969) described the special-

ized cells on the edge of the sclerenchyma in which

the silica bodies occurred as ‘stegmata.’ It is quite

likely that stegmata are ontogenetically derived from

parenchyma cells that have become lignified.

The distribution on the above character states

throughout the African Restionaceae is summarized

in Table 4.

Phyletic implications of anatomical variation

The distribution of the various character states in

the non-African Restionaceae is summarized in

Table 3. From this table, character states unique to

the African Restionaceae can be identified by

outgroup comparison. In addition, some estimate of

the degree of convergent evolution can be made,

allowing the establishment of levels of confidence in

various characters.

The results of these analyses are summarized in

the cladogram in Fig. 20. The cladogram is discussed

in detail below. For convenience the terminal groups

have been labelled alphabetically.

Group A

This anatomically fairly homogeneous group is

delimited by the possession of two epidermal layers.

Although there is variation in the development of a

central cavity in the central ground tissue, the width

of the parenchyma layer, the position of the guard

cells in the stomata and the shape of the epidermal

cells, no discrete subgroups are evident. In general,

species attributed to Elegia frequently have a central

cavity and tend to have square epidermal cells, but

there are exceptions. Elegia also usually has fewer

protective cells per substomatal cavity.

The position of the guard cells is difficult to

H. P. LINDER

35

Fig. 15. — Variation in the organization of the chlorenchyma. A, chlorenchyma in two scarcely differentiated layers, Restio foliosus,

x 100; B, chlorenchyma in three mdifferentiated layers, individual cells oblong, Hypodiscus alternans, x 250; C,

chlorenchyma in two layers, the inner ayer lignified, Staberoha remota, x 100; D, chlorenchyma in two layers, cells of the

outer layer much bigger than in the inner layer, lignified, Leptocarpus rattrayi, x 100; E, chlorenchyma in two layers with the

inner layer very reduced and ‘dumb-bell-shaped’, Restio rottboellioides, x 100; F, chlorenchyma in two layers with the outer

layer lignified, Willdenowia stokoei , x 100.

36

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

H. P. LINDER

37

Fig. 17. — Variation in the sclerenchyma cylinder. A, small humps alternating with the vascular bundles, Hypodiscus alternans, x

100; B, sclerenchyma ring simple, Restio mahonii, x 100; C, ridges reaching to the epidermis, Phyllocomos insignis, x 100; D,

sclerenchyma ridges not reaching to the epidermis, Hypodiscus neesii, x 100; E, sclerenchyma ridges terminating in false pillar

cells, Willdenowia sulcata, x 100; F, sclerenchyma girders present, Anthochortus ecklonii, x 100.

38

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Fig 18. — Variation in the organization of the central ground tissue. A, vascular bundles evenly scattered, no cavities present,

Chondropetalum aggregatum, x 16; B, vascular bundles evenly scattered, small cavities scattered, Thamnochortus comptonii,

x 40; C, vascular bundles in a ring with a central cavity, Hypodiscus binatus, x 40; D, stem flattened, no cavities, Restio

cascadensis, x 60; E, stem flattened, central core with satellites, Hypolaena graminifolia , x 60.

H. P. LINDER

39

Fig. 19. Distribution and nature of silica, A, bodies in the parenchyma layer, Restio setiger, x 250; B, bodies in the chlorenchyma,

Restio rottboellioides, x 250; C, bodies on the edge of the sclerenchyma, Hypodiscus neesii, x 250; D, granular silica in the

central ground tissue, Restio pedicellatus, x 250.

quantify in the group as, in addition to being on the

outer or inner wall of an epidermis cell, it also varies

from being on either the outer or the inner

epidermal cell layer. There is much more variation

in this character than is usual in the other groups.

Generally, where a central cavity is formed, it

appears to develop by tearing through the vascular

bundles, which are almost always evenly distributed

through the central ground tissue. In one or two

species of Elegia (i.e. E. neesii ) the outer epidermal

layer is reduced to occasional tubercle-like cells (Fig.

12).

Within Chondropetalum two groups can be

recognized by the length of the chlorenchyma cells

and the shape of the epidermal cells.

TABLE 3.— Distribution of anatomical character states in the non-African Restionaceae

40

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

< 5

os

IS

H <

a> £

fi S

> <

& S

.2 S

c t!

<L> §

<D O

s

Q <

C

M <1*

a> ^

^5 o

s-g

0$ 3

a *

£ <

« -a" £ ~ a

* Ik. C3 k

O M M ^

a « g <u <u s 22

~ c £■ e e c- c

=3 O Q HI O t3 O

< Z a; o o ^ z

#o ^

*3

C-g

O ^

-o a

-S ^

5 .a

6 °

3 3-

•s •« -a

111

1^-

0 0 2

&

« o. -o o

s b §

^3 -2

« O q

0 0(0

■SO.

> . ■*>

o .« 3

OG *3

.s .s «

SS2$go

3 o -a ss £

o. a <“ 4J S"r<3

Cfl w G ? ^ O

•g 3

'Eog.

S‘8

2-5

(D

cd

G 02

73 9-

<u "5

O O

13

S> 4>

oz < z

d.o

Oh ^3

a ^

B

■i

g. §

S* G *

5 S r o

. nj Cl<J (73

•as d . «

c ■a ,g •§ 73

■g S J g. 8

o. c So c- c

O 4> ta 0>

^ o ^ai o

+2 K

5? *3

*1

•a §.

13

St

| -a

|.S

II

$5 O

1-2

3 ^ "S T3

s € .a « S

— . s; C — i w

2 a ■« 2 S'

<u 'S >Z d Si <u

C g <3 C 2 -o

U g S <U O -S

OU^OZ^

c g

=3 =3

cd ctf

& &

13

c g 2 ^ ^

5 g § .a .y

<L> ,0

—a CT3

O H B- < ifl < <

C Miljj 5-

O m Om

4> JO

5 3

a 5

O j-

O M

Oh O

Oh GJ

o <

« 6

>» o

0) O

i-h TlJ

ed ^

Oh U

e s "3

w £ <

■M i

o 2

ca O

•S2 o

E a

X)

a e

Q. 3

I!

2 o

to cu

3 E

•s: 3

8 Z

4: £

e o

w e a 3

|| 11

£0 8 >

$ «7J

to Z

TABLE 4. — Distribution of anatomical character states in the African Restionaceae

<N

cooacQQaQSCQoapaQapaaaeQQafflCQCQCQCQcQCQQaeQDacQoaDOOQCQCQCQCQ

•O ^ </2

•< < < <t| <

U O U U

z z z z

oooo«ifioooooon'Oooooooooo\ooooo<trno»ooooooooo

ooooooooooooooooooooooooooooooo

<o

(NtS<S(NNNr)(SM(N(NN(NM(N(SMM(S(N(NN(S(N | (N N M (S M (N

<N

^nr/if/iPQ^f/innnnOOOOOO

00v000OOOV0OO00OOO00Ov0<0>n00v0O00

a.a,00tt-0000ft.000fl-00000fl<00

(S(N(S(SfSN(N(S(S(N(N(N(NMNM(N(N(N(N(N(N

VO O O VO vo

- U U U U

“■zzzz

^ (N _ (N

OwOOOOOOOOOOOOOOOOOOOw on on c/2 c/2 c/2

+ + + + + + + + + + + + + + + + + + + + + + + + + -f + + -f + +

SSSSSHHSHSh H S HfflHHcQhhSSHHHhhhfflSS

>o(N co <o ro <n <n men wn m (N <o <n co co <o ■*}-

I I <N <N | | | | | CJrlNO | | <N | <N <N | | C- | | N-. | | | |

<o — ndN""" <n cn <o — » , — h <n — ri (s N m

+

«"•«"• + + + + + + + <>• + + + + + + + + + -I- + + -I- + + + + -I- + +

+ «"• + + + + *-+ + + + + + + + + +

<-• C” c-> c-- c** c- <-•

+ +

**• + + ='•

+

ao-s

2^)

00 s;

.L. 3

3 .£- O

£ ts P

:g 1 ! | § | 3

o cy

a 3

r* •*** ^ <0 <L> v ?-

^

•5 s

S? £

a-

S «5 S

$>•, a •?

II

» & 3

a o

c s ■ o

a.

D

O

- - . - . - ~ —

O 5q ki kj ki kj kj ki ki ki ki ki ki ki ki ki ki kj ki kj kj 6 U U U U U U U U U U

ST/J3 ^ K

„ , s p' 5 2 |

■5-S^f | §|

as 8.S 2

-V C>

0 O P

-* vj x.

O 3)

5

:~ | e-

•s « 8

s S? p

~ vj

£ -3

fiP g

+ +

+

+ + + + H — (- + + + + + + + H — (- + + + + +

SHhhll-'hhhhHShf-HhhhhHSH

<o on m

I ^ <N I CN | I I I I I Iwwl I — <N | CO <N CN

<N <n — ro — — • -* <n -* — • cn -* <n

+ + + + + + + + + + + + + + + + + + + + + +

£

e x.

CQ .3

a

o

cu

D *=

O 5

06 g

O K

p .a

£?-S

- - 5 g

alaS-i-

g | O S K -C

g" *>• «o

-S 3 3

ji 3 !g a d a .a

■SP G '3 -S ^ £ a

3 -3 -3

3 5 ^ -

‘C .sp o' ’5 -5 -5 O £

60 E* Cl"3-3 £ a -5 a £ O a 33 Sb a

£ C

a. 2

1 3 <j) o

:^£-§

o <5> v

CQ 05 03 CO *S

<N 04 <N

+ + + + +

P)

Q. c

s I

cu

D

o 'C O 3 3 3

S ts O 5 O !o

DC

O O' O' O' O' O'

<<<CQCQ <<<<<<<<<OJ<<<<

<<<CQ<<<<QQ

« — (N _H _ — — — (N — ^ ^ (S — — — — — — — — <N — — — — ^ _ (N (N — — — ~ _ fS —

SSSSSSSS^^SS'^S'^’S0000'^00'^’'0'0'^’'^"10'^'1000000 0,,1'i'0'f®o0'o^oo'0(ioooix)ooovoo«'on

OOOOB.OOOO.OB.OOOOO O O O O z z O O ^ O O O O O O 0- o O O O z O O a, a, a, o O O O O O O O O

(N

<N<N<N(N(N<N<N<N<N<N(N<N<N<N<N<N<N | CJn(N--(SMn(N(Nn(N(N (N(N(N(N(N(N-(N(N(NM(N(N(N(N(N(N(N(N(NMMrJ

>>>>>°>c/3cni/5>>><P5^^0ana)onOaoon(/20000 0on0 0cn0^000coonco^^^00ononon0^0onco

c- + +

+

+ + + + +

+

+ + + + +

hhhhhhf-hhhHhhhHWhhhfflHhhhl-'HHh HHH hhhhhhhhhhHHhHHhHhhfflHhh

m m m n N

<N (N <N

1 1 1 1 1 r.«cc-"rnw«s«N"r«M--7c««‘f'f--'rci-M cM-??77b'aatjcM7'i,T7-7-77«N-«

(N(N(N«« — - <N (-) — — — — — — rstN — <n — — — — (S-NM J, J, r!)

m cn Tt m tN fs m

+ + + + ++ + ++ + «"• +4- + + + + + + «"•

+ + + + + + + + + + + + + + + + ^- + + + + + + + + + + + + + + + + + c-

+ + + + + + + + + + + + + + + + + + + + + + + + + +

+ + +<>■ + + (-.

+ + + + + + ++ + + + + e>~ + + + + + + +

a

a a

a 000 a a

a a o a

a o

000

a

E2. 5

■§ 3

I 5 3

c 5 s E

S ^ | £ 3

II. 2 s '§>•§^<2. 8 J-§5

^ § §--2 2

- cl; aoo'sa.o - ~

cl *

s 8=aSi

S5 1 ^ ^ i a a

2So's§?-ca

>> 'K. !U .3 J3

o 60

g;'So g a.-* •§ -g

S n cn Q t3 <3

bo

VT.

•3

S 3

« 3

«■§ S"3

5 3

2 =s S

05

cl -c> cl ^ ^-^3 d

.S3

CJ,<U 'i

CS Nk C/

Cr ^

•S £

3 •=

3- £

2 f.

v' p -

•3 ^2

a. "3

Cr ^ ^

0 3 P

£ -v

P O <u o

£ £

■a 3 c 5

c ^ o S

>> 3 a. 2 o

S £ §■ -SP

O

DC ^

O O U tj <0 CJ (2 O' c< O' Ce; o' o; o; o; o; a; qq q; O' o; o' a* a; aj a; c< o' o' o' < a; < o ' < q; dq' q; o' oj c<' <i L L L -4 o' o' o' o' o' o'

a.

< < < < <

u

z ~

VO VO 00 00 VO

O CC CL, -4 Oh

CN<N<NCN<N

o o ^ o n

> > 00 > °

hhhhH

ro m <n m <n

I I I I I

<N <N « <N — i

+ + + + +

3 ^--H

lf,S

las

p

2 ^ *

Co Co Co I

<<<<<<<<<<<<<<<<<<<<

(N«-<<N-i-i(*)^rt222^N — , <N — • <N — i

Tj-OOOcocoOOfNOOO^ooioioO^iooo

a,OavOvOOa<QQQQQQtLa,OBvOaHO

»OOOOi«Om<« iO^OOmwOOmO

C/2

^ + + + + + +

+ + + +

+

+ + + +

+ + +

--a. Ci<N

+ + + + + + «"• + =- + + + + + + + + + + +

+

+++++ ++++ ++ ++++++

a a a a a

OQCQCQCQpaCQCQCQ CQ CQCQOQCQOQCQCQCQ

33

S. 2

a 3

■3

l, .§ 3

3 3 «

1 O Q

D o js

0 'S o

a; 2

a i< o'

s 2 o s

•s? &

•S3 C\ o ~ .»

o ^3 a. c £ s

^ §• 2 -c o g

•3 If ^ ^

S al -2

^ a

c<(^k;i^^cc;c<ik;k:k;c<c<c^c<c<c<cL;k:

V'; ; rTrrr-' ; ■

00

<N<N<N<N<N<N<N<N<N<N<N<N<N<N<N<N<N<N<N<N

o

o o

H H CQ

<N _ <N

+ + +

5

O =3

£ 5 §

3 .5

2

I Si Si

<N <N

o o

+ 7

<N

O co

H H

+ «"•

+ +

so

• 4j tc;

<C<03^

O O O 00 o

000

<N<N<N<N<N

o o < o

o n o

> 0 >

+ + e-

CQ

hhh I h

S

_ _ „ _ <N

+ + + +

+

+ + + +

+

3 ? ^

•2 S

L. -3

.s

cl o

P ^

2 s

H <r>

S S' S' s s

u <

+ +

K d

I -2

a. ^

>>

3: LT. ^

O to I<

nOOOO

0 > > > >

H H s H H

+ + + + +

+ + c- +

+ /

CQ CQ m CQ CQ

C>0

S 3

^ -a 2

O .5 ... 2

§ § 8 1 -8

3 3 ^3 a 3

o <3 u <0 u u

’

< < < < <

VO O O O VO

OOOOO

<N<N<N<N<N

O O

CQCQOQOQOQOQCQCQCQCQCQCQCQ

OOOOOOOOOOOOO

<N<N<N<N<N<N<0<N<N<N<N<N<N

MM^MWOO^WMonwO

+

+ +

+++++++++++++

++++++++ ++++

+++++++ ++++

+

QacQoacQoacQoaoaoacQoQcQCQ

Co .Co

a a o

3 -s a

3 3 33

Si -3 O

s o p s

§. -2 § -s S ^ S §

a

I' Hi c<

a 32

^335

S3 s? si tij

+

+ +

z 3

o ^ S;

si

O ^

C< 3

O X

■■■

CQ CQ 05 CQ

u

1 z

<0 VD KI vf

o 000

<N <N <N <N

C/2 C/2 O O C/2

H H H H H

+ + + + +

+ + + +' +

+ + + + +

CQ CQ CQ CQ CQ

3 o 3

3 O a 3

£5 e .g s « i

O 23 4 ® i °

2 a ^ a o a

o S S nj t?

Cell shape: S = square; O = oblong; VO = very oblong.

Degree of differentiation between outer and inner chlorenchyma layers: O = none; P = partial; D = inner dumbbell-shaped.

Protective cell type, see text.

H. P. LINDER

43

H31DVHVH3

'A JflOHO

r <moao

I dflOMO

o dnoao

h dnoao

j dnoao

3 anoao

a arioso

3 dnoao

a df!030

v dflOMO

■O 't i-

O 0/j

r„ g .£

8 v !2 £ g

J>~ «> % E

> - 2 w

“5 ea 00 Cu —

u o . ► (S

g^s S.S* ss

|/s|^ S3

& 8 2 § «

■C O U’SS |

C c O „ S'

“5 a £ c g

to ■> = >> g

* u ^"5 Si 5

•o S & g 2- a

S~«2s = »

s" , O. « J3

gf- C -C "

Si § ‘.s'~ 3 .£

2 .s a> s «.

oo > .£ =5 •-

2 _ g 5^-5

<a co '■> u g o

to E o .* j? x.

g 8

o ' ^ ~ 2

° o 2 T> u .

.2 a J £ -J c

1 1 s .§> § I

^ g sg ~ g

m E \s O — ca

2 i- c ca c

a O-o = c

-o c g 2 u S'

_ <u 2 E op-g

ca a. £ S' ^2 c

-« 00"2 Q3

E — -3 C <3 «

ou S " E o.

« ° C ,2 J? 2

C .52 <U rr o

^ C <-> o c

oS^ugC

4= <L> ^ W 5 ^

" S2 c 3 o <u

c a. <u o <« :=

o «-> -^ . ’2

_ - c o ^ 2

2 • - J« ■ - iT £0

5S1 g §:l

; ^ « o. S 2

I S' D.^3 T3

liSagc-

! g = X> " S

! “ ° »u g,

■ fo -o *? o <r

: o- S E 2 c

i ^ 3 <l> £ a>

; o „ 'S.^ *5

i >- <r> g* U ^

ic-O o 2 M

O 'O

£ s §

£ <u

OJ • —

1-. *->

a. -

S <*

^ ^ c - ^

CO « J£ £

fg 2 -S H •?

44

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Group B

This group includes Thamnochortus and a few

species of Restio. It is distinct by the peculiar

epidermal cells with thickened wavy side walls and a

very small rounded thin-walled section at the base

{Thamnochortus- type epidermal cells), and by

scattered cavities throughout the cgt caused by the

breakdown of small clusters of thin-walled cells.

Several species in the group lack one of these two

characters, but are included on the strength of

having the other character. In addition, similar

states occur outside the group, but never quite as

clearly developed. Less clear characters are the very

long and slender chlorenchymatous cells, with the

inner and outer layers usually identical.

Variation within the group does not form any

patterns. One subgroup could be two species with

peculiar unicellular hairs. (Thamnochortus argenteus

and T. fruticosus). Other variation affects the

number of cell layers in the parenchyma and the

position of the guard cells.

There is some difficulty in delimiting the group in

the species attributed by Pillans (1928) to Restio, but

it seems fairly clear that R. fruticosus, R. foliosus

and R. rhodocoma belong here, even though the

latter lacks the thamnochortoid epidermal cells.

Group C

This rather distinct group has flattened culms and

guard cells placed at the very inner surface of the

epidermal cells. The overall aspect of the culm

anatomy is very uniform. In addition, in most of the

species the chlorenchyma has been reduced to a

single layer (at least in the compressed portions of

the stem) and in four of the six species included the

parenchyma is a single cell-layer wide. This is clearly

a very distinct group, which shows only weak

affinities to some subgroups of Group D.

Group D

This group is here one of convenience, including

54 taxa and a range of anatomical structure.

Although several groups may be distinguished, there

are no synapomorphies for the whole group.

One large group of species, belonging to Restio

and Chondropetalum, is remarkably similar to

Group A, but lacks the double epidermis. This

group also includes a set of Restio species which have

shorter and squatter chlorenchyma cells and may

sometimes differ in general facies. Although there is

general similarity in appearance, this group does not

have any unique synapomorphies.

Restio madagascariensis and R. mahonii are

distinct by the single layer in the chlorenchyma. This

character links them to Group C, but they have

superficial guard cells. Although they are both

rather peculiar (R. madagascariensis has peculiar

pimples on the epidermal cells), they cannot be

placed on anatomical data.

Restio quadratus has rectangular stems with

sclerenchyma ridges at the angles extending to the

epidermis. These are clearly not homologous with

the ridges found in Group H and the species is

anatomically isolated.

Another large group of species have granular

silica, usually in the ground tissue, frequently a

central cavity in the central ground tissue and the

vascular bundles arranged approximately in a ring.

The phylogenetic value of the granular silica is

difficult to assess. Its relative uniqueness may

suggest that it is a derived character, but the

presence and distribution of silica in the African

Restionaceae is complex, and I do not understand

the pattern. The other characters are plesiomor-

phous. The group shows a subgroup based on the

presence of well-developed tubercles. However, R.

stokoei in the first subgroup also has tubercles.

Within the group, there is further variation in the

organization of the central ground tissue, the shape

of the chlorenchyma cells, the number of layers in

the parenchyma and the position of the silica

granules. On the whole, it is an unsatisfactory group,

but it is difficult to resolve it.

The genus Leptocarpus falls into two groups; one

has scattered cavities in the ground tissue and

pseudo-thamnochortoid epidermal cells, and the

other has a central cavity in the ground tissue. Both

generally lack silica. The distinction between the two

groups is blurred by the difficulty in deciding

between large not-so-scattered cavities and a single

small central cavity. It is not possible to place the

taxa with any more certainty, consequently they are

also included in Group D.

Groups A — D

The complex of Groups A—D, which in the

cladogram are shown as a polychotomy, are linked

by the even distribution of vascular bundles

throughout the central ground tissue — a character

which holds for Groups A— C, but only for about

half of Group D. In general, silica does not occur in

the group, except again, for part of Group D (largely

the part which has the vascular bundles arranged

more or less in a ring). Silica bodies never occur.

The parenchyma layer is generally several cells wide,

but there are numerous exceptions.

It is clear that Groups A-C are well corrobor-

ated, but that Group D is problematic. Although a

few subgroups may be taken out of Group D, there

is still an unresolved body of species. Consequently

these have been placed together (Group D), but no

meaning can be attached to this ‘group.’

Group E

Staberoha does not have any distinctive synapo-

morphies, although the culms all show the same

general facies. In 5. remota the inner layer of the

chlorenchyma is lignified, while in S. ornata very

peculiar unicellular pimples occur. However, the

group as a whole cannot be placed on the cladogram.

Group F

Restio p.p. (the species of Restio with two styles)

is anatomically distinctive by the presence of silica

bodies in the parenchyma of the majority of the

species. The other distinctive structures, i.e. a

central cavity in the central ground tissue, a single

layer in the parenchyma and type A protective cells,

are probably symplesiomorphies. Consequently it is

not possible to place the group on the cladogram.

H. P. LINDER

45

unless the silica bodies in the parenchyma are

interpreted to be homologous with the silica bodies

in the sclerenchyma of some other taxa. However, it

is debatable whether silica bodies can be regarded as

an apomorphy. A peculiar feature of many species

of this group is the dramatic diversification in the

shape of the inner and outer chlorenchyma layers. In

its extreme form, the inner layer is very reduced and

in transverse section the cells look like dumb-bells.

Group G and I

These groups have few distinctive features, but

are linked to Groups H— P by the presence of silica

bodies on the margins of the sclerenchyma, a single

parenchyma layer and a central cavity. Group G is

distinct by the presence of short chlorenchyma cells

and type C or type B protective cells.

Willdenowia stokoei is somewhat anomalous

within group I, as it has very weakly developed

sclerenchyma ridges alternating with the vascular

bundles, which links this species to Group M. In

addition, the outer chlorenchyma layer is lignified,

but without any other differentiation from the inner

chlorenchyma layer and there is no trace of the

protective cells.

Group H

This small group also lacks sclerenchyma ridges.

However, the chlorenchyma is like that of group O,

very short and squat, while the protective cells are

type C. In addition, the culms of Hypolaena

graminifolia are flattened, with two subsidiary

sclerenchyma rings flanking the main ring.

Group J

The two species in this group show peculiar

thickenings on the inner walls of the parenchyma,

alternating with the vascular bundles, which links it

to Group M. However, both taxa lack silica bodies.

It is possible that these two species are better

grouped with Group D.

Groups K and L

These groups have sclerenchyma ridges which are

variable, similar to the situation in Group M. They

are distinct from Group M by the long slender

chlorenchyma cells and the type A protective cells.

Leptocarpus rattrayi, the only member of Group

L, has a lignified outer layer of the chlorenchyma,

and the lignified cells are almost twice as large as the

cells of the inner layer.

Group M

This group is defined by small humps in the

sclerenchyma, which, although they interrupt the

parenchyma, do not penetrate into the chlorenchy-

ma. In addition, the chlorenchyma cells are squat

and the protective cells are of type B. The facies of

all the included taxa are rather similar, except for

Willdenowia arescens and W. luceana. In Hypodiscus

neesii the humps are somewhat better developed,

and H. willdenowia has a compressed culm, with

ridges from opposite vascular bundles reaching the

epidermis.

Group N

The taxa included in this group all have rather

different facies, but in all of them the sclerenchyma

KEY TO THE GROUPS

1. Ridges in sclerenchyma cylinder alternating with the vascular bundles 2

Ridges absent, or when present, opposite vascular bundles 9

2. False pillar cells present Group P

False pillar cells absent 3

3. Ridges reaching the epidermis 4

Ridges not reaching the epidermis 6

4. Ridges developed into girders Group 0

Ridges simple 5

5. Culms flattened, long ridges opposite vascular bundles Hypodiscus willdenowia

Stems terete , all ridges alternating with vascular bundles Group N

6. Ridges present as lignification of the parenchyma layer Group J

Ridges present as humps in the sclerenchyma 7

7. Outer chlorenchyma layer lignified Group L

Outer chlorenchyma layer not lignified 8

8. Chlorenchyma cells more than 5 x longer than wide Group K

Chlorenchyma cells less than 5 x longer than wide Group M

9. Silica bodies present in the sclerenchyma layer 10

Silica bodies absent from the sclerenchyma 12

10. Protective cells in two rows, usually reaching the epidermis Group I

Protective cells in a single row 11

11. Protective cells lignified only on the outer surface Group H

Protective cells bone-shaped Group G

12. Silica bodies present in the parenchyma Group F

No silica bodies present 13

13. Epidermis with two layers Group A

Epidermis with a single layer 14

14. Epidermis cells of the Thamnochortus type; central ground tissue with scattered cavities Group B

Not with the above combination of characters 15

15. Stems compressed, guard cells at base of stomata Group C

Not with the above combination of characters Group D & E

46

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

ridges reach the epidermis. The group shows

similarity to Groups M, O, and J.

Group O

Only Anthochortus ecklonii is included in this

group, defined by the peculiar sclerenchyma girders,

which are developments from the ridges. The

peculiarity of this group was already noticed by Gilg

(1891).

Group P

This group includes elements of Willdenowia,

Hypodiscus and Hypolaena and is defined by the

presence of false pillar cells. However, this

distinctive feature may be variable within taxa, e.g.

Willdenowia affinis (Van Greuning & Van der

Schijff, 1973). Taxa such as Hypodiscus neesii and

H. sulcatus may provide a link to Groups N and M.

Groups E to P are linked by the presence of silica

bodies in the sclerenchyma, and show variation in

the development of sclerenchyma ridges and the

shape of the chlorenchyma and protective cells.

However, the variation does not correlate well,

consequently either one large variable group can be

recognized, or numerous small groups. The latter

approach was followed here, with the smaller groups

formed into larger groups with the recognition of

convergence. The final determination of the

characters showing homoplasy can only take place in

a cladogram employing a larger data set.

PALYNOLOGY

Introduction

The pollen morphology and ultrastructure of the

Restionaceae has previously been studied either in

relationship to the intercontinental delimitation of

the genera (Chanda, 1966), or to the rich fossil

record of the group (Van Zinderen Bakker, 1953;

Ladd, 1977; Hochuli, 1979; Muller, 1981 and

references therein). Van Zinderen Bakker (1953)

gives superficial descriptions of 17 species. Chanda

(1966) provided a light microscope survey of the

Centrolepidaceae, Restionaceae (including Anar-

thriaceae and Ecdeiocoleaceae) and Flagellariaceae.

He reviewed the sparse literature up to 1965, and

studied acetolyzed pollen of 114 species of

Restionaceae, of which 59 were African. He

recognized the two major apertural types, centrole-

pidoid and graminoid, and suggested a phylogenetic

sequence from centrolepidoid to graminoid pollen

apertures. Chanda & Rowley (1967) published

detailed studies on the wall anatomy of a few

species. Ladd (1977) studied the surface morphology

of theRestionaceae, Flagellariaceae and Centrolepi-

daceae, and published a series of scanning electron

micrographs illustrating the range of surface

features, and established a terminology to describe

them.

This study is an attempt to establish the

contribution of pollen data to the determination of

the phylogeny of the African Restionaceae. It is part

of an ongoing programme to revise the classification

of the African Restionaceae.

Materials and methods

Pollen material was taken from the Bolus

Herbarium, University of Cape Town (BOL) and

the Herbarium, Royal Botanic Gardens, Kew (K).

Material of 93 species was studied by light

microscope and a smaller sample by SEM and TEM

(see Appendix 4).

Pollen was acetolyzed for 10 minutes according to

the method described by Erdtman (1969). A portion

of the acetolyzed pollen was transferred into glycerol

(for temporary slides) or glycerine jelly (for

permanent slides) for light microscopic study. The

material studied was photographed with a Zeis

photomicroscope and a Nikon LKe light micro-

scope. The remaining part of the sample was used

for scanning and transmission electron microscopy

(SEM and TEM). Material for SEM was either

suspended in distilled water, transferred onto

double-sided ‘sellotape’, and the water evaporated

off, or critical point dried (CPD) and dusted onto

double-sided ‘sellotape’. CPD was found to be a

more satisfactory technique. The stubs were coated

with gold or platinum and examined with a Jeol T20

SEM and a Model MSM-4 Hitachi Abashi SEM.

Material was prepared for TEM by fixing

acetolyzed exines with 2% osmium tetroxide,

prestaining with uranyl acetate and embedding in

Eponaraldite (Skvarla, 1966; Skvarla & Kelley,

1968; Skvarla & Pyle, 1968). Sections were cut with

a diamond knife and post-stained with uranyl acetate

and Reynold’s lead citrate and examined with Jeol

100S TEM.

Size measurements were made by light micro-

scope, accurate to 1 pm. However, the sample

measured per species varies from 1 to 10 grains

(Table 5).

General pollen morphology and ultrastructure

The pollen grains are shed as monads. The shape

varies from depressed ovoid to spheroidal, with the

most common state being broadly ovoid. The

longest axis varies from 20-65 pm, the shortest axis

(from the centre to the aperture) from 17-50 pm.

Ladd’s (1977) measurements refer chiefly to

Australian taxa. Measurements published by Chan-

da (1966) and Van Zinderen Bakker (1953) differ

from the measurements published here. This may

indicate that size is not a reliable characteristic or

that the sample sizes were not adequately large.

lnterapertural wall

The interapertural wall is 1—3 pm thick and under

LM the sexine and nexine can be readily disting-

uished. TEM studies show that only an ectexine is

present, consisting of tectum, columellae and foot

layer. The tectum and foot layer appear to be of

about the same thickness. The columellae are

generally short, well-formed pillars, rather sparse

and form a layer usually narrower than the tectum or

the foot layer. The endexine is absent.

The exine is pierced by channels ranging in

diameter from 0,01 pm to 0,4 pm. Under LM only

the larger channels are visible and have generally

been termed scrobiculi (Van Zinderen Bakker,

1953; Chanda, 1966; Chanda & Rowley, 1967; Ladd,

H. P. LINDER

47

1977; Chanda & Ferguson, 1979). The smaller

channels have been termed puncta (Ladd, 1977).

Chanda & Rowley (1967) suggested that the

distinction between scrobiculi and puncta is that the

former penetrate both the tectum and the foot layer,

whereas the latter only pierce the tectum. This

difference is readily observed in thin sections (both

with LM and TEM). SEMicrographs of the inside of

the wall show the scrobiculi emerging on the inside.

Ladd (1977) proposed that scrobiculi be defined as

having a diameter of 0,2-1 pm, and puncta as being

smaller than 0,2 pm. This allows them to be

separated in surface view by SEM, and there is

probably a good correlation with Chanda &

Rowley’s definition of the terms. In this study,

scrobiculi are taken to be larger than 0,1 pm, as in

several taxa the larger channels are between 0,1 and

0,2 pm in diameter.

Ladd (1977) and Chanda & Ferguson (1979)

determined the densities of scrobiculi, puncta and

surface ornamentations. These data were not

collected in this study, as subjective assessments of

densities was thought to be adequate.

The surface ornamentation is variable and difficult

to describe. Ladd (1977) suggested a terminology,

which was followed by Chanda & Ferguson (1979)

and which is used here in an amended form. The

following tectal projections have been recognized:

(a) microverrucae — small, isodiametric, blunt or

acute projections (i.e. Fig. 28J). Ladd (1977)

recognized two size categories, but as their size was

found to vary continuously, the size is given directly.

Similar structures in the Poaceae are termed spinules

(Watson & Bell, 1975; Kohler & Lange, 1979);

(b) spinulae — taller than wide, very acute, the

apex extending into a slender mucro (i.e. Fig. 25 I, J,

G);

(c) microbaculae — taller than wide, with a more

or less truncate apex (i.e. Fig. 23K).

The spacing of the tectal projections is given as

dense (less than 1 diameter apart), sparse (1-3

diameters apart) and very sparse (more than 3

diameters apart).

The surfaces are adequately described by the

terms proposed by Ladd (1977). A smooth surface is

flat, a verrucate-subrugulate surface is gently

undulating, with the scrobiculi in the hollows, while

a rugulate surface consists of square blocks

separated by steep-sided grooves much narrower

than the blocks. There appear to be occasional

intermediates between smooth surfaces with dense

verrucae and rugulate surfaces (i.e. Restio bifurcus).

The distinction between smooth and verrucate-

subrugulate surfaces can occasionally also be

difficult.

Apertural regions

The grains are ulcerate.

Chanda (1966), Chanda & Rowley (1967) and

Ladd (1977) recognize two basic types of apertures:

graminoid and centrolepidoid. Chanda (1966) did

not clearly define his two types and recognized

transitional forms. Ladd (1977) defined graminoid

apertures as being 3—4 pm in diameter, circular,

either with or without an annulus or operculum,

whereas the centrolepidoid apertures are 7— 8pm in

diameter, irregular in outline, with loose granules

around the margin and without an annulus. It is clear

that all African taxa have graminoid pollen, as

suggested by Johnson & Briggs (1981), but the size

of the aperture criterion should be dropped.

TEMicrographs show that the tectum, foot layer

and colummellae continue over the border region

and that the tectum and foot-layer fuse at the edge of

the aperture. Both the tectum and the foot-layer

may be variously modified: the foot-layer may be

much thickened (as is typical of the Poaceae)

(Erdtman, 1971), or somewhat reduced. The tectum

is usually less modified, but may also be reduced in

thickness. Generally, the surface of the tectum is

smooth, but tectal projections still occur as on the

interapertural walls.

An endexine occurs in the border region and is

manifested as either a granular or laminated layer.

Kress & Stone (1982) suggest, contra Guedes

(1982), that the laminated layer may in the monocots

originate from the foot-layer. This endexine appears

to be continued under the apertures and is usually

burst by acetolysis. Usually a ring-like remnant still

remains, but in a few micrographs it can be seen to

extend over the entire aperture. The endexine also

shows a distinct endo-sculpture (vide Van Campo,

1978). Not enough samples were seen to determine

whether or not there is any variation in the

endo-sculpturing patterns.

In most taxa, granules may be found around the

edge of the aperture, lying on the endexine.

TEMicrographs published by Chanda & Rowley

(1967) and Chanda & Ferguson (1979) show that

these granules consist of tectum, columellae and

foot-layer, very similar to the general wall structure.

SEMicrographs show the presence of puncta and

tectal projections on some of the larger granules.

This suggests that they are exine fragments.

Two different types of aperture border can be

recognized:

(1) the tectum is not modified, but the foot-layer

is thickened up to x3 its normal thickness and may

be lamellated. This is obvious under LM as a raised

area with a finer denser texture. Under SEM it is

only visible as a raised area. There appear to be no

granules in the aperture and the aperture diameter is

3—6 pm. This organization is similar to that

illustrated by Chanda & Rowley (1967) for

Ecdeiocolea monostachya, or by Christensen &

Horner (1974) for Sorghum bicolor (Poaceae). The

light microscope appearance is the same as for

Lyginia barbata, Hopkinsia anoectocolea and Anar-

thria gracilis, i.e. it is the ‘typical’ graminoid pore

(see also the comment in Chanda 1966 : 378). It is

the African graminoid type (of Restio subverticella-

tus) referred to in the palaeo-palynological literature

(Hochuli, 1979) (i.e. Figs 24A, J,C, & 25A, B, H).

(2) the tectum is not modified or is slightly

reduced in thickness, while the foot-layer is scarcely

thickened, but may be slightly thickened or reduced.

The tectum is usually raised to form a raised border

'w>

48

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Fig 21. — A, Restio tuberculatus, x 700; B, R. tuberculatus, x 700; C, R. tuberculatus, SEM, X 1450; D, R. tuberculatus, SEM of

border and general surface, x 7250; E, R. madagascariensis, SEM of border and interapertural surface, x 7250; F, R.

madagascariensis, SEM, X 1450; G, R. dispar, TEM, x 1400; H, R. dispar, TEM of border, showing endexine and granule in

aperture, x 7000; I, R. dispar, x 700; J, R. dispar, X 700; K, R. dispar, SEM, x 1450; L, R. dispar, SEM of border, granules

and interapertural surface, x 5438.

H. P. LINDER

49

Fig. 22.— A, Restio tetragonus, SEM, x 1450; B, R. tetragonus , SEM of border and interapertural surface, x 5438; C, R. tetragonus,

SEM of fracture showing the border of the aperture and a granule, x 7250; D, R. bifidus, SEM of fracture across the aperture,

showing the border, granules and endosculpturing, x 3625; E. R. bifidus, SEM of aperture with granules, border and

interapertural surface, x 5438; F, R. bifidus, SEM, x 1450; G, R. cascadensis, SEM of border and interapertural surface, x

7250; H, R. mahonii, SEM of interapertural surface, x 12000; I, R. major, x 700; J, R. tetragonus, x 700; K, R. tetragonus, x

700; L, R. major, x 700.

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

50

Fig. 23.— A, Leptocarpus asper, SEM, x 2536; B, L. asper, SEM of aperture and interapertural surface, x 7250; C, Restio

filiformis, x 700; D, R. debilis, x 700; E, R. debilis, X 700; F, Leptocarpus gracilis, X 700; G, L. gracilis, x 700; H, L. asper,

x 700; I, L. asper, x 700; J, Restio pedicellatus, SEM of aperture and interapertural surface, x 4200; K, R. pedicellatus, detail

of surface, showing scrobiculi and baculi, SEM, x 1200; L. R. brachiatus, SEM of aperture and surface, with granules in

aperture with puncta and microverrucae, x 7250; M, R. debilis, SEM of aperture and surface, x 5438; N, R. similis, SEM of

aperture and surface, x 5438.

and is usually, but not always, curved inwards at the

actual edge of the aperture. The foot-layer is either

raised with the tectum, or continues straight, so

forming a ring-like cavity around the aperture (i.e.

vestibulate). Under LM this type is clearly distinct

from the preceding types by the lighter texture of the

border in front view, and from the side it is obvious

that the walls are not thickened. Often the

vestibulate ring-like structure is very visible. There

are usually, if not always, granules in the aperture.

Aperture diameter is 3 — 12 pm.

Type 2 apertures can be further subdivided, but it

is not always possible to place every species into one

of the subdivisions.

H. P. LINDER

51

Fig. 24. — A, Restio curviramis, SEM, x 1450; B, R. curviramis, SEM of aperture and general surface, x 7250; C, R. curviramis,

TEM of aperture showing thickened foot-layer in the border region, x 4200; D, R. curviramis, TEM of whole grain, X 1400;

E, R. ocreatus, SEM, x 1800; F, R. monanthus, SEM of surface of aperture and interapertural regions, x 6000; G, R.

monanthus, SEM, x 2400; H, R. distractus, x 700; I, R. distractus, x 700; J, R. ocreatus, x 700; K, R. ocreatus, x 700; L, R.

tenuissimus, x 700.

(2a) Border region very slightly raised, foot and

tectum slightly separated, hardly differentiated.

Under LM this appears as an undifferentiated

border region, but under SEM the wall can be seen

to be slightly raised. This is similar to Chanda’s

(1966) ‘transitional’ type. (i.e. Fig. 25L, M).

(2b) Border region more or less sharply raised,

forming a very regular ring. The regularity of the

ring is also obvious under LM. Granules in the

aperture tend to be sparser and smaller (i.e. Figs 26

& 27).

(2c) Border region raised, forming an irregular

ring, manifested under LM as apparent cracks in the

border. Under SEM it can be seen that the border is

quite variable (i.e. Fig. 21H, F, B).

Discussion

The variation in the morphology of the pollen is

summarized in Table 5, and illustrated in Figs

21—28. The assessment of the taxonomic implica-

tions of these data is complicated by the fact that

Fig. 25. — A, Staberoha aemula, x 700; B, 5. aemula, x 700; C, Thamnochortus argenteus, x 700; D, T. dichotomus, X 700; E, T.

serpens, x 700; F, Staberoha aemula , SEM of border region, x 12000; G, 5. distachya, SEM of surface, x 6000; H, S. aemula,

TEM of whole grain, x 1400; I, 5. aemula, TEM of interapertural wall, showing the spinulae, x 10000; J, 5. aemula, SEM of

surface, x 1200; K, Thamnochortus paniculatus, SEM of border and interapertural region, x 7250; L, T. dichotomus, TEM of

apertural region and interapertural wall, showing granules and scrobiculi, x 4200; M, T. dichotomus, SEM, x 2536; N, T.

dichotomus, SEM of border, x 7250.

H. P. LINDER

53

Fig. 26. — A, Hypolaena diffusa, x 700; B, H. diffusa, x 700; C, H. digitata, SEM, x 1450; D, H. digitata, SEM of aperture and

interapertural surface, x 5438; E, H. digitata, TEM of aperture, x 4200; F, H. digitata, TEM of interape rtural wall showing

scrobiculi, x 4200; G, H. digitata, TEM of whole grain, x 1400; H, Restio obtusissimus, x 700; I, R. obtusissimus, x 700; J.

Cannomois acuminata, x 700; K, C. acuminata, x 700; L, Hypodiscus binatus, SEM, x 2400; M, Restio obtusissimus, SEM of

border and interapertural surface, x 7250; N, Cannomois acuminata, SEM, x 1800; O, Restio obtusissimus, SEM, x 1450.

54

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Fig. 27. — A, Willdenowia humilis, SEM of border and interapertural surface, x 7250; B, W. humilis , SEM, x 1450; C, W.

fimbriata, SEM of fracture through aperture, X 3625; D, IV. striata, surface of border and interapertural wall, SEM, x 3625;

E, W. esterhuyseniae, TEM of aperture, x 2800; F, W. esterhuyseniae, TEM through aperture showing reduced foot-layer and

granules, x 5800; G, Anthochortus ecklonii, SEM of surface of border and interapertural wall, X 12000; H, A. ecklonii, SEM,

x 2400; I, Willdenowia argentea, x 700; J, W. argentea, x 700; K, IV. esterhuyseniae, X 700; L, W. esterhuyseniae, x 700.

H. P. LINDER

55

Fig. 28. — A, Chondropetalum microcarpum , x 700; B, C. nudum, x 700; C, C. nudum, x 700; D, C. albo-aristatum, x 700; E, C.

albo-aristatum, x 700; F, Elegia capensis, x 700; G, E. capensis, x 700; H, Chondropetalum nudum, SEM, x 1800; I, C.

nudum, SEM of border and interapertural wall, x 6000; J, C. ebracteatum, TEM of interapertural wall, x 4200; K, C.

ebracteatum, TEM of aperture showing reduced foot-layer in the border region, x 2800; L, Elegia equisetacea, SEM of border

and interapertural surface, x 12000; M, E. equisetacea, SEM, x 2400; N, E. glauca, SEM, x 2536; O, E. glauca, SEM of

border and interapertural surface, x 5438.

56

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

only 90 of the approximately 320 species have been

investigated.

Table 5.

Phenetically, the following groups may be

recognized. It is not possible, however, to assign

every species with certainty to one of the groups.

(1) Annulus type 1, raised, the foot-layer much

thickened, walls not separated; surface smooth,

covered in dense spinulae obscuring puncta and

scrobiculi; tectum thinner than foot-layer; granules

absent from aperture.

Staberoha aemula, S. distachya, S. remota.

(2) Annulus type 1, raised, foot-layer somewhat

thickened, walls not separated, no granules in

aperture. Surface verrucate-subrugulate; microver-

rucae 0,08—0,14 pm in diameter, usually very

sparse, rarely sparse.

Restio curviramis, R. distractus, R.

monanthus, R. ocreatus, R. sieberi, R.

tenuissimus, R. virgeus.

(3) Annulus type 2a, slightly raised, wall not

thickened, slightly separated, appearing undifferen-

tiated under LM, granules present in the aperture;

surface smooth to rarely verrucate-subrugulate,

microverucae usually dense, rarely sparse, generally

larger than 0,25 pm.

Thamnochortus argenteus, T. dichoto-

mus, T. erectus, T. glaber, T. paniculatus,

T. platypteris, T. serpens, Restio tetrago-

nus. R. rhodocoma, R. quinquefarius, R.

egregius, R. fruticosus.

Note: 1. The border varies from undifferentiated

to slightly ‘blown out.’ These changes may,

however, be related to the effect of the acetolysis

treatment.

2. R. quinquefarius and R. egregius are

intermediate to group 6.

(4) Annulus type 2b, sharply or rarely gradually

raised in a regular ring, walls not thickened, well

separated or scarcely separated, then both lifted;

granules usually present in aperture; surface

verrucate-subrugulate, microverrucae less than 0,2

pm diameter, sparse; puncta sometimes absent (i.e.

only 1 size channel present).

Restio obtusissimus, Hypolaena digitata,

H. diffusa, H. purpurea, Hypodiscus

aristatus, H. albo-aristatus, H. binatus,

H. neesii, H. striatus, H. willdenowia,

Cannomois acuminata, C. virgata, Will-

denowia humilis, W. argentea, W.

luceana, W. sulcata, W. teres and Phyllo-

comos insignis.

(5) Annulus type 2b, sharply or rarely gradually

raised in a regular ring, walls not thickened, more or

less separated; granules usually present in aperture;

surface smooth, microverrucae dense, more than 0,2

pm in diameter at the base.

Willdenowia esterhuyseniae, W. fim-

briata, W. striata and Anthochortus

ecklonii.

Note: W. fimbriata is intermediate to group 4,

with sparse microverrucae more than 0,2 pm in

diameter.

(6) Annulus type 3c, walls raised into an irregular

ring, or spout, usually slightly separated, occasional-

ly developed into a ring; granules usually present in

aperture; surface rugose, microverrucae generally

more than 0,2 pm in diameter at the base, usually

sparse, rarely dense with the rugose structure partly

developed, annulus generally partially rugose.

Restio compressus, R cascadensis, R.

mahonii, R. bifidus, R. bolusii, R.

bifurcus and R. major.

Note: 1. R. mahonii differs by the microverrucae

smaller than 0,09 pm.

2. R. egregius and R. quinquefarius from

group 3 may belong here.

(7) Annulus type 3c, raised slightly, exine

fragments present; surface smooth, with dense

microbaculae.

Restio pedicellatus.

(8) Annulus type 3c raised into an irregular ring,

often strongly raised, appearance variable usually

slightly separated, occasionally forming a ring

structure; granules present in aperture; surface

verrucate-subrugulate, microverrucae sparse,

0,15— 0,25pm in diameter at the base.

Restio filiformis, R. madagascariensis, R.

dispar, R. similis, R. brachiatus, R.

tuberculatus, R. debilis, Chondropetalum

albo-aristatus, C. andreaeanum, C. eb-

racteatum, C. insigne, C. microcarpum.

KEY TO THE POLLEN PHENETIC GROUPS

1. Foot-layer thickened:

2. Surface smooth; spinulae dense Group 1

2. Surface verrucate-subrugulate; microverrucae sparse Group 2

1. Foot-layer not thickened:

3. Surface rugulate Group 6

3. Surface smooth or verrucate-subrugulate:

4. Annulus a regular, well-formed ring:

5. Microverrucae dense, more than 0,25 pm in diameter at base Group 5

5. Microverrucae sparse, generally less than 0,25 pm at base Group 4

4. Annulus irregular or weakly developed:

6. Surface with dense microbaculae Group 7

6. Surface with microverrucae:

7. Surface usually smooth; annulus low Group 3

7. Surface usually verrucate-subrugulate, annulus prominent Group 8

TABLE 5. — Pollen morphology and ultrastructure of some African Restionaceae

oo

vn

■

r~~

in

c o

<D k

•C 52)

o

ea O

O

******«*• .o ^ .o *<t **««*♦ vo no vo oo so vo » oo g » g vo v© »<» oo oo r- | ^ M oo | |co I co « m n « « S? J?

ggggggggg-g ,8-8«8»«8'-2»— |J« i—8S88g 'SS8s288»8«S8»8»8S ‘gg-gffgggg

on oo

O O

m m

oo oo oo oo

t— r / — -\

oo oo

Q Q oo oo

m «

I I I I I I

-> -> -> •> C~-

°0 OO OO OO OO I O'- oo oo I oo oo

CQ

oo oo oo oo oo oo

O 1 § § 00 § § § § 1 00 T1 ^ ^ i9°?

2 o _ o

. m ,

OQ C/3 c/5

"> m O'- -> e« ©--

'S ex

S £

CQ

a

3

.a C

.o 5

o -S

in -t3

,O.O,r>TO,£).Q-0.OX>.a

<N(N(NMM(N(N(SNfv|

1 m S m m m S 58 m m m m c^ <8 m <8 on ^ m <8 c^£ m ^ <8 m on m £ <3 <3 on on m <3 <3 1-El

' N N (M (M

a- ® a- a- vo a-

ON ON^ ^H CN^ <N — I ON CO

o o o' o' o' o o' o

ON ON

o' o'

CN ON t-h •— h

o' o" o" o'

N" V£>

ON H

ON Ol ON

ON 00 00

ON ON ON

<D O' O o"

ON ON

o' o'

cooo r— <o oo r— vo vo -t vi ^ oo n-

'-frTcrIfvlcrf0i<Nir^cr2>cri,r~ 1 to 'N co on -h on on co

oooooo'ooo'oo'o'oo'ooooo'

o o o o

r— oo oo <o oo oo r~~

I '"'i I ON CO ON ON CN ON ON —I I

o o o' o' o' o' o' o' o

IN 00 VO Tt- VO VO

onon^hcnonoo h on -—i

o o o o o cT o' o' o'

OOOOOCJOC _

<NON<NONONONON<N<N

VO ON VO VO VO VO 00

r"^ *“H rH t-H i-H t— ( t-H r-H

O O O o' o' o' o' O

Q Q

00 rr\

> “

W CO t/3 (/]

> > > >

CO C/3

> >

Q > > w

C/3

C/3 C/3

WWMOW^QMMQMMMCOMMQMm^ I C/3 C/3 C/3 C/3 Q I Q Q C/3 C/3

t/3Mi'>c/'^Q QSc/2£>c/3> lc/5§

> I wQO I^^cococo^co&oq

P O

O C

o <U c

£ S .2

N coS *±

3 3 £

N- <N t 00 Ov oo CO CO OO

<N CO I ^OOOOOOCN

o o o o o cT o' cT cT o'

N- r-

r—i r-H

o' o'

o o o o

>>>>>>>> |>> |>>>>

Tj- VO

o' o'

I > I > >

S li 5 152 ^ 2? 00 oo oo vo on oo to oo <o oo

rP. eN 1 ^ ^ OO ’ ^ CN^ O ON^ CN^ <N^ <N^ CN^ ON t-h < ON ON <— < ON O

*— >t— ><— ^ ’OOOOOOOOOO cT o' o' cT cT o'

>>>>>>>>>>>>>J>>.>;>->;>;>;>

VO oo vo VO 3- M vo I — vo

1 ON OO) ON N

00000 o' cT o' CD

> > > > « !>>>> | >

r? ^2 2 °9 °°, ® o ;ov 00 - 00 00 <0 <o vo 00

'“t °i <N ON I ON I y-HT-HCQoOT-HT-HT-HONO

o o o o o o' <o o' o' ^ o' o' o' o' o' o' o'

<— r <— r o

vo to a-

VO OO VO VO 00 ON

-“t <N | O T-H

000 o o cT o' <dT o' cd" o' cdT

>>>>>>>>>>>> p> ^ ^ ^ I > > ;> I >>>>>>> > >

o<

Pt&aS&aiPlpiiiX

>>>>>>>>

& & 05 o< oS oi c2

>>>>>>>

a! a;

> >

on od 05 c2

50 > > >

o5 05

> >

05

05 05 05 05 05

05

>

I I

o5o5o5o5o5o5o5o5c/3

>>>>>>>>

I VO CO */3 Hj- CO vo

P) 't

10 IN

VH ^ — '

Tf 1/3

*— S

C<3 Tt

cn

I

o

I I

VO O

ON CO

<tOO\

to CO

: 1 1 1

O h OO

CO) CO

00

00 O 00 O CO

NcOHvt

1 I I I ,TJ

t-~ t- VO Tj-

(N ON N C3 OV

ON

N 3-

<N CC)

N to to) O o)

C3 ON ON C3 <<3

h-s /— s

N

ON CN

VO

I

Ti-

vo ON

m co

I I

m 00

co <N

I I

o t — i^a-

vOOhnioOcciN-

CO -vl- ON rj- w ff) ■

: 1 1 1 1^1

CO CO 1 — | CO O ON vo 00

ON CO

CJ-V t-h o NO

ON TJ- Tj- CO

O O CO

CO Tf CO

00 r- 00

ON CO ON

ONcNONtJ-COoNONCOCOCOOONOOhcO 1 <N CO CN ON

Co

fe a

8 |

rv

•S *5®

ts "S

O' O'

S -3

69 S

s .S

.3 5

•E5

a

'E? I

2 I

'c .E

cos**

g § § ? g |S 3 | § a =a a J

bo 5^ ’r** ^ O ^3 b

=H O Ci.'CS N3

3 O ^ .«

d, a, « -q

K c? S

vo

vo r- vo t c —

| vo vo vo r~' vo vo

vt t<3 1/) vo vo

00 00 o

00 CO CO T}-

00 on ov vo

CO ON CO

VO CO c — •

*a- -*a-

ioococr\ior^<-HHj-

■3" *<t CO Tf Tj- Tj- Tfr

O vo CO

*a- co *a-

ON O

CO ^r

ZN O OO O t— < O O

3 <u

&-S

^ §

■s^a

i a “

o

3

3 .s .a

« e ai

s ^ s

3

CO -cf

00 c» c^ vo

w \ ^ s — -

VO I VO VO vo

' ' /— H /— S ^ — s H- H

vo vo VO vo ■'S-

OO vo

^ — '

VO vo

r — N /— s

VO

- (J\

' O ’ — < N* p-

IN 00 IN

' ' H-N \

vo VO VO

H OO 00 VO O P .

V' ' — ' W ' — - , N-

O P Mo 1 VO

t-H ■— h ^ — t /— > I

/->vo vo a- 00 a- 'e

on <0 vo -t

1 vo CO

t /—v /— S H-S

00 CO N- OO

^r'7 2l’voootTTv?v?^77

O vo N-vo'OVON-rt'covo

2, I I

<0 M Tj-

UT

m *a-

vo on 00 on r-

<N O CO Tf CO CO CO

ON Vt

MOO

00 Tf <0

2iSi£i23:2lolov^>0(N'=>o'^afooovioM^F'oN

<0 00 00 o 00

Tf OO CO Tf </) <N

*3-

O M

N CO Oh 00' CN

OO OO OO OO OO

OlOoo

on co a-

O VO ON VO

N- <0 00 co

M 00 ON M

ON ON ON ON

CO O C — VO vo

CO OO VO vo ON ON ON

' ' ' ON Hj-

co

VO

I I

N- O VO

00 OO CO

vo CO ON ON OO VO

CO CO CO CO ON

vo VO CO CO 00

ON ON CN CO ON

COO.^T-

a- on co co a- h

ON CN CN CN CN

vo vo co vo

3- CO CO VO

CO O CO CO CN ON CN O ON O

N-COCO'—'CNCO'-— 'N-0NCO

1 1 1 S 1 1 m 1 1 1

voooioXoo^pcnc-

COCNCNOCNCO-va-COCNCN

CO CN

OO T-H I — T-H

Cl O rt VO K) Tf Tf

CN vt

VO vo vo

■vj- t}- vo

VO

^ VO 00 03 t-h

oriSx§x£7 1

cn co vo 1 co cn t co 00 rt m oo rr- Q 5T- C'^P5T'5rS^rN|cr''~'r^or)'-H^^

CON- Tl- co CO CO CO *0 CO N" CO r^!2L:3:r0Srr‘OOr^<^>'^00<N

wwwww WWW WWW WWW jo N- CO CO co COCO ON CN CN

OOOO bo JnJnct^

VOVON-N-VOCOVON-N-CTVOOCOCOCN

N- ir'co >^^CN Tj-.www

IX I^VO I | O O VO VO M

CO CO J . I CO CO ON CO CO

10 CO M 00

vo rf N- <0

o

N"

OOvoCTvON-COOvOCOCNCNcO

co'tvo'tco'H'iNvi-y'a- co’co

CO CO v<0 10 ov /— «, 00 O 00 CO 00 loo

CN N- N-/CO CNOCNCOVOCOCNCN CO

CO

CN

rfim22222HOOONONONOw<Nn(siooo

OON-OcOOCNMOcON-OOOcOt-hN-

■vfcNCOcON-OcON-OcON-cO

■I

5“ -a

o E;

't5

to a; 05 as as as as as as ^ t*i 5ci ^ tsj ttj aj aj aj a U I a** ^ ^ J 3* ^ J

c

I

CX) gj

Co ^ Co

S «>3 >s

S°|v,

O 5 5£, 3 :-

« X ts -o

^ c< O'

Co co

Hi

.3 -§> s“

a ^

a s->

2 §

*- .VH

a a

o ^

V)|g:a

S 3 ’i) 0 ■§ §

a. £> -C -3

’es

c

^ s

e s

Q co

«§. sa

O S a 3

M's s a

3

*§ a

a a

3 ^3

a a a § a a.

•8 3

| 3“ I

3 ^tO

P-.S a

3 T^

a. 3- v

S

,&o 3 ~a

a ^

.a I §

E5 t3 tS

^^^^^OiQ'C^C^Q'oioiQ'Q'O'oioio' O' O' O'

fc>

3

Co 5

*9 a |

i o

ag

3 .3 55

■c *• *a

a a a

a

c-a Or® s-^a.^ 3-~ s*-£f

a^-s §5 a s s-o s

>3733boKi.a.i*ia,'-a5<)£i,Coa3-aa kskk^vjvu

a; O' o' f-, 1-5 f-5 aJ f-J f-5 t— " n n -4 *4 2 S O' O' O" O' O' O" 5.J O Lq tq Cq’

a- a

&

3

to

3

e*

:a a

a 3

3 .Vo

3

o

-a

3

a

<x

§■

a

3

a 3

a

3

a o

o

s 5

£ J

3 ^2

3

C3 3

S 3

C 3

3 -2 3

&

'C

S .«

Q .*S

- o .C

fi-ga

53 -Bo a

3 C

V, Qj

1 1 g-f

e I §5 .1

o.a.25

t»o

tq* tq Cq' Cq' Cq' Cq' Cq'

<D

d5

>

J3

t cn O

O -q ctj

O 3

G 0 d

K « E

CO O

'S

<N

II II

V IH

t! m

^ o>

CO

.

S3 .-2

CX O

GS

TJ COS

O

-o OJ

§ e

S3 II

B>

n M %

N 3 .C II

“-S « o

Oh CO PQ

O - ^

!«^I

p Q o

3 C .. So

3 O £3 3:

’5 c £•

cb o

"8 I

•S H

8 0

5 <D

Q ^

7. Surface usually verrucate-subrugulate, annulus prominent

Group 8

H. P. LINDER

59

C. mucronatum, C. nitidum, C. nudum,

Elegia asperiflora, E. capensis, E. equise-

tacae, E. galpinii, E. glauca, E. muirii, E.

parviflora, E. prominens, E. racemosa,

E. stipularis, Leptocarpus burchellii, L.

gracilis and L. asper.

Note: This is a fairly variable group, which cannot

be readily further divided, but Elegia and Chon-

dropetalum may be slightly distinct and may also

contain two subtypes. The reduction of the thickness

of the foot-layer in C. ebracteatum suggests a link to

group (4).

In order to determine the phylogenetic import of

the pollen data on the monophyly of the phenetic

groups, the distribution of the characters in the

outgroup is discussed below.

(a) Aperture type

Chanda (1966) suggested that the graminoid

aperture is more specialized than the centrolepidoid

aperture, and by implication, derived from it.

Johnson & Briggs (1981) hold the opposite opinion,

on the grounds that it correlates with the other

characters which they used in constructing their

phylogenetic tree. If the Poaceae are accepted as the

sister group for the Restionaceae, as implied by

Dahlgren & Clifford (1982), the graminoid aperture

can be established as being plesiomorphous by

outgroup comparison.

In the typical graminoid aperture, the foot-layer is

much thickened, as it is too in the outgroup of the

African Restionaceae. Clearly then, the aperture

type found in Aperture type 2 (see above) is derived.

This analysis, unfortunately, does not allow us to

proceed to determine polarity among the three

subtypes of type 2.

(b) Surface types

Of the three surface types which Ladd (1977)

recognized, only smooth and verrucate-subrugulate

surfaces occur in the Australian Restionaceae. He

recorded rugulate surfaces from the Centrolepi-

daceae, which may not be closely related to the

Restionaceae (Cutler, 1969). In addition, it is

possible that the rugulate surfaces in the Centrolepi-

daceae may differ in their ontogeny from the

rugulate surfaces recorded for the African Res-

tionaceae, where they appear to form from the

fusion of microverrucae on a smooth surface, rather

than from a modification of a verrucate-subrugulate

surface. They look rather similar to the insulae

recorded in the Poaceae (Watson & Bell, 1975).

As both smooth and verrucate-subrugulate sur-

faces occur in both African members and the rest of

the family, no polarity can be assigned to these

features and they may well be homoplasious.

(c) Ornamentation

Microverrucae occur widespread in the Res-

tionales and Poales (Ladd, 1977; Page, 1978;

Watson & Bell, 1975; Kohler & Lange, 1979).

Spinulae and microbaculae have, in the Restionales,

only been recorded in the African Restionaceae (see

data in Ladd, 1977), and are therefore likely to be

derived features.

It is tempting to suggest a similar argument for

dense microverrucae of diameter larger than 0,25

pm. They appear to occur occasionally in the

Australian Restionaceae. They are likely to be

derived, but not likely to be uniquely so.

These data are presented in the cladogram in Fig.

29. The basic structure of the cladogram is supported

by the well-corroborated character of the reduction

of the foot-layer and the further branches depend on

the weaker character of the various subdivision of

aperture type 2. Three of the groups are not strictly

speaking shown to be monophyletic (marked with an

asterisk, see Blackmore, 1982). Group 2 is very

distinct and most likely a monophyletic group,

whereas groups 4 and 8 may well not be

monophyletic.

There is little congruence between the phyletic

and phenetic results and the existing classifications

1

2

3

4

Sa

5b

5c

6

Fig. 29. — Cladogram based on

pollen data. Only synapo-

morphous characters are in-

dicated. Groups with aste-

risks are not supported by

autapomorphies, and there-

fore not demonstrated to be

monophyletic. Characters

used are (plesiomorphies list-

ed first, synapomorphies se-

cond); 1, microverrucae

sparse/dense; 2, microbacu-

lae absent/present; 3, surface

smooth or rugulate-verru-

cate/rugulate; 4, spinulae

absent/present; 5(a), aper-

ture type 2a; 5(b) aperture

type 2b; 5(c) aperture type

2c; 6, aperture type 1/2.

60

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTION ACE AE

of Pillans (1928) and Gilg-Benedict (1930).

The genus Restio L. (even restricted to Africa)

appears to be polyphyletic. Restio obtussisimus

belongs to the Willdenowia-Hypodiscus - Hypolaena

group. Pillans (1928) placed it at the end of his

treatment of Restio, possibly implying that it did not

comfortably fit into that genus. One group of Restio

species (my group 2) is very distinct from Restio

s.str. Morphologically, these species belong to the

group of species with two styles, as compared to the

remainder of the genus, which has three styles.

Pillans (1928) also placed these taxa together.

Within the remainder of the genus, 4 groups may be

recognized [my groups 7, 8, 9 and 3 (partially) ].

Placing the first three groups together would not

result in a paraphyletic group and the entire group 3

could also be added to the taxon.

Hypolaena, Hypodiscus, Cannomois, Willdeno-

wia, Phyllocomos and Anthochortus all have rather

similar pollen grains and their claim to monophyly is

rather weakly supported by the aperture type. A

small segregate, including Willdenowia esterhuyse-

niae, W. striata and Anthochortus ecklonii, differs

from the rest by the presence of dense large

microverrucae and may form a subgroup. However,

there is no character holding the residue together, so

that it is difficult to assess the taxonomic implica-

tions of this group in isolation.

The position of Leptocarpus is difficult to resolve,

as the aperture type appears to be intermediate

between the Willdenowia type and the Restio type,

and no detailed data of the border structure are

available.

Thamnochortus appears to be distinct by aperture

structure, although no good quantified data are

available. However, some species of Restio have

pollen very similar to Thamnochortus and have been

included in group 3 with Thamnochortus. On these

data, it would appear that either Thamnochortus be

included in Restio, or these species be transferred

from Restio, to Thamnochortus.

Elegia and Chondropetalum have somewhat

variable pollen grains, but in general they cannot be

distinguished from Restio s. str. and these genera

have been included in group 8.

PHYLOGENETIC ANALYSES

In this section, the data on palynology, phytoche-

mistry, macro-morphology, anatomy and seed-coat

structure presented above are synthesized into a

phylogeny by the methods suggested by Hennig

(1979), Bremer & Wanntorp (1978) and Wiley

(1981). The argumentation scheme presented by

Hennig (1979) may not be the most sophisticated

method, but it does allow continuous interaction

between hypotheses of phylogeny and data as

data-collection proceeds in the absence of the

computing facilities required by the more sophisti-

cated computer-based systems.

The data set used is given in Table 6. This data set

is based on the empirical data presented in the

previous sections, and on the homologies deter-

mined in these sections. However after the

completion of the initial analyses, the polarity of

further data sets was determined by using functional

outgroups (Watrous & Wheeler, 1981) (i.e. Stabe-

roha and Ischyrolepis as functional outgroup to the

rest of the African genera). In constructing the

cladogram presented in Fig. 30, the most parsimo-

nious cladogram was understood to be the one with

the least homoplasy.

The terminal groups used are approximately

equivalent to genera, except in a few cases, where

they are more or less equivalent to subdivisions of

genera. A rigorous criterion for terminal groups

TABLE 6. — Characters used in the overall cladogram. The first state indicated is plesiomorphic, the second is apomorphic.

Characters indicated by an asterisk are conditional synapomorphies

1. Seed coat types

2. Pollen surface smooth of verrucate-subrugulate / rugulate

3. Pollen surface with sparse / dense microverrucae

4. Pollen surface with microbacula absent / present

5. Pollen surface with spinulae absent / present

6. Pollen aperture type 2a

7. Pollen aperture type 2b

8. Pollen aperture type 2c

9. Pollen aperture type 1/2

10. Number of layers in the epidermis 1/2

11. Epidermal cell type normal / thamnochortoid

*12. Central ground tissue with the vascular bundles in a ring /

scattered

*13. Central ground tissue with a central cavity or no cavity /

several scattered cavities

*14. Culms terete / compressed

*15. Guard cells superficial / sunken

16. False pillar cells absent / present

17. Girders absent / present

18. Ribs to the epidermis absent / present

19. Sclerenchyma ridges alternating with the vascular bundles

absent / present

20. Outer chlorenchyma layer not lignified / lignified

*21. Protective cells type A / B

*22. Protective cells type A / C

*23. Chlorenchyma cells slender / squat

24. Silica bodies absent / present

25. Myricetin derivatives absent / present

*26. Culms simple / branching

*27. Sheaths tightly convoluted / loosely convoluted

*28. Sheaths persistent / deciduous

29. Sheath margins simple / fimbriate

30. Male inflorescences spicate / thyrsoid

31. Male spikelets erect / pendulous

32. Male bracts cartilaginous or smaller than the perianth /

hyaline or larger than the perianth

33. Female bracts larger than or equal to the perianth / smal-

ler than the perianth

34. Female bracts larger than or equal to the perianth, coria-

ceous / larger than the perianth, hyaline

35. Male perianth segments overlapping / spreading, not

touching

36. Anthers exserted at anthesis / included at anthesis

*37. Female perianth lobes cartilaginous, conduplicate / hya-

line, flat

*38. Female perianth lobes cartilaginous, conduplicate / carti-

laginous, nitid, flat

39. Fruit dehiscent / woody nut

40. Fruit dehiscent / parenchymatous nut

41. Nut sessile / pedicellate

42. Styles 3 / 2

43. Styles 3/1

44 Styles free / fused

45. Locules 3/2

46. Locules 3/1

1

2

3

4

51

29

34

43|

14

15

17

20

22

27

44

45

401

25

10

33

36

28

38

11

6

13

311

26

8

12

30

32

35

23

16

18

21

37

39

41

19

46[

42

24

7

9

H. P. LINDER

61

at the approximately generic level for the African Restionaceae. The character numbers are related to the

in Table 6. Bars indicate the apomorphous condition. The different kinds of shading are merely to facilitate

logram. Half-filled blocks indicate the presence of the apomorphy in some species of the terminal unit.

62

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTION ACE AE

bigger than species is, that they must be monophyle-

tic and a determined attempt was made to use only

groups for which the hypotheses of monophyly are

most likely to survive attempted falsification. In two

groups ( Restio and Hypodiscus) it has not been

possible, on the existing data-base, to provide data

corroborating the hypotheses of monophyly, but

circumstantial evidence makes me confident that

given the existing data, no better hypothesis can be

produced. In the Willdenowia-Hypodiscus-

Hypolaena group several conflicting groupings are

possible. One grouping would be based primarily on

the anatomical data base, the other largely on the

macro-morphological data base. Here two criteria

were used in delimiting the groups. The first

criterion is parsimony, but this depends to some

extent on the character delimitations. Consequently,

groups were also delimited in such a way as to avoid

the formation of residue groups, into which species

lacking any distinguishing features are deposited.

Anagenesis was not taken into account when

forming groups.

The basic division in the cladogram is based on the

structure of the pollen aperture. This datum is not

congruent with the distribution of silica bodies, or of

two styles. In addition, the basal trichotomy is of

importance, because it delimits the functional

outgroup for the remainder of the African Res-

tionaceae. In a simple parsimony analysis, the

division suggested by the silica body distribution

should be followed, as it is supported by the

distribution of two styles. However, silica bodies

also occur in the outgroup and there is little

independent data supporting the hypothesis of

homology of silica bodies. The same situation

pertains to the distribution of two styles. Aperture

type 2 for the pollen is unique to the African

Restionaceae and is consequently regarded with

more confidence than the other two characters. The

various possible analyses are given in Fig. 31.

The basic trichotomy results in two isolated

genera, Ischyrolepis and Staberoha, whereas the

third branch includes the rest of the species,

arranged into 19 terminal groups, which in turn form

four major clades: (1) Ceratocaryum - Willdenowia

clade, (2) Thamnochortus clade, (3) Restio clade

and (4) Elegia clade.

The Ceratocaryum-Willdenowia clade

This clade is based on three weak characters: two

styles, silica bodies in the outer sclerenchyma cells

and pollen aperture type 2b. The first two characters

do not fare well in outgroup comparison, even in

conditional outgroup comparison, while the third

character is operationally difficult. Yet the group is

very distinct by ridges in the sclerenchyma

alternating with the vascular bundles and large

woody nuts, usually on fleshy pedicels. Unfortu-

nately, neither character occurs in all taxa. The

group is consistent with both the pollen data and

with the anatomy data if anatomy Group J is

excluded.

The sclerenchyma ridges can be developed into

girders, or reach the epidermis, or may be

terminated by false pillar cells. Cutler (1969)

a, pollen aperture type 1/2; b, styles 3/2;

c, silica bodies absent /present

a, styles 3/2; b, silica bodies present /lost;

c, pollen aperture type 1 /2

a, styles 3/2; b, silica bodies absent /present;

c, pollen aperture type 1/2

Fig. 31. — Analysis of the basal trichotomy of the cladogram in

Fig. 30.

regarded these characters as being of taxonomic

importance. However, groupings based on these

characters are not congruent with the groupings

based on the structure of the inflorescence or the

habits of the plants, and these curious anatomical

features are consequently regarded as homopla-

sious. This implies that the important evolutionary

step was the evolution of the ridges, and that further

developments from this step occurred in parallel.

The only species in this group from which the ridges

H. P. LINDER

63

are absent is one species of Anthochortus, several of

Willdenowia and all the species of Ceratocaryum. It

is tempting to regard these absences as secondary

losses and so to include the sclerenchyma ridges as a

synapomorphy for the whole group, but at present

there are not sufficient data supporting such an

elegant answer.

The large (5 — 10mm long) woody nuts have a

more restricted distribution within the group and

could never be a synapomorphy for the whole group,

as Nevillea, with its bilocular dehiscent fruits, could

never have had woody nuts. The nuts are peculiar by

the large size, the enormously strong walls

constructed out of stone-cells and the frequent

occurrence of fleshy pedicels. These nuts are also

absent from Anthochortus and Hydrophilos where

the nuts are parenchymatous.

The arrangement of the genera in this group is not

very satisfactory and it is possible that future data

may indicate that Nevillea is basal to the group, if it

can be demonstrated that the absence of sclerenchy-

ma ridges from Ceratocaryum is due to secondary

loss.

The Thamnochortus clade

This clade is based on anatomical, palynological

and inflorescence data. The anatomical and in-

florescence characters regarded as synapomorphous

for the group also occur occasionally outside the

group. A further, non-quantified character is the

habit of simple fertile culms with lacerated tightly

convoluted sheaths and the often copious production

of sterile culms from the nodes of the fertile culm.

The relationship to Restio is not completely clear,

as Rhodocoma shows strong similarities to Restio.

However, the most parsimonious grouping is to

place Rhodocoma with Thamnochortus.

The Restio clade

This clade is based entirely on a conditional

synapomorphy, the branching culms. To weaken the

character further, it has also revolved independently

in several other groups. This clade is big, including

over 100 species,, but is singularly lacking in derived

characters. Most of the features of the plants are

plesiomorphous.

Within the group, variation in culm anatomy, seed

surfaces, fruit type and overall morphology is not

correlated, except in a few taxa. Consequently the

subdivision of the clade will be very difficult.

The Elegia clade

As in the Ceratocaryum clade, this clade is

delimited by rather weak characters — sheath

caducous and anthers included in the. perianth at

anthesis, both conditional synapomorphies. In

addition, the structural importance of the floral

bracts is more or less reduced relative to the

importance of the perianth and the perianths are all

more or less of the same type.

The subdivision of the group is based on a series of

remarkably correlated characters, all of which are

excellently based on outgroup comparison — the

presence of myricetins, the number of layers in the

epidermis, the reduction of the floral bracts and the

structure of the testa.

‘ Lamprocaulos’ is very curious for the remarkable

convergence to Willdenowia. This is manifested by

the reduction in the number of styles to two, the

formation of a semi-lignified nut, the reduction of

the female perianth segments, the persistence of

sheaths and the loss of male spikelets. Fortunately,

the anatomy of the two groups is completely distinct,

although in two species of ‘Lamprocaulos’ the outer

layer of the epidermis is reduced to the occasional

tubercle-like cell.

SYNOPSIS OF THE GENERA

Except in two cases, the genera are taken to be

equivalent to the terminal units on the cladogram, so

that this section doubles as a discussion of the

terminal units, as well as presenting a generic

classification of the African Restionaceae.

1. Staberoha Kunth, Enum. 3: 442 (1841).

Lectotype species: S. distachya (Rottb.) Kunth.

Phillips (1951) did not advance any arguments for

selecting S. distachya as lectotype.

This group is as delimited by Pillans (1928) and

Dyer (1976), and includes six species. There is

strong similarity to Thamnochortus, particularly in

the pendent male inflorescence and the often winged

female perianth. As a result, the genus has been

included in Thamnochortus by Masters (1878, 1897)

and placed next to it by Pillans (1928) and

Gilg-Benedict (1930). However, the pollen data,

reinforced by the flavonoid data, indicate an isolated

position. The anatomy is quite unspecialized.

Morphologically, Staberoha is distinguished by the

simple culms and the unilocular flattened nuts with 3

(-1) styles, whereas Thamnochortus is disting-

uished by its single style.

Anatomy group E, pollen group 1.

2. Ischyrolepis Steud., Syn. PI. Glum. 2: 249

(1855). Type species: Ischyrolepis subverticellatus

Steud.

This group of about 50 species agrees with

Pillans’s (1928) informal group of two-styled species

of Restio, excluding R. obtusissimus Steud. Al-

though the group was recognized as early as 1841 by

Kunth, it was never formally recognized i.e. given a

Latin name, either at generic or any other rank. It is

a rather uniform group, distinguished by a distinct

pollen type, by several anatomical features, and

macromorphologically by a two-locular or rarely

one-locular dehiscent ovary with two more or less

fused styles.

The exact relationships of the genus are difficult to

determine, as many of the characters are plesiomor-

phous.

Anatomy group F, pollen group 2, seed groups 1,

2, 3 and 4.

3. Elegia L., Mantissa PI. Altera 162 (1771).

Type species: Elegia juncea L.

Lamprocaulos Mast., D. C., Monogr. Phaner. 1:

64

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

349 (1878). Lectotype species: Lamprocaulos gran-

dis (Kunth) Mast. = Elegia grandis Kunth.

Masters originally included Elegia grandis and E.

neesii in his new genus, Lamprocaulos. There has

been no previous lectotypification, and either

species is equally suitable.

Elegia includes about 35 species and is circums-

cribed here in the same way as by Pillans (1928) and

Dyer (1976). The genus is distinct possessing a

double epidermis, myricetin derivatives and an

indehiscent unilocular nut. In addition, the floral

bracts are shorter than the perianth. It is clearly

closely allied to Chondropetalum, with which it

shares many of the above features.

Masters (1878, 1897) recognized a segregate,

Lamprocaulos, based on persistent sheaths and

branching culms. Gilg-Benedict (1930) also main-

tained the segregate, as she found anatomical data

supporting it. However, analysing a larger sample

showed that the various characters supporting the

group are not correlated. In addition, if Lampro-

caulos were recognized, Elegia s.str. could not be

demonstrated to be monophyletic.

Anatomy group A (partially), pollen group 8.

4. Chondropetalum Rottb., Programmate 11

(1772). Lectotype species: Chondropetalum deustum

Rottb.

Rottboell (1772) included C. nudum and C.

deustum in his new genus. Of the former, he only

had female material and of the latter he only had

male material. However, the latter is selected as the

lectotype, as Farr et al. (1979) list it as lectotype.

They cite Thunberg (1788) as authority, but there is

no trace of implicit or explicit lectotypification in

Thunberg’s (1788) publication.

The genus, as circumscribed here, is a segregate of

Chondropetalum as used by Pillans (1928), Gilg-

Benedict (1930) and Dyer (1976). It only includes

those species in which the bracts are shorter than the

perianth and unbranched (but it includes C.

microcarpum). It comprises 12 species and is closely

allied to Elegia by the joint possession of a double

epidermis, myricetin derivatives and bracts shorter

than the perianth, but may be distinguished by the

usually dehiscent, two or three locular ovaries.

Chondropetalum microcarpum has several aber-

rant morphological features. These are interpreted

as anagenesis. Removal of C. microcarpum would

probably result in Chondropetalum being para-

phyletic.

Anatomy group A, pollen group 8, seed group 6.

5. Dovea Kunth, Enum. 3: 457 (1841). Lecto-

type species: Dovea macrocarpa Kunth.

This lectotype is selected from the three species

listed by Kunth after his initial description: Dovea

macrocarpa, D. ebracteata and D. microcarpa. As

the last species is somewhat different from the first

two and Kunth listed its characters as being ‘variant’,

it is not considered a candidate for lectotype.

According to the Rules, both D. macrocarpa and D.

ebracteata are equally good candidates. So D.

macrocarpa is here selected as lectotype purely on

the grounds of convenience. If D. ebracteata is

selected as lectotype, Dovea will be a synonym of

Chondropetalum (where D. ebracteata is included),

and D. macrocarpa will need a new generic name.

This genus is monotypic. It is very similar to

Chondropetalum, differing in the absence of

myricetin derivative and the possession of a corky

seed-coat. Macro-morphologically it can be recog-

nized by the enormous female flowers and the

branching culms. It cannot be included in any of the

related genera without forming paraphyletic groups.

Anatomy group A, pollen not known, seed group

12.

6. Askidiosperma Steud., Syn. PI. Glum. 2: 257

(1855). Type species: Askidiosperma capitatum

Steud.

Both Masters (1878, 1897) and Gilg-Benedict

(1930) recognized this genus, but did not include all

the species that belong here. Possibly Gilg-Benedict

missed the other species of the group because she

did not investigate them anatomically.

This genus of 11 species is separated from

Chondropetalum, where its species were included by

Pillans (1928), by the absence of a double epidermis

and myricetin derivatives and by bracts longer than

the perianth. It is unique by its possession of hyaline

bracts, often lacerated, that are taller than the

perianth. It is linked to the Chondropetalum group

by the inflorescence structure and the caducuous

sheaths.

Anatomy group D (partially), pollen group 8,

seed group 5.

7. Platycaulos Linder, genus novum, Restio

Rottb., Thamnochorto Berg, et Chondropetalo

Rottb. affine, sed culmis compressis, stomatibus

depressis et testa alba, rugosa et retifoveata differt.

Type species: Platycaulos compressus (Rottb.)

Linder, comb, nov., basionym: Restio compressus

Rottb., Program. 11 (1772).

Culms branching, not rhizomatous, more or less

compressed. Sheaths closely convoluted, often

green, persistent, often with large awns. Male and

female inflorescences similar, spicate, with large

spathes, small sessile spikelets and with most bracts

fertile. Perianth cartilaginous, outer lateral segments

conduplicate, carinate, villous on the carina, all

segments imbricate. Male flowers somewhat smaller

than the female flowers, anthers exserted from the

perianth at anthesis. Female flowers with stami-

nodes, gynoecium with three free styles and fruit a

two-locular dehiscent capsule. Anatomy group C,

pollen group 8, seed group 7.

The name refers to the compressed culms.

The species included in this genus have to date

been included in Restio Rottb., but they are

consistently different by the compressed culms,

sunken stomata, a chlorenchyma which is generally

one cell layer wide and the peculiar seed coats.

Seven species are to be included in this group. The

genus is probably closely related to Restio Rottb.

8. Restio Rottb., Programmate 9 (1772), nom.

H. P. LINDER

65

cons. Lectotype species: Restio triticeus Rottb.

(McVaugh, 1968).

Craspedolepis Steud., Syn. PI. Glum. 2: 264

(1855). Type species: Craspedolepis verreauxii

Steud., = Restio filiformis Poir.

This genus of about 80 species agrees with the taxa

which Pillans (1928) included in his informal group

of Restio species with three styles, excluding the taxa

separated in Platycaulos and Rhodocoma. Pillans’s

delimitation of Restio was based on a symple-

siomorphy, the dehiscent fruit. Gilg-Benedict (1930)

appreciated that Restio included several groups, but

made no attempt to break the genus up. The groups

that could be demonstrated to be monophyletic have

here been removed from this exclusively African

genus, (i.e. Ischyrolepis, Nevillea, Platycaulos and

Rhodocoma). However, it is still not possible to

demonstrate that Restio, as delimited here, is

monophyletic. Nor are there any data indicating that

it may be paraphyletic or polyphyletic. Consequent-

ly the group is recognized.

Seed coat data, anatomical data and macro-

morphology suggest various subgroups, but the

data-sets do not correlate well. It may be possible to

recognize infra-generic groups within the genus.

Anatomy group D and J, pollen group 8, seed

groups 8, 9, 11 and 13.

9. Calopsis Beauv. ex Desv. in Ann. Sc. Nat.

13: 44, t.3 (1828). Type species: Calopsis paniculatus

(Rottb.) Desv. = Restio paniculatus Rottb.

This group of about 28 species was previously

included in Leptocarpus R.Br. (Masters, 1878,

1897; Pillans, 1928, Dyer, 1976), but it was

considered to constitute a separate genus by

Gilg-Benedict (1930). Calopsis is here, as also by

Masters, Pillans and Gilg-Benedict, separated from

Restio by the possession of an indehiscent fruit. This

is not, however, a very satisfactory character and the

exact boundary between Restio and Calopsis is not

yet resolved. This will probably only be clarified

when the internal phylogeny of Restio is better

understood.

Anatomy group D, pollen group 8.

10. Thamnochortus Berg., Descr. Plant. Cap.

353 (1767). Type species: Thamnochortus fruticosus

Berg.

This genus of some 32 species is used in the same

sense as by Pillans (1928), Dyer (1976), Gilg-

Benedict (1930) and Masters (1878, 1897). It is a

distinct genus possessing scattered cavities in the

central ground tissue, curiously thickened anti-clinal

walls in the epidermal cells, pendulous male flowers,

and female fruits which are flattened, unilocular and

with a single style. There is some macro-

morphological similarity to Staberoha, but the

anatomy and pollen structure is different.

Anatomy group B, pollen group 3.

11. Rhodocoma Nees in Lindl., Nat. Syst. Bot.,

edn 2:450 (1836). Type species: Rhodocoma capense

Nees ex Steud.

Nees (1836) cited neither a specific name nor a

specimen after his description of Rhodocoma. Kunth

(1841) repeated Nees’s description verbatim.

Steudel (1855) listed Rhodocoma capense, which he

attributed to Nees, after his generic description.

There is no indication of the type of the species.

Masters (1868) described a Restio rhodocoma,

which he claimed to be synonymous with ‘Rhodoco-

ma equisetum,’ which he alleged Nees published in

1836, but which he may have found on a herbarium

sheet. Restio rhodocoma is clearly a superfluous

name for Rhodocoma capense, which must stand as

the type species.

This group of three species has generally been

included in Restio on the basis of two symplesiomor-

phies, namely three free styles and dehiscent

capsules. It shares several synapomorphies with

Thamnochortus, such as the scattered cavities in the

central ground tissue and the pollen type, the

peculiar epidermal cells and simple culms with

pendulous male spikelets. Thamnochortus is kept

distinct from Rhodocoma by the unilocular,

single-styled fruits, and Rhodocoma is distinct by the

type of seed-coat.

Anatomy group B , pollen type 3 and seed type 10.

12. Ceratocaryum Nees in Lindl., Nat. Syst.

Bot., edn 2: 451 (1836). Type species: Ceratocaryum

argenteus Kunth.

Nees (1836) based his generic description on a

specimen named ‘Restio argenteus’ in the Willdenow

herbarium, and on a female specimen collected by

Ecklon. He did not, however, make the necessary

combination. In 1841 Kunth described Ceratocary-

um argenteus, using the specimen in the Willdenow

herbarium as his type. This species then constitutes

the type of the genus.

Ceratocaryum includes five species, which Pillans

(1928) included in Willdenowia. They differ from

typical Willdenowia by the absence of any ridges in

the sclerenchyma ring, by the possession of slender

chlorenchyma cells and by the culms being simple.

The male inflorescence has the flowers aggregated

into spikelets, but the perianth and bracts are

similar, obscuring the spikelet organization.

Masters (1897) and Gilg-Benedict (1930) also

recognized Ceratocaryum, but delimited it from

Willdenowia by having sessile nuts. However, this

character is not correlated with the above suite of

characters, and is probably homoplasious.

Anatomy group I (excluding W. stokoei), pollen

type 4 (partially) and 5 (partially).

13. Cannomois Beauv. ex Desv. in Ann. Sc.

Nat. 13: 43 t.3 fig. 1 (1828). Type species:

Cannomois cephalotes Desv. = Cannomois virgata

(Rottb.) Steud.

Cucullifera Nees in Lindl., Nat. Syst. Bot., edn 2:

451 (1836). Type species: Cucullifera durus Nees,

nom. illeg. , based on Restio acuminatus Thunb.,

nom. illeg. = Cannomois acuminata (Thunb.)

Pillans, nom. illeg.

Mesanthus Nees in Lindl., Nat. Syst. Bot. edn 2:

451 (1836). Type species: Mesanthus macrocarpus

66

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Nees, nom. illeg. , based on Willdenowia compressa

Thunb. = Cannomois virgata (Rottb.) Steud.

This group is circumscribed in the same way as by

Pillans (1928) and Dyer (1976). It is a distinct group

having compressed shiny black nuts and bony bracts.

The female spikelets often have several flowers, as

compared with the single flower found almost

invariably in the rest of the taxa with hard nuts. It is

very difficult to establish the species limits in this

genus which comprises between five and eight

species, several of which are not well-known.

Anatomy group K, pollen group 4.

14. Nevillea Esterhuysen & Linder, genus

novum, quoad anatomiam ad Cannomodem Desv.

accedit, sed ovario biloculari differt.

Types species : Nevillea obtusissimus (Steud.)

Linder, comb, nov., basionym: Restio obtusissimus

Steud., Syn. PI. Glum. 2: 252 (1855).

Culms simple, sheaths closely convoluted. Male

inflorescences of one or several cone-like spikelets,

with numerous fertile, obtuse, densely imbricate

bracts. Perianth membranous, outer laterals condu-

plicate, carinate. Female inflorescence of 2—5

spicate spikelets, each obscured by its spathe and

3—5 flowered. Bracts tightly convoluted. Perianth

lobes membranous, glabrous, the outer laterals

conduplicate. Ovary bilocular, indehiscent, styles 2,

fused at the base. Anatomy group M, pollen group

4.

This genus is named after Neville Pillans.

The two species in this group are related to

Master siella and Cannomois by their anatomy. The

flowers are quite different, and the bilocular

dehiscent ovary seems to indicate that they must be

primitive compared to the rest of the genera with

sclerenchyma ridges.

15. Hydrophilus Linder, genus novum. Hypo-

disco Nees et Cannomodi Desv. affine, sed spiculis

femineis floribus numerosis et nuce perianthio

parviore differt.

Type species •. Hydrophilus rattrayi (Pillans)

Linder, comb, nov., basionym: Leptocarpus rattrayi

Pillans in Trans. R. Soc. S.Afr. 29: 347 (1942).

Culms branching, sheaths tightly convoluted.

Male and female inflorescences similar, of several

spikelets. Spathes and bracts densely imbricate,

chartaceous, mostly fertile. Perianth segments

glabrous, the outer cartilaginous, with the laterals

conduplicate and the inner membranous. The male

flowers are slightly smaller than the female flowers.

The female flowers have an indehiscent unilocular

ovary, with two styles which are shortly connate at

the base. Anatomy group L, pollen not known.

The name is derived from the Greek ‘hydro’ =

water and ‘philos’ = love, reflecting the preference

which this genus has for wet seepages.

There is only one species in this genus. It is

anatomically quite peculiar, but shows all the

features of the Cannomois-Willdenowia group.

Pillans (1942) placed it in Leptocarpus, but noted

that it is ‘a very distinct and somewhat anomalous

species’. He made this comment largely because the

species has two styles, which is rather characteristic

of the Ceratocaryum-Willdenowia group. Within this

group, Hydrophilus is peculiar because it has

numerous flowers per female spike let.

16. Anthochortus Nees in Lindl., Nat. Syst.

Bot., edn 2 : 451 (1836). Type species: Anthochortus

ecklonii Nees.

Phyllocomos Masters in Bot. Jb. 29 Beih. 66: 19

(1900). Type species : Phyllocomos insignis Mast.

Anthochortus, as redefined here, includes two

monotypic genera that Pillans (1928) and Dyer

(1976) recognized, as well as about half of

Master siella Gilg-Benedict (= Hypolaena sensu

Pillans, 1928). This results in a group of six species,

which in some respects (especially anatomically) is

quite variable, but which share several characters

which are difficult to quantify: slender, much

branched culms, spicate inflorescences with spikelets

with few or often single flowers with no or few sterile

bracts, and membranous perianths. The nuts are

unlignified, often greenish. The plants are often

stoloniferous. Frequently, in young growth the leaf

awn is developed as a blade. Although it is difficult

to demonstrate on the data analysed above, this

group appears to be monophyletic.

Anthochortus has been maintained by all previous

authors as a highly distinct genus, largely because of

its peculiar anatomy (Gilg, 1891; Cutler, 1969). An

as yet undescribed species, which Ms E. Esterhuysen

found on Table Mountain at Cape Town, clearly

belongs to the same genus and has false pillar cells.

This indicates that the anatomy can be variable.

Phyllocomos has also been regarded as a peculiarity,

as for a long time only the type collection was known

and this is bisexual. However, later collections are

all dioecious. Obviously, individual species in this

genus show extensive anagenesis.

Anatomy groups H, N and O; pollen group 4.

17. Mastersiella Gilg-Benedict in Nat. Pflan-

zenfam. edn 2, 15A:25 (1930). Lectotype species:

Mastersiella digitata (Mast.) Gilg-Benedict.

After the description, Gilg-Benedict listed six new

combinations, which constitute the syntypes. Of

these, two have been shown to belong to Restio, one

is Calopsis, two are placed in Anthochortus and the

remaining species, M. digitata (Mast.) Gilg-Benedict

is placed in a new genus with two species which were

described later. As there is no nomenclatural reason

why any one of the six species should be preferred as

the lectotype, M. digitata is chosen. This means that

the name Mastersiella can be used for the new genus.

The three species here retained in Mastersiella

were placed in Hypolaena R. Br. by Masters (1897)

and Pillans (1928). It has been shown that the

African species of Hypolaena are not related to the

non-African members of Hypolaena, and conse-

quently Gilg-Benedict described the new genus

Mastersiella for the African species of Hypolaena.

Mastersiella, as originally delimited by Gilg-

Benedict, is here divided: four species are assigned

to Anthochortus and the remaining three, which are

distinct by the reduction in the outer perianth whorl

H. P. LINDER

67

in the male flowers and the woody shiny black nuts'

and many-flowered male spikelets, are retained in

Master siella.

Anatomy group M and pollen group 4.

18. Hypodiscus Nees in Lindl., Nat. Syst. Bot.,

edn 2: 450 (1836), nom. cons. vs. Lepidanthus

Nees. Type species Hypodiscus aristatus (L.) Nees

(vide Masters, 1868).

Leucoploeus Nees in Lindl., Nat. Syst. Bot., edn

2: 450 (1836). Type species: Restio argenteus

Thunb. = Hypodiscus argenteus (Thunb.) Mast.

Boeckhia Kunth, Enum. 3: 448 (1841). Lectotype

species: Boeckhia laevigata Kunth = Hypodiscus

albo-aristatus (Nees) Mast.

Lepidanthus Nees Linnaea 5: 665 (1830), nom.

rej. Type species: Lepidanthus willdenowia nom.

illeg., based on Willdenowia striata Thunb. =

Hypodiscus willdenowia Mast.

Kunth listed two species after his description of

Boeckhia, namely B. striata and B. laevigata. The

latter fits the generic description better and is

consequently chosen as lectotype of Boeckhia.

Hypodiscus includes about 15 species and is

circumscribed here in the same way as by Pillans

(1928), Masters (1897) and Dyer (1976). It is quite a

variable genus and Pillans delimited it by the

symplesiomorphy of the male flowers being aggre-

gated into spikelets. I have also failed to demon-

strate any unique characters, that would support the

hypothesis of monophyly for this group. There is

extensive variation in the anatomy of the group and

some variation in the macro-morphology. This

suggests that two subgroups could be recognized,

based on the presence of false pillar cells and the

lack of differentiation between male and female

spikelets. However, there is insufficient correlation

among the characters, so that it seems more sensible

to leave it as a single group.

Anatomy group M and P, pollen group 4.

19. Willdenowia Thunb. in K. svenska Veten-

sAkad Handl. 11: 28 (1790). Lectotype species:

Willdenowia striata Thunb.

Nematanthus Nees in Linnaea 5: 661 (1830). Type

species: Nematanthus ecklonii Nees = Willdenowia

striata Thunb.

Thunberg originally included three species in

Willdenowia: W. striata, W. teres and W. compressa.

The last species has been transferred to Canno-

mois, into the synonymy of C. virgata. There is no

compelling reason to choose either of the first two

taxa, so W. striata is selected. Thunberg (1790) used

the spelling Wildenowia, which he repeated in 1794.

In 1811 he used both Willdenowia and Wildenowia.

Willdenow himself (1806) used the spelling Willde-

nowia. Farr et al. (1979) use Wildenowia and regard

Willdenowia as an orthographic variant. However,

since Willdenowia has been used consistently since

1811, and Thunberg used it in 1811, and Willdenow

himself used it, too, I shall carry on using it, and

consider that Thunberg made an error using the

spelling ‘Wildenowia. ’

This group of about ten species is a segregate of

Willdenowia as used by Pillans (1928). It includes all

the species in which the male inflorescence is truly

thyrsoid and the perianth segments are linear. As

such it is a very distinct genus, which agrees

approximately with the generic delimitation of the

genus by Masters (1897) and Gilg-Benedict (1930).

There are two distinct subgroups within the genus.

One group of eight species has branching culms and

loosely convoluted sheaths, while the remaining two

species have simple culms, closely convoluted

sheaths and elongated ornamented nuts. The

anatomy of the group is variable, and parallels that

of Anthochortus and Hypodiscus.

Anatomical groups G, M and P; pollen groups 4

and 5.

CONCLUSION

None of the data-suites agrees completely with the

final cladogram. Homoplasy is postulated for the

pollen surface features, in some anatomical features,

in the inflorescence structures and in the gynoecia

and fruits. Homoplasy cannot be demonstrated for

the seed coats, as seed coats are uninformative on

the structure of the cladogram. Some of the

convergence demonstrated can be very striking, such

as between ‘Lamprocaulos’ (in Elegia) and Willde-

nowia involving numerous characters. Homoplasy is

also postulated for a few apparently unique

characters, such as the false pillar cells in

Anthochortus, Hypodiscus and Willdenowia. It is

clear that it would be very risky to base a

classification on a single data set, as the homoplasies

could then not be detected.

The resulting cladogram is quite well corroborated

and the only section that may change upon the

addition of one extra datum is the sequence of

genera from Ceratocaryum to Willdenowia, where

several other groupings would be almost equally

parsimonious. This does not affect the delimitations

of the terminal taxa. It would be quite possible to

use the cladogram for structural-functional interpre-

tations, especially as regards the perianth and

inflorescence structure, and the dispersal biology.

Both are related to the important field of the

reproductive biology of the taxa — a field as yet little

investigated.

The phylogeny presented here, with the associ-

ated generic classification, differs strongly from the

classifications presented by Pillans (1928) and

Gilg-Benedict (1930).

Pillans (1928) recognized 12 genera. Of these,

only five remain essentially unaltered. Five genera

are split into two or more genera, and two genera

are fused. The classification here presented has 19

genera. It is likely that the Pillansian classification is

so different from the present classification because

(a) it was based only on macro-morphological data

and (b) Pillans recognized groups based on

symplesiomorphies.

There is much more agreement with the

classification of Gilg-Benedict (1930), partially

because she distinguished smaller genera than

Pillans, but also because she recognized the

68

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

differences between the African and the Australian

taxa probably due to her use of anatomical data.

However, her delimitations of the African genera

still left much to be desired, possibly because she

had only investigated the anatomy of a few species.

Her grouping of the genera also differs dramatically

from the groupings adopted here.

Given the much wider data-base available in this

study, when compared to the data available to

Pillans and Gilg-Benedict, it is not surprising that

the genera have again changed so much. It remains

to be seen what effect future additional data will

have on the classification.

ACKNOWLEDGEMENTS

The work for this paper was performed at the

Bolus Herbarium and the Botany Department of the

University of Cape Town, the Botanical Research

Institute in Pretoria, and the Herbarium and Jodrell

Laboratories at the Royal Botanic Gardens, Kew. I

am greatly indebted to Ms Elsie Esterhuysen of the

Bolus Herbarium for her encouragement and help,

for allowing me access to her very rich collections of

Restionaceae and for freely sharing her knowledge

with me. Dr Keith Ferguson prepared the TEMico-

graphs of the pollen and gave much helpful advice;

Mrs S.M. Perold took hundreds of SEMicrographs

of seed and pollen at PRE; Dr David Cutler allowed

Aie the use of the anatomy slide collection at the

Jodrell Laboratories and assisted me greatly in

understanding Restioid culm anatomy and Peter

Gasson showed me many anatomical tricks. I thank

Dr Paula Rudall, Peter Jaeger and various others for

much stimulating discussion.

UITTREKSEL

Die generiese grense van die genera van die

Restionaceae van Afrika word hersien. Drie nuwe

genera (Nevillea Esterhuysen & Linder, Platycaulos

Linder en Hydrophilos Linder) word beskryf, die

grense van Restio Rottb., Chondropetalum Rottb.,

Mastersiella Gilg-Benedict, Willdenowia Thunb. en

Anthochortus Nees word ingrypend verander, en

Ischyrolepis Steud., Dovea Kunth, Askidiosperma

Steud., Rhodocoma Nees en Ceratocaryum Nees

word weer in gebruik geneem. Hierdie hersiening is

op ’n kladistiese analise van makro-morfologiese,

saadhuid-morfologiese, anatomiese, fitochemiese en

palinologiese gegewens gebaseer.

Drie nuwe kombinasies word gemaak, naamlik

Platycaulos compressus (Rottb.) Linder, Nevillea

obtusissimus (Steud.) Linder en Hydrophilos rattrayi

(Pillans) Linder.

REFERENCES

Bentham.G. & Hooker, J. D., 1883. Genera Plantarum, Vol. 3.

London: Reeve.

Bergius, P.J., 1767. Plantae Capenses. Stockholm.

Blackmore, S., 1982. Palynology of subtribe Scorzonerinae

(Compositae: Lactuceae) and its taxonomic significance.

Grana 21 : 149—160.

Borwein, B., Coetzee.M.Z. & Krupko.S., 1949. Development

of the embryosac of Restio dodii and Elegia racemosa. Jl S.

Afr. Bot. 15 : 1-12.

Botha, D. J., 1971. Taksonomies—morfologiese ondersoek van

die genus Elegia L. M.Sc. thesis. University of Pretoria.

Botha, D.J. & Van der Schijff, H.P., 1976. The structure and

ontogeny of the chlorenchyma in the stems of Elegia L.

(Restionaceae). Jl S. Afr. Bot. 42: 57-62.

Botha, D.J., Van der Schijff, H.P. & VanTonder, E.M.A.,

1972. The position, structure and ontogeny of the stomata in

the stems of Elegia vaginulata. Tydskr. Natuurwet. 12 :

193-199.

Bremer, K. & Wanntorp, H.-E., 1978. Phylogenetic systematics

in Botany. Taxon 27 : 317—329.

Briggs, B.G., 1963. Chromosome numbers in Lepyrodia and

Restio in Australia. Contr. N.S. W. natn. Herb. 3 : 228—232.

Briggs, B.G., 1966. Chromosome numbers of some Australian

monocotyledons. Contr. N.S.W. natn. Herb. 4 : 24-34.

Chanda, S., 1966. On the pollen morphology of the Centrolepi-

daceae, Restionaceae and Flagellariaceae, with special

reference to taxonomy. Grana Palynologica 6 : 355-415.

Chanda, S. & Ferguson, I.K., 1979. Pollen morphology of

Calorophus and Empodisma (Restionaceae) and its taxono-

mic significance. Kew Bull. 33 : 411-416.

Chanda, S. & Rowley, J. , 1967. Apertural types in pollen of the

Restionaceae and Flagellariaceae. Grana Palynologica 7 :

16-36.

Cheadle, V.I., 1955. The taxonomic use of specialization of

vessels in the metaxylem of Gramineae, Cyperaceae,

Juncaceae and Restionaceae. J. Arnold Arbor. 36 :

147-157.

Cheadle, V.I. & Kosakai, H., 1975. Vessels in Juncales: II.

Centrolepidaceae and Restionaceae. Am. J. Bot. 62 :

1017-1026.

Christensen, J.E. & Horner, H.T., 1974. Pollen pore

development and its spatial orientation during microspo-

rogenesis in the grass Sorghum bicolor. Am. J. Bot. 61 :

604-623.

Cutler, D.F., 1965. The taxonomic significance of the anatomy of

Restionaceae. Ph. D. thesis. Imperial College, London.

Cutler, D. F., 1966. Anatomy and taxonomy of the

Restionaceae. Notes Jodrell Lab. 4.

Cutler, D.F., 1969. Juncales. In C.R. Metcalfe, Anatomy of the

Monocotyledons. Oxford: Clarendon Press.

Cutler, D.F., 1972. Vicarious species of Restionaceae in Africa,

Australia and South America. In D.H. Valentine, Tax-

onomy, phytogeography and evolution. London: Academic

Press.

Dahlgren, R. & Clifford, H.T., 1982. The monocotyledons: a

comparative study. London: Academic Press.

Desvaux, N.A., 1828. Observations sur quelque families des

plantes monocotyledones, d'apres les manuscrits de feu le

Baron Palisot de Beauvois. Annls Sci. nat. 13 : 37-52.

Dyer, R.A., 1976. The genera of southern African flowering

plants, Vol. 2. Pretoria: Government Printer.

Eldredge, N. & Cracraft, J., 1980. Phylogenetic patterns and

the evolutionary process. New York: Columbia University

Press.

Erdtman, G., 1969. Handbook of palynology. Copenhagen:

Munksgaard.

Erdtman, G., 1971. Pollen morphology and plant taxonomy.

New York: Hafner.

Farr, E.R., Leussink, J.A. & Stafleu, F.A. (eds.), 1979. Index

Nominum Genericorum (Plantarum). Utrecht & The

Hague.

Gilg. E., 1891. Beitraege zur vergleichenden Anatomie der

xerophilenFamilie der Restionaceae. Bot. Jb. 13:541—606.

H. P. LINDER

69

Gilg-Benedict, C., 1930. Restionaceae. In A. Engler, Die

Natuerlichen Pflanzenfamilien edn 2, Vol. 15a : 8—27.

Leipzig: Wilhelm Engelmann.

Goldblatt, P., 1978. An analysis of the flora of southern Africa:

its characteristics, relationships and origins. Ann. Mo. bot.

Gdn 65 : 369-436.

Guedes, M., 1982. Exine stratification, ectexine structure and

Angiosperm evolution. Grana 21 : 161 — 170.

Harborne, J., 1979. Correlations between flavonoid chemistry,

anatomy and geography in the Restionaceae. Phytochemis-

try 18 : 1323-1327.

Harborne, J. B. & Clifford, H.T., 1969. Flavonoid patterns of

the Restionaceae. Gossypetin in Restio and a new flavone in

Hypolaena. Phytochemistry 8 : 2071—2075.

Hennig, W., 1979. Phylogenetic systematics edn 2. Urbana:

University of Illinois Press.

Hochuli, P.A., 1979. Ursprung und Verbreitung der Res-

tionaceen. Vjschr. naturf. Ges. Zurich 124 : 109—131.

Johnson, L.A.S. & Briggs, B., 1981. Three old southern families

— Myrtaceae, Proteaceae and Restionaceae. In A. Keast

Ecological biogeography of Australia, 427—469. Utrecht:

Junk.

Johnson, L.A.S. & Cutler, D.F., 1973. Empodisma, a new

genus of Australian Restionaceae. Kew Bull. 28 : 381—385.

Jordaan, P.G., 1946. Die anatomie van die wortel van Restio

paludosus en Leptocarpus vimineus. Jl S. Afr. Bot. 12 :

115-120.

Keng, H., 1978. Orders and families of Malayan seed plants.

Singapore: Singapore University Press.

Koehler, E. & Lange, E., 1979. A contribution to distinguishing

cereal from wild grass pollen grains by LM and SEM. Grana

18 : 133-140.

Kress, W.J. & Stone, D.E., 1982. Nature of the sporoderm in

monocotyledons, with special reference to the pollen grains

of Canna and Heliconia. Grana 21 : 129-148.

Krupko, S., 1962. Embryological and cytological investigations in

Hypodiscus aristatus Nees (Restionaceae). Jl S. Afr. Bot. 28

: 21-44.

Krupko, S., 1964. Macrosporogenesis and embryo sac develop-

ment in Chondropetalum hookerianum (Mast.) Pillans

(Restionaceae). Acta Soc. Bot. Pol. 32 : 171-190.

Krupko, S., 1966. Some varied embryological and cytological

observations on members of Restionaceae family. Bull. Soc.

Amis Sci. Lett. Poznan 7 : 59—67.

Kunth, C.S. 1841. Enumeratio Plantarum Vol. 3. Stuttgard:

Cotta.

Ladd, P.G., 1977. Pollen morphology of some members of the

Restionaceae and related families, with notes on the fossil

record. Grana 16: 1-14.

Larsen, K., 1972. Restionaceae FI. Thailand 2, 2 : 172-174.

Linnaeus, C., 1767. Systema Naturae edn 12, Vol. 2 : 735.

Stockholm: Laurentii Salvii.

Linnaeus, C., 1771. Mantissa Plantarum altera. Stockholm:

Laurentii Salvii.

Masters, M.T., 1865. Observations on the morphology and

anatomy of the genus Restio L. together with an

enumeration of the South African species. J. Linn. Soc.,

Bot. 8 : 211-255.

Masters, M.T., 1869. Synopsis of the South African Res-

tionaceae. J. Linn. Soc., Bot. 10 : 209 —279.

Masters, M.T., 1878. Restiaceae. In A. & C de Candolle,

Monographiae Phanerogamarum Vol. 1 : 218—398. Paris:

Masson.

Masters, M.T., 1897. Restiaceae. In W.T. Dyer, Flora Capensis

Vol. 7. London: Reeve.

Masters, M.T., 1900. Restionaceae novae capenses herbarii

Berolinensis, imprimis Schlechterianae. Bot. Jb. 29 Beih. 66

: 1-20.

McVaugh, R., 1968. Report of the committee for Spermatophy-

ta. Conservation ofgeneric names, VIII. Taxon 17 : 85-87.

Moore, L.B. & Edgar, E., 1970. Flora of New Zealand Vol. 2.

Wellington: Government Printer.

Muller, J., 1981. Fossil pollen records of extant angiosperms.

Bot. Rev. 47 : 1-142.

Nees von Esenbeck, C.G., 1830. Beitrag zur Kenntniss der

Familie der Restiaceen in Ruecksicht auf Gattungen und

Arten. Linnaea 5 : 627—666.

Nees von Esenbeck, C.G., 1836. In J. Lindley, Natural system of

Botany edn 2. London: Longman, Rees, Orne, Brown,

Green & Longman.

Netolitzky, F., 1926. Anatomie der Angiospermen-Samen. In

K. Linsbauer, Handbuch der Pflanzenanatomie Vol. X.

Berlin: Gebrueder Borntraeger.

Noel, A.R.A. , 1959. The stem anatomy of Restio triticeus Rottb.,

Bobartia indica L., and Cadaba juncea (Sparm.) Harv. Jl S.

Afr. Bot. 25 : 357-370.

Page, J.S., 1978. A scanning electron microscope survey of grass

pollen. Kew Bull. 32 : 313-319.

Pfitzer, E. , 1869- 1870. Beitraege zur Kenntnis der Hautgewebe

der Pflanzen. II. Ueber das Hautgewebe einiger Res-

tionaceen. Pringsheims Jb. wiss. Bot. 1 : 561—584.

Phillips, E.P. , 1951. The genera of South African flowering plants

edn 2. Pretoria: Government Printer.

Pillans, N.S., 1928. The African genera and species of

Restionaceae. Trans. R. Soc. S. Afr. 16 : 207 —440.

Pillans, N.S., 1942. New species of South African Restionaceae.

Trans. R. Soc. S. Afr. 29 : 339—356.

Pillans, N.S., 1945. New and hitherto imperfectly known species

of African Restionaceae. Trans. R. Soc. S. Afr. 30 :

245-266.

Pillans, N.S., 1952. New and hitherto imperfectly known species

of South African Restionaceae. Jl S. Afr. Bot. 18 :

101-122.

Pizarro, C.M., 1959. Sinopsis de la flora chilena. University of

Chile.

Rottboell, C.F., 1772. Descriptiones plantarum rariorum

iconibus lustrandas programmate indicit. Copenhagen.

Skvarla, J.J., 1966. Technique of pollen and spore microscopy.

Part I. Staining, dehydration and embedding. Okla. Geol.

Notes 26 : 179—186.

Skvarla, J.J. & Kelley, A.G. 1968. Rapid preparation of

pollen and spore exines for electron microscopy. Stain

Technol. 43 : 139—144.

Skvarla, J.J. & Pyle, C.C., 1968. Techniques of pollen and

spore electron microscopy. Part II. Ultramicrotomy and

associated techniques. Grana Palynologica 8 : 255-270.

Slingsby, P. & Bond, W., 1981. Ants, friends of the fynbos. Veld

Flora 67 : 39-45.

Specfit, R.L., 1981. Major vegetation formations in Australia. In

A. Keast, Ecological biogeography of Australia. The Hague:

Junk.

Stearn, W.T., 1973. Botanical Latin edn 2. Newton Abbot:

David & Charles.

Steudel, E.G., 1855. Synopsis Plantarum Glumacearum Vol. 2.

Stuttgard: Metzler.

Thunberg, C.P., 1788. Dissertatione de Restione. In Disserta-

tiones Academicae Upsaliae Habitae Vol. 2 (1800) :

298-316. Gottingen: Dieterich.

Thunberg, C.P., 1790. Beskrifning pao Wildenowia, et

synnerlight och nytt Graos-Slag. K. svenska Vetensk. Akad.

Handl. 11 : 26-32.

70

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Thunberg, C.P., 1794. Prodromus Plantarum Capensium.

Uppsala: Edman.

Thunberg, C.P. 1811. Flora Capensis Vol. 1. Uppsala.

Ueberfeld, M., 1925. Beitraege zur Kenntnis des sexuellen

Dimorphismus der Restionaceen. Bot. Jb. 60 : 176-206.

Van Campo, M., 1978. La Face interne de l’exine. Rev.

Palaeobot. Palynol. 26 : 301-311.

Van Greuning. J.V., 1972. Morphological study of Hypodiscus

and Willdenowia. M.Sc. thesis. University of Pretoria.

Van Greuning, J.V. & Van der Schijff, H.P., 1973. Chloren-

chyma and sclerenchyma in the culms of Willdenowia

Thunb. and Hypodiscus Nees. Ber. dt. bot. Ges. 86 :

537-550.

Van Greuning, J.V. & Van der Schijff, H.P., 1975. External

morphology of the genera Hypodiscus Nees and Willdeno-

wia Thunb. and a key to the genera and species based on

external morphological properties. Kirkia 9 : 331-347.

Van Zinderen Barker, E.M., 1953. South African pollen grains

and spores Part 1. Amsterdam: Balkema.

Vrba, E., 1^79. Phylogenetic analysis and classification of fossil

and recent Alcelaphini Mammalia: Bovidae. Biol. J. Linn.

Soc. 11 : 207-228.

Watrous, L.E., & Wheeler, Q.D., 1981. The outgroup

comparison method of character analysis. Syst. Zool. 30 :

1-11.

Watson, L., & Bell, E.M., 1975. A surface-structural survey of

some taxonomically diverse grass pollens. Aust. J. Bot. 23 :

981-990.

Weberling, F., 1981. Morphologie der Bleuten und der

Bleutenstaende. Stuttgart: Ulmer.

Willdenow, C.L., 1806. Species Plantarum edn 4, Vol. 4,2.

Berlin.

APPENDIX 1. Vouchers for the scanning electron microscope study of seed surface micro-morphology. Names of new taxa in inverted

commas are not yet published

H. P. LINDER

71

72 A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

APPENDIX 2. Vouchers for the light microscope study of seed coats

H. P. LINDER

73

74

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

H. P. LINDER

75

sulcata Mast.

teres Thunb.

xerophila Pillans

Acocks 24485 K

Esterhuysen 7472 K

Burchell 787 K

Esterhuysen 23143 K

Parker 3527 K

Esterhuysen 30655 K

APPENDIX 4. Vouchers for the palynological study. LMt indicates that only temporary slides were made; Lm indicates permanent slides,

and SEM that material was studied by SEM

76

A PHYLOGENETIC CLASSIFICATION OF THE GENERA OF THE AFRICAN RESTIONACEAE

Bothalia 15, 1 & 2: 77-88 (1984)

Revision of the genus Myrsiphyllum Willd.

A. A. OBERMEYER*

Keywords: Africa, Liliaceae, Myrsiphyllum, revision

ABSTRACT

The genus Myrsiphyllum Willd. (Liliaceae — Asparageae) is revised. Twelve species are recognized, one of

which is new, namely, M. alopecurum Oberm. Eight new combinations are made. A key is provided for

distinguishing Myrsiphyllum from Protasparagus Oberm.

MYRSIPHYLLUM

Myrsiphyllum Willd.** in Ges. naturf. Freunde

Berl. Magazin 2: 25 (1808); Kunth, Enum. PL 5: 105

(1850). Type species: M. asparagoides (L.) Willd.

Hecatris Salisb., Gen. PI. 66 (1866). Type species:

H. asparagoides (L.) Salisb.

Asparagus, section Myrsiphyllum (Willd.) Bak. in

J. Linn. Soc. 14: 597 (1875); FI. Cap. 6: 258 (1896);

Jessop in Bothalia 9: 38 (1966); Dyer, Gen. 2: 943

(1976)7

Perennial, innocuous, glabrous climbers or erect,

usually chamaephytes. Rhizome cylindrical, often

not lignified; cataphylls small or vestigial. Roots

placed radially on the often long, creeping rhizome,

or irregularly dorsiventral on a compact rhizome;

forming fusiform tubers crowded on rhizome or

distant from it, filled with soft tissue saturated with

aqueous fluid; in M. ovatum (Salter) Oberm. and M.

undulatum (L.f.) Oberm. new plants may evolve on

distal side of a tuber. Stems voluble or erect.

Phylloclades solitary or 2— 3-nate, placed in axils of

scale-leaves, the latter not forming spines. (A/.

fasciculatum (Thunb.) Oberm. forms an exception in

that the cladodes are many in each fascicle and small

spines are developed.) Flowers 1— 3-nate, bisexual,

pendulous on short to long pedicels or stalks (viz. a

combination of pedicel and pericladium, separated

by a disk). Tepals white, usually with a green central

band, connivent at base, forming a cup or tube, free

lobes recurved above (except in M. ramosissimum

and M. scandens where they remain erect). Stamens

erect, usually connivent around gynoecium with

filaments flattened, attenuate above, widened

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

** Origin of name: the phylloclades of M. asparagoides (L.)

Willd. resemble the leaves of the myrtle.

t In Clifford & Dahlgren’s study The Monocotyledons (1982) H.

Huber divided the classical family Liliaceae of Kunth, Baker and

others into two Orders, Asparagales and Liliales. The former

contains 24 families, the latter eight. His family Asparagaceae is

represented by six genera; two of these, Protasparagus and

Myrsiphyllum are African.

For the present it was decided to leave these two African

genera in the Liliaceae sensu lato to conform with the Flora of

Southern Africa pattern.

below, where they may form two extended spurs;

anthers introrse, yellow, orange or red. Ovary

3-locular; ovules 6—12 in each locule, biseriate;

styles 1 or 3, stigmas 3, papillate. Berry globose or

ovoid-apiculate, red, yellow or orange; seeds

globose, black.

Species 12, recorded from the Winter Rainfall

Region, with M. asparagoides and M. ramosissimum

extending along the eastern Escarpment to the

Transvaal; the former also spreading northwards to

tropical Africa and southern Europe. Recently

recorded as a troublesome adventive in Australia.

The genus Myrsiphyllum, separated from Aspar-

agus by Willdenow in Ges. naturf. Freunde Berl.

Mag. 2: 25 (1808), was upheld by Kunth in his

Enum. Plant. 5 (1850). Baker in J. Linn. Soc., Bot.

14 (1875), however, placed it as a section of

Asparagus and so did subsequent taxonomists. H.

Huber in The Monocotyledons, a comparative study

by Dahlgren & Clifford (1982), resuscitated the

genus and I agree.

1. M. multituberosum (R. A. Dyer) Oberm.,

comb. nov.

Asparagus multituberosus R. A. Dyer in Bothalia 6: 442 (1954).

Type: Ceres, Karoopoort, Marloth 9006 (PRE, holo.!; STE!).

A. asparagoides sensu Jessop in Bothalia 9: 81 (1966).

Scandent with thin stems up to c. 0,4 m tall, or

short, erect and compact (in dry areas in Nama-

qualand). Rhizome thin, horizontal, up to 0,5 m

long, covered by a cylinder of overlapping small,

fusiform, pointed tubers c. 10 x 3 mm, occasionally

some continuing below as normal thin roots but

scattered long filiform roots are also produced,

sometimes one of these producing a tuber distant

from rhizome. Phylloclades ovate to cordate, c.

15-25 mm long, apiculate, many-veined, margin

papillate. Flowers 1-3 in axils of membranous

scale-leaves; stalk c. 5 mm, curved, articulated

below perianth. Tepals c. 7 mm, forming a wide tube

below, spreading above. Stamens as long as tepals,

filaments expanded at base, flat. Ovary ovoid, with

10—12 ovules in each locule. Styles 3, curved

outwards; stigmas apical, papillate. Berry not seen.

Figs. 1.1, 2.1 & 3.1.

Only known from the western Cape in montane

vegetation. Flowering July-September. Fig. 4.

78

REVISION OF THE GENUS MYRS1PHYLLUM WILLD.

KEY TO GENERA PROTASPARAGUS (ASPARAGOPSIS KUNTH) $ AND MYRSIPHYLLUM WILLD.

Perianth with 6 free spreading tepals; stamens free, spreading; flowers erect, fascicled or racemose, rarely

solitary: spines present (absent in P. virgatus and P. pendulus ); rhizome dorsiventral; cataphylls

present 1113 Protasparagus

Perianth with the tepals cohering in lower half, forming a tube, spreading to recurved in upper half; stamens

cohering with filaments forming a tube around gynoecium (exceptions are M. scandens and M.

ramosissimum where flowers resemble those of Protasparagus). Flowers pendulous, solitary or rarely

2-3 placed next to one another on a node; spines absent; roots emerging radially from the cylindrical

rhizome, rarely dorsiventral 1113a Myrsiphyllum

KEY TO SPECIES OF MYRSIPHYLLUM

Phylloclades solitary:

Scandent plants (but cf. M. multituberosum which may be erect when stunted);

Phylloclades ovate to cordate, 12—20 mm broad:

Root-tubers attached close to rhizome:

Styles 3: root-tubers cylindrical, c. 10-12 x 3 mm, overlapping on long rhizomes

1 . M. multituberosum

Style 1: root-tubers fusiform, c. 40-60 x 10-15 mm, radiating from

rhizome 2. M. asparagoides

Root-tubers distant from rhizome 3. M. ovatum

Phylloclades narrowly ovate to broadly linear, c. 4—6 mm broad:

Phylloclades narrow ovate, midrib distinct; root-tubers fusiform, firm 4. M. kraussianum

Phylloclades narrowly elliptic without a distinct midrib, soft; root-tubers oblong-globose

5. M. volubile

Erect plants:

Plants branching freely; phylloclades narrowly ovate-acuminate, strongly ribbed 6. M. undulatum

Plants ‘fox-tail-shaped’, viz the erect stems bearing numerous contracted branches forming a leafy

cylinder around stem; phylloclades narrowly linear:

Phylloclades 1—2 mm broad with smooth or minutely ciliate margins: roots with fusiform tubers

some distance away from rhizome 7. M. juniperoides

Phylloclades 0,5 mm broad; margin fimbriate; root-tubers densely packed on rhizome

8. M. alopecurum

Phylloclades in fascicles of 3, or very numerous, linear to filiform; scandent, shade-loving:

Phylloclades in fascicles of 3, linear to filiform:

Branches declinate (viz forming recurved loops); branchlets short with overlapping phylloclades; berry

ovoid, shortly pointed above and below, many-seeded 9. M. declinatus

Branches and branchlets spreading at right angles to stem; berry globose, 1-few-seeded:

Phylloclades spreading in one plane, shallowly S-shaped, c. 2 mm wide, one opposing two others and

one shorter than other two 10. M. scandens

Phylloclades laxly spreading, c. 1 mm wide, with a prominent midrib; stems, branches and branchlets

angled 11. M. ramosissimum

Phylloclades in dense fascicles (15-30-nate), linear, arcuate, c. 4-8 mm long 12. M. fasciculatum

Vouchers

Marloth 9006; Bayer 3454; Schlechter 8069; Marloth 12778;

Rosch & Le Roux 01218 (01211).

It is the only Myrsiphyllum species in which the

three styles have not fused into a single column.

2. Myrsiphyllum asparagoides (L.) Willd. in

Ges. naturf. Freunde Berl., Mag. 2: 25 (1808);

Kunth, Enum. Plant. 5: 105 (1850); Hook. f. in

Curtis’s bot. Mag., t. 5584 (1866).

Medeola asparagoides L., Sp. PI. 339 (1753); Mant. alt. 370

(1771); Del. in Red., Liliac. , t. 442 (1816). Hecatris asparagoides

(L.) Salisb., Gen. Plant. 66 (1866). Asparagus asparagoides (L.)

Wight in Century Dictionary II: 845 (1909); Jessop in Bothalia 9:

81 (1966), nom. illegit. Asparagus asparagoides (L.) Druce in

Rep. both Exch. Club Brit. Isl. 1913, iii: 414 (1914). Type: Tilli,

Cat. Plant. Horti Pisani, 1. 12, f. 1 (1723) as ‘Asparagus Africanus,

scandens Myrti folio’ (icono.).

Medeola angustifolia Mill., Diet, edn 6 (1768); Ait., Hort. Kew.

edn 1: 490 (1789). Myrsiphyllum angustifolium (Mill.) Willd. l.c.

2: 25 (1808). Asparagus medeoloides (Thunb.) Bak., var.

angustifolius (Mill.) Bak. in F.C. 6: 273 (1896). Type: Tilli, Cat.

Plant. Horti Pisani, t. 12, f. 2 (1723), as ‘Asparagus Africanus,

scandens Myrti folio angustiore’ (icono.).

t Kunth’s name Asparagopsis (Enum. PI. 5: 76, 1850), is a

homonym as it was used by Montagne in 1841 for a genus in the

Rhodophyceae.

Dracaena medeoloides L.f. , Suppl. 203 (1781). Asparagus

medeoloides (L.f.) Thunb., Prodr. 66 (1794), FI. Cap. edn 2: 333

(1823); Bak. in J. Linn. Soc., Bot. 14: 627 (1875); F.C. 6: 272

(1896); Marloth, FI. S. Afr. 4, t. 20 f (1915). Type: Cape of Good

Hope, without precise locality, Thunberg (UPS, holo.; BOL!;

PRE!, photo 8453).

Elide Medicus (1791) by lectotypification, cf. Farr, Leusink &

Stafleu, Index Nom. Gen. 1139 (1979).

Myrsiphyllum falciforme Kunth, Enum. PI. 5: 107 (1850);

Saund., Ref. Bot., t. 47 (1869). Type: Cape, without precise

locality, Drege 2704a in ‘Herb. Luc.’ (K, iso.). Asparagus

medeoloides Thunb. var. falciformis (Kunth) Bak. in F. C. 6: 273

(1896).

Asparagus kuisibensis Dinter in Feddes Repert. 29: 270 (1931).

Type: South West Africa/Namibia at the Kuiseb River, Tjuezu in

Herb. Dinter 4698 (B, holo.; PRE, photo!).

Scandent, much branched perennials with shiny

green ovate phylloclades, deciduous or semideci-

duous. Rhizome cylindrical, bearing numerous

fusiform root-tubers c. 40—60 x 10—20 mm, radially

arranged, variable in size, close to rhizome. Stems

twisting, wiry, smooth or ridged, up to 2 m tall.

Branches usually short, with beaded ridges. Phylloc-

lades variable in size and shape, ovate-acuminate, c.

25-40 x 8-20 mm, flat or folded and curved,

many-nerved but mostly with 3 more pronounced on

each side; margin smooth or minutely denticulate.

Flowers on pedicels c. 10 mm long, articulated below

perianth. Tepals 6-10 mm long, forming a tube in

lower half, reflexed above. Stamens erect, con-

A. A. OBERMEYER

79

R I'lcrlcsrtjj'

Fig. 1. — 1, root system of Myr-

siphyllum multituberosum, x

1; 2, root system of M.

asparagoides, x 3, M.

ovatum showing a young

plant appearing from a root,

reduced [after Fig. 1. on p.

167 in J1 S. Afr. Bot. 6: 166

(1940)]; 4, a root of M.

fasciculatum, x ); 5, flower

of M. ovatum [t. 1 146 in

Flower. PL Afr. 29 (1943) by

C. Letty, x V6],

nivent, expanded below into 2 small spreading teeth;

anthers red. Ovary pear-shaped, stipitate, narrowed

into a style as long as ovary; ovules c. 6 in each

locule; stigmas short, spreading, ciliate. Berry

globose, c. 10 mm in diam., usually many-seeded.

Fig. 1.2, 2.2, 3.2, 5a & 5b.

Widespread and common in southern and eastern

Cape forests and coastal valley bushveld; inland in

wooded areas and along riverbanks in the eastern

parts of Natal and Transvaal. Also recorded from

South West Africa/Namibia and further north to

tropical Africa. Naturalized locally in southern

Europe. An invader in Australia where it has

become troublesome. Flowering July— September.

Fig. 6.

Vouchers

Goldblatt 2638; Archibald 3610; Rosch & Le Roux 457; Ward

4998; Merxmuller & Giess 32272; Purcell 7.

Introduced in England in 1702 by the Duchess of

Beaufort. It was in cultivation in the hortus in Pisa,

Italy and illustrated in Tilli’s Catalogue of Plants in

1723, on tab. 12, f. 1. The plant pictured on the

right-hand side has narrower, somewhat smaller

phylloclades, which led Tilli, Miller and others to

regard it as a separate species but the phyllodes vary

in size and width. The origin of the plants, figured by

Tilli, is unknown.

3. Myrsiphyllum ovatum (Salter) Oberm.,

comb. nov.

Asparagus ovatus Salter in J1 S. Afr. Bot. 6: 167 (1940); in FI.

Cape Penins. 174 (1951); Dyer in Flower. PI. Afr. 29: t. 1146

(1943). Type: Cape, Rugby, Salter 8214 (BOL, holo.!; NBG!;

PRE!). Asparagus asparagoides sensu Jessop in Bothalia 9: 81

(1966).

Scandent, with twining branches up to 1 — 1,5 m

tall. Rhizome compact scaly, woody with long roots

extending in all directions, bearing numerous hard,

swollen, fusiform tubers c. 50—100 x 20 mm, far

removed from rhizome; much smaller in young

plants; tubers may bring forth young plants at their

proximal ends. Phylloclades deciduous, ovate, 30 x

15 mm but variable in size, many-nerved, shiny.

Flowers 1-3, beside the base of a phylloclade;

pedicels c. 10 mm, articulated near base of flower.

80

REVISION OF THE GENUS MYRSIPHYLLUM WILLD.

Fig. 2. — Seed coats of Myrsiphyllum species (SEM photographs). 1, M. multituberosum , x 600 ; 2, M.

asparagoides, x 600; 3, M. ovatum, x 240; 4, M. kraussianum, x 600; 5, M. volubile, x 600; 6, M.

undulatum, x 600.

Perianth c. 6 mm, tepals re flexed near centre.

Stamens with filaments flattened, erect, bearing a

small basal spur on each side. Ovary oblong, with c.

8—10 ovules in each locule; style and stigmas just

exserted from staminal column. Berry globose, 10

mm in diam., red. Figs 1.3, 1.5, 2.3 & 3.3.

.Confined to the Cape Province, usually along the

coastal belt from the south-western Cape to the

eastern Cape, but also inland. Fig. 7.

Vouchers

Dyer 5338; Mauve & Hugo 243; Mauve 5347; Leach & Bayliss

12647.

4. Myrsiphyllum kraussianum Kunth, Enum.

PI. 5: 107 (1850); (as ‘krausianum’). Type: Cape,

Wynberg, Constantia, Krauss 1333 (B, holo.; in

errore Krause).

Asparagus krausianus (Kunth) MacBride in Contr. Gray Herb.

Harv. 56: 17 (1918); Salter in FI. Cape Penins., 174 (1950). A.

krausii Bak. in F.C. 6: 272 (1896).

A. asparagoides sensu Jessop in Bothalia 9: 81 (1969).

? Myrsiphyllum gramineum Kunth, Enum. Plant. 5: 108 (1850).

Type: Cape of Good Hope, collector unknown. (B).

Scandent, 1-2 m tall, resembling M. asparagoides

but with straight, narrower phylloclades and shorter

pedicels. Rhizome bearing a congested cluster of

sessile fusiform tubers. Stems and branches laxly

curved, angled, ridged. Phylloclades narrowly

cordate-acuminate, c. 15-30 x 5 mm, flat, striate,

A. A. OBERMEYER

81

Fig. 3. — Phylloclades of Myrsiphyllum species. 1, M. multitub ero sum, x 1; 2, M. asparagoides, x 1;

3, M. ovatum, x 1; 4, M. kraussianum: a, typical form; b, elongated form, x 1; 5, M. volubile, x

1; 6, M. undulatum, X 1; 7, M. juniperoides: a, x 1; b, x 4; 8, M. alopecurum: a, X 1; b, x 4; 9,

M. declinatum showing declinate branching, x 1; 10, M. scandens, x 1; 11, M. ramosissimum x

1; 12, M. fasciculatum, x 1.

with a distinct midrib, apiculate. Flowers pendulous,

axillary; pedicels c. 3-5 mm long, articulated below

perianth. Tepals c. 5 mm, white with a green midrib.

Stamens with flattened filaments bearing two basal

spurs; anthers orange. Ovary with 6 ovules in each

locule. Berry 6—9 mm in diam., red. Figs 2.4, 3, 4a

& 4b.

Southern Cape; recorded from the Cape Peninsu-

la to the south-western and south-eastern Cape, in

strandveld and fynbos. Fig. 8.

Vouchers

Purcell 68; Taylor 10235; Barker 10227; Strey 521; Van Breda &

Joubert 1960; Acocks 17992.

A form of this species, bearing longer, narrower

phylloclades c. 35—45 mm long, has been collected

around the south-western Cape, viz Ceres, Worces-

ter, Laingsburg and Swellendam.

Vouchers

Bayer 3153, 3344; Bond 247; Mauve & Hugo 240.

82

REVISION OF THE GENUS MYRSIPHYLLUM WILLD.

Fig. 5b. — Tilli’s Catalogue Horti Pisani, Tab. 12, figs 1 & 2 of

Myrsiphyllum asparagoides.

CATALOGUS

PLANTARUM HORTI PISANI

AUCTORE

MICHAELE ANGELO TILLI

E CASTRO FLORENTINO

In Pisano Athcnxo Simplicium Lectorf. Ordinario

ET EJUSDEM HORTI CUSTODE

NEC NON

Regie Societ Londini-nsis , ac Acad. Botanical Florentine Socjo

SUB iAUSPICIIS *R. C.

C O S M I III.

M. E. D.

T L O R i, N T ' I ;V M. D C C % X 1 1 L

Typis Regis Cclfitudinis. Apud Tartinium & Franchium. Safer. Permijfu.

Fig. 5a. — Tilli’s Catalogue Horti Pisani, frontispiece.

5. Myrsiphyllum volubile (Thunb.) Oberm.,

comb. nov.

Asparagus volubilis Thunb., Prodr. 66 (1794), FI. Cap. edn 2r

332 (1821); Schult. f., Syst. 7: 347 (1829); Bak. in J. Linn. Soc.,

Bot. 14: 628 (1875), F. C. 6: 262 (1896). Type: Cape, without

locality, Thunberg 8469 (UPS, holo; PRE, photo).

IRuscus volubilis Thunb., Prodr. 13 (1794); Kunth, Enum. PI.

5; 276 (1859). Type as above.*

Asparagus asparagoides sensu Jessop in Bothalia 9: 81 (1966);

non Druce.

Scandent with wiry twisted stems and short leafy

branches, c. 1 — 1,5 m tall, often with a garlic scent.

Rhizome terete, bearing swollen bulbous tubers c.

300 — 500 mm long, closely attached to it. Stems

bearing short branches 40 — 80 mm long.

Phylloclades broadly linear, abruptly acute above

and below, c. 15 - 20 x 3 mm, without a distinct

midrib, flat, soft, pale green. Flowers on short

pedicels 4-5 mm long, articulation below perianth;

tepals fused below, spreading above, c. 5 mm long,

cream. Stamens with orange anthers, typical. Ovary

ovoid; ovules 6 in each locule; style short. Berry not

seen. Figs 2.5 & 3.5.

Recorded from the south-eastern to north-eastern

Cape, in fynbos, coastal scrub or in forest

undergrowth. Fig. 9.

Vouchers

Mauve & Hugo 124, 174, 241, 138; Archibald 4557; Smook

3819; Ecklon & Zeyher 4; Jessop 623; Geldenhuys 639, 515; Sim

4082; Thode Mill.

Fig. 4. — Distribution of Myrsiphyllum multituberosum.

* Dracaena volubilis L.f. , Suppl. 204 (1781). The nan^e is

mentioned on the microfiche edition, Nr 435/4 but no specimen is

shown. Baker in F.C. 6: 272 (1896) identified it as Asparagus

scandens Thunb.

A. A. OBERMEYER

83

Fig. 6. — Distribution of Myrsiphyllum asparagoides.

Fig. 7. — Distribution of Myrsiphyllum ovatum.

Fig. 8. — Distribution of Myrsiphyllum kraussianum.

Fig. 9. — Distribution of Myrsiphyllum volubile.

M. volubile can be distinguished from M.

kraussianum by its softer phylloclades, which are

rounded at the base and do not show a midrib.

‘Giving off a garlic odour which scents the air

around’ (Tyson). Milk becomes tainted when cows

feed on this plant.

6. Myrsiphyllum undulatum (L.f. ) Kunth,

Enum. PI. 5: 109 (1850). Type: Cape of Good Hope,

without precise locality, Thunberg (UPS, 8446;

PRE, photo!).

Dracaena undulata L.f., Suppl. 203 (1781); Roem. & Schult.,

Syst. Veg. 7: 346 (1829). Asparagus undulatus (L.f.) Thunb.,

Prodr. 66 (1794); Bak. in J. Linn. Soc., Bot. 14: 628 (1875); FI.

Cap. 6: 273 (1896); Marloth, FI. S. Afr. 4: t. 20 E; fig. 14a (1915);

Salter in FI. Cape Penins. 173 (1950); Solch, Beitr. FI.

Siidwest-Afr. 39 (1961); Jessop in Bothalia 9: 84 (1966).

Asparagus klinghardtianus Dinter in Feddes Rep. 29: 270

(1931). Type: South West Africa/Namibia: Klinghardtgebirge,

Dinter 3998 (B, holo.! PRE!).

Fig. 10. — Distribution of Myrsiphyllum undulatum.

84

REVISION OF THE GENUS MYRSIPHYLLUM WILLD.

Fig. 11. — Seed coats of Myrsiphyllum species (SEM photographs). 7, M. juniperoides, x 600; 8, M.

alopecurum, x 600; 9, M. declinatum, x 600; 10, M. scandens, X 600; 11, M. ramosissimum, X 1000;

12, M. fasciculatum, x 600.

Erect rigid bushes (chamaephytes) 0,3 -0,5 m tall,

with long, simple spreading branches bearing

regularly placed phylloclades. Rhizome compact,

scaly, bearing numerous roots, some of these with

large fusiform tubers continued below as normal

long thin roots; occasionally they may produce a new

shoot. Stem erect, ridged, with spreading long,

usually simple branches. Phylloclades ovate-

acuminate, up to 25 mm long (smaller on new

growth), leathery, striate, often folded above.

Flowers pendulous, 1-3 at a node; pedicels c. 5-9

mm long, with disk below perianth. Tepals broadly

linear, c. 6 mm long, purple or green with cream

margins, re flexed in upper half. Stamens with flat

attenuate filaments bearing 2 small basal spurs;

anthers orange. Ovary with 8—10 ovules in each

locule. Berry globose, c. 5 mm in diam., red. Figs

2.6, 3.6.

Recorded from the Cape Peninsula along the west

coast to South West Africa/Namibia, usually in the

strandveld with its sandy soil. Flowering in spring.

Fig. 10.

7. Myrsiphyllum juniperoides (Engl.) Oberm.,

comb. nov. Type: South West Africa/Namibia, Aus,

Marloth 1538 (B, holo.).

Asparagus juniperoides Engl, in Bot. Jb. 10: 3 (1889); Schinz in

Bull. Herb. Boissier. 1, 4, app. Ill: 44 (1896); Solch, Beitr. FI.

Siidwest-Afr. 37 (1961); Jessop in Bothalia 9: 63 (1966), pro

parte. Type: South West Africa/Namibia, Aus, Marloth 1538 (B,

holo.; PRE, photo!).

A. A. OBERMEYER

85

Fig. 12. — Myrsiphyllum juniperoides, showing a soft branch

which developed after the plant was placed in a conservatory

at the Botanical Research Institute. J. J. Lavranos 21242,

Rosh Pinah, SWA/Namibia.

Erect chamaephytes with stems up to 0,5 m tall,

covered by a ‘cylinder’ of short ‘leafy’ branches (viz.

phylloclades) c. 40 mm in diam. (‘fox tail’ habit).

Rhizome dorsiventral, creeping, bearing numerous

long roots which form fusiform swellings c. 30 mm

long towards root-tips, up to 120 mm away from

rhizome, densely covered by a velamen of root hairs

when young. Branches short, erect, c. 10—30 mm

long, overlapping, placed in axil of a short

membranous erect scale-leaf. Phylloclades solitary,

linear, c. 10 x 2 mm, apiculate, margin smooth,

bright shiny green. Flowers axillary, solitary, on

peduncles c. 1 mm long. Tepals narrowly oblong, c.

6 mm long, forming a short tube below, curved

outwards in upper half, outer ciliate. Stamens with

outer shorter than inner, filaments fused to tepals in

lower half, erect. Ovary obovoid with c. 10 ovules in

each locule; style about as long as ovary; stigmas

apical. Berry c. 5 mm in diam., several-seeded, red.

Figs 3.7, 11.7, 12.

A psammophyte recorded from southern South

West Africa/Namibia and northern Namaqualand,

with a few records further inland. Fig. 13.

Vouchers

Dinter 3723; Merxmiiller & Giess 3418; Muller 818; Giess

13004; Verdoorn 1835.

8. Myrsiphyllum alopecurum Oberm., sp. nov.,

M. juniper oidi (Engl.) Oberm. affinis sed rhizomati-

bus elongatis multituberosis foliis fimbriatis differt.

Type. — Western Cape 2917 (Springbok): 22 km

W by S of Springbok (— DB). Acocks 19268 (PRE,

holo.!).

Chamaephytum. Rhizoma elongata, tuberis mul-

tis tecta. Tubera allantoidea, ad 40 mm longa.

Caules erecti ad 0,5 m alti. Rami brevissimi

aggregati. Phyllocladia numerosa anguste linearia,

c. 10 mm longa dense ciliata adscendentia. Flores

solitarii, breviter pedicellati; tepala anguste oblon-

ga, ad basin tubum brevem formantes, lobi recurvati

externi ciliolati. Stamina typica. Ovarium ovoideum

4-ovulatum. Bacca plurisperma.

Chamaephyte with erect annual stems c. 0,5 m

tall, slender, attenuated above, covered by a

feathery ‘cylinder’ of much abbreviated densely

leafy branches, appearing ‘fox-tail’ shaped. Rhizome

creeping, long, densely covered by numerous

allantoid, shortly stalked root-tubers, placed in all

directions, up to 40 x 10 mm, continued below as

normal thin roots, densely covered by a velamen of

root hairs. Stems ridged, scale-leaves persisting.

Branches c. 30 mm long, ascending, overlapping.

Phylloclades narrowly linear, c. 12 x 0,5 mm,

glossy, yellow-green, margin densely ciliate, apex

mucronate. Flowers solitary on the short branches;

pedicels 1 mm long with the disk below perianth.

Tepals free, connivent, forming a tube, curved

outwards above, linear, obtuse, c. 7 mm long, white

with a green midrib, margin of outer tepals densely,

shortly ciliate, inner smooth. Stamens with fi-

laments fused to tepals in lower half, outer

somewhat shorter. Ovary oblong-ovoid with 4

ovules in each locule; style as long as ovary, stigmas

small, curved outwards. Berry several-seeded, c. 7

mm in diam., red? Figs 3.8 & 11.8.

Apparently endemic to the north-western Cape

from Namaqualand to Clanwilliam, in sandy places

in Namaqualand Broken Veld. Flowering in winter.

Fig. 13.

Vouchers

VanderWesthuizen272; Marloth 6719, 7794; Schlechter 8028.

Marloth 6719 from Garies-Okiep and Andreae

409 from Klawer show a sudden lengthening of the

upper side-branches giving the plants a plumose

apex. This has also been observed in other species

adopting the cylindric habit.

Fig. 13.— Distribution of ■ , Myrsiphyllum juniperiodes; ■, M.

alopecurum

86

REVISION OF THE GENUS MYRSIPHYLLUM WILLD.

Schlechter on some of his collections gave it the

name ‘ alopecurus’ , (as ‘ alepocurus ’ ), meaning fox

tail.

9. Myrsiphyllum declinatum (L.) Oberm.,

comb. nov.

Asparagus declinatus L., Sp. PI. 313 (1753). Thunb., Prodr.

Cap. 66 (1794), FI. Cap. edn 2: 333 (1823). Type: Herbarium van

Royen in Herb. Lugd. Bat. 913 62 567 (L, holo.!; PRE, xerox

copy!).

A. crispus Lam., Encyc. 1: 295 (1783); Roem. & Schult., Syst.

Veg. 7: 326 (1829); Kunth, Enum. PI. 5: 73 (1850); Bak. in J.

Linn. Soc., Bot. 14: 607 (1875); FI. Cap. 6: 259 (1896); Marloth,

FI. S. Afr., t. 20 (1915); Salter in FI. Cape Penins. 174 (1950);

Jessop in Bothalia 9: 80 (1966). Type: ‘L'lle de France’

(Mauritius), collector unknown (P, holo.; PRE, photo!).

A. flexuosus Thunb., Prodr. 66 (1794); Kunth, Enum. PI. 5: 74

(1850). Type: Cape of Good Hope, without precise locality,

Thunberg (UPS, 8450, holo.; PRE, photo!).

A. decumbens Jacq., Hort. Schoenbr. 1: 51, t. 97 (1797). Type:

Cape of Good Hope, without precise locality, collector unknown.

Asparagopsis decumbens (Jacq.) Kunth, Enum. PI. 5: 77 (1850).

Soft deciduous copiously branched scramblers up

to c. 1 m high. Rhizome creeping, covered by

numerous thick fusiform root-tubers up to 50 mm

long. Stems annual, angled, thin, green, declinate;

branches likewise; branchlets short, bearing over-

lapping ternate cladodes, or sometimes these more

laxly arranged. Cladodes linear-acuminate, c. 5—15

mm long, somewhat arcuate. Flowers solitary on

pedicels c. 5-11 mm long, articulated below

truncate base of flower; tepals broadly Unear,

reflexed in upper half, c. 5 mm long, white to pale

pink. Stamens with filaments bearing 2 basal spurs;

anthers orange or red. Ovary oblong-ovoid, tapered

into a persistent style, dark green; locules with 2—4

ovules. Berry oblong-ovoid, attenuated below,

apiculate, c. 12 mm long, whitish, semi-transparent,

3-9-seeded. Figs 3.9, 11.9 & 14.

Recorded from southern South West Africa/Na-

mibia to Namaqualand, the Cape Peninsula and east

as far as Riversdale; in fynbos or coastal scrub,

usually on rocky outcrops. Flowering May-Septem-

ber. Fig. 15.

Vouchers

Oliver 3430; Compton 14640; Hall 3640; Bos 168; Taylor 8621;

Merxmiiller & Giess 32218.

10. Myrsiphyllum scandens (Thunb.) Oberm.,

comb. nov.

Asparagus scandens Thunb., Prodr. 66 (1784); Roem. &

Schult., Syst. Veg. 7: 325 (1829); Saund. in Ref. bot., t. 21 (1869);

Bak. in J. Linn. Soc., Bot. 14: 622 (1875); FI. Cap. 6: 268 (1896);

Salter in FI. Cape Penins. 174 (1950); Jessop in Bothalia 9: 78

(1966). Syntypes: Cape of Good Hope without precise locality,

Thunberg s.n. (UPS, 8461, 8462; BOL! PRE! photos).

Asparagus pectinatus Delile in Red., Lil. , t. 407 (1813). Type:

TOrangerie du jardin la Malmaison’, t. 407 (iconotype).

Asparagopsis scandens (Thunb.) Kunth, Enum. PI. 5: 78

(1850).

Perennial herbaceous, scandent, up to 2 m tall,

probably evergreen. Rhizome small, covered by

fusiform roots 20—40 mm long, tapering towards

root-tips. Stems twining, wiry. Phylloclades ar-

ranged close together on one plane on both sides of

branchlets, the whole resembling the frond of a fern,

ternate on alternating sides, one cladode somewhat

longer than other 2, shallowly S-shaped, apiculate,

1-nerved, c. 8—17 x 1 mm. Scale-leaves minute,

forming fringed white membranous tufts at nodes.

Flowers pendulous, 1—3 at a node, on pedicels up

to c. 10 mm long, articulated below attenuated

pericladium. Tepals spreading, c. 4 mm long, white.

Stamens attenuated towards apex; anthers green or

yellow. Ovary globose with 3—4 ovules in each

locule; style erect short, with 3 spreading stigmas.

Fig. 15. — Distribution of Myrsiphyllum declinatum.

A. A. OBERMEYER

87

Fig. 16. — Distribution of Myrsiphyllum scandens.

Fig. 17.— Distribution of Myrsiphyllum ramosissimum.

Berry globose, c. 8 mm in diam., orange or red,

1-2 seeded. Figs 3.10 & 11.10.

A Cape endemic. Recorded from the west coast at

St Helena Bay to the Cape Peninsula and eastwards

as far as Tsitsikamma. Also inland as far as

Worcester; a shade-loving plant, frequent in wooded

areas, needing moisture all the year round,

apparently evergreen. Flowers August -October.

Fig. 16.

Vouchers

Fries, Norlindh & Weimarck 540; Mauve & Hugo 232;

Marloth 4517; Ecklon & Zeyher s.n.; Werdermann & Oberdieck

355.

M. scandens (Thunb.) Oberm. and the next

species, M. ramosissimum (Bak.) Oberm. have

perianths similar to those of Protasparagus but

agree in all other respects with Myrsiphyllum. They

are considered more primitive.

11. Myrsiphyllum ramosissimum (Bak.)

Oberm., comb. nov.

Asparagus ramosissimum Bak. in Gdnrs’ Chron. n.s. 2: 6

(1874); J. Linn. Soc., Bot. 14: 622 (1875); FI. Cap. 6: 268 (1896).

Type: Eastern Cape, Pirie Forest near King Williamstown,

Flanagan 2235 (PRE, neo!).

A. scandens Thunb. var. deflexus Bak. in J. Linn. Soc., Bot.

14: 622 (1875); FI. Cap. 6: 268 (1896). Type: Somerset East,

Boschberg, MacOwan 1771 (BOL; SAM).

Scandent, 1—2 m tall, evergreen perennial.

Rhizome creeping, woody, scaly, c. 20 mm in

diam., producing fusiform root-tubers terminating in

long thin roots. Stems green, angled, ridges

minutely papillate. Branches and branchlets

spreading. Phylloclades usually 3-nate, linear, flat

above, keeled below, 8—12 x 0,7 mm, apiculate,

margin minutely papillate. Flowers solitary, similar

to Protasparagus type; stalk filiform, c. 10 mm,

articulated in lower half, pendulous. Tepals

spreading, oblong-attenuate, c. 3-4 mm, outer with

a small callosity at base, white. Stamens with

linear-acuminate filaments, anthers yellow or red,

apiculate. Ovary ovoid, attenuated into a short

style, stigmas with short spreading ciliate lobes;

ovules c. 6 in each locule. Berry red or orange, 1—3

seeded. Figs 3.11 & 11.11.

Recorded from montane areas in the eastern

Transvaal, Natal and eastern Cape to the southern

Cape; in forests, kloofs and riverbank vegetation, in

moist shady places. Flowering in summer. Fig. 17.

Baker described this species from a living plant,

collected by Cooper in the eastern Cape and

cultivated by Wilson Saunders in his garden at

Reigate. It was evidently not preserved, [cf Jessop in

Bothalia 9: 80 (1966)].

Vouchers

Kerfoot, Gooyer & Eastman 286; Galpin 14531; Devenish 206;

Strey 6292; Compton 28084; Killick 1105; Galpin 10173;

Dieterlen 707; Dahlstrand 1851; Compton 30552.

12. Myrsiphyllum fasciculatum (Thunb.)

Oberm., comb. nov.

Asparagus fasciculatus Thunb., FI. Cap. edn 2: 329 (1823);

Jessop in Bothalia 9: 54 (1966). Type: Cape, without precise

locality, Masson s.n. (UPS, sub Thunberg 8447, holo.; BOL;

PRE!, photos).

Asparagopsis consanguinea Kunth, Enum. PI. 5: 76 (1850).

Asparagus consanguineus (Kunth) Bak. in J. Linn. Soc., Bot. 14:

615 (1875); FI. Cap. 6: 260 (1896). Type: Cape, without precise

locality, Drege s.n. (Kiel, holo., PRE, photo!; K, PRE, photo!).

Note: On the Kew sheet N. E. Brown gave Drege’s locality as

‘between Krom River and Pietersfontein on the Piquetberg,

under 1 000 ft, July’. Asparagopsis schlechtendalii Kunth, Enum.

Plant 5: 90 (1850). Type: Cape; at the Olifants River, Krebs s.n.

(March), B, holo?

Scandent or prostrate, non-spinous (rarely with

some small spines on stem below), much branched

with deciduous cladodes in feathery fascicles,

bearing solitary flowers. Rhizome woody with

numerous swollen roots, c. 300—600 mm long and c.

20 mm in diam., smaller in younger plants. Stems

spineless, or with some brittle spines below,

glabrous, reddish brown. Branches widely spaced,

laxly zigzagging, in turn producing numerous short

branchlets bearing the cladode-fascicles. Cladodes

unequally long, 15— 30-nate, c. 5 — 10 mm long,

smooth, apiculate, somewhat curved, greyish green,

deciduous. Flowers apical, solitary or rarely

2— 3-nate, beside a cladode-fascicle; pedicel c. 5 mm

long with disk below perianth. Tepals c. 5 mm,

broadly linear, white with a brown central stripe,

88

REVISION OF THE GENUS MYRSIPHYLLUM WILLD.

Fig. 20. — Distribution of Myrsiphyllum fasciculatum.

forming a tube, recurved above. Stamens erect,

filaments flat, attenuated above, abruptly narrowed

below, forming basal spurs; anthers orange or

yellow. Ovary oblong-ovoid, shortly stipitate, with

6—8 ovules in each locule, style long, stigmas short,

spreading. Berry several-seeded, c. 10 mm in diam.,

pinkish red. Figs 1.4, 3.12, 11.12 & 18.

Common in the western Strandveld up to

Namaqualand and the southern Karoo as far as

Laingsburg. Not recorded from the Peninsula. The

young cladodes appear in March— April; flowers in

May— June; fruits ripen in September. Fig. 20.

Vouchers

Marloth 9049; Ecklon & Zeyher 1675; Esterhuysen 1805;

Thompson & Le Roux 112; Schlechter 7890; Bolus 13882;

Werger 414; Leistner 3390.

The fascicles containing numerous cladodes are

unusual for this genus, but the pendulous flower,

including the true pedicel, are typical of Myrsiphyl-

lum. In old plants the root-system develops into a

huge mass of long cylindrical roots, each 0,3— 0,6 m

long and c. 30 mm thick. Marloth (his No. 9049)

counted up to 120 ‘finger thick’ roots. Bayer 3287

shows a similar mass (photo). Fig. 19.

ACKNOWLEDGEMENTS

We thank the director and staff of the Rijksher-

barium, Leiden for the xerox copy of the type of

Asparagus declinatus L.; Dr H. P. Linder at Kew

for assisting with types; Mrs S. Perold for the SEM

photographs of the seedcoats and Mr M. B. Bayer of

the Karoo Botanic Garden, Worcester, for most

generous co-operation.

UITTREKSEL

Die genus Myrsiphyllum Willd. (Liliaceae — As-

parageae) word hersien. Twaalf spesies word erken,

waarvan een ’n nuwe spesie is, naamlik, M.

alopecurum Oberm. Agt nuwe kombinasies word

gemaak. Daar is ook 'n sleutel om Myrsiphyllum

Willd. van Protasparagus Oberm. te onderskei.

Fig. 18. — Myrsiphyllum fasciculatum. Marloth 9049.

Fig. 19. — Myrsiphyllum fasciculatum showing large rootsystem.

Bayer 3287. Photo M. B. Bayer.

Bothalia 15, 1 & 2: 89 - 100 (1984)

New taxa of Aneilema R. Br. (Commelinaceae) from southern and

tropical East Africa

R. B. FADEN*

Keywords: Aneilema, Commelinaceae, new taxa, southern Africa, tropical East Africa

ABSTRACT

Four new species of Aneilema are described: A. indehiscens Faden, with subsp. indehiscens (Kenya, Tanzania)

and subsp. lilacinum Faden (Zimbabwe, Mozambique, South Africa); A. arenicola Faden (Mozambique, South

Africa); A. brunneospermum Faden (Mozambique, Swaziland, South Africa); and A. tanaense Faden (Kenya). A

new subspecies, Aneilema dregeanum Kunth subsp. mossambicense Faden (Mozambique), is also described, and

A. johnstonii K. Schum. is lectotypified.

INTRODUCTION

A revision of the genus Aneilema for the Flora of

Southern Africa has included a detailed study of a

group of taxa which were only superficially

examined during a previous investigation (Faden,

1975). The completed research has revealed the

presence in southern Africa and adjacent floral

regions of three undescribed species and a new

subspecies, which are described below: A. arenicola

Faden, A. brunneospermum Faden, A. indehiscens

Faden (with two subspecies), and A. dregeanum

Kunth subsp. mossambicense Faden. Aneilema

tanaense Faden, from tropical East Africa, is also

described because of the need to validate the name

for a forthcoming publication. The problem of

typifying A. johnstonii K. Schum. is discussed, and a

lectotype is selected.

ANEILEMA DREGEANUM KUNTH AND A. SCHLECH-

TERI K. SCHUM.

Two related southern African species, Aneilema

dregeanum Kunth and A. schlechteri K. Schum.,

were not clearly differentiated until Brenan (1961)

studied them. They have continued to be confused

with an unrelated but sympatric species that is

described below as A. indehiscens Faden. In the

discussion under A. schlechteri, Brenan (1961) noted

that the Kew isotype ( Schlechter 11748) differed

from other specimens by having the cells of the outer

capsule wall epidermis more or less isodiametric as

opposed to longitudinally elongate. He considered

that difference as probably due to the immature

state of the capsules on the type, and he could find

no reason to separate this collection taxonomically

from the others. When I examined the specimens at

Kew in 1974, I came to the same conclusion.

As part of my investigations on Aneilema for the

Flora of Southern Africa I have examined a much

greater number of specimens than have previous

workers. I was also able to do field work in South

Africa in 1974. These studies have shown that both

A. dregeanum and A. schlechteri are taxonomically

more complex than Brenan (1961) or I had realized.

Aneilema dregeanum has been found to be separable

into two subspecies: A. dregeanum subsp. dregea-

num and A. dregeanum subsp. mossambicense which

is described below. Studies of the holotype and nine

additional isotypes of A. schlechteri, as well as two

collections from Zimbabwe which agree with

them, have revealed that (1) the shape of the cells of

the capsule wall on Schlechter 11748 is not a function

of the developmental stage of the capsules, and (2)

these specimens exhibit further characters which do

not fall within the range of the other collections that

are usually treated as A. schlechteri. Aneilema

schlechteri has proven to be a rare species which is

known in southern Africa from only two collections

(including the type). The much more common

species lacks a specific name and is described below

as A. brunneospermum Faden.

My field work in northern Natal yielded two

collections of a plant that is clearly related to A.

schlechteri and A. brunneospermum but differs from

them in several significant characters. Several

additional collections from the same region and from

southern Mozambique were subsequently found

among the specimens on loan from various

institutions. This plant represents a distinct species

which is described below as A. arenicola Faden.

The two subspecies of A. dregeanum and three

species in the A. schlechteri group may be separated

by the following key:

* Department of Botany, National Museum of Natural History,

Smithsonian Institution, Washington DC 20560, USA.

90

NEW TAX A OF ANE1LEMA R. BR. (COMMELINACEAE) FROM SOUTHERN AND TROPICAL EAST AFRICA

Perennials; capsules oblong-elliptic to obovate-oblong or oblong, (3,5-)5-7,5 mm long, locules 2-seeded

(sometimes 1-seeded by abortion):

Maximum leaf width 16-45 mm; cincinnus peduncles l-9,5(— 13) mm long; bracteoles spaced

0,5— 2(— 3) mm apart; stamen filaments inconspicuously bearded with white hairs less than 0,3 mm

long; seeds with testa foveate-scrobiculate, with hilum in a groove less than 1/3 the width of the

seed A. dregeanum subsp. dregeanum

Maximum leaf width 8—22 mm; cincinnus peduncles (3 — )4— 15 mm long; bracteoles spaced (0,6-)l-5,6

mm apart; stamen filaments conspicuously bearded with yellow hairs more than 0,5 mm long;

seeds with testa rugose to scrobiculate, with hilum in a groove 1/3— 1/2 the width of the seed

A. dregeanum subsp. mossambicense

Annuals; capsules broadly elliptic to obovate or obovate-orbicular, 2,8-4,5(-5) mm long, locules

1-seeded:

Pedicels 1,5—3 mm long, erect or only slightly further recurved in fruit; capsules transversely wrinkled at

maturity, valves relatively planar, cells of the outer wall transversely elongate; seeds buff or light

brownish orange A. arenicola

Pedicels 2-8 mm long, strongly recurved (120°-270°) in fruit; capsules not transversely wrinkled at

maturity, valves strongly convexo-concave, cells of the outer wall longitudinally elongate or

isodiametric; seeds dark brown or pale pinkish grey:

Inflorescences composed of (6— )10 - 20(-29) cincinni; pedicels puberulous for up to half their length;

lateral stamen filaments (4,8-)5,5-9,5 mm long; cells of the outer capsule wall longitudinally

elongate; seeds dark brown A. brunneospermum

Inflorescenses composed of 3-13 cincinni; pedicels puberulous for more than half their length; lateral

stamen filaments c. 3,2-4, 5 mm long; cells of the outer capsule wall ± isodiametric to slightly

longitudinally elongate; seeds pale pinkish grey A. schlechteri

AneUema dregeanum Kunth, Enum. PI. 4: 73

(1843). Type: South Africa, Pondoland: between

Umtata River and St Johns River, 1839, Drege 4471

(B,holo.!;B!,BM,FHO!,G!,K!,MO!,P!,S!,iso.).

(a) Aneilema dregeanum Kunth subsp. dregea-

num

Subsp. dregeanum is endemic to South Africa,

occurring in Natal and eastern Cape Provinces. It

grows in moist or mesic situations, most commonly

in forest, but also in bush, along streams and (rarely)

in grassland, from about sea level to 940 m altitude,

usually in partial or dense shade. Flowering

specimens have been seen in all months except

October, June and July, although the main flowering

period is December to April. The flowers open in

the morning and fade by 13h30.

In addition to the type collection, the following

specimens have been seen:

Cape. — 3129 (Port St Johns): Insinuka near St Johns River

mouth ( — DA), Bolus (in Bolus Herb.) 10348 (BOL; MO);

Insinuka near Port St Johns (-DA), Flanagan 2504 (BOL; PRE)

& Flanagan 2607 (PRE; Z); Port St Johns, west banks (—DA),

Galpin 3194 (BOL; PRE). Grid Reference Uncertain: 3 mi.

inland of Umgazi River mouth, Wells 3493 (K; PRE).

Natal. — 2732 (Ubombo): Lake Sibayi (-BC), Balsinhas 3191

(MO; PRE); Ubombo-Sordwana Bay road, 1 km before entrance

to Sordwana Bay National Park and crossing of Ngobeseleni

(Sordwana) River, c. 27°31'S, 32°40'E (-DA), Faden & Faden

74/214 (BR; EA; K; MO; NH; NU; PRE; US; WAG). 2831

(Nkandla): Empangeni road, 8 mi. out from Eshowe, Lawn 184a

(NH); Ntambanana Region, Mtunaini, Venter 807 (NH). 2832

(Mtubatuba): Bahene Forest, False Bay (-AB), Oatley BA2

(PRE); Lower Umfolozi, S of Lake Umsingazi, near Richards

Bay (-CC), Oatley 13 (K; NH); Dukuduku Forest (-AD), Strey

6990 (NH). 2930 (Pietermaritzburg): Umgeni Valley near Nagle

Dam (-DA), Cheadle 622 (PRE); Inanda Falls (-DA), Dodds

132 (NU); Table Mtn, Pietermaritzburg (-DA), Killick 246

(NU); Nagle Dam (-DA), Wells 1051 (NU; PRE); Inanda,

Umzinyati ( = Inanda?) Falls (-DB), Wood 1220 (BOL; NH).

2931 (Stanger): Mapumulo District, Oqaqeni (-AA), Edwards

1838 (NU); Umhlanga Bush (-CA), Huntley 63 (MO; NH; NU;

PRE) & Huntley 88 (MO; PRE); Hawaan Forest, S bank of

Umhlanga River (-CA), Ross & Moll 2302 (NH). 3030 (Port

Shepstone): near Mehlomnyama (-CB), Acocks 13312 (K;

PRE); Elliott’s Farm, Paddock, side ravine to Evongo (-CC),

Strey 7161 (K; NH; NU); Warner Beach (-BB), Strey 9484 (EA;

K; NH; NU; PRE), Strey 9383 (=9484?) (SRGH) & Strey 9928

(NH); Gibraltar (-CB), Strey 9579 (K; NH; NU; S). Grid

References Uncertain: Palmiet, Evans 540A (NH); Natal,

without precise locality, Gerrard s.n., July 1865 (K) & Mrs

Saunders s.n., June 1881 (K); Zululand, without precise locality,

Mrs McKenzie s.n., June 1882 (K) & Wylie in Wood 8772 (NH);

Umhlovi Beach, Graham 13 (NU); Pietermaritzburg District,

Oribi Aerodrome, Moll 2367 (NH); Umgeni, Rehmann 8571 (Z);

near Umgeni, Schlechter 3072 (Z); Near Umkoruangi (Umkoruo-

ruzi?) River, Schlechter 6711 (BOL; K; P; Z).

(b) Aneilema dregeanum Kunth subsp. mossam-

bicense Faden, subsp. nov.

Ab subspecie dregeano laminis angustioribus et

plerumque magis breviter petiolatis, inflorescentiis

pedunculis cincinnorum longioribus [(3— )4— 15

mm], bracteolis distantioribus [(0,6— )1— 5,6 mm],

floribus filamentis staminum lateralium pilis longio-

ribus et luteis barbatis, seminibus testa rugosa ad

scrobiculatam, hilo in sulco latior dispositis differt.

Type. — Mozambique, Nampula (Mozambique)

District: Na estrada de Nampula para Corrane, a c.

de 7 km de Nampula, 13 April 1961, Balsinhas &

Marrime in Balsinhas 383 (PRE!, holo.; COI!, K!,

LISC!, LMA!, SRGH!, iso.).

Subsp. mossambicense is known only from the

northern half of Mozambique (c. 14°S— 18°S

latitudes). It is recorded from open deciduous forest

and woodland from about sea level to 30 m altitude.

Flowering specimens have been seen in October,

November, January and March to May.

Mozambique. — Ilha: De Memba a Nacala (Ke), Torre 1497

(COI; LISC). Nampula: Na estrada de Nampula para Corrane, a

c. de 7 km de Nampula, Balsinhas & Marrime in Balsinhas 383

(COI; K; LISC; LMA; PRE; SRGH); Arredores de Nampula,

proximo do acampamento do CICA, Barbosa & Balsinhas 5259

(LMA); 37 mi. W of Nampula, Leach & Rutherford- Smith 10980

(K; LISC; PRE; SRGH); Antonio Enes, Matangula (Makan-

gula?) Praia area. Moss 32414 (LISC; SRGH); Nampula, Pedro &

Pedrogao 3204 (EA; LMA). Zambezia: Entre Naburi e o rio

Ligonha, a 11,8 km de Naburi, Barbosa & Carvalho 4327 (K;

LMA); Moebede Road, Lugela, Faulkner 194 (BR; K) &

Faulkner 194(2) (BR; COI; EA; K; S; SRGH); Quelimane, Sim

20618A (PRE) & Sim 20837 (PRE); Entre Mocuba e lie

(Errego), Torre 5016 (LISC).

R. B. FADEN

91

Plants of subsp. mossambicense are different in

aspect from those of subsp. dregeanum because of

their narrower, commonly more coriaceous, and

usually more shortly petiolate leaves, as well as their

laxer-appearing inflorescences due to the longer

cincinnus peduncles and more widely spaced

bracteoles. The differences in stamen filament hair

length and colour are quite striking, even in the

dried specimens, in which they seem to have been

preserved with unexpected frequency. It is probable

that further floral characters will be found when

living material of subsp. mossambicense can be

obtained.

The cells of the outer capsule wall in both

subspecies are transversely elongate. In subsp.

mossambicense these cells are arranged in regular

files; in subsp. dregeanum they are less well-ordered.

The seeds of the two taxa are quite distinct.

Although there is almost complete overlap in seed

size, the seeds of subsp. dregeanum (1,95—3 mm

long) tend to be longer than those of subsp.

mossambicense (1,9— 2,3 mm long). They are also

consistently darker in colour, and more deeply and

finely pitted, with the hilum in a much narrower

groove (less than J the width of the seed vs | |

the width of the seed) than the seeds of subsp.

mossambicense. Furthermore, the seeds of subsp.

dregeanum typically have the farinaceous material

confined to the testa depressions, with the edges of

the separate patches granular, whereas the seeds of

subsp. mossambicense are commonly completely

covered by matted farinaceous material which is not

at all granular.

In view of the suite of differences between these

two taxa, and in particular the consistent and

significant dissimilarities in seed and stamen filament

hair characters, treating these taxa as distinct species

was considered. Because our knowledge of both

taxa, especially the Mozambique plant, is incomple-

te, it was deemed best to describe them as subspecies

at this time.

Brenan (1961) recorded the capsule locules of A.

dregeanum as 2— 3-seeded. I have been unable to

find more than two seeds per locule in any of the

specimens of either subspecies that I have examined.

Aneilema schlechteri K. Schum., A. brunneosper-

mum Faden and A. arenicola Faden comprise a

group of closely related species. They may be

distinguished by the above key. Further contrasting

characters are given in Table 1.

Aneilema schlechteri K. Schum. in Bot. Jb. 33:

376 (1903). Type: South Africa, Transvaal: Komati

Poort, 15 December 1897, Schlechter 11748 (B!,

holo.; BM!, BOL!, BR!, COI!, G!, K!, NSW!,

PRE!, S!, Z!, iso.).

This species is confined to south-eastern Zimbab-

we and eastern Transvaal. It is recorded from

mopane woodland, Acacia nigrescens Community,

and as a weed in irrigated sugarcane. It is noted as

growing in basaltic soil or heavy, black turf.

Flowering specimens have been seen in December

and January, and a fruiting specimen in May.

In addition to the type collection, the following

specimens have been seen:

Zimbabwe. — Chiredzi: Hippo Valley Estates, Section 19,

Taylor 234 (MO). Nuanetsi: between Chikombedzi and Chipinda

Pools, Drummond 7845 (BR; K: SRGH).

Transvaal — 2431: Kruger National Park, S of Banguterritory

(-BB), Nel 5570 (PRE, only photocopy seen, capsule characters

checked by A. A. Mauve).

Aneilema brunneospermum Faden, sp. nov.

Aneilema dregeanum Kunth var. galpinii C. B. Clarke in

Thiselton-Dyer, F. C. 7; 13 (1897); Brenan in Kew Bull. 15: 216

(1961), pro syn. Type: Transvaal, Barberton (details of specific

localities differ on all three specimens), 16 December 1890,

Galpin 1187 (K!, holo.; NH!, PRE!).

Aneilema schlechteri sensu Brenan in Kew Bull. 15: 216 (1961).

Aneilema dregeanum sensu Compton, FI. Swaziland, 83 (1976),

non Kunth (1843).

Herbae annuae caespitosae. Folia laminis plerum-

que longe vel breviter petiolatis, lanceolatis ad

lanceolato-ellipticas, ovato-ellipticas vel ovatas,

2,5 — 10( — 14) cm longis (cum petiolo),

(0,6 — )1— 3,5(— 6) cm latis. Inflorescentiae thyrsi pro

parte maxima terminates, ovoidei ad ellipsoideos

parce densi ad parce laxos, (2— )2,5— 6(— 7,5) cm

longi, (1,5 — )2— 4,5(— 6) cm lati, cincinnis

(6— )10— 20(-29) pro parte maxima altemis ascen-

dentibus compositi, bracteolis (0,8— )1— 3(— 3,5) mm

distantibus. Flores perfecti et staminati, (9—) 11 — 15

mm lati, pedicelis (2,2— )2,5— 7(— 8) mm longis,

tempore fructigero recurvatis 120°— 180°(— 270°),

supra medium puberulis. Sepala (2— )2,5— 3,6(— 4,3)

mm longa, puberula, glandibus subapicalibus bilo-

batis. Petala postica lavandula vel pallide lilacina,

5,3— 7,5(— 9) mm tonga, 3—7 mm lata, petalum

anticum 2,5— 3,5 mm longum. Stamina lateralia

filamentis dense barbatis, 4,8— 9,5 mm longis. Stylus

4,5—6 mm longus. Capsulae biloculares, obovatae,

late obovatae vel obovato-orbiculares,

(2,8— )3— 4,5(— 5) mm longae, (3,3— )3,8— 4,5(— 4,8)

mm latae, loculis monospermatis, cellulis superficiei

capsulae longistrorum elongatis. Semina plerumque

elliptica, 2— 2,5(-2,8) mm longa, 1,7-2,05 mm lata,

1,35 — 1,85 crassa, testa brunnea, foveolato-

reticulata ad foveolato-scrobiculatam vel foleolato-

rugosam.

TYPE. — South Africa, Natal, Zululand: 2732

(Ubombo): foothills of Lebombo Mountains, c. 5

km on road to Ndumu from the junction of the

Ingwavuma-Ndumu and Ingwavuma-Ubombo

roads, c. 27°07'S, 32°04'E, 18 February 1974

(— AA), Faden, Faden & Pooley 74/209 (US, holo.;

B, BR, K, MO, P, PRE, WAG, iso.).

Tufted annual to c. 60 cm tall (habit type IC of

Faden, 1975). Roots thin, fibrous, produced only at

the base and lower nodes. Shoots ascending. Leaves

spirally arranged, sheaths (0,3— )0,5 — 1,5(— 1,8) cm

long, puberulous, ciliate at the apex; blades longly to

shortly petiolate (rarely some sessile), lanceolate to

lanceolate-elliptic, ovate-elliptic or ovate,

2,5 — 10(— 14) cm long (including the petiole),

(0,6-)l-3,5(— 6) cm wide, apex acute to acumi-

nate, base cuneate (to rounded), adaxial surface

92

NEW TAX A OF ANEILEMA R. BR. (COMMELINACEAE) FROM SOUTHERN AND TROPICAL EAST AFRICA

usually scabrid (rarely not), puberulous, abaxial

surface scabrid or not, puberulous. Inflorescences

thyrses, terminal and sometimes axillary from the

upper leaves, moderately dense to moderately lax,

ovoid to ellipsoid, (2— )2,5— 6(— 7,5) cm long,

(1,5— )2— 4,5(— 6) cm wide, with (6-)10-20(-29)

alternate (or a few subopposite), ascending cincinni.

Penduncles (1,5— )2,5— 6( — 10) cm long, puberulous

with hook-hairs either uniform or of two sizes.

Cincinni to 3,5 cm long and 13-flowered. Cincinnus

bracts ovate to lanceolate, c. 1— 3(— 9) mm long,

usually glabrous (rarely sparsely puberulous).

Cincinnus peduncles ± uniform within the inflore-

scence or, more commonly, those of the middle

cincinni the longest and those of the lowermost

cincinni the shortest, 3 — 11( — 14) mm long, green to

purple, puberulous with hook-hairs of two sizes,

Bracteoles spaced (0,8— )1— 3(— 3,5) mm apart,

asymmetrically cup-shaped, usually perfoliate,

(1— )1,3— 2(— 2,3) mm long, (usually prominently)

glandular subapically (also margin occasionally

glandular-thickened), glabrous or sparsely puberu-

lous basally. Flowers perfect and staminate, very

faintly scented, (9 — )11 — 15 mm wide. Pedicels

TABLE 1. — Comparison of three related species of Aneilema from southern Africa

A. brunneospermum A. schlechteri A. arenicola

R. B. FADEN

93

(2,2-)2,5-5,5(-6,5) mm long in flower, 3— 7(— 8)

mm long in fruit, ascending in flower, recurved

either uniformly or just near the apex

120°-180°(— 270°) in fruit, puberulous above the

middle. Sepals convexo-concave, ovate,

(2 — )2,5 — 3,6( — 4,3) mm long, green with violet

margins, hooded apically, with subapical glands

bilobed, sparsely puberulous (rarely glabrous) with

uniform length hook-hairs. Paired petals

5,3-7,5(— 9) mm long, 3-6 mm wide, limb ovate,

4-7 mm long, pale lilac to lavender (RHS colours:

87C-D, 87D), apex rounded to truncate, claw 1-2

mm long, white, glabrous. Medial petal elliptic to

obovate, hooded apically, 2,5— 3,9 mm long,

1,8 -2, 5 mm wide, white with a medial or subapical

reddish purple spot. Filaments free. Medial stamino-

de entirely yellow, c. 1—1,5 mm long, antherode

bilobed, lobes sessile or subsessile, obovate to

elliptic, connective elongate. Lateral staminodes

with filaments 2-3 mm long, entirely yellow,

antherodes bilobed, lobes stalked, obovate, 0, 3-0,5

mm long, 0,3—0,75 mm wide. Lateral stamens with

filaments dorsiventrally flattened, parallel or slightly

divergent below the middle, strongly divergent

above, 4, 8-9, 5 mm long, greenish yellow, some-

times shading to white apically, densely bearded

for 1-2 mm above the middle with patent, white,

uniseriate hairs c. 0,25—0,4 mm long, anthers facing

the floral midplane and somewhat forward, ovate to

ovate-elliptic, 0,8-1, 2 mm long and c. 1 mm wide,

anther sacs blue-black or violet-black, pollen dirty

yellow. Medial stamen with filament 4—4,5 mm long,

arcuate-decurved, greenish yellow, anther shield-

shaped, ovate or broadly ovate, c. 0,9—1 mm long,

0,9 -1,4 mm wide, connective broad, yellow, anther

sacs orange-yellow, pollen orange-yellow to pale

yellow (changing to white?), discolorous with lateral

anther pollen. Ovary sessile, broadly elliptic,

0,75 — 1,1 mm long, 0,85-1,3 mm wide, green,

densely covered with forward-pointing, colourless,

glandular-capitate hairs except middorsally (where

glabrous) and midventrally (where sometimes

glabrous), apex rounded, dorsal locule suppressed,

ventral locules each 1-ovulate; style (3-) 4,5-9 mm

long, arcuate-decurved, then slightly recurved near

apex, tapering apically, greenish yellow in basal half

to three-quarters, contrastingly violet above, stigma

small or slightly capitate, white. Capsules sessile to

substipitate, broadly elliptic to obovate, broadly

obovate or obovate-orbicular, dehiscent, bivalved,

bilocular, (2,8 — )3 — 4,5( — 5) mm long,

(3,3— )3,8— 4,5(— 4,8) mm wide, tan or grey-tan to

brown (becoming stramineous with age), lustrous,

sparsely puberulous, apex truncate to emarginate

(rarely rounded), base cuneate to rounded, valves

persistent, not wrinkled, strongly convexo-concave,

dorsal locule suppressed, ventral locules 1-seeded,

cells of the outer capsule wall longitudinally

elongate. Seeds elliptic (to ovate-elliptic or

reniform-elliptic), rounded to truncate at both ends,

2— 2,5(— 2,8) mm long, 1,7—2,05 mm wide,

1,35 — 1,85 mm thick, hilum dark brown, strongly

raised but ± not in a groove, longly extended onto

apical and basal surfaces, embryotega dark brown

(concolorous with testa), testa dark brown (rarely

warm brown), foveolate-reticulate to foveolate-

scrobiculate or foveolate-rugose, sparsely white (and

also sometimes dark brown) farinose in the

depressions, the farinose material not easily detach-

ed. Fig. 1.

Fig. 1. — Aneilema brunneosper-

mum Faden. 1, flower, front

view; 2, flower, side view; 3,

capsule; 4, cells of the capsu-

le wall; 5, seed, ventral view;

6, seed, dorsal view. Bar = 1

mm except in Fig. 1.4 where

bar = 0,1 mm. (From Faden,

Faden & Pooley 74/209.)

94

NEW TAX A OF ANEILEMA R. BR. (COMMELINACEAE) FROM SOUTHERN AND TROPICAL EAST AFRICA

Aneilema brunneospermum ranges from southern

Mozambique to Swaziland and north-eastern South

Africa (Natal and Transvaal). It occurs in scrub or

forest (rarely grassland or poolsides), often in rocky

places, and usually in partial shade, from 150 — 950

m altitude. Flowering specimens have been seen

from October to February, April and May. The

flowers open about sunrise and fade in the late

morning.

Mozambique. — Lourengo Marques: Namaacha, Goba,

proximadades do fonte dos Libombos, Balsinhas 1250 (LMA);

same locality, Barbosa & Lemos in Barbosa 8261 (COI; K; LISC;

LMA; SRGH); Goba, Exell, Mendonga & Wild 556 (K; LISC;

SRGH); Namaacha, Gomes & Sousa 420 (K; LMA); Namaacha,

Goba, prox. da caseata, Mendonga 1672 (LISC); Namaacha,

Goba, prox. do ponte, Torre 2057 (LISC).

Swaziland. — District uncertain: 2632 (Bela Vista): Farm

Mlawula, south of Umbuluzi Poort, Rhino Pool (-AA),

Culverwell 0099 (PRE). Hhohho: Mdimba, Kemp 685 (MO).

Piggs Peak: Komati Bridge, Compton 26824 (K; NH; PRE); same

locality, Compton 30048 (K; PRE).

Natal. — 2732 (Ubombo): Lebombo Mountains, Pongolapoort

lookout between Jozini Dam and Mkuke, c. 27°30'S, 32°00'E,

Faden & Faden 741210 (K; MO; PRE); foothills of Lebombo

Mountains, c. 5 km on road to Ndumu from junction of

Ingwavuma-Ndumu and Ingwavuma-Ubombo roads, c. 27°07'S,

32°04'E ( — AA), Faden , Faden & Pooley 741209 (B; BR; K; MO;

P; PRE; US; WAG); Lebombo Mountains N of Josini (-AC),

Strey 8140 (K; NH; NU; PRE). Without Grid References:

Hluhluwe Game Reserve, Hitchins 55 (PRE); Lebombo

Mountains, Strey 4674 (K, NH); Mtunzini, Ngoya Mountain,

Venter 1419 (NH); Babanango District near Old Gold Mine in

valley. Venter 2935 (PRE); Hlabisa District, Hluhluwe Game

Reserve, Ward 1928 (NH; NU); Mkuzi Game Reserve, west

facing krantz above Mkuzi River, Ward 3982 (NU; PRE); Inanda,

Wood 1220 (K); same locality. Wood s.n. (K); Near Nanoti River,

Wood s.n. (MO).

Transvaal. — Without Grid References: Barberton, Galpin

1187: valley near Edwin Bray Battery (K); Kaap River Valley

(NH); Queens River Valley (PRE). Barberton, Thorncroft 9620

(PRE); same locality, Thorncroft 11310 (K); Krokodilbrug,

Kruger-wildtuin, (Komatirivierpoort on PRE sheet), Van der

Schijff 3992 (PRE; PUC); Malelane District, Tlalaberge, Van der

Schijff 4205 (PRE).

Aneilema arenicola Faden, sp. nov.

Herbae annuae ramosissimae foliis distichis,

plerumque sessilibus vel breviter petiolatis laminis

lanceolato-ellipticis ad ovato-ellipticas vel ovatas,

1.5— 4,5(— 7) cm longis, 0,7— 2( — 2,7) cm latis.

Inflorescentiae thyrsi terminales ovoidei parce densi

ad parce laxos, 2—5 cm longi et lati, cincinnis

(2— )5 — 13 pro parte maxima alternis ascendentibus

compositi, bracteolis 2-5,2 mm distantibus. Flores

perfecti, 6,5— 8, 5(— 9,5) mm lati, pedicelis 1,5—3

mm longis, tempore fructigero plus minusve erectis.

Sepala puberula (1,5— )2— 3,5(— 3,8) mm longa,

valde cucullata, glandibus subapicalibus bilobatis.

Petala postica pallide lilacina, 3,5-5 mm longa,

2,7— 3,5 mm lata, petalum anticum 2— 2,5(-2,8) mm

longum. Stamina lateralia filamentis dense barbatis,

2.5- 3 mm longis. Stylus 1,5— 2,3 mm longus.

Capsulae biloculares, late ellipticae ad elliptico-

orbiculares vel obovato-orbiculares, 2,8—4 mm

longae, 2,6-3,75 mm latae, loculis monospermatis,

cellulis superficiei capsulae transverse anguste

elongatis. Semina elliptica, 2,2— 3,1 mm longa,

1,5 — 1,9 mm lata, 0,95-1,3 mm crassa, testa

foveolato-reticulata.

Type. — South Africa, Natal, Zululand: 2732

(Ubombo): Ubombo-Sordwana Bay road, 2,6 km

towards Sordwana Bay from Shongwe, 27°26'S,

32°24'E, 19 February 1974 (-AD), Faden & Faden

741211 (US, holo.; B, BR, EA, K, MO, NH, NU, P,

PRE, UPS, WAG, iso.).

Densely branched annual to 30 cm tall (habit type

IB of Faden, 1975). Roots thin, fibrous. Shoots

ascending to decumbent, densely branched, rooting

at the lower nodes. Leaves distichous (except on the

primary shoot on which spirally arranged), sheaths

0,5 — 1 cm long, green, puberulous, sparsely ciliate at

the apex, blades shortly (rarely longly) petiolate

(lower leaves) to sessile (upper leaves), lanceolate-

elliptic to ovate-elliptic or ovate, l,5-4,5(— 7) cm

long (including the petiole), 0,7— 2(-2,7) cm wide,

apex acute, base broadly cuneate, adaxial surface

dull, pale green, slightly scabrid, puberulous-

hirsute, abaxial surface lustrous, puberulous. Inflo-

rescences thyrses, terminal on the main and,

ultimately, all lateral shoots, also sometimes axillary

from the upper leaves, moderately lax to moderately

dense, ovoid, 2-5 cm long and wide, with (2— )5 — 13

alternate (or a few subopposite), ascending cincinni.

Peduncles 1,5—5 cm long, puberulous with hook-

hairs of two sizes. Cincinni to 4,5 cm long and

9-flowered. Cincinnus bracts ovate to ovate-elliptic

(or lanceolate), 1— 3(— 4,2) mm long, glabrous to

sparsely puberulous. Cincinnus peduncles of increa-

sing length from lower to upper cincinni, exceeding

the cincinnus bracts (except occasionally the

lowermost), 1,5—10 mm long, green, puberulous

with hook-hairs of two sizes. Bracteoles spaced

2-5,2 mm apart, asymmetrically cup-shaped, usual-

ly perfoliate, 1,2— 1,5(— 2,2) mm long, prominently

glandular subapically, glabrous or sparsely puberu-

lous basally. Flowers all perfect (very rarely

staminate), odourless, 6,5— 8,5(-9,5) mm wide.

Pedicels 1,5—2 mm long in flower, to 3 mm long in

fruit, horizontal to erect in flower, usually erect

(recurved c. 70°— 120°(— 180°)) in fruit, puberulous

for more than half their length. Sepals strongly

convexo-concave, ovate (to ovate-elliptic or oblong-

elliptic), (1,5— )2— 3,5(— 4) mm long, (1,3— )1, 5— 2,2

mm wide, green except for the hyaline margin,

strongly hooded and thickened apically, with

subapical bilobed glands, sparsely puberulous with

uniform length hook-hairs. Paired petals 3—5 mm

long, 2,7— 3,5 mm wide, limb ovate to ovate-

orbicular, 2,5— 3,5 mm long, pale lilac (RHS

colours: 76C— D, 76D), apex rounded to truncate,

sometimes slightly hooded, claw 1 — 1,5 mm long,

whitish, glabrous. Medial petal elliptic to obovate,

hooded apically, 2— 2,5(— 2,8) mm long, 1,4— 1,8

mm wide, white or greenish white, usually tinged

with pink. Filaments free or the stamen filaments

very shortly fused basally. Medial staminode entirely

yellow, filament c. 0,5 mm long, antherode bilobed,

lobes sessile or subsessile, elliptic to obovate,

0,3 -0,4 mm long. Lateral staminodes with filaments

1 — 1,5 mm long, entirely yellow, antherodes

bilobed, lobes shortly stalked, obovate-cuneate,

(0,15— )0, 3— 0,4 mm long, 0,25-0,5 mm wide.

Lateral stamens with filaments dorsiventrally flat-

tened, parallel, then converging near the apex, 2,5

-3 mm long, greenish yellow, densely bearded for

R. B. FADEN

95

Fig 2. — Aneilema arenicola Faden. 1, habit; 2, inflorescence; 3, flower, front view; 4, flower, side view; 5, medial staminode; 6,

lateral staminode; 7, gynoecium; 8, medial stamen; 9, medial stamen anther; 10, lateral stamen, dorsal view; 11, lateral

stamen, ventral view; 12, stamen filament hairs; 13, capsule; 14, cells of the capsule wall; 15, seed, ventral view; 16, seed,

dorsal view. Bar = 1 mm in Figs. 2.3—2.11, 2.15 and 2.16; bar = 10 mm in Figs 2.1 and 2.2; bar = 0,1 mm in Figs 2.12 and 2.14.

(Fig. 2.2 from Faden & Faden 74/211; all others from Faden et al. 74/204.).

96

NEW TAX A OF ANEILEMA R. BR. (COMMELINACEAE) FROM SOUTHERN AND TROPICAL EAST AFRICA

0,5 -0,8 mm above the middle (and broadened in

this region) with patent, white, uniseriate (some

terminally hooked) hairs (attached ventrally and

laterally to the filament) to c. 0,5(— 0,7) mm long,

anthers facing the floral midplane and contacting

one another, ovate-elliptic, 0,5-0,75 mm long,

0,3— 0,8(— 0,9) mm wide, anther sacs blue-black or

blackish violet (sometimes pale), pollen yellow.

Medial stamen with filament 1,6-2 mm long,

undulate or straight, recurved near the apex,

greenish yellow, anther ovate to obovate, 0,3—0,65

mm long, 0,4— 0,9 mm wide, entirely yellow (the

anther sacs somewhat brighter than the connective),

pollen yellow (concolorous with lateral anther

pollen). Ovary sessile, suborbicular, 0,8 -1,1 mm

long and wide, green, densely covered with

forward-pointing, colourless, glandular-capitate

hairs except midventrally (where sparsely hairy) and

middorsally (where glabrous), apex rounded, dorsal

locule suppressed, ventral locules each 1-ovulate;

style 1,5— 2,3 mm long, abruptly distinct from ovary,

medial in flower, arcuate-decurved, tapering apical-

ly, green or greenish yellow basally, yellow above,

stigma capitate, white or greenish yellow, making

contact with the anthers. Capsules subsessile,

broadly elliptic to elliptic-orbicular or obovate-

orbicular, dehiscent, bivalved, bilocular, 2,8—4 mm

long, 2,6-3,75 mm wide, tan to greenish tan,

lustrous, sparsely puberulous, apex truncate to

emarginate, base broadly cuneate to truncate, valves

persistent, transversely wrinkled and relatively

planar, spreading about 180°, dorsal locule suppress-

ed, ventral locules each 1-seeded, cells of the outer

capsule wall transversely elongate. Seeds elliptic,

rounded to rounded-truncate at both ends, 2,2— 3,1

mm long, 1,5— 1,9 mm wide, 0,95 — 1,3 mm thick,

hilum dark brown, straight, raised within a shallow

groove, not at all to very slightly extended onto

apical and basal surfaces, embryotega chocolate

brown, testa buff to light brownish orange,

foveolate-reticulate, densely farinose in some or all

of the depressions and frequently also around the

hilum and embryotega, the farinose material often

coalescing and becoming sheet-like. Fig. 2.

Aneilema arenicola is restricted to southern

Mozambique and extreme northern Natal. It is

found on roadsides and hillsides with partially open

woodland. It grows in sandy soil in full sun from

10-60 m altitude. Flowering specimens have been

seen in November, December, and February to

April. In the field the flowers were observed to fade

at lOhOO. In cultivation they open about two hours

after sunrise and remain open for approximately

two-and-a-half hours.

Mozambique. — Lourengo Marques: Maputo, proximo da

ponte do rio Futi, Correia & Marques 776 (US; WAG); s. loc.,

Borle 364 (PRE). (A specimen with this collection number at G is

A. indehiscens Faden subsp. lilacinum Faden.)

Natal.— 2632 (Bela Vista): Ndumu Game Reserve, 1 km E of

main Rest Camp, 26°55'S, 32°19'E (-BD) Faden etal. 741204 (K;

MO; NH; NU; PRE; US); Ndumu Hill, between camps

Ukondo-Ndumu Game Reserve (-CC), Pooley 1679 (MO;

PRE). 2732 (Ubombo): 8 mi. E of Pongola River on road to

Maputa (-AB), Moll 4621 (EA; K; PRE); Ubombo-Sordwana

Bay road, 2,6 km towards Sordwana Bay from Shongwe, 27°26'S,

32°24'E (- AD), Faden & Faden 74/211 (B; BR; EA; K; MO;

NH; NU; P; PRE; UPS; US; WAG); Mkuzi Game Reserve, near

Bube hide (-CA or -CB), Stewart 1698 (MO).

ANEILEMA INDEHISCENS FADEN AND A. TANAENSE

FADEN

Aneilema indehiscens Faden and A. tanaense

Faden belong to Aneilema section Lamprodithyros

which is centred in tropical East Africa (Faden,

1975). One subspecies of A. indehiscens occurs

within the Flora of Southern Africa area, so that

species is described below. Seeds of A. tanaense

have been used in a series of germination

experiments that are soon to be published.

Therefore, although that species is endemic to

Kenya, it is described herein. The accounts of both

species have been largely adapted from Faden

(1975). Because of the urgent need to validate the

names, these descriptions are being published in

advance of a monograph of section Lamprodithyros.

Aneilema indehiscens Faden, sp. nov.

Aneilema petersii (Hassk.) C. B. Clarke in DC., Monogr. Phan.

3: 225 (1881), pro Kirk s.n.; in Thiselton-Dyer, FI. Trop. Africa 8:

70 (1901), pro Kirk s.n.

Aneilema dregeanum sensu Compton, FI. Swaziland, 33 (1966),

p.p., non Kunth (1843).

Herbae perennes caulibus florentibus ad 60 cm

altas. Folia spiraliter disposita laminis anguste

lanceolatis, lanceolato-ellipticis, lanceolato-ovatis

vel ovato-ellipticis, (2,5-)3— 10(— 13) cm longis,

(0,7— )1— 2,5(— 3,5) cm latis. Inflorescentiae thyrsi

ovoidei ad late ovoideos, (2— )2,5— 5(— 8) cm longi,

(1,5)2— 5(— 7) cm lati, cincinnis (1— )3— 9 compositi.

Pedicelli tempore fructigero uniformiter recurvati c.

180°. Petala Candida vel pallide lilacina, medio

cupulato stamina lateralia raro retinenti antheSis

initio. Stamina lateralia filimentis plerumque haud

cruciatim dispositis. Capsulae dehiscentes vel inde-

hiscentes, triloculares, castaneae, murinae vel

pallide brunneae fuscobrunneis guttatae,

(4-)4,5-6(-6,8) mm longae, (1,9— )2,3— 3(— 3,4)

mm latae, loculo dorsali plerumque monospermato,

loculis ventralibus uterque plerumque 2-spermato.

Semina loculorum ventralium 1,5— 2,2 mm longa,

1,3— 1,8 mm lata.

Type. — Kenya, Tana River District: Garsen-

Malindi road, 1,5 km towards Malindi from turnoff

to Oda, 2°32'S, 40°07'30'E, 22-24 July 1974, Faden

& Faden 74/1184 (US, holo.; BR, EA, FI, K, MO,

PRE, WAG, iso.).

Perennial herbs (habit type IIA3 of Faden, 1975).

Roots fibrous. Vegetative shoots sparsely branched,

trailing and often looping along the ground,

occasionally rooting at the nodes, sometimes

straggling through shrubs, to 3 m long (or longer?),

flowering shoots produced irregularly, unbranched

or sparsely branched, erect to ascending, to c. 60 cm

tall (reaching a greater height when straggling

through shrubs). Leaves spirally arranged, laminae

shortly petiolate, gradually reduced towards the

terminal inflorescence, narrowly lanceolate to

lanceolate-elliptic, lanceolate-ovate or ovate-

elliptic, rarely ovate, (2,5— )3 — 10( — 13) cm long,

(0,7-)l-2,5(— 3,5) cm wide, both surfaces lustrous,

R. B. FADEN

97

puberulous, veins pale on the adaxial surface.

Inflorescences thyrses, terminal and frequently

axillary from the upper leaves on the flowering

shoots, lax to moderately dense, ovoid to broadly

ovoid, (2— )2,5— 5(— 8) cm long, (1,5— )2— 5(— 7) cm

wide, with (1 — )3 — 9 cincinni, ascending (the lower

sometimes patent), mostly alternate (frequently

some subopposite, rarely some subverticillate).

Cincinni up to 7,5 cm long and 27-flowered.

Bracteoles spaced 1— 3,5(— 4,5) mm apart, ±

herbaceous, eccentrically cup-shaped, usually perfo-

liate, 1,3-2, 6 mm long, to 1 mm high, green, with a

prominent subapical gland, puberulous at least

basally or medially, frequently also with 1-several,

long, uniseriate hairs on or near the fused edge,

margin sometimes slightly thickened (glandular?)

near the fused edge. Flowers perfect and staminate,

odourless, (9 — )13 — 17,5 mm wide. Pedicels

3,8-6(-8) mm long in flower, to 10 mm long in

fruit, erect or ascending in flower, ± uniformly

recurved in fruit, usually c. 180°, greeri, puberulous.

Sepals glandular near the apex, puberulous; medial

sepal 2,4— 4, 3(— 4,9) mm long, subapical gland ±

distinctly bilobed; lateral sepals 2,6-4,3(— 4,6) mm

long, subapical gland usually unlobed (rarely

bilobed). Paired petals 7,3— 9,5 mm long, 6-8,5 mm

wide, limb broadly ovate to ovate-deltate, white or

pale lilac (RHS colours; 76C, Faden & Faden 74/202;

84D, Faden & Faden 74/208), claw white or whitish.

Medial petal cup-shaped, obovate (occasionally

ovate or suborbicular), 6-8 mm long, 4— 6(— 7,5)

mm wide, 3—5 mm deep, concolorous with the limbs

of the paired petals. Stamen filaments fused basally.

Medial staminode with filament (1,5 — )2, 7— 4,6 mm

long, antherode (rarely absent) bilobed, yellow,

lobes sessile to shortly stipitate, obovate-cuneate to

sickle-shaped and decurved. Lateral staminodes with

filaments 4—5,6 mm long, antherodes bilobed,

yellow, generally similar in size and form to that of

the medial staminode. Lateral stamens with fila-

ments usually ± parallel or slightly divergent for

their entire length, or sometimes convergent

apically, 7, 7 -8, 5 mm long, gently S-shaped, anthers

ovate to ovate-elliptic or occasionally elliptic to

lanceolate-elliptic, 0,65 — 1,3 mm long, pollen yellow

to orange or dirty white. Medial stamen with

filament 5—7 mm long, anther ovate to ovate-

elliptic, saddle-shaped, 1,5— 2,4 mm long, pollen

yellow to orange-yellow, concolorous with the

pollen of the lateral anthers or different in colour.

Ovary substipitate, densely and uniformly covered

with patent, glandular hairs (very rarely mixed with

a few hook-hairs), dorsal locule prominent, sub-

equal to the ventral locules or distinctly smaller than

them, 1— (or rarely 2-)ovulate, ventral locules each

2-(or very rarely 3-) ovulate; style 8-9,3 mm long,

straight or gently arcuate-decurved for most of its

length and strongly curved laterally out of the floral

midplane, stigma capitate. Capsules subsessile to

stipitate, obovate-elliptic to obovate-oblong, oblong

or oblanceolate, dehiscent or indehiscent, when

dehiscent, bivalved (occasionally partially trival-

ved), trilocular, (4— )4,5— 6(— 6,8) mm long,

(1,9— )2,3— 3(— 3,4) mm wide, chestnut brown or

mottled dark and light brown or grey-brown,

lustrous, puberulous, apex emarginate, valves

persistent, dorsal valve truncate to rounded apically

or sometimes terminating in a narrow ridge and

subequal to the ventral valve, dorsal locule

prominent, 1-seeded or, by abortion, empty (very

rarely 2-seeded), ventral locules each 2-(or, by

abortion, l-)seeded (very rarely 3-seeded); cells of

the capsule wall transversely elongate. Seeds elliptic,

2—2,9 mm long, 1,35 — 1,65( — 1,9) mm wide (dorsal

locule seed) or ovate to trapezoidal, l,5-2,2(-2,5)

mm long, 1,3 — 1,8 mm wide (ventral locule seeds),

0,65-1 mm thick, testa usually orange-buff (rarely

buff or orange-brown), very shallowly scrobiculate,

with white farinose granules sparse to dense around

the hilum, sparse around the embryotega and very

sparse or lacking in the depressions.

Key to the subspecies

Petals white (rarely faintly tinged with pink); capsules often

indehiscent, dorsal valve terminating in a narrow ridge

subsp. indehiscens

Petals pale lilac; capsules dehiscent, dorsal valve rounded to

truncate apically, not terminating in a ridge subsp. lilacinum

(a) Aneilema indehiscens Faden subsp. indehi-

scens

The typical subspecies is confined to eastern

Kenya and north-eastern Tanzania. It grows in

bushland and thickets of varied species composition,

in sandy or clayey soils, usually in partial shade, at c.

10 — 1 050( — 1 250) m altitude. Flowering specimens

have been seen in January, from March to May, July

and October. In the field the flowers open

06h00-06h30(-c. 06h45) and fade Ilh00-12h30.

Kenya. — Tana River: Garsen, 2°16'S, 40°07'E, Faden & Faden

74/1066 (B; EA; FI; K; MO; P; PRE; UPS; US; WAG); 0,8 km

towards Garsen from turnoff to Kibusu on Malindi-Garsen road,

2°21'S, 40°07'E, Faden & Faden 74/1173 (BR; C; EA; K; LISC;

MO; PRE); Garsen-Malindi road, 1,5 km towards Malindi from

turnoff to Oda, 2°32'S, 40°07'30”E, Faden & Faden 74/1184 (BR:

EA; FI; K; MO; PRE; US; WAG). Taita: Maungu Hills, 3°38'S,

*38°44'E, Faden, Evans & Githui 701158 (EA; K); Mile Post

»Xaveta 36/Voi 36 on Taveta-Voi road, c. 3°25'S, 38°10'E, Faden

Evans & Siggins 69/318 (EA; FI; K; MO); Tsavo National Park

East, Buchuma (Bachuma) Gate, 3°40'S, 38°56'E, Faden &

Faden 72/72 (EA; cultivated Missouri Botanical Garden: BR;

EA; K; MO; PRE); 28 km towards Taveta on Voi-Taveta road

from turnoff on Nairobi-Mombasa road, 3°30'S, 38°19'E, Faden

& Faden 74/489 (BR; EA; FI; K; MO); 18 km towards Taveta on

Voi-Taveta road from turnoff on Nairobi-Mombasa road, c.

3°30'S, 38°24'E, Faden & Faden 74/532 (EA; MO); 26 km

towards Taveta on Voi-Taveta road from Nairobi-Mombasa road

turnoff, c. 3°31'S, 38°21'E, Faden & Faden 74/536 (EA; K; MO);

11,7 km towards Mombasa past Maungu Station on Nairobi-

Mombasa road, 3°37'S, 38°50'E, Faden & Faden 74/1284 (MO); 3

km E of Bura Railway Station, Gillett 19562 (EA; K; MO); Voi,

Napier 973 (EA; K).

Tanzania. — Bagamoyo: 4,5 km towards Mbwewe from

crossing of Milgoji River on Korogwe-Dar es Salaam road, c.

5°57'S, 38°12'E, Faden & Faden 74/380 (MO; cultivated Missouri

Botanical Garden: DSM; EA; K; MO); Mbwewe, Faulkner 4471

(K). Lushoto: Mazinde, Drummond & Hemsley 2337 (K). Tanga:

Magunga Estate, Faulkner 1160 (K).

Zanzibar. — Chumbuni, Vaughan 1851 (EA; K).

(b) Aneilema indehiscens Faden subsp. lilacinum

Faden, subsp. nov.

Ab subsp. indehiscenti petalis pallide lilacinis,

capsulis semper dehiscentibus valva dorsali apice

rotundata vel truncata non crista terminata differt.

98

NEW TAXA OF ANEILEMA R. BR. (COMMELINACEAE) FROM SOUTHERN AND TROPICAL EAST AFRICA

Type. — South Africa, Natal: Ingwavuma-Ndumu

road, 15,5 km towards Ingwavuma from junction

with Ndumu-Maputa road, c. 27°06'S, 32°12'E, 16

February 1974, Faden & Faden 741202 (US, holo.;

BR, EA, K, LISC, MO, NH, NU, PRE, WAG,

iso.).

Subsp. lilacinum occurs from southern Zimbabwe

and southern Mozambique to northern Transvaal,

Swaziland and northern Natal. It grows in open

forest, woodland, thickets, lowveld bush and edges

of marshes, in sandy or clayey soils, usually in partial

shade, at c. 10—550 m altitude. Flowering

specimens have been seen in all months except

August. The flowers open about sunrise and begin to

fade after four to six hours.

Zimbabwe. — District unknown: Bank of Lundi River, Bayliss

BS7216 (MO).

Mozambique. — Gaza: Vila de Joao Belo, Chipenhe, Barbosa

& Lemos 8431 (COI; K; LISC; LMA; PRE). Inhambane:

Benguerua Isle, central ridge, Mogg 28886 (SRGH). Lourengo

Marques: Costa do Sol, Barbosa 655 (LISC); Without precise

locality, Borle 364 (G); Near Lourengo Marques town (Costa do

Sol), Gomes & Sousa 3441 (BR; K-2 sheets); Maputo, Sep 1930,

Gomes & Sousa s.n. (LISC); Ricatla, Acajou wood, Junod 493

(LISC; PRE-2 sheets); Inhaca Island, 23 mi. E of Lourengo

Marques, Mogg 27469 (K); Inhaca, Picada Estagao-Hotel, Moura

et al. 399 (US); Costa do Sol, Pedro 109 (LMA); Matolla, Quintas

64 (COI-2 sheets). Manica e Sofala: Mouth of River Melambe,

Zambesi Delta, 8 Jul 1861, Kirk s.n. (K).

Swaziland —Hlatikulu: Ingwavuma Poort, Compton 28610

(K; PRE).

Transvaal. — 2230 (Messina): 18 mi. NE of Sibasa on road to

Sambandou, Codd 6891 (K; PRE); 2431 (Acornhoek): Manyeleti

Game Reserve, Albatross koppie, (-CB), Bredenkamp 1795

(PRE).

Natal. — 2632 (Bela Vista): Ndumu Game Reserve, near main

Rest Camp, c. 26°55'S, 32°19'E (-CD), Faden & Faden 741208

(K; MO; NH; NU; PRE); Grid reference only (-CD), Moll 4152

(EA; K; NH; NU; PRE); Ndumu Game Reserve, Ndumu Hill

(—CD), Oatley C6 (PRE); Ndumu Game Reserve, E of Polwe

Pan (-CD), Pooley 1399 (NU). 2732 (Ubombo): Ingwavuma-

Ndumu road, 15,5 km towards Ingwavuma from junction with

Ndumu-Maputa road, c. 27°06'S, 32°12'E (— AA), Faden &

Faden 741202 (BR; EA; K; LISC; MO; NH; NU; PRE; US;

WAG); Ubombo Flats (-AB), Strey 10326 (EA; K; NH; NU;

PRE); Lake Sibayi (-BC/D), Vahrmeijer 693 (PRE); Mahatini

Flats, Vahrmeijer & Tolken 192 (PRE). 2831 (Nkandla): Eshowe,

above reservoir (-CD), Fawn 1289 (NH).

Aneilema indehiscens is most closely related to A.

petersii (Hassk.) C. B. Clarke with which subsp.

indehiscens is sympatric. Although quite distinct in

the field, dried specimens lacking capsules may be

difficult or impossible to determine. Such specimens

can sometimes be distinguished from A. petersii by

the form of the antherodes and bracteoles.

In the field or with more complete specimens or

detailed collector’s notes, A. indehiscens is readily

distinguishable from A. petersii on the basis of its

vegetative shoots long-trailing, antherode lobes

often falcate, connectives usually slightly elongate,

lateral stamen filaments usually not crossing,

capsules narrow, often indehiscent, and dorsal and

ventral locule seed dimorphism only slight. Aneile-

ma indehiscens is consistently tetraploid and A.

petersii regularly diploid (Faden, 1975, 1983).

Although the two subspecies are separated

geographically by almost 1 500 km, few characters

distinguish them unequivocally. These are given in

the key above. However, other tendencies are

shown by the subspecies which are sometimes useful

diagnostically. Subsp. indehiscens usually has fewer

uniseriate hairs on the bracteoles. It also tends to

have shorter fruiting pedicels [4,5— 6,5(— 8) mm]

than subsp. lilacinum [(5-)6-10 mm]. The shape of

the antherodes apparently will also separate all or

nearly all of the specimens, although further living

material is required to determine the extent of the

variation in subsp. lilacinum. In subsp. lilacinum the

anther sacs and/or sutures of all three anthers are

blue-black, while in subsp. indehiscens those of the

lateral anthers are entirely or partly yellow and those

of the medial anther are wholly yellow or

orange-yellow. In subsp. indehiscens mature cap-

sules are usually chestnut brown; in subsp. lilacinum

they are grey-brown or mottled light and dark

brown.

In southern Africa A. indehiscens subsp. lilacinum

has been overlooked or confused with the unrelated

A. dregeanum or A. brunneospermum. The Swazi-

land collection of A. indehiscens listed in the above

exsiccatae is cited by Compton (1976) as A.

dregeanum, a species which probably does not occur

in that country. (Compton has also included A.

brunneospermum in his A. dregeanum.) Similarly,

Ross (1972) omits any reference to A. petersii or a

related species in Natal, although A. indehiscens is

quite frequent in the northern part of that province.

He, too, may have included this species in A.

dregeanum or A. brunneospermum.

Aneilema tanaense Faden, sp. nov.

Aneilema clarkei Rendle, J. Linn. Soc., Bot. 30, PI. 34, Fig. 8

tantum. Fig. 7 & 9-12 et descr. excl. (1895).

Aneilema calceolus Brenan, Kew Bull. 15: 223 (1961), pro

Gregory s.n.

Herbae annuae. Inflorescentiae grandiores thyrsi

1— 2(— 3) cm longi, 1,5— 3(— 5) cm lati, cincinnis ad 8

compositi; inflorescentiae parviores cincinnis uno ad

aliquot fasciculatis compositae. Cincinni ad 2,2 cm

longos, bracteolis 1— 2(— 2,5) mm distantibus. j

Pedicelli (4— )5,5 — 10(— 11) mm longi, tempore

fructigero uniformiter recurvati 180°— 270°(— 360°).

Petalum medium calceolatum. Staminodium medium

nullum vel vestigiale. Capsulae (2,4— )2,7— 3(— 3,4)

mm longae, (1,1— )1, 5— 2,1 mm latae, valva dorsali

decidua. Semen loculi dorsalis hemisphaericum

1,1 — 1,4 mm longum, 1—1,2 mm latum, testa laevi

testacea. Semina loculorum ventralium subtriangu-

laria, testa scrobiculata non profunda, grisea.

Type. — Kenya, Tana River District: Garissa-

Malindi road, 16 km N of junction for Garsen, c.

2°08'S, 40°04'E, 15 January 1972, Gillett 19528 (US!,

holo.; B!, BR!, EA!, FI!, K!, MO!, PRE!, iso.).

Annual (rarely perennial) herbs (habit types IB,

IC of Faden, 1975). Roots thin, fibrous, produced

only at the base and lower nodes. Main shoot erect

or ascending, much branched at the base, 15-35 cm

tall, lateral shoots decumbent, or prostrate initially

and then ascending. Leaves spirally arranged on

main shoot, distichous (at least initially) on lateral

shoots, laminae sessile or shortly petiolate, gradually

reduced towards the inflorescence on the main

R. B. FADEN

99

shoot, lanceolate or lanceolate-elliptic to ovate,

1-6,5 cm long, 0,8-2,5(-3) cm wide, both surfaces

lustrous, puberulous. Inflorescences terminal on the

main and major lateral shoots and on very reduced

lateral shoots, ultimately produced from nearly all

nodes; reduced lateral shoots frequently perforating

the sheaths; larger inflorescences thyrses, moderate-

ly dense, broadly ovoid, 1 — 2( — 3) cm long,

1.5 — 3( — 5) cm wide, with up to 8 cincinni,

subopposite or subverticillate (occasionally some

alternate), ascending; smaller inflorescences con-

sisting of 1-several, clustered cincinni, lacking a

distinct axis and not clearly thyrses. Cincinni up to

2,2 cm long and 10-flowered (to 3,5 cm long and

17-flowered in cultivation). Bracteoles spaced

l-2(-2,5) mm apart, symmetrically or eccentrically

cup-shaped, perfoliate, 1,4— 1,8 mm long, promi-

nently glandular near the apex and with smaller

glands along the margin, puberulous in the basal 1/2

or, more commonly, only at the base. Flowers

perfect and staminate, odourless, (9— )10— 14,5 mm

wide. Pedicels (4— )5,5 — 10(— 11) mm long, erect to

slightly arcuate in flower, ± uniformly recurved in

fruit for their entire length 180°— 270°(— 360°), often

spirally twisted as well, puberulous. Sepals promi-

nently glandular near the apex, puberulous except

for glabrous margins; medial sepal 2,5-3 mm long,

with subapical gland distinctly bilobed, with smaller

glands also generally present along the margin near

the base; lateral sepals 2,8-3 mm long with

subapical gland unlobed, lacking marginal glands.

Paired petals (4,2— )6,5— 8 mm long, 4,8-7 mm

wide, limb ovate, pink or pale lilac (RHS colours:

77D-78D, Faden & Faden 74/1053; 84B— C, Faden

& Faden 74/1185), claw white. Medial petal

slipper-shaped, obovate-elliptic to suborbicular or

subquadrate, 4,7-6 mm long, 3-4,7 mm wide,

3-3,6 mm deep, concolorous with the limbs of the

paired petals. Stamen filaments fused basally. Medial

staminode usually absent (rarely vestigial). Lateral

staminodes with filaments 3,3-4 mm long, anther-

ode bilobed, yellow. Lateral stamens with filaments ±

parallel in the basal 1/2, then sharply divergent,

5.5 — 6,5 mm long, S-shaped, glabrous, anthers

elliptic to ovate, 0,7-1, 2 mm long, pollen yellow or

orange-yellow. Medial stamen with filament 3,5-4

mm long, anther ovate, saddle-shaped, 1 — 1,5 mm

long, pollen yellow or orange-yellow (concolourous

with that of the lateral anthers). Ovary substipitate,

densely and uniformly covered with patent, glandu-

lar hairs (mixed with hook-hairs along the lateral

sutures), dorsal locule 1-ovulate, ventral locules

each 2-ovulate; style 5, 5-6,5 mm long, arcuate-

decurved, then recurved near the apex, also gently

curving out of the floral midplane (rarely not),

stigma capitate. Capsules substipitate to shortly

stipitate, obovate (to ovate), dehiscent, bivalved,

trilocular, (2,4 — )2,7 — 3( — 3,4) mm long,

(1,1 — )1 ,5 — 2,1 mm wide, lustrous, puberulous, apex

emarginate, dorsal valve deciduous, dorsal locule

very prominent, often with a seed, ventral locules

each 2-(or, by abortion, l-)seeded; cells of the

capsule wall transversely elongate. Dorsal locule

seed hemispherical, 1,1 — 1,6 mm long, 1 — 1,4 mm

wide, 0,85—1 mm thick, embryotega whitish, testa

tan, smooth, lacking farinose granules and hypha-

like filaments except around the hilum. Ventral

locule seeds subtriangular, 1,2 — 1 ,4( — 1,7) mm long,

1,2— 1,3(— 1,5) mm wide, 0,8—0,95 mm thick,

embryotega dark brown to greyish brown, testa grey

or greyish tan, shallowly scrobiculate on all surfaces,

sparsely white-farinose in many of the depressions

and around the embryotega, densely so around the

hilum, frequently some hypha-like filaments present

among the farinose granules.

Kenya. — Kwale: Mombasa-Nairobi road, 2,5 km towards

Mombasa from turnoff to Maji ya Chumvi Railway Station,

3°49'S, 39°20'E, Faden & Faden 771582 (BR; EA; F; FI; K; MO;

P; PRE; US; WAG); Lungalunga-Ramisi road, 1 km before

turnoff to Kinango, 4°32'30"S, 39°05'30''E, Faden & Faden

771738 (EA; F; K; US); 5 km Maji ya Chumvi-Mackinnon Road

(Kilifi District on label), 3°48'S, 39°20'E, Gilbert & Rankin 4834

(EA). Tana River: 105 km N of Malindi on Garsen road, Andrews

in EA15070 (EA); Galole-Garsen road, 8 km towards Garsen

from turnoff to Wenje, 1°52'S, 40°05'E, Faden & Faden 74/1053

(C; EA; F; K; MO; P); Garsen, 2°16'S, 40°07'E, Faden & Faden

74/1064 (EA; MO); Malindi-Garsen road, 0,8 km towards Garsen

from turnoff to Kibusu, 2°21'S, 40°07'E, Faden & Faden 74/1171

(BR; EA; K; MO; PRE; WAG); Garsen-Malindi road, 1,6 km

towards Malindi from crossing of Lugga Buna, 2°23'S, 40°07'E,

Faden & Faden 74/1179 (EA; K; MO); Garsen-Malindi road, 1,5

km towards Malindi from turnoff to Oda, 2°32'S, 40°07'E, Faden

& Faden 74/1185 (B; BR; EA; FI; K; MO; US); Garissa-Malindi

road, 16 km N of junction for Garsen, c. 2°08'S, 40°04'E, Gillett

19528 (B; BR; EA; FI; K; MO; PRE; US); 105 km N of Malindi

on Garsen road (Kilifi District on label), Gillett 19532 (EA; K);

Galole-Malindi road, 16 km S of spot height 106, 2°14'S, Gillett

19973, cultivated at Missouri Botanical Garden (BM; EA; K;

MO); Lake Dumi, 13 February 1893, Gregory s.n. (BM).

This species is confined to coastal and subcoastal

Kenya where it occurs in deciduous or semi-

evergreen bushland and thickets at 10 - 250 m

altitude. In Tana River District it grows in a

seasonally waterlogged, grey-brown, clayey alluvi-

um with patches of sand. In Kwale District, where

both perennial populations have been collected, the

plants grow in a better drained soil. Flowering

occurs (December-) January to March and July to

August. In the field flowers open 08h30-09h00 and

fade 13h00-13h30.

The taxonomic confusion of this species with A.

clarkei is due to the publication by Rendle (1895) of

six figures (his PI. 34, Figs 7—12) which accompany

the type description of A. clarkei. Fig. 8 clearly does

not belong to that species. Through correspondence

with Brenan — and subsequent examination of the

specimens at the British Museum (Natural History)

— the writer determined that Gregory had made two

separate, unmixed collections of Aneilema with the

same label data. One of them is the type of A.

clarkei which, significantly, has on it all of the

drawings published by Rendle except Fig. 8. The

second sheet is the one cited by Brenan (1961) as A.

calceolus but is here treated as A. tanaense.

Aneilema tanaense is most closely related to A.

benadirense Chiov. of Somalia and A. calceolus

Brenan of Kenya and Tanzania. It may be

distinguished from A. benadirense by its generally

shorter leaves, smaller, often non-thyrsiform inflo-

rescences with fewer cincinni, shorter cincinnus

peduncles, less widely spaced bracteoles, puberulous

sepals, hook-hairs on the ovaries and capsules, and

smaller capsules. It can be separated from A.

calceolus by its less prostrate habit, more widely

100

NEW TAX A OF ANE1LEMA R. BR. (COMMELINACEAE) FROM SOUTHERN AND TROPICAL EAST AFRICA

spaced bracteoles which have marginal glands,

medial sepals generally with marginal glands, lateral

anther pollen yellow, and smaller capsules.

LECTOTYPIFICATION OF ANE1LEMA JOHNSTONII

Three collections from Tanzania were cited by

Schumann (1895) when he described Aneilema

johnstonii K. Schum.: Johnston s.n., Von Hohnel

159 and Volkens 2146. Among the syntypes Johnston

s.n. and Volkens 2146 belong to A. johnstonii, as it is

usually interpreted, while Von Hohnel 159 is A.

hockii De Wild. Although most of the description

applies equally well to all three collections, the

colour of the petals is taken from the Volkens

specimen, and the description of the capsule from

the Volkens and Johnston specimens. Because these

characters clearly separate A. johnstonii from A.

hockii, there can be no confusion as to which

element the name A. johnstonii should be applied.

The Von Hohnel collection cannot be the lectotype

for A. johnstonii.

Of the remaining two collections, the Berlin

specimen of Volkens 2146, which was presumably

seen by Schumann, has survived, whereas that of

Johnston s.n. has not. I am therefore designating

Volkens 2146 (B!) as the lectotype of A. johnstonii.

An isolectotype is at BM(!).

ACKNOWLEDGEMENTS

I wish to thank Alice Tangerini for preparing the

figures, except Fig. 2.1 which was drawn by Audrey

J. Faden; the Government of Kenya for permission

to do research in that country; Mrs J. Stewart and

Mr J. Lavranos for facilitating the field work in

South Africa; the Botanical Research Institute,

Pretoria for drying and shipping the specimens

collected in South Africa; the National Science

Foundation for Doctoral Dissertation Improvement

Grant GB-40817 which supported field work in

Africa by the author; the Missouri Botanical Garden

for providing travel funds for Audrey J. Faden; the

directors of the various institutions cited in the text

for the loan of specimens; and my wife, Audrey J.

Faden, for assistance with many facets of this study.

UITTREKSEL

Vier nuwe Aneilema-spes/es word beskryf : A.

indehiscens Faden, met subsp. indehiscens (Kenia,

Tanzanie) en subsp. lilacinum Faden ( Zimbabwe ,

Mosambiek, Suid-Afrika); A. arenicola Faden

(Mosambiek, Suid-Afrika); A. brunneospermum,

Faden (Mosambiek, Swaziland, Suid-Afrika ); en A.

tanaense Faden (Kenia). ’n Nuwe subspesie, Aneile-

ma dregeanum Kunth subsp. mossambicense Faden

(Mosambiek), word ook beskryf, en die lektotipe van

A. johnstonii K. Schum. is aangewys.

REFERENCES

Brenan, J. P. M., 1961. Notes on African Commelinaceae. III.

Kew Bull. 15: 208-228.

Compton, R. H., 1976. The flora of Swaziland. Jl S. Afr. Bot.,

Suppl., 11: 1-684.

Faden, R. B., 1975. A biosystematic study of the genus Aneilema

R. Br. ( Commelinaceae ). Ph. D. dissertation, Washington

University, St Louis.

Faden, R. B., 1983. Isolating mechanisms among five sympatric

species of Aneilema R. Br. (Commelinaceae) in Kenya.

Bothalia 14: 673-678.

Rendle, A. B., 1895. A contribution to the flora of Eastern

Tropical Africa. J. Linn. Soc., Bot. 30: 373-435.

Ross, J. H., 1972. Flora of Natal. Mem. bot. Surv. S. Afr. 39:

1-418.

Schumann, K., 1895. Commelinaceae. In A. Engler, Die

Pflanzenwelt Ost-Afrikas und der Nachbargebiete, Theil C.

Berlin: Reimer.

Bothalia 15, 1 & 2: 101-106 (1984)

Notes on the genus Rubus in southern Africa

C. H. STIRTON*

Keywords: distribution, leaf variation, Rubus, subspecies, taxonomy

ABSTRACT

The species Rubus ludwigii Eckl. & Zeyh. is re-examined and a key is provided to the subspecies recognized, A

new subspecies spatiosus C. H. Stirton is described. R. rigidus Sm. var .longepedicellatus C. E. Gust, is raised to

specific rank, viz R. longepedicellatus (C. E. Gust.) C. H. Stirton.

INTRODUCTION

This short paper is the third in a series of papers

on the genus Rubus in southern Africa (Stirton

1981a, 1981b).

1 Rubus ludwigii Eckl. & Zeyh.

Rubus ludwigii, Ludwig’s bramble, is perhaps the

most distinctive of the indigenous species of Rubus

in southern Africa. It is a widespread species,

extending from Beaufort West in the south-west of

its range to as far north as Lydenburg in the

Transvaal.

Throughout its range it is quite variable. A closer

study of this variation suggests that it should be

divided into two subspecies.

Rubus ludwigii Eckl. & Zeyh., Enum. 2: 263

(1836); Harv. in Harv. & Sond., FI. Cap. 2: 287

(1862); Focke in Bibl. Bot. 72, 2: 178, f. 178 (1911);

Gustafsson in Arkiv. for Bot. 26A, 7: 13 (1933); non

Sudre (1913). Type: Inter saxa strata (altit. IV)

laterum montium ‘Sturmberge’ prope ‘Witte — et

Zwartekeyrivier’ (Tembukiland). (PRE!; S!).

Rubus rhodacantha E.Mey. in Drege, Zwei Pfl. Donkum. 1:217

(1843), nom. nud.

Primocanes 1,0 -1,5 m long, erect and arching or

sprawling, covered in a white bloom; axis terete,

eglandular; sparsely or densely covered in 1-4 mm

long prickles, often with fine acicles at the base.

Floricanes reddish, glabrous, with or without a

bloom; sparsely covered in 3— 5 mm long, patent and

partly deflexed prickles or densely armed with

robust curved prickles. Leaves 5—9 partite on

floricanes; 3-7 partite on primocanes; lacerate,

laciniate or incised; upper surface glabrous or

sparsely pubescent on the blade, pubescent along

the sunken veins; lower surface fine to coarsely

woolly, pale greenish white to silver-white. Stipules

inserted 1-5 mm above the base of the petiole,

persistent, linear to lanceolate-falcate; 4-15 mm

long, 0,5 -2,0 mm wide, acute, entire, glabrous

inside, pubescent, tip glandular. Petioles of primo-

cane leaves 12—40 mm long, of floricane leaves

12—40 mm long. Synflorescence a few to many

flowered frondose-bracteose thyrse; rhachis with or

without prickles, glabrous to pubescent; pedicels

with or without prickles. Flowers 10-13 mm wide,

4-7 mm deep. Hypanthium saucer-shaped, 4-5

mm across, finely pubescent outside, glabrous inside

except near the base of the torus. Sepals 5-6 equal,

spreading and upcurving at the tips, becoming erect

after anthesis; teeth triangular, 5-10 mm long, 2-3

mm wide, pubescent. Petals pale pink to crimson,

5. 5— 7,0 mm long, 4, 0-5,6 mm wide, held vertically.

Stamens 100—120, glabrous, upcurving; filaments

2.5— 5,0 mm long, held in two rows. Pistils 40—45,

2.5— 3,5 mm long, variously pubescent below.

Collective fruit ovoid, 6—10 mm across, densely

white pubescent.

ludwigii.

* Botanical Research Institute, Department of Agriculture,

Pretoria. Current address: The Herbarium, Royal Botanic

Gardens, Kew, Richmond, Surrey , TW9 3AE, United Kingdom.

102

NOTES ON THE GENUS RUBUS IN SOUTHERN AFRICA

KEY TO SUBSPECIES

Synflorescence type I, (Fig. 1) 5 — 13 flowered; venation of lower surface of leaflets scarcely pubescent,

brown on dried specimens, easily discernible against the greenish-white or cream finely woolly blade;

tertiary venation obscure; terminal leaflets of floricane leaves lacerate; primocanes liberally armed with

robust recurved prickles subsp. ludwigii

Synflorescence type II, (Fig. 2) 20—40 flowered; venation of lower surface of leaflets thickly pubescent,

scarcely discernible against the silver to greenish white coarsely woolly blade; tertiary venation raised,

prominent; terminal leaflets of floricane leaves incised, rarely laciniate; primocanes sparingly armed

with slender, straight or partially deflexed prickles subsp. spatiosus

subsp. ludwigii

Primocanes up to 1 m long, sprawling or

creeping, slender, densely covered in stiff patent or

recurving 3-4 mm long prickles, extending along

the entire cane, covered in a white bloom. Floricanes

(Fig. 3) reddish, glabrous, with or without a bloom,

armed with robust hooked prickles. Leaves

7-partite on primocanes, 5(-7)-partite on flori-

canes; range of variation shown in Fig.4A

discolorous below, upper surface of leaflets deep

Fig 2. — Synflorescence type II found in Rubus ludwigii subsp.

spatiosus.

green, sparsely pubescent on the blade, more dense

along the sunken veins; lower surface finely

appressed woolly, greenish- white to cream, venation

scarcely pubescent, brown on dried specimens and

contrasting with the blade, tertiary venation

obscure. Stipules inserted, 1-3 mm above the base

of the petiole, persistent, linear-lanceolate, 4-7 mm

long 0,5— 2,0 mm wide, much narrower and shorter

on the floricanes. Petioles of primocanes 12-25(32)

mm long, of floricane leaves 12-20(25) mm long.

Synflorescence a 5-13-flowered frondose-bracteose

thyrse, not yet found on primocanes; rhachis

glabrous with a bloom or sparsely pubescent,

prominently armed with robust hooked prickles;

pedicels densely pubescent and armed with narrow

prickles. Flowers up to 11 mm wide, 4-6 mm deep.

Hypanthium saucer-shaped, 4 mm across, pubes-

cent at the base of the torus. Sepals 5-6, teeth 7-9

mm long, 2-3 mm wide. Petals pale pink to

crimson, 5, 5-6,0 mm long, 4,0-5, 0 mm wide,

narrowly elliptic to broadly ovate, margin ciliate and

erose. Stamens ± 100-120, filaments 2, 5-3,0 mm

long, held in two rows. Pistils 40-45, 2, 5 -3,0 mm

Fig. 3. — Upper section of the floricane of Rubus

ludwigii subsp. ludwigii ( Burgher s.n., PRE

23252).

C. H. STIRTON

103

4, Subsp. LUDWIGII

4,4b 4 4> 4

4.4$

®

FLORICANE.

♦ 4**4

4 4

PRIMOCANE

®

4 4 ♦ 4 ^ %!?'

Subsp. SPATIOSUS

$ 4 4

®

* * i

4Hi

. PRIMOCANE

• 44*4

4*4' ^

#4#4

Fig. 4. — Range of variation of leaflets in Rubus ludwigii subsp. ludwigii (A) and in Rub us ludwigii subsp . spatiosus (B),from:l,

terminal leaflet of floricane leaf; 2, lowest lateral leaflet of floricane leaf: 3, terminal leaflet of primocane leaf and; 4, lowest

lateral leaflet of primocane leaf.

long, pubescent; stigma bilobed. Collective fruit

6-8 mm across, southern populations smaller;

sepals erect during fruit development. Fruits

densely pubescent, greyish pink when young; lightly

packed.

Subspecies ludwigii is found most commonly in

rocky outcrops on grassy hillsides. From the few

ecological notes that are available it seems to favour

sandstone outcrops. In Lesotho this species some-

times forms tangles on the outskirts of woodlands

and is said not to pose a problem (Jacot Guillarmod,

1971). This subspecies is distributed to the west and

south of subspecies spatiosus (Fig. 5). It flowers

from October to December, but has a peak in

November.

Transvaal— 2629 (Bethal): Nooitgedacht (-DB), Henrici

1132 (PRE).

O.F.S. — (Senekal). Doornkop (-BC), Goossens 911 (PRE);

Westbury (-DD), Galpin 13939 (K; PRE). 2828 (Bethlehem):

Bethlehem area (-AB), Phillips 3175 (K; PRE); Mt Morkel

(~CA), Repton 6513 (PRE); Dunelm Farm, Fouriesburg

(~CA), Potts 3101 (PRE); Clarence (— CB), Van Hoepen s.n.

(PRE 18221); Wodehouse Plateau, Golden Gate Nature Reserve

(-DA), Roberts 3391 (PRE), Strey 2881 (PRE). Nelson’s Kop

(-AB), Cooper 850 (K). 2927 (Maseru): Ladybrand (-AB),

Rogers 5157 (PRE). 3026 (Aliwal North): Gaatje (-CC), Drege

s n. (P); 34 km from Aliwal North on road to Rouxville via

Kalkfontein (-DB), Herman 454 (PRE).

Natal.— 3029 (Kokstad): farm Vaalfontein (-AD), Comins

1906 (PRE).

Lesotho.— 2828 (Bethlehem): Leribe (-CC), Dieterlen 19

(PRE; SAM; STR). 2927 (Maseru): near Guilbeault Hall

(-AD), Schmitz 96 (PRE); Roma (-BC), Ruch 1792 (PRE),

Schmitz 4530 (PRE). 2929 (Underberg): Tsoelikana, between

the dam site and Khubetsoana (-CC), Hoener 1702 (PRE). 3027

(Lady Grey): 30 km from Mohales Hoek on road to Mafeteng

(-AB), Marais 1071 (K, PRE).

Transkei— 3127 (Lady Frere): Cala (-DA), Pegler 1644

(PRE).

Cape.— 3027 (Lady Grey): Jouberts Pass (-CA), Werger 1810

(PRE); Waterval farm in Saalboomspruit Catchment area

(-CD), Muller 754 (PRE); Barkly East (-DC), Gerstner 166

(PRE). 3124 (Hanover): Carolus Poort (-BB), Burchell 2757

(K); Compasberg (-DC), Drege s.n. (P). 3126 (Queenstown);

Molteno (-AD), Kuntze s.n. (K); between Garipina and

Zuurepoort ( — AD), Ecklon and Zeyher s.n. (PRE); Broughton

(-AD), Flanagan 1584 (PRE); near Queenstown (-DD),

Galpin 2189 (PRE). 3222 (Beaufort West): Kookfontein (-AB),

Burger s.n. (PRE); Nieuveld Mts (-BA), Marloth 8306 (PRE);

Mountain View Farm (— BD), Gibbs Russell, Robinson, Herman

and Downing 108 (PRE). 3224 (Graaff-Reinet): Graaff-Reinet

(-BC), Bowker 31 (K); 3326 (Fort Beaufort): Great Winterberg

(-AD), Ford s.n. (PRE). 3327 (Stutterheim): Windvogelberg

(-AC), Drege 3565 (P); Hogsback (-CA), Rattray s.n. (PRE

55600).

Without Precise Locality: Jacot Guillarmod 654(PRE),

Drege s.n. (K; PRE, ex Herb. Hennecart and Herb. Drake);

Drege s.n. (PRE, ex Herb. Steudel and P); Cooper 1110, 2292

(K, ‘Basutoland’); Atherstone s.n. (K); Prior s.n. (K).

104

NOTES ON THE GENUS RUBUS IN SOUTHERN AFRICA

Fig. 5. — Known distribution of Rubus ludwigii subsp. ludwigii.

Focke’s Fig. 73 in Bibl. Bot. 72,2: 178 (1911) is a

photograph of this subspecies. He places R. ludwigii

in subgenus Ideobatus, section Ideanthi, series

Pinnatifidi. With it he includes R. thibetanus

Franch, an Asian species. The diagnosis that he

gives for series Pinnatifidi, ‘Rami fertiles foliosi,

apice flores fere 5 — 12 capitato-racemoso ferentes.

Flores mediocres. Foliola pinnatifida, subtus albo-

tomentosa’, would have to be amended to include

subsp. spatiosus. In any case, I do not believe that

R. ludwigii and the Tibetan species are at all

related.

Fig. 6. — Upper section of the primocane of Rubus ludwigii

subsp. spatiosus (Trauseld 565).

Phillips (Ann. S.A. Mus. 16: 88-89, 1917) records

that the Lesotho call this subspecies ‘Monokotsoai

oa basali’ , the mulberry or raspberry of the women.

subsp. spatiosus C. H. Stirton, subsp. nov.,

primocannis parce armatis aculeis angustis rectis vel

partim deflexis; floricannae foliarum foliolis termi-

nalibus incisis; venulis tertiariis prominentibus venis

dense pubescentibus contra laminam argenteam vel

alboviridem grosse lanatam vix visilibus distinctus.

Type. — Natal, 3029 (Kokstad): 10 km from

Kokstad to Franklin (—AD), Stirton 8144 (PRE,

holo.!).

Primocanes (Fig. 6) up to 1,5 m tall, erect,

arching, green, covered in a white bloom; axis

terete, eglandular, base densely covered in fine

acicles and stiff, patent 1—2 mm long, brown-tipped

prickles, becoming fewer towards the apex.

Floricanes (Fig. 7) reddish, glabrous, with or

C. H. STIRTON

105

Fig. 8. — Known distribution of Rubus ludwigii subsp. spatiosus.

except for a few scattered hairs along the sunken

veins; lower surface coarsely woolly, silvery-white to

greenish white. Stipules inserted 4—5 mm above the

base of the petiole, persistent, lanceolate-falcate,

10-15 mm long, 1,5— 2,0 mm wide, edges of the

outer face, tip glandular. Petioles of primocane

leaves 30-40 mm long, of floricane leaves 30-40

mm long. Synflorescence a frondose-bracteose

thyrse, 30-40-flowered, fewer flowered on primo-

canes, axillary on floricane, terminal on primocanes,

leaves reducing apically to bracts; finely pubescent,

rhachis scattered with small prickles; pedicels only

rarely armed. Flowers 10—13 mm wide, 6—7 mm

deep. Hypanthium saucer-shaped, 5 mm across,

finely pubescent outside, glabrous inside except near

the base of the torus. Sepals 5-6, equal, spreading

and upcurving at the tips, becoming erect after

anthesis; tips green with sessile glands; teeth

triangular, 5-6(10) mm long, 3 mm wide; pubescent

inside and outside. Petals bright pink (Bisalski

colour Kll-4— 3,5), 6-7 mm long, 5,0— 5,6 mm

wide, with claw 1,25-3,0 mm wide and sharply

reflexed, erect covering androecium and gynoecium,

forming a false corolla tube, margin erose,

Hermannia- like, persistent, withering at fruit set.

Stamens ± 120, glabrous; curving inwards, arched

from the base; filaments 4-5 mm long, flushed pink

just below the yellow anthers, held in two rows.

Pistils 40—45; ovary 1 mm long, pubescent, style

3,0— 3,5 mm long, sparsely hairy in basal half,

grooved; stigma truncate, somewhat flared. Collec-

tive fruit ovoid, up to 10 mm across; sepals enclosing

fruit. Fruits pubescent; endocarp rugose, lateral

faces glabrous.

Subspecies spatiosus has been found in the

Themeda triandra Grassveld, in and around bush

clumps in open grassveld, near or among dolerite

outcrops, on the edges of forest and alongside

streams. It flowers between late October and

January with a peak in December. This subspecies

ranges from East Griqualand to the southern

Transvaal (Fig. 8).

Transvaal. — Lydenburg (-AB), Wilms 457 (K.) 2529

(Witbank): 35 km from Stoffberg to Lydenburg via Roossenekal

(-BB), Stirton 9769 (PRE). 2530 (Lydenburg): 15 km S of

Lydenburg (-AB), Codd 8291 (PRE; K); 4 km from Dullstroom

to Belfast ( — AC), Stirton 6727 (PRE); 1 km from Dullstroom to

Lydenburg (—AC), Stirton 6729 (PRE). 2629 (Bethal): Spitzkop

(-BD), Pott 4943 (PRE); Ermelo (-DB), Leendertz 7870

(PRE). 2630 (Carolina): Leeupoort (-AA), Burtt Davy 7417

(PRE). 2730 (Vryheid): Wakkerstroom (—AC), Pole-Evans

19652 (PRE), Galpin 9871 (PRE), Watt and Brandwyk 1629

(PRE); Oshoek (-AD), Devenish 184 (PRE); 20 km north of

Belfast (-DA), Prosser 2036 (PRE, K); Belfast (—DA), Burtt

Davy 1242 (PRE), Leendertz 2733 (PRE).

O.F.S. — 2828 (Bethlehem): Witzieshoek ( — DB), Junod 17421

(PRE).

Natal. — 2730 (Vryheid): 10 km SNE of Groenvlei (-AD),

Thode 1156 (PRE). 2829 (Harrismith): Nolens Volens (-AC),

Van der Zeyde 146 (PRE); Cathedral Peak Forest Station

(-CC), Killick 1331 (PRE; K). 2830 (Dundee): Buffalo River

(-BC), Rogers 5065 (PRE). 2929 (Underberg): Cathkin Peak

St

(id

Fig. 9. — Habit and habitat of

Rubus ludwigii subsp. spa-

tiosus ( Stirton 6727).

106

NOTES ON THE GENUS RUBUS IN SOUTHERN AFRICA

(-AB), Galpin 11872 (PRE); Giants Castle Game Reserve

(-AB), McGregor 5 (PRE ), Trauseld 565 (PRE); Ntabamhlope

Research Station (-BA), Acocks 10033 (PRE); road from

Giants Castle to Kamberg (-BA), Henderson 42 (PRE),

Estcourt (— BB), Wedermann & Oberdieck 1214 (PRE; K);

Rosetta (— BD), Brown s.n. (PRE 56005); Underberg (—CD),

McClean 773 (PRE; K); Carters Neck ( — DA), Henderson 47

(PRE); Shafton Plantation ( — DD), G. Hemm s.n. (PRE), 2930

(Pietermaritzburg): near Howick (-AC), Hutchinson 1852 (K;

PRE), Marr & Scotney 5 (PRE); 1 km S of Curry’s Post (- AC),

Moll 1188 (K; NH; PRE); 30 km SW of Nottingham Road

(-AC), Codd 8532 (PRE); Nottingham Road (—AC), McClean

826 (PRE); Lidgetton (-AC), Mogg 6742 (PRE); Balgowan

(-AC), Mogg 3914 (PRE); Tweedie (-AC), Mogg 1061

(PRE); Cedara (— CB), Sim 19967 (PRE); Keerom, Byrne

(-CC), Strey 10849 (PRE); Deepdene, near Richmond (-CD),

Sanderson 855 (K). 3029 (Kokstad): 14 km from Swartberg to

Underberg (-AB), Stirton 8154 (PRE); 10 km from Kokstad to

Franklin (-AD), Stirton 8144 (PRE), Umzimhlava River

(-AD), Schlechter 6544 (K); Kokstad (-CB), Tyson 869, 1137

(PRE; K).

Cape. — 3029 (Kokstad): Mt Currie ( — AD), Goossens 231

(PRE).

Without Precise Locality: Wood 1856 (K).

Rubus ludwigii subsp. spatiosus is a more robust

plant than subsp. ludwigii (Fig. 9).

There is a photograph of a Beauvens specimen of

this subspecies in G with the manuscript name var.

tomentosus. As far as I can determine, this is a

manuscript name only.

2 Rubus longepedicellatus (C.E. Gust.) C. H.

Stirton

The recent discovery of a hybrid swarm between

two species of Rubus, one indigenous and one

introduced, called for a re-evaluation of the correct

identity of the indigenous element. Previously

recognized as a forma of R. rigidus, the indigenous

element is now considered a distinct species.

Rubus longepedicellatus (C.E. Gust). C. H.

Stirton, comb, et stat. nov.

Rubus rigidus Sm. var. longepedicellatus C.E. Gust, in Arkiv.

Bot. 26A, 7: 59 (1933). Type: Transvaal, Duiwelskloof, Galpin

10107 (PRE, iso.!; K!).

R. rigidus Sm. var. rigidus f. subinermis C.E. Gust, in Arkiv.

Bot. 26A, 7: 59 (1933). Lectotype: Transvaal, New Agatha,

McCullum 887-88 (PRE!).

This species occurs along the margins of natural

forests and plantations and in open grassland. It

hybridizes with R. sp. cf. cuneifolius Pursh,

forming extensive swarms in a few areas of the

eastern Transvaal. It is predominantly a bramble of

high rainfall upland areas.

Transvaal. — 2330 (Tzaneen): Duiwelskloof (-CA), Galpin

10107 (PRE; K); New Agatha (-CC), McCullum 887-88

(PRE); between Tzaneen and Duiwelskloof (-CC), Stirton 5783

(PRE); Tzaneen (— CC), Stirton 5756 (PRE); 10 km from

Tzaneen to Duiwelskloof (— CC), Stirton 5755 (PRE). 2430

(Pilgrims Rest): Mariepskop ( — DB), Van der Schijff 4562, 6249

(PRE, PRU); Bourkes Luck (-DB), Viljoen 27 (PRE); 15 km

south of Bourkes Luck (-DB), Gaum & Henderson 32 (PRE);

14 km from Graskop to Vaalhoek (-DB), Meeuse 10011 (PRE);

MacMac ( — DB), Smuts & Gillett 2255 (PRE); Pilgrims Rest

(-DD), Burtt Davy 505 (PRE); Graskop (-DD), Gttuni &

Henderson 22 (PRE), Stirton 9865 (PRE); 3 km from MacMac

to Graskop (-DD), Stirton 9862 (PRE). 2431 (Acornhoek): 8,6

km from Graskop-Bosbokrand-Hazyview T-junction (— CC),

Stirton 9808 (PRE). 2531 (Komatipoort): Klipkoppies (-AC),

Stirton 9855 (PRE).

Rubus longepedicellatus is easily separated from

R. rigidus Sm. by its well-developed, much ramified

bracteose thyrse, the consistently 3-foliolate

floricanes and the 7-foliolate primocanes. The petals

in this species exceed the sepals, whereas in R.

rigidus the petals are equal to or shorter than the

sepals.

UITTREKSEL

Die spesies Rubus ludwigii Eckl. & Zeyh. is

ondersoek en ’n sleutel tot die twee erkende subspesies

word verskaf. ’n Nuwe subspesies spatiosus C. H.

Stirton word as nuut beskryf. R. rigidus Sm. var.

longepedicellatus C. E. Gust, word tot spesifieke rang

verander, nl. R. longepedicellatus (C.E. Gust.) C.

H. Stirton.

REFERENCES

Jacot Guillarmod, A. F. M. G., 1971. Flora of Lesotho

( Basutoland ) 474 pp. Lehre: Cramer.

Stirton, C. H., 1981a. Notes on the taxonomy of Rubus in

southern Africa. Bothalia 13: 331—332.

Stirton, C. H., 1981b. New records of naturalized Rubus in

southern Africa. Bothalia 13: 333-337.

Bothalia 15, 1 & 2: 107-116 (1984)

The genus Thunbergia in southern Africa*

E. RETIEF* * and W. F. REYNEKE+

Keywords: Acanthaceae, revision, southern Africa, taxonomy, Thunbergia

ABSTRACT

The genus Thunbergia Retz. in southern Africa is revised and eleven species are recognized.

THUNBERGIA

Thunbergia Retz. in Phys. Saellsk. Handl. 1: 163

(1780, non ‘1776’) nom. cons, non Thunbergia

Montin in K. svenska Vetenskakad. Handl. 34: 288

(1773), nom. rejic. ; L.f., Suppl. 292 (1781);

Cothenius, Disp. 8: Mai-Jun. (1790) as ‘Thunberga’;

Poiteau in Revue hort., Paris 2,4: 409 (1845) as

‘ Thumb er gia’ ; Nees in DC., Prodr. 11: 54 (1847);

Benth. & Hook, f., Gen. PI. 2: 1072 (1886); Baill.,

Hist. PI. 10: 423 (1891); Lindau in Nattirl. PflFam.

4,36: 291 (1895); Burkhill & Clarke in FI. Trop. Afr.

5: 8 (1899); C.B. Cl. in FI. Cap. 5,1: 3 (1912);

Bailey, Cycl. Hort. 3337 (1929); Phill. , Gen. 700

(1951); Chittenden in Diet. Gard. 2105 (1951);

Heine in FI. W. Trop. Afr. 2 (edn 2): 392 (1963);

Melchior in Engl., Syll. PflFam. 2 (edn 12): 458

(1964); Heine in Flora Gabon 13: 58 (1966); Benoist

in FI. Madag. 182: 9 (1967); P.G. Meyer in Prodr.

FI. S.W. Afr. 129 (1968); Long in J. Arnold Arbor.

51: 273 (1970); Ross, FI. Natal 321 (1972); R.A.

Dyer, Gen. Sthn Afr. PI. 1: 585 (1975); Compton,

FI. Swaz.: 546 (1976). Type species: T. capensis

Retz.

Diplocalymma Spreng., Neue Entd. 3: 30 (1822).

Heksacentris Nees in Wall., PI. Asiat. Rar. 3: 78

(1832).

Endomelas Rafin. in FI. Tellur. 4: 67 (1836).

Pleuremidis Rafin. in FI. Tellur. 4: 67 (1836).

Schmidia Wight, in Ic. PI. Ind. Orient.: 1. 1848

(1852).

Herbs or herbaceous creepers; rootstock present

or absent; epigaeous organs sparsely to densely

hairy. Trichomes unbranched, uni- or multi-cellular;

glandular trichomes stipitate with heads panduri-

form or pinshaped. Stems laterally grooved, square,

cylindric or ribbed; unbranched or monopodially

branched. Leaves simple, decussate, sessile or

petiolate; petioles sometimes winged, basal part

somewhat swollen; blade linear, lanceolate, ovate or

elliptic, apex emarginate, rounded, obtuse or

attenuate, base hastate, auriculate, cordate.

’Partly based on an M.Sc. thesis, University of Pretoria.

* ’Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

’Department of Botany, University of Pretoria, Hillcrest,

Pretoria 0002.

rounded or obtuse or acute, margin entire, lobed or

toothed, veins looped or parallel to apex, main vein

and some lateral veins ending in a mucro (Fig. 1).

Stipules absent. Flowers axillary, zygomorphic,

bisexual, pedunculate, subtended by 2 bracts.

Flower bracts 2, sessile, opposite, caducous or

persistent , enlarged sometimes to protect fruit , ovate ,

boat-shaped, keeled, base cordate. Calyx persistent,

connate at the base, usually with 10—16 subuliform

lobes, unequal in length. Corolla gamopetalous,

tube cylindric at base, infundibuliform in the upper

part; 5-lobed, lobes obovate or sometimes nearly

triangular, contorted; inner and outer surface of

lobes and outer surface of tube white, shades of

yellow, orange or purplish-blue; inner surface of

tube shades of yellow, yellowish green or dark

purplish black; one-celled trichomes sometimes

present, as well as stipitate glands. Stamens 4;

included, didynamous, inserted at the base of the

infundibuliform part of the tube. Filaments broad at

the base, linear dorsifixed, connective ending in an

apiculum, multicellular hairs present at base.

Anthers 2-thecous; thecae of long stamens equal,

spurred, with trichomes nearly along the whole

length of the cleft and concentrated at the base;

thecae of short stamens unequal, the longest one

with trichomes along the whole length of the cleft,

spurred, the other one with trichomes only at the

base, spurs absent or rudimentary; spurs un-

branched in exceptional cases, branched at apex

(Fig. 2). Gynoecium superior, included, ovary

2-locular with two axile ovules per locule; style

cylindric; stigma funnel-shaped or 2-lobed, usually

with one lobe above the other (Fig. 3). Disc present,

annular, fleshy. Fruit a capsule, globose, beaked,

loculicidal dehiscent, trichomes present or absent;

fruit stalk, sometimes swollen (Fig. 4). Seed

semi-globose with a cavity on the side of the hilum;

testa brown to greyish black, pubescent, trichomes

homogeneously distributed or adhering to form

different patterns on the testa (Figs 5 & 6).

In 1773 Montin described the genus Thunbergia in

Rubiaceae. In 1780 Raetzius, aware of Thunbergia

Montin, used the same generic name for describing

species of Acanthaceae collected by Thunberg at the

Cape, hence Thunbergia Retz. was a later homonym

and therefore illegitimate. Later authors such as the

younger Linnaeus (1781) did not attribute Thunber-

gia to Retzius, and Lindau (1893) and C.B. Clarke

(1912) actually attributed the name to Linnaeus fil.

Fortunately the problem has been resolved by^

108

THE GENUS THUNBERGIA IN SOUTHERN AFRICA

conservation of Thunbergia Retz. against Thunbe-

rgia Montin with T. capensis Retz. as the type

speciesofthe conserved name (ICBN,p. 396, 1978).

The presence of conspicuously large bracts

concealing the saucer-shaped calyx with its tooth-

like lobes, panduriform glandular hairs, the structu-

re of the corolla, the fruit being a globose capsule

with a prominent, ensiform beak and the absence of

retinacula are important features characterizing the

genus Thunbergia.

The species of Thunbergia, indigenous in southern

Africa, occur in grassveld, thorn veld, bushveld,

along the margins of forests or in false fynbos as

creepers, shrublets or herbs. Although certain

herbaceous creepers like T. dregeana, T. amoena

and T. purpurata are usually pgjt of the undergrowth

of forests and wooded areas, the species are not

Fig. 1. — Variation of leaf base in

Thunbergia, x 0,5. 1, T.

natalensis (Retief 220); 2, T.

purpurata (Retief 213); 3, T.

amoena ( Retief 178); 4, T.

capensis (Retief 254); 5, T.

aurea (Van Greuning 246); 6,

T. venosa (Stirton 347).

restricted to shady and moist places, but do occur in

the open.

1. Thunbergia natalensis Hook, in Curtis’s bot.

Mag. 24: t.5082 (1858); Anderson in J. Linn. Soc.,

Bot. 7: 18 (1864); Wood in Trans. S. Afr. phil. Soc.

204 (1909); C.B. Cl. in FI. Cap. 5,1: 4 (1912); Bews,

FI. Natal & Zululand 192 (1921); Phill. in Flower.

PI. Afr. 4: t.147 (1924); Bailey, Cycl. Hort.: 3339 '

(1929); Chittenden in Diet. Gard. 2106 (1951);

Hulme, Wild Flow. Natal t.5, f. 10 (1954); Eliovson,

S. Afr. Flow, for the Garden 274 (1955); Batten &

Bokelmann, Wild Flow. E. Cape 135 (1966); Ross,

FI. Natal 323 (1972); Kruger, Borne, Struike &

Rankplante 514 (1973); Agnew, Upland Kenya Wild

How. 578 (1974); Gibson, Wild Flow. Natal t.97, f. 4

(1975); Jeppe, Natal Wild Flow. 93 (1975);

Compton, FI. Swaz. 547 (1976). Type: Natal,

collector unknown (K, holo.!; PRE, photo!).

Fig 2. — Different anther types in Thunbergia, x 6. 1, part of a

short stamen with anther locules different in length, T. alata;

2, part of a long stamen with anther locules more or less the

same length, T. alata; 3, spur; 4, part of a long stamen with

anther locules more or less the same length, T. natalensis; 5,

part of a short stamen with anther locules different in length,

T. natalensis.

Fig. 3. — Stigmas and styles in Thunbergia, x 2,5. 1, T. natalensis,

stigma funnel-shaped; 2, T. capensis, stigma lobed; 3, T.

alata, stigma lobed.

E. RETIEF AND W. F. REYNEKE

109

Fig. 4. — Flower parts of Thunbergia purpurata; A, carpel; B,

bract; C, part of fruit stalk; D, seed.

Dwarf shrub, herbaceous, ± 0,75—1 m high;

stems puberulous. Leaves sessile or with petioles up

to 28 mm long; blade elliptic to ovate, (21 -) 55-110

(-159) x (15-) 35-70 (-82) mm, apex acute, base

cordate, margin entire, lobed or toothed. Flower

bracts 17—27 x (10—) 12—15 (—20) mm, light green

with conspicuous venation; with pin-like glands, ± 1

mm long, caducous. Calyx 1,5-6 mm long;

connated at the base, unequally shortly lobed;

unicellular trichomes present. Corolla with lobes

12-17 x 13—20 mm, lobes light blue to purplish

blue; tube 23-31 mm long and 2—18 mm in

diameter, inside yellow to white with purplish blue

to light blue stripes running from the lobes, outside

light blue to purplish blue. Filaments 7-10 mm long,

with stipitate glands, anther thecae 2,2— 3,2 mm

long, spurs conspicuously curved. Style 13-17 mm

long with pin-like glands. Stigma funnel-shaped with

pin-like glands and lateral clusters of multicellular

trichomes. Fruit pubescent; 21-30 mm long,

globose base 8-17 mm in diameter; stalk (26—)

35—76 (—89) mm long. Seed reddish-brown; surface

ridged; 4,5— 6,5 mm in diameter.

Found in shade along the margins of forests and

woodland or pine plantations, rarely in direct

sunlight. Recorded from the Transkei, Natal,

Swaziland, Transvaal and northwards to Kenya

(Fig. 7).

T. natalensis is characterized by its growth form,

purplish blue corolla, the curved spurs of the anthers

and a funnel-shaped stigma. The species flowers

from October to March.

The ‘Natal Blue Bell’ is found growing in gardens.

Bor & Raizada (1941) mention that T. natalensis

occurs in gardens throughout India. According to J.

Erens (pers. comm., 1976) the species can easily be

propagated by means of cuttings.

2. Thunbergia venosa C.B. Cl. in FI. Cap. 5,1: 6

(1912); Bews, FI. Natal & Zululand 192 (1921);

Hulme, Wild Flow. Natal t.12, f.ll (1954); Ross, FI.

Natal 323 (1972). Types: Natal, Itafamasi, Wood 643

(K, lecto.!; BOL!; NH!; PRE, photo!; SAM).

KEY TO SPECIES

Herbaceous shrublets or single-stemmed herbs:

Shrublet, 0,75-1 m; corolla purplish blue; stigma funnel-shaped 1. T. natalensis

Shrublet, 0,25— 0.4 m or a single-stemmed herb, 0,09-0, 3 m; stigma 2-lobed,one lobe below the other:

A single-stemmed herb; leaf blade narrowly ovate 2. T. venosa

A herbaceous shrublet; leaf blade ovate to broadly ovate 3. T. atriplicifolia

Herbaceous climbers (sometimes rather a scrambler):

Leaves with petioles not winged; fruit glabrous or sparsely pubescent:

Leaves sessile or with petioles less than 8 mm; anther thecae not spurred 4. T. capensis

Leaves with petioles usually longer than 8 mm; anther thecae spurred (if not, then leaf base hastate):

Leaf base hastate:

Leaf with veins on abaxial side dark purple 5. T. purpurata

Leaves not as above 6. T. pondoensis

Leaf base cordate:

Flowers white 7. T. neglecta

Flowers dull orange or lemon-yellow:

Leaves densely pubescent, flowers lemon-yellow (Transvaal) 8. T. amoena

Leaves sparsely pubescent, flowers dull orange (Natal, Cape Province) 9. T. dregeana

Leaves with petioles winged; fruit densely pubescent:

Inner surface of corolla tube dark purplish black, corolla orange; seeds with trichomes adhering in

wing-like protuberances 10. T. alata

Inner surface of corolla tube dull white, corolla dull orange; seeds with trichomes adhering in

papillae-like protuberances 11. T. aurea

110

THE GENUS THUNBERGIA IN SOUTHERN AFRICA

Fig. 5. — Part of the seed surface of Thunbergia pondoensis where

the trichomes adhere to form ridges, x 42.

Herb, sparsely to densely pubescent; trichomes

on stem, yellowish brown. Leaves sessile; blade

linear to lanceolate, (14—) 30-77 x (4-) 7-30 mm,

apex acute, base cordate, margin entire or with a few

teeth, main vein and lateral veins raised on under

surface. Flower bracts (18— ) 20— 25 x (8—) 16—20

mm, longer than mature fruit. Calyx covered with

stipitate glands having one stalk cell and multicellu-

lar trichomes; lobes 2,5 -4,8 mm long. Corolla with

inner and outer surface of lobes and outer surface of

tube yellow, inside of tube dull yellow; lobes 9,5—14

x 11—20 mm, tube 23—21 mm long and 2,4—11 mm

in diameter. Filaments 3—4,6 mm long; anther

thecae 3—4,2 mm long. Style 11 — 13 mm long;

unicellular trichomes present on upper part. Fruit

glabrous; 15—25 mm long; globose base 6—10 mm in

diameter; stalk 45-60 mm long. Seed reddish

brown; surface with prominent ridges; 2,6— 4,6 mm

in diameter.

T. venosa is found in grassland where it occurs in

open sunlight. The species is endemic in Natal

(Fig. 8).

It is distinguished from related taxa by its single

stem with slender, sessile leaves, densely crowded in

the upper part. The species flowers from September

to April.

A decoction obtained from the minced roots of T.

venosa is used by the Zulu people as a love-potion

(Hulme, 1954).

3. Thunbergia atriplicifolia E. Mey. ex Nees in

DC., Prodr. 11: 56 (1847); Anderson in J. Linn.

Soc., Bot. 7: 20 (1864); Sim, FI. Kaffraria 61 (1894);

Wood in Trans. S. Afr. phil. Soc. 203 (1909); Wood,

Natal Plants t.594 (1912); C.B. Cl. in FI. Cap. 5,1: 7

(1912); Bews, FI. Natal & Zululand 192 (1921).

Hulme, Wild Flow. Natal 1. 12, f. 11 (1954); Letty,

Wild Flow. Trans. 313 (1962); Batten & Bokelmann,

Wild Flow. E. Cape 140 (1966); Van der Schijff,

Checklist Kruger Nat. Park: 86 (1969); Gledhill, E.

Cape Veld Flow. 214: t.57, f. 7 (1971?); Ross, FI.

Natal 323 (1972); Gibson, Wild Flow. Natal t.97, f. 2

(1975); Jeppe, Natal Wild Flow. 92 (1975);

Fig. 6. — Part of the seed surface of Thunbergia aurea with patches

where no trichomes occur, x 56.

Fig. 7. — •, Thunbergia natalensis.

Fig. 8. — •, Thunbergia venosa.

E. RETIEF AND W. F. REYNEKE

111

Compton, FI. Swaz. 546 (1976). Type: Cape

Province, Kei River near Komga, Drege s.n. (K.,

lecto. ! ; PRE, photo!).

T. aspera Nees in DC., Prodr. 11: 56 (1847); Sim, FI. Kaffraria

61 (1894); Wood in Trans. S. Afr. phil. Soc. 18: 203 (1909); C.B.

Cl. in FI. Cap. 5,1: 8 (1912); Bews, FI. Natal & Zululand 192

(1921); Ross, FI. Natal 323 (1972). Type: Cape Province,

Transkei, Gill s.n. (K, holo.!; PRE, photo!).

T. aspera Nees var. parvifolia Sond. in Linnaea 23: 90 (1850);

C.B. Cl. in FI. Cap. 5,1: 8 (1912). Type: Transvaal, Magaliesberg,

Zeyher 1418 (K, lecto.!; PRE, photo!).

T. bachmannii Lindau in Bot. Jb. 17: 94 (1893); Lindau, l.c.

(Beibl. 41): 38,41 (1893); C.B. Cl. in FI. Cap. 5,1: 8 (1912);

Compton, FI. Swaz. 546 (1976). Type: Transkei, sin. loc.,

Bachmann 1267 (K, lecto.!; PRE, photo!).

T. baurii Lindau in Bot. Jb. 24: 312 (1898). Type: ‘Cape

Province’, Baziya, Baur 169 (K, holo.!; PRE, photo!).

T. flavohirta Lindau in Bot. Jb. 24: 311 (1898). Type: ‘Cape

Province’ Baziya, Baur 58 (K, holo.!; PRE, photo!).

T. galpinii Lindau in Bot. Jb. 24: 310 (1898); C.B. Cl. in FI.

Cap. 5,1: 6 (1912). Type: Transvaal, Saddleback Mountain,

Galpin 1211 (K, holo.!; PRE!).

T. xanthotricha Lindau in Bot. Jb. 24: 311 (1898); C.B. Cl. in

FI. Cap. 5,1: 8 (1912); Bews, FI. Natal & Zululand: 192 (1921);

Van der Schijff, Checklist Kruger Nat. Park 86 (1969). Type:

Transvaal, Barberton, Galpin 496 (K, holo.!; BOL!; PRE!;

SAM!).

T. atriplicifolia Nees var. kraussii C.B. Cl. in FI. Cap. 5,1: 7

(1912). Type: Natal, sin. loc., Krauss 405 (K, holo.!; PRE,

photo!).

T. bachmannii Lindau var. minor C.B. Cl. in FI. Cap. 5,1: 8

(1912). Type: sin. loc.. Cooper 893 (K, holo.!; BOL!; PRE,

photo!).

T. cordibracteolata C.B. Cl. in FI. Cap. 5,1: 7 (1912). Type: sin.

loc., Cooper 895 (BOL, iso.!; PRE, photo!).

T. hirtistyla C.B. Cl. in FI. Cap. 5,1: 7 (1912); Bews, FI. Natal

& Zululand 192 (1921); Ross, FI. Natal 323 (1972). Type: Natal,

sin. loc., Gerrard 1274 (K, holo.!; NH!; PRE, photo!).

Shrublet, 0,25—0,4 mm high, sparsely to densely

pubescent. Stems with a tendency to twine. Leaves

sessile or with petioles up to 4 mm long; blade ovate,

elliptic to lanceolate, (10 -) 23-62 (-80) x (6-)

12—35 (—55) mm, apex acute, obtuse or emarginate,

base cordate, truncate or cuneate, margin entire or

toothed, usually only with two teeth at the base.

Flower bracts persistent, (13—) 17—27 mm long.

Calyx 3-5 mm long and (8-) 10-17 mm broad.

Corolla with inner and outer surface of lobes and

outer surface of tube dull white, inner surface of

tube dull yellow; lobes (10-) 13-17 (—20) x (11-)

15-29 mm, tube 20—27 (-32) mm long and (1,7—)

2-10 mm in diameter with unicellular trichomes and

stipitate glands. Filaments 4—6,5 mm long; anther

thecae 3—4 mm long. Style 11-16 mm long, upper

part pubescent or glabrous. Fruit (13-) 15-19 mm

long with globose base (6—) 7-10 mm in diameter;

stalk (26-) 32-78 (-104) mm long. Seed reddish or

greyish brown with trichomes usually evenly

arranged; 3,8-6 mm in diameter.

T. atriplicifolia is found as a subordinate species in

grassland where it flourishes in both loam and sandy

soil. The species has a wide distribution, occurring in

the Transvaal, Swaziland, Natal, Transkei and the

Cape Province (Fig. 9).

Fig. 9. — #, Thunbergia atriplicifolia.

The species is very variable and occurs in different

habitats: it is not only found inland, but also in

grassveld close to the sea.

Lindau (1898) who described T. bachmannii, T.

xanthotricha and T. flavohirta mentions that these

three species might be varieties of one species,

namely T. atriplicifolia and that they needed further

investigation. Clarke (1912) in his revision of the

genus also indicates that it is an open question

whether it is not better to regard the species T.

bachmannii, T. xanthotricha and T. aspera as

conspecific. In 1864 Anderson cited T. aspera as a

synonym of T. atriplicifolia. Clarke (1912) used leaf

shape, pubescence of the style and the form of the

flower bracts, to distinguish T. atriplicifolia, T.

aspera, T. cordibracteolata, T. bachmannii, T.

hirtistyla and T. xanthotricha. The range of variation

within this species complex is very marked.

However, as extreme forms are linked by numerous

intermediates no formal subdivisions are upheld

here. Variation of the leaf blade within the species is

shown on Fig. 10.

Where the distribution areas of Thunbergia

atriplicifolia and T. capensis overlap, plants occur

pointing to hybridization of gene-flow between the

two species. Such specimens were in the past

identified either as T. atriplicifolia or T. capensis.

The leaves of the plants look like those of T.

capensis only larger, whereas the flowers resemble

those of T. atriplicifolia. The following specimens

are examples of such intermediate plants:

Transkei. — 3228 (Butterworth): Mazeppa Bay (-BC), Story

4463 (PRE); Kei Mouth (-CB), Flanagan 552 (GRA, PRE,

SAM); Bulura River Mouth (— CC), Acocks 15799 (PRE).

According to Batten & Bokelmann (1966), the

Xhosa people prepare a love-potion from the roots

and leaves of the plants and Watt & Breyer-

Brandwijk (1962) mention that the leaves of T.

atriplicifolia are used to make a hairwash. The

^species is known as ‘Natal Primrose’ or ‘Cape

Primrose’.

4. Thunbergia capensis Retz. in Phys. Saellsk.

Handl. 1, 3: 164 (1776); L.f., Suppl. 292 (1781);

Fig 10. — Variation of leaf shape in Thunbergia atriplicifolia, x 0,75. 1, De Sousa 431; 2, Retief 192; 3, Retief 120; 4, Relief 222; 5,

Pott 5656.

Thunb., Nov. Gen. PI. 1: 21 (1781); Thunb., Prodr.

106 (1800); Lam., Encycl. 7: 637 (1806); Lam., Tabl.

Encycl. 3: 97, t.549 (1819); Thunb. FI. Cap. (edn 2):

488 (1823); Lodd., Bot. Cab. t.1529 (1829); Nees in

Linnaea 15: 351 (1841); Drege, Zwei Pfl. Doc. 137,

226 (1843); Krauss, Flora: 136 (1846); Nees in DC.,

Prodr. 11: 55 (1847); Anderson in J. Linn. Soc., Bot.

7: 20 (1864); Lindau in Bot. Jb.: 17 (beibl. 41): 36,

39 (1893); Sim, FI. Kaffraria 61 (1894); C.B. Cl. in

FI. Cap. 5,1: 5 (1912); Marloth, FI. S. Afr. 3: 170

(1932); Fourcade, Check list Flow. PI. George,

Knysna, Humansdorp & Uniondale 6 (1941);

Chittenden in Diet. Gard. 2105 (1951); Martin &

Noel, FI. Albany & Bathurst 100 (1960); Batten &

Bokelmann, Wild Flow. E. Cape 135 (1966). Type:

Cape Province, Thanberg in Herb. Thunb. 14601

(UPS, lecto.; PRE, microfiche!; LUND!).

Scrambler, greyish green, pubescent. Stems

prostrate, twining or erect. Leaves sessile or shortly

petiolated; petioles up to 7,5 mm long; blade ovate

to reniform, 15-33 (—50) x 14-38 (—47) mm, apex

acute, obtuse or emarginate, base cordate, cuneate

or truncate, margin lobed or emarginate, trichomes

on upper surface orientated in the same direction.

Flower bract 10- 16 x 8—13 mm. Calyx covered with

short stipitate glands and unicellular trichomes;

lobes 2,5 -4,4 mm long. Corolla with the inner and

outer surface of the lobes and the outer surface of

the tube as well as part of the inner surface of the

tube yellowish green, bottom part of tube on inner

surface dull yellowish green; lobes 7-9 (—18) x 3—6

( — 15) mm, tube 12,3 — 13,5 (—20) mm long and 2—7

(-10) mm wide. Filaments 3—5,5 ( — 15) mm long;

anther thecae without spurs or only one locule of

long stamens spurred, thecae 1,5—2 (—4) mm long.

Style 7,5-8 ( — 12) mm long. Fruit 11,3-24 mm long

with bulbous base 6-11 mm in diameter; stalk (18—)

25-62 (-98) mm long. Seed brown; some trichomes

adhere to form comb-like protuberances; 3,2—5 mm

in diameter.

T. capensis flourishes not only on sandy soil in

grassland along the coast, but occurs also in stony

soil of false fynbos as found in the Ecca Pass area. T.

capensis is endemic to the Cape Province and the

Transkei (Fig. 11).

Fig. 11. — •, Thunbergia capensis; ▲, T. purpurata.

The plant is characterized by broadly ovate leaves

with apices acute, obtuse or emarginate, the

greenish yellow colour of the corolla and the anther

thecae not being spurred. The flowering period of T.

capensis is from October to April.

The species is suspected of hybridizing with T.

atriplicifolia where their areas overlap (see T.

atriplicifolia ). Specimens of T. capensis with erect

stems and flowers nearly double the normal size

were collected near Hayes railway halt between

Grahamstown and Port Alfred. On investigation

they were found to have a triploid chromosome

number.

E. RETIEF AND W. F. REYNEKE

113

Leaves of this species mixed and crushed with

those of a species of Clematis and Ziziphus

mucronata Willd. subsp. mucronata are applied to

glandular swellings (Batten & Bokelmann, 1966).

5. Thunbergia purpurata Harv. ex C.B. Cl. in

FI. Cap. 5,1: 5 (1912); Bews, FI. Natal & Zululand

192 (1921); Ross, FI. Natal 323 (1972). Type: Natal

sin. loc., Sanderson 442 (Herb. Hook. K, lecto.!;

PRE, photo!).

A dark green, herbaceous creeper, sparsely

pubescent. Stems purplish green where internodes

thicken; nodes hairy, a few trichomes present on

nodes. Leaves with veins on abaxial side dark

purple; petiolate, petioles (11—) 18—47 mm long;

blade lanceolate (21—) 46-78 (-90) x (11—) 18—37

(-47) mm, apex usually acuminate, sometimes

obtuse or emarginate; base hastate, margin entire.

Flower bracts 13-23 x 8-20 mm. Calyx with

unicellular trichomes and stipitate glands; lobes

1,3-4, 7 mm long. Corolla with inner and outer

surface of lobes and outer surface of tube white,

tube dull yellow on inner surface; lobes 9—13 x

10.5— 13 mm, tube 20-25 mm long, 2,5—8 mm in

diameter. Filaments 4-6,5 mm long; anther thecae

3. 5- 3, 7 mm long, spurs not strongly developed.

Style 11-14 mm long. Fruit 22-25 mm long, globose

base 8—10 mm in diameter; stalk 30—52 mm long.

Seed reddish brown; some trichomes adhere to form

comb-like protuberances; 4,5-7 mm in diameter.

T. purpurata is usually found along the margins of

forests or in open parts in the tropical coastal forests.

The species occurs only along the coast of Natal and

the Transkei as far south as the Kei River (Fig. 11).

It is distinguished from related species by its dark

green leaves with conspicuously dark purplish veins

on the under surface. The species flowers from

November to March.

In 1827 Hooker published a drawing and a

description of T. angulata Hils. & Boj. ex Hook.,

endemic in Madagascar, with a separate drawing of

an anther that was from a Thunbergia species of the

Cape (Clarke, 1912). T. angulata is not found along

the coast of southern Africa as maintained by

Hooker and also Anderson (1864) and Lindau

(1893). The anther could, according to Clarke

(1912), belong to T. purpurata described later on.

The two species can easily be distinguished by the

anther thecae which, in T. purpurata, are not

spurred or weakly so, the leaf base of T. purpurata is

hastate and not cordate, whereas the corolla of the

species is white and not purple.

6. Thunbergia pondoensis Lindau in Bot. Jb. 17:

93 (1893); Lindau l.c. (beibl. 41): 41 (1893); Wood,

Natal Plants 4: t.340 (1906); Wood in Trans. S. Afr.

phil. Soc. 18: 204 (1909); Bews, FI. Natal &

Zululand 192 (1921); Ross, FI. Natal 323 (1972).

Type: Natal, Backbeach, Durban, Bachmann 1265

(Bt); Inyesane, Wood 4015 (NH, lecto.!; PRE,

photo!).

Herbaceous creeper, pubescent. Leaves petiolate,

petioles 10—52 mm long; blade ovate, 25—90 x

14-70 mm, apex acute, sometimes obtuse, base

hastate, margin entire, lobed or toothed. Flower

bracts persistent; 14—24 x 9—20 mm. Calyx with

lobes 2—7 mm long, lobes subuliform. Corolla with

inner and outer surface of lobes and outer surface of

tube white, inner surface of tube orange yellow;

lobes 10—15 x 14—23 mm, tube 18—20 mm long,

4—7 mm in diameter. Filaments 4—7 mm long,

anther thecae 3-5 mm long. Style 8-12 mm long,

unicellular trichomes present on upper part. Fruit

15—20 mm long, globose base 7—10 mm in

diameter. Seed dark brown; trichomes adhere to

form ridges on testa; 2,5— 4,5 mm in diameter.

T. pondoensis is found as a twiner in the

undergrowth under trees in woodland. The species

occurs in the Transvaal and Natal (Fig. 12).

Fig 12. — - #, Thunbergia pondoensis; ▲, T. neglecta.

Characterized by its hastate leaf base, the species

flowers from December to March.

The specific epithet pondoensis refers to the area,

Pondoland (eastern part of the Transkei), but the

species is not known from this area. The locality

given by Clarke (1912) and Wood (1909) for Wood

4015, the lectotype, is Entumeni in Natal and not

Inyesane as stated on the label of the specimen in

NH.

7. Thunbergia neglecta Sond. in Linnaea 23: 89

(1850); Anderson in J. Linn. Soc., Bot. 7: 20 (1864);

C.B. Cl. in FI. Cap. 5,1: 9 (1912); Van der Schijff,

Checklist Kruger Nat. Park 86 (1969); Ross, FI.

Natal 323 (1972); Compton, FI. Swaz. 547 (1976).

Type: Transvaal, Magaliesberg, Zeyher 1420 (Herb.

Benth. K, lecto.!).

T. hirta Sond. in Linnaea 23: (1850); Lindau in Bot. Jb. 17

(Beibl. 41): 36, 38 (1893). Type: Orange Free State, Groot Vet

Rivier, Zeyher 1419 (Herb. Benth. K, iso.; PRE, photo!).

T. dregeana sensu Van der Schijff, Checklist Kruger Nat. Park

86 (1969); sensu Compton, FI. Swaz.: 546 (1976).

Scrambler, pubescent. Leaves petiolate, petioles

9-30 mm long; blade ovate, (12—) 29-47 (—57) x

(13—) 20—47 (—55) mm, apex acute, obtuse or

emarginate, base cordate, margin entire, lobed or

toothed. Flower bracts persistent; 11—20 x (5 — )

114

THE GENUS THUNBERGIA IN SOUTHERN AFRICA

8—14 mm. Calyx with lobes 2—6 mm long; lobes

subuliform, sometimes toothed, stipitate glands and

a few unicellular trichomes present. Corolla with

inner surface and outer surface of lobes and outer

surface of tube dull white, inner surface of tube dull

yellow; lobes 9,5 — 14,5 x 13—23 mm; tube 17-27

mm long, 2—3 mm in diameter. Filaments 4—6 mm

long; anther thecae 2,4-3 mm long. Style 7-12 mm

long, unicellular trichomes present on upper part.

Fruit 13—20 mm long, globose base 6—10 mm in

diameter. Seed reddish brown; trichomes adhere

together to form comb-like protuberances on testa;

3.5— 5,0 mm in diameter.

Woodland, thornveld and grassland are natural

habitats of T. neglecta. The plants usually occur in

loam soil under trees and shrubs, but they do grow in

direct sunlight between the trees and shrubs. T.

neglecta is the only species of the genus Thunbergia

occurring in the Orange Free State. It is also found

in Botswana, Transvaal, Swaziland, Natal and the

Cape Province (Fig. 12).

In 1850 Sonder described two new species,

namely, T. hirta and T. neglecta. Anderson (1864)

regarded them as conspecific and chose the epithet

neglecta for the species, therefore, according to

Article 57.2 of the International Code of Botanical

Nomenclature (1978), his choice must be followed.

8. Thunbergia amoena C.B. Cl. in FI. Cap. 5,1:

9 (1912). Type: Transvaal, Houtbosch Mountain,

Nelson 498 (K, lecto.!; PRE!).

Creeper with a densely pubescent, velvet ap-

pearance. Leaves petiolate, petioles (8—) 12—48

(-65) mm long; blade ovate, (31—) 40—64 (—79) x

(28—) 36—62 (-76) mm, apex acute to obtuse, base

cordate, margin entire or toothed. Flower bracts

persistent; 14-20 x 10-20 mm. Calyx covered with

short stipitate and pinlike glands; lobes 3-6 mm

long. Corolla with inner and outer surface of lobes

and outer surface of tube lemon yellow, inner

surface of tube dull yellow; lobes 10-16 x 14-17

mm, tube 19-29 mm long and 2—10 mm in

diameter. Filaments 5-6 mm long, anther thecae

2. 5- 3, 2 mm long. Style 11,4-13 mm long;

Fig. 13.— •, Thunbergia amoena; ▲ , T. dregeana.

unicellular trichomes present on upper part. Fruit

16-18 mm long, globose base 6-7 mm in diameter;

stalk 16-24 mm long. Seed brown; trichomes adhere

to form ridges on surface; 3-5 mm in diameter.

T. amoena occurs mainly along the margins of

inland forests, but is also found along the margins of

pine plantations and against rocky slopes where the

forest changes to mountain grassland. It fQrms dense

mats on the pine needles and rocks. T. amoena is

endemic to the Transvaal (Fig. 13).

Characteristics of T. amoena are the dense

pubescence of the plant and the lemon-yellow colour

of the corolla. The species flowers from November

to February.

9. Thunbergia dregeana Nees in Linnaea 15: 352

(1841); Presl, Bot. Bemerk. 94 (1844); Nees in DC.,

Prodr. 11: 58 (1847); Drege in Linnaea 20: 200

(1847); Anderson in J. Linn. Soc., Bot. 7: 20 (1864);

Sim, FI. Kaffraria 60 (1894); Wood, Natal Plants 3:

t.280 (1902); Wood in Trans. S. Afr. phil. Soc. 18:

203 (1909); C.B. Cl. in FI. Cap. 5,1: 10 (1912);

Bews, FI. Natal & Zululand 192 (1921); Fourc.,

Check List Flow. PI. George, Knysna, Humansdorp

& Uniondale 6 (1941); Chittenden in Diet. Gard.

2106 (1951); Batten & Bokelmann, Wild Flow. E.

Cape 141 (1966); Ross, FI. Natal 323 (1972);

Gibson, Wild Flow. Natal t.97, f. 33 (1975); Jeppe,

Natal Wild Flow. 92 (1975). Type: Cape Province,

Fort Beaufort, Drege s.n. (K, lecto. ! ; PRE, photo!).

T. fragrans sensu E. Mey. in Drege, Zwei Pfl. Doc. 141, 226

(1843); Nees in DC., Prodr. 11: 58 (1847). Type: Cape Province,

sin. loc., Eckl. 38b (B).

Herbaceous, pubescent, creeper. Leaves petiola-

te, petioles (2—) 15—60 mm long, blade ovate,

13—53 (—70) x 14—50 (—70) mm, apex acute; base j

cordate, margin toothed. Flower bracts caducous;

13-19 x 8-20 mm. Calyx covered with short

stipitate glands; lobes 3,5-6 mm long, sometimes

branched at the apex. Corolla with inner and outer

surface of lobes and outer surface of tube dull

orange yellow, inner surface of tube dull yellow,

lobes 13-25 x 13—27 mm, tube 16-26 mm long and

2,6-6 mm in diameter. Filaments 4-5,7 mm long;

anther thecae 3,6-4, 6 mm long. Style 10—13 mm

long. Fruit 15—24 mm long, base 7 — 10 mm in

diameter; stalk (19-) 39-70 (-100) mm long. Seed

greyish brown; trichomes adhere to form ridges on

surface; 4-6 mm in diameter.

T. dregeana is found along streams and the

margins of forests. The species occurs along the

eastern coast of Natal and in the Cape Province as

far south as the Gamtoos River (Fig. 13).

This species is characterized by a dull orange-

yellow corolla with large lobes. It flowers from

November to February. A strong decoction of T.

dregeana is used by the Zulu people as a remedy for

venereal diseases (Watt & Breyer-Brandwijk, 1962).

Medley Wood (1902) mentions that local Asian

women, like the Zulus, prepare a hairwash from the

green fruits of T. dregeana.

10. Thunbergia alata Sims in Bot. Mag. t.2591

(1825); Lodd., Bot. Cab. t.1045 (1825); Hook.,

E. RETIEF AND W. F. REYNEKE

115

Exot. FI. t.177 (1827); Nees in DC., Prodr. 11: 58

(1847); Klotzsch in Peters’ Reise Mossamb. 196

(1862-1864); Anderson in J. Linn. Soc., Bot. 7: 19

(1864); Lindau in Bot. Jb. 17 (beibl. 41): 40 (1893);

Lindau in Engl., Pfl. Ost. Afr. C. 366 (1895);

Burkhill in FI. Trop. Afr. 5: 16 (1899); Wood, Natal

Plants 3: 6, 300 (1902); C.B. Cl. in FI. Cap. 5,1: 10

(1912); Bews, H. Natal & Zululand 192 (1921);

Bailey, Cycl. Hort. 3338 (1929); Chittenden in Diet.

Gard. 2105 (1951); Eliovson, The Complete

Gardening Book for S. Afr. 36, 131 (1960); Benoist

in FI. Madag. 182: 10 (1967); Ross, FI. Natal 323

(1972); Kruger, Borne, Struike en Rankplante 514

(1973); Jeppe, Natal Wild Flow. 92 (1975); Plowes &

Drummond, Flow, of Rhodesia 129, t.170 (1976).

Type: Zanzibar, Bojer s.n. (K, holo.!, PRE,

photo!).

Herbaceous creeper; pubescent. Leaves petiolate,

petioles ( 10 -) 22-50 mm long, winged; blade ovate

(16-) 21-43 (-63) x (10-) 17-32 (-48) mm, apex

acuminate, acute or obtuse, base cordate or hastate,

margin entire or toothed. Flower bracts 13—19 x

10—16 mm. Calyx with sessile and stipitate glands.

Corolla with inner and outer surface of lobes and

outer surface of tube orange, inner surface of tube

dark purplish black with white stripes at base; lobes

7,5—13,5 x 11—20 mm. Filaments 4—5,7 mm long;

anther thecae 3,6— 4,6 mm long. Style 8,5 — 11 mm

long; stipitate glands present. Fruits 15-18 mm

long, globose base 7,5-9 mm in diameter; stalk

(13-) 35—64 mm long. Seed brown; some trichomes

adhere to form wing-like protuberances; 2,4—4 mm

in diameter.

T. alata is found along margins of forests where it

is humid and the soil is humic, but also in open

sunlight and more sandy soil. The species is

indigenous to east Africa and occurs southwards to

the eastern Transvaal and Natal (Fig. 14).

The species is characterized by the colour of the

corolla, winged petioles and the trichomes of the

seed surface forming wing-like protuberances. T.

alata flowers nearly the whole year round.

Fig. 14. — •, Thunbergia alata; ▲ , T. aurea.

T. alata is known as the Black-eyed Susan,

because the inner surface of the corolla tube which is

usually a purplish black colour, surrounded by the

bright orange corolla lobes, reminds one of an eye.

It is cultivated in many tropical and subtropical

parts of the world where the species often becomes

naturalized. Different colour variations occur. In

1839 Paxton described a species known as T.

aurantiaca, darker in colour and with larger flowers

than T. alata. The author mentions that the seed of

this species cultivated in England was from the Cape

of Good Hope. As far as known, only the form with

flowers having an orange corolla with the inner

surface of the corolla tube purplish black, is

indigenous in southern Africa. Invisible ultra violet

patterns on the fresh corolla of T. alata but visible as

fluorescent patterns in pressed herbarium speci-

mens, were observed by Eisner (1973).

11. Thunbergia aurea N.E. Br. in Kew Bull.

1909: 127 (1909); P.G. Meyer in Prodr. FI. S.W.

Afr. 129 (1968). Type: Botswana, Kwebe-Hills,

Lugard 114 (K, lecto.!; PRE, photo!).

Perennial herb with erect or trailing stems,

pubescent. Leaves petiolate, petioles (7-) 24-43

mm long, winged; blade lanceolate, (29-) 38—97

(-114); x (15—) 37—52 (—75) mm, apex acute or

acuminate, base hastate or truncate, margin entire.

Flower bracts 15-21 x 8-17 mm, papery thin with a

conspicuous venation and longer than the mature

fruit. Calyx 1,4— 4,2 mm long; sessile and stipitate

glands as well as uni- and multicellular trichomes

present. Corolla with inner and outer surface of

lobes and outer surface of tube dull orange-yellow,

inner surface of tube dull white; lobes 1,9— 4,5 mm

x 3—5,3 mm, tube 9-17 mm long and 1,4-5 mm in

diameter. Filaments 3,4-5 mm long, anther thecae

1,4— 2 mm long. Style 5,5— 6,5 mm long. Fruit 14-20

mm long, globose base 6-9 mm in diameter; stalks

(13—) 22—58 mm long; densely pubescent. Seed

brown; testa distinctly visible in spaces between

trichomes, trichomes adhere to form papillae-like

protuberances; 2,4-4 mm in diameter.

T. aurea occurs in three of the veld types

distinguished by Giess (1971) in South West

Africa/Namibia, namely: (i) mountain savanna and

karst veld, (ii) thorn savanna; and (iii) tree savanna

and dry forest. It is found in sandy soil under trees

and shrubs or against the slopes of dolomite hills.

The species is also known from Botswana and the

Transvaal (Fig. 14).

The species is characterized by its small size and

the dull orange-yellow colour of the corolla, a

winged petiole like T. alata and the trichomes of the

seed surface adhering to form papillae-like protube-

rances. T. aurea flowers from January to March.

According to De Winter (specimen no. 4159) the

leaves of T. aurea are used as spinach, and Rodin

(specimen no. 9105) mentions that the leaves are

dried and then boiled. The decoction obtained, is

used as a remedy against coughing.

116

THE GENUS THUNBERGIA IN SOUTHERN AFRICA

IMPERFECTLY KNOWN TAXA

T. humilis Eckl. & Zeyh., Cat. Sem. PI. Cap.: no

type.

T. capensis Retz. var. grandiflora Nees in DC.,

Prodr. 11: 55 (1847): the type has not been found.

T. stenophylla C.B. Cl. in FI. Cap. 5,1: 6 (1912):

the type specimen in the British Museum (Natural

History) is very poor and throws no light on the

identity of the taxon.

UITTREKSEL

’n Hersiening van die genus Thunbergia in

suidelike Afrika is onderneem. Elf spesies word

erken.

REFERENCES

Anderson, T., 1864. An enumeration of the species of

Acanthaceae from the continent of Africa and the adjacent

Islands. J. Linn. Soc., Bot. 7: 13-54.

Batten, Auriol, & Bokelmann, Hertha, 1966. Wild flowers of

the eastern Cape. Cape Town: Books of Africa.

Bor, N.L. & Raizada, M.B., 1941. Some beautiful Indian

climbers and shrubs. J. Bombay Nat. hist. Soc. 42: 685-697.

Clarke. C.B., 1912. Acanthaceae. In T.-Dyer, FI. Cap. 5,1:

1-92.

Eisner, T., 1973a. Flower markings that lure bees now seen by

man. Bioscience 23: 380.

Eisner. T., 1973b. Plant taxonomy: ultraviolet patterns of flowers

visible as fluorescent patterns in pressed herbarium

specimens. Science 179: 486-487.

Giess, W., 1971. ’n Voorlopige plantegroeikaart van Suidwes-

Afrika. Dinteria 4: 5-114.

Hulme, Mairn. 1954. Wild flowers of Natal. Pietermaritzburg:

Shuter & Shooter.

Lindau.G., 1893a. Acanthaceae africanae. Bot. Jb. 17: 89-113.

Lindau, G., 1893b, Ubersicht liber die bisher bekannten Arten

derGattung Thunbergia L.f. Bot. Jb. 17 (beibl. 41): 31-43.

Lindau, G., 1898. Acanthaceae africanae. Bot. Jb. 24: 310—312.

Linnaeus fil., C., 1781. Supplementum plantarum.

Braunschweig.

Montin, L., 1773. Thunbergia. K. svenska VetenskAkad. Handl.

34: 288.

Paxton, J., 1839. Thunbergia aurantiaca. Paxt. Mag. Bot. 6: 269.

Retzius, A.J., 1780. Thunbergia. Phys. Saellsk. Handl. 1:

163-164.

Sonder, W., 1850. Thunbergia hirta. Linnaea 23: 88.

Sonder, W., 1850. Thunbergia neglecta. Linnaea 23: 89.

Watt, J.M. & Breyer-Brandwuk, Maria, 1962. The medicinal

and poisonous plants of southern and eastern Africa.

Edinburgh: Livingstone.

Wood, J.M., 1902. Thunbergia dregeana. In J.M. Wood, Natal

Plants 3, t.280.

Wood, J.M., 1909. Revised list of the Flora of Natal. Trans. S.

Afr. phil. Soc. 18: 203-204.

APPENDIX: SPECIMENS EXAMINED

The specimens are listed alphabetically according

to the name of the collector. The figures in brackets

refer to the number of the taxon in the text.

Acocks 9898(2), 10799(7), 10966(5), 11756(3), 12554(4),

13049(3), 13298(3), 13317(9), 13584(4), 17883(4), 21962(3);

Archibald 5572(4), 6083(4). Bachmann 1267(3); Bandert 41(3);

Barker 1504(4), 2732(4), 2876(3), 6136(3), 7909(9); Barrett

154(3); Baur 58(3), 169(3); Bayliss 1706(3), 2518(1); Blenkiron

sub J 16128(3) ; Bokelmann 1(1); Bos 1263(8); Botha & Van Wyk

1063(3); Bourquin & Fakude 17(3); Bredenkamp 368(7), 526(7),

1282(7), 1308(7), 1324(7); Bremekamp & Schweickerdt 155(7);

Breyer sub TRV 17895(6); Brown & Shapiro 450(7); Buitendag

731(3), 1026(3), 1027(6); Burn Davy sub PRE 10637(3). Codd

1804(3), 4166(11), 4784(7), 6059(6), 7954(8); Codd & De Winter

3081(1), 4950(3); Codd & Dyer 9188(8); Coetzee 1523(3); Comins

368(9); Compton 17652(4), 17659(4), 23803(3), 27102(3),

28327(1), 28382(1), 28424(7), 30902(7); Cooper 893; Crawford

291(3), 344(3); Curson 953(11). Davidson 1591(5); Dekker 18(2);

De Sousa 431(3), 506(3), 641(3); Devenish 115(3), 289(3); De

Winter 2227a(7), 2911(11), 3008(11), 4187(11); De Winter & Wiss

4159(11); Dinter 878(11), 2439(11); Dodds 70(3); Downing

598(7), 606(7); Du Plessis 614(3); Du Preez 369(3); Duthie

744(4); Dyer 4287(1). Edwards 1703(1), 1769(9), sub J17545(3);

Elan-Puttick 204(3), 227(7). Fakude 89(3); Flanagan 814(4),

1130(5), 1747(9); Fourcade 3673(9). Galpin 217(4), 496(3),

1277(3), 3399(1), 7689(3), 7747(9), 9662(3), 10973(3), 11439(9),

12038(7), 12086(3), 12252(3), 14817(7), sub BOL 31279(3), sub

BOL 31280(7), sub BOL 31282(8), sub BOL 31284(1), sub TRV

14387(7), 14770(7); Geldenhuys 601(5); Gemmell 6569(3);

Germishuizen 923(3), 1711(1), 1799(3); Gerrard 1274(3);

Gerstner 2506(10); Gertenbach 5076(7); Giess 9959(11); Gies &

Muller 11820(1); Giffen 812(9); Gilmore 2335(3); Goossens

1640(7); Gordon-Gray 1033(1), 1273(5). Haagner sub TRV

10749(7); Hansen 3515(7); Harrison 188(3); Hemm 304(1);

Henderson , Brokenshoe & Collins 72(9); Henrici 1693; Hilliard

4720(9); Hilliard & Burn 5648(5), 5920(3), 7387(3); Hitchins

78(3), 387(3), 492(3), 624(3), 881(3); Holt 208(8), 346(1); Huntley

732(1), 1363(3). Ihlenfeldt 2099(7). Jacobsen 77(1), 1195(1),

1225(3), 1664(7), 2135(7); Jacobsz 778(3), 1785(3); Jarman &

Guy 153(9), 208(5); Jenkins sub TRV 6752(7); Johnson 383(3),

805(4); Junod 2360. Kelaole A86(7); Kemsley 202(4); Kerfoot

7269(9); Kers 841(11); King 283(3), 326(3); Kinges 2955(11);

Kluge 312(1), 1084(1); Krantz 6850(7); Krauss 405(3). Lang sub

TRV 31081(7); Lawn 1321(1), 622(10); Leendertz 754(3),

5961(7); Lewis 4469; Liebenberg 8015(3); Lindahl 2(3); Lugard

107(11), 114(11). Maguire 2305(11); Markotter sub STE 19956(3),

sub STE 19991(8); Mauve 1056(3), 4932(9); McDonald 100(3),

160(3), 421(3); Meeuse 9176(1), 9227(8), 9407(3); Meldrum 97(7);

Mogg 3437(2), 8943(7), 12666(3), 13516(9); Moll 1039(3),

1487(3), 2131(3), 2141(3), 2456(9), 2853(1), 2926(9), 3100(9),

3218(9), 4136(7), 4728(3); Moore 12(4); Morris 776(9); Morris &

Engelbrecht 16(7); Muller 1643(7), 2063(3), 2097(1), 2206(2). Nel

227(3); Nelson 498(8); Nicholson 1714(9), 1790(9), 1801(3). Oates

94(7); Obermeyer 367(6), sub TRV 3042; Oliver 67(1), 583a(3);

Olivier 1989(4); Onderstall 299(6). Papendorf 263(1); Pentz

521(9); Pole Evans 15802(3); Pooley 136(3), 1073(3); Prosser

1513(3), 1654(3). Repton 197(3), 832(7); Relief 100(3), 103(3),

104(3), 110(3), 111(3), 112(3), 116(3), 117(1), 118(10), 120(3),

124(3), 142(3), 151(7), 152(3), 154(1), 163(8), 164(3), 178(8),

179(3), 180(1), 182(8), 189a(l), 190(3), 190a(3), 191(3), 192(3),

200(3), 201(9), 205(3), 209(3), 210(3), 212(3), 213(5), 214(3),

215(3), 217(3), 218(5), 220(1), 222(3), 227(3), 228(1), 236(3),

237(3), 244(9), 245(9), 254(4), 862(9), 1143(6); Relief & Herman

126(7); Rodin 9105(11); Rogers 198(4), 18042(8), 18951(8), sub

BOL 11738: Ross 1468(3); Ross & Moll 2271(9); Rudatis 27(3),

479(1), sub STE 2397(5); Rump sub NH 20323(2). Salberg 10(3);

Sandeman sub NU 58797; Sanderson 442(5); Scheepers 89(8),

850(3), 1033(1); Schlechter 2908(2), 6561(3), 21342(7); Schlieben

9382(3), 9551(10); Schlieben & Strey 8301(8); Schonland 3820(4),

3903(3); Schweickerdt 971(3); Seydel 2149(11); Shirley 169(3), sub

NU 31686(3); Sidey 131(3); Sim 1391(4); Smuts 380(7), 394(3);

Smuts & Gillen 3053(7), 3219(8), 3393(3); Stephen 2631(7);

Stewart 112(3); Steynberg 724(8); Stirton 111(3), 347(2), 376(3),

727(3), 1010(3), 1023(3), 1025(3), 1092(3), 1095(3), 1103(3),

1128(3), 1116(3), 1135(3), 1136(3), 1137(3), 1209(3), 1211(3),

1215(3), 1218(2), 1220(3), 1232(2), 1245(2), 1261(3), 1277(3),

1302(3), 1318(3), 1327(3), 1347(3), 1349(3), 1354(3), 1360(3),

1370(3), 1380(3), 1386(3), 1399(3), 1425(1), 1435(1), 1714(3),

1734(3), 5143(3), 5615(3); Story 585(3), 4985(3), 6305(3); Strey

4195(5), 4300(9), 5060(9), 5620(3), 6208(1), 6403(3), 7131(1),

8015(3), 9192(1), 9392(3), 9418(1); Sutton 8103(3). Taylor

5253(3); Theron 1078(3), 1598(9), 1797(9), 1993(1); Thode

A273(l), sub STE 2745(10); Thunberg sub Herb. Thunb. 14601;

Tyson 1206(3), 2559(9). Vahrmeijer 607(9), 1577(7); Vahrmeijer

& du Preez 2495(11); Vahrmeijer & Tolken 277(3); Van der Schijff

8(1), 1222(7), 3255(7), 5895(1), 5916(1); Van der Walt 321(3);

Van Greuning 263(11); Van Hoepen 1738(3); Van Jaarsveld

1079(8), 1098(6); Van Vuuren 1332(3), 1835(3); Van Wyk

1402(7), 5094(1); Verdoorn 140(3). Walters sub PRE 10759(7);

Ward 571(9), 942(9), 3517(7); Watmough 402(9); Wells 2018(3),

3124(4); Werdermann & Oberdieck 1738(7), 2013(7), 2124(1);

Wild & Drummond 6971(11); Williams sub TRV 7677(7); Wood

1(3), 284(1), 543(9), 643(2), 696(2), 1218(5), 4015(6), 6376(9),

sub SAM 18933; Wylie sub NH 28009(2), sub Wood 8478. Young

303(3), 372(1), 489(3), A 545, A 655. Zeyher 1418(3), 1419(7),

1420(7); Zwanziger 686(11).

Bothalia 15, 1 & 2: 117-124 (1984)

Studies in the liverwort genus Riccia (Marchantiales) from the

south-west Cape

O. H. VOLK* and S. M. PEROLD**

Keywords: dorsal epithelium, Pilifer, Riccia species, spores, sporophytes

ABSTRACT

A new species of Riccia, R. parvo-areolata Volk & Perold, as well as sporophytes and spores of R. villosa Steph.

ex Brunnthaler, are described. Earlier descriptions of R. villosa were based on sterile plants. The unique structure

of the dorsal epithelium consisting of loose cell pillars in these two species (and also present in allied species

grouped together in the section Pilifer) is described and illustrated.

Riccia parvo-areolata Volk & Perold, sp. nov.

Dioica (?), perennis, singulares vel subgregaria.

Gametophytum mediocra ad 10 mm longa, ad 5 mm

lata, ad 1,3 mm crassa, oblonge-ovata, in sicco

subalbido-viridis, fissurata-tomentosa, in tumido

pallide-viridis, subnitida, velutina; furcis brevibus,

divergentibus, lata rotundato-truncatis breviterque

emarginatis, apice sulcatis, ceterum convexis margi-

nibus deflexis, costa subplana, viridis vel rare

purpureo-maculata, in sicco apicibus contractis,

squamis obvelatis breviterque sulcatis ceterum lata

concavis, marginibus elevatis, subtus lunatis. Squa-

mae excedentes hyalinae, ad 600 pm magnae,

semiorbicularae, basi plus minusve rare purpureae,

confertae, integerrimae, cellulae longiusculae, ad

120 x 50 pm magnae, 5—7 angulosae, parietibus

rectis. Frondis adultis sectio transversalis fere (2—)

3-4-plo latioribus quam altis, stratum piliferum

(epithelium) ca x/6, stratum aeriferum 3/6— 4/6,

stratum penarium 2/6 altitudine frondis; pili liberi,

hyalini, conferti, (3-) 4 cellulae tumidae seriatim,

ad 200 pm longi, 60-90 pm crassi, cellulae

terminales obconici vel mamillati; columnae strati

aeriferi ad 600 pm longae, 8—12 cellulae seriatae;

canalabis aeriferibus 4—7 (—8) angulati, perpendi-

cularibus vel gradatim extrorsis inclinatis et apice

versus oblique adscendentes, canalos aeriferos

includentes. Antheridia et archegonia non visa.

Sporangia matura libera, 400-900 sporae continen-

ts; illae fuscae, subtriangularo-globosae, 70-85-96

pm in diametro, alae angustae, minutissime crenula-

tae, poribus inconspicuae, distaliter parvo-areolata,

inde nomen, 18—24 foveoli in diametro, anguli

reticuli minute papillatae, proximaliter tres facettae

parvo-areolatae subdistinctae limitatae. Chromoso-

mata n=8 (Bornefeld) Riccia albomarginatae subsi-

milis, sed differet inter aliis ab colore pallido- viridis,

cellulae pilorum crassiorae et ornatio sporarum

parvo-areolata. Habitat in pascius (?) temporaliter

humidis, subarenosis, acidis, expositis radiatonis.

TYPE. — Cape, 3318 (Vanrhynsdorp): near Do-

ringbaai, W of Vredendal, Kliphoek farm, gravel on

sandstone (—CD), J. M. Perold 23 (PRE, holo.).

* Botanische Anstalten d. Univ., Wurzburg D 8700, Germany.

Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

Dioecious?, perennial, single or in loose patches.

Gametophyte medium-sized, segments up to 10 mm

long, 5 mm broad, 1,2 mm thick; shape broadly

oblong-ovate; seldom branching, and then branches

short, divergent; dorsal surface when wet pale

green, slightly glistening, velvety, apex roundly

truncate and shortly emarginate, grooved apically,

otherwise convex with downward sloping margins

(Fig. 1.2, 1.3); in dry state dorsal surface

greenish-white, felt-like, apex inflexed and groove

covered by scales, otherwise broadly concave with

margins ascending, nearly acute (Fig. 1.1); ventral

surface slightly convex, green, occasionally flecked

with violet. Scales semi-circular (Fig. 1.10) projec-

ting above margin, size up to 600 pm hyaline, base

flecked with wine-red, imbricate, erect, margin

smooth, with cells elongate, 5—7 sided, straight-

walled, size up to 120 x 50 pm. Thallus in transverse

section (2-) 3-4 times broader than high,

epithelium about 1/6, assimilation tissue 3/6 to 4/6,

storage tissue 2/6 the height of the thallus (Fig. 1.3);

dorsal cell pillars (Fig. 2.2, 2.5, 2.7) free, hyaline,

about 200 pm long with (3-) 4 inflated cells 60—90

pm wide, terminal cell mamillate or bluntly conical;

assimilation tissue about 600 pm thick, columns

8—12 cells deep, enclosing 4—7 (— 8)-sided air

canals, rising vertically or frequently sloping towards

the sides or the apex (Fig. 1.3); the canals in the

centre of the thallus narrow (Fig. 1.8) and enclosed

by four cell columns, while those nearer the margin

are broader and enclosed by 5-8 cell columns (Fig.

1.9); storage tissue cells often with oil droplets.

Antheridia and archegonia not seen. Sporangia

when ripe he freely in the decaying thallus and

contain about 400-900 spores. Spores (Fig.

3. 1-3. 6) rounded, triangular-globular, brown, size

70-80-90 pm with narrow crenulate wing and

inconspicuous pores at marginal angles, distal face

with ornamentation finely areolate (hence its name),

18—22 areolae across the diameter, the corners of

the areolae with short papillae; proximal face with

three finely areolate facets, not sharply demarcated

(Fig. 3.1).

Chromosome number: n=8, (Fig. 1.11, Bome-

feld). Their shapes are similar to those of R.

albomarginata and R. albovestita (habil T. Bome-

feld, pers. comm).

Riccia parvo-areolata colonizes exposed open

118

STUDIES IN THE LIVERWORT GENUS RICCIA (MARCH ANTI ALES) FROM THE SOUTH-WEST CAPE

2 mm

i »

100 pm

Fig 1. — Riccia parvo-areolata ( Perold 23, PRE). Structure of the thallus. 1, two dry thalli; 2, fresh thallus; 3, cross section of the

thallus; 4, thallus apex from below; 5, thallus apex from above; 6, epithelium marginally; 7, epithelium centrally; 8, air canals

at centre of thallus (chloroplasts adjacent to canal walls) seen from above; 9, air canals at margin of thallus; 10, ventral scale;

11, chromosomes; 12, distribution of Riccia parvo-areolata. (1, 2, 4, 5, 6-10 by O. H. Volk; 3 by R. Holcroft; 11 bv habil T.

Bornefeld.)

O. H. VOLK AND S. M. PEROLD

119

Fig. 2. — Riccia parvo-areolata (Perold 23, PRE). Epithelium and assimilation tissue. 1, thallus; 2—5, different shapes of the terminal

cells of the epithelium; 6, air canals widening by the inclusion of new cell pillars; 7, epithelium on cross section. (1-5 SEM

micrographs by O. H. Volk; 6—7 LM photographs by O. H. Volk.) Scale bar = 50 pm.

stands in veld grazed by sheep, well-watered during

infrequent rainfall, on weakly acid (pH 5,2— 5,8)

shallow sandy loam soil, together with moss species

Bryum argenteum Hedw., Bryum bicolor Dicks,

Oedipodiellia australis (Wager & Dix.) Dix.

Apparently the species is endemic to the western

Cape. Together with R. albovestita Volk, R.

albomarginata Bisch. emend. Sim, R. villosa Steph.,

R. concava Bisch. emend. S. Amell and other as yet

undescribed species ( R . sarcosa ined., R. duthieae

ined.) it belongs to a group of species which Volk

(1983) has grouped together in the section Pilifer.

They are characterized by a dorsal covering

epithelium of free hyaline hair-like cell pillars (Figs

1.6, 1.7 & 2.7). This epithelium is mostly very fragile

and can hardly be preserved. It is therefore rarely

reconstituted in herbarium material and then only by

using chloral hydrate, lactophenol or KOH. Even

when fresh material is transferred to dilute alcohol,

these cell pillars collapse. The cell walls are also not

easily wettable and the resistance to water remains

even after fixation in a mixture of alcohol-acetic acid

(Volk, 1984). The composition of the cell walls is

unknown. The zinc chloride-iodine reaction for

cellulose is negative.

These free dorsal pillars are lacking in species of

the Lichenoides, where the dorsal epidermal cells

are not free standing and almost always arranged in

parallel lines — an arrangement which is revealed in

species of the Pilifer group only on horizontal

sections through the chlorenchyma (Fig. 1.8, 1.9).

The structure of the epithelium was studied with the

light microscope (LM) as well as with the scanning

electron microscope (SEM) using living material

120 STUDIES IN THE LIVERWORT GENUS RICCIA (MARCHANTIALES) FROM THE SOUTH-WEST CAPE

Fig. 3. — Riccia parvo-areolata ( Perold

23, PRE). Spores. 1, proximal face;

2, apex; 3, distal face; 4, side. (SEM

micrographs by S. M. Perold); 5,

side; 6, distal face. (LM photo-

graphs by O. H. Volk.) Scale bar =

50 pm.

prepared by the usual procedure of critical point

drying. At low magnification tangled, furry dorsal

cell pillars are seen from above (Fig. 2.1). At high

magnifications cells of the pillars of R. parvo-

aerolata are found to be extraordinarily variable in

shape (Figs 1.6, 1.7 & 2. 2-2. 7). The terminal cells

of the variably wide (25-90, but mostly 40-50 pm)

pillars are mamillate (Fig. 2.2), vermiculate (Fig.

2.3) to conically inflated (Fig. 2.5). Measurements of

a large number of terminal cells of the pillars from

different thalli revealed a length ; breadth ratio of

1,2: 1 , where the length varied between 25 — 49 — 75

pm, and the breadth between 24-40-77 pm

(average value italicized). The rest of the cells of the

epithelium have similar dimensions (Table 1).

Riccia parvo-areolata is somewhat similar to R.

albomarginata but differs from it in the pale green

colour, the size of the epithelial cells and the finely

areolate ornamentation of the spores.

Older collections of a number of R. parvo-areolata

specimens were incorrectly identified as R. concava.

This is not surprising as in the dry state the thallus is

pronouncedly concave (Fig. 1.1, 1.4) (Sim, 1926). R.

parvo-areolata differs from R. concava in that it

lacks radially arranged ridges in the ornamentation

of the outer face of the spore, nor do the epithelial

cells swell out again once they have collapsed.

Cape. — 2917 (Springbok): Hester Malan Res. nr Springbok

(-DB), Schelpe 7759, 7776 (BOL). 3118 (Vanrhynsdorp):

Nortier Exp. Farm nr Doringbaai (-AB), Perold 14, 15 (PRE);

Kliphoek Farm, nr Doringbaai (-CD), Perold 19, 22 -26 (PRE).

3219 (Wuppertal): Wuppertal Miss. Sta. (-AC), Malherbe &

Davies 5375 (BOL). 3220 (Sutherland): nr Sutherland ( — BC),

Duthie 5407 (BOL). 3319 (Worcester): Op-die-Berg (-AB),

Stirton 9169 (PRE). 3418 (Simonstown): Knorhoek (-BB),

Duthie 5414 (BOL).

Riccia villosa Steph. ex Brunnthaler in

Denkschr. Kais. Akad. Wiss. Math.— Nat. Kl, 88:

TABLE 1. — Measurements of the cells of the epithelium of Riccia parvo-areolata (in pm) ( Perold 23)

O. H. VOLK AND S. M. PEROLD

121

Fig 4. — Riccia villosa. Structure of the thallus. 1, thallus; 2, cross section; 3, ventral scale; 4, apex of scale; 5, toothed margin of

scale and striolate cuticle; 6, epithelial cells; 7, ornamentation of distal spore face (LM); 8, chromosomes; 9, distribution of

Riccia villosa. (1, 2, 3 from Stephens 24726, BOL (by R. Holcroft); 4, 5 from leg. unknown 63, BOL (by O. H. Volk); 6, 7 from

Perold 20, PRE (by O. H. Volk); 8, from Perold 20, PRE (by habil T. Bornefeld).

122 STUDIES IN THE LIVERWORT GENUS RICCIA (MARCH ANTI ALES) FROM THE SOUTH-WEST CAPE

Fig 5. — Riccia villosa (Perold 20, PRE). Epithelium and assimilation tissue. 1, end on view of thallus; 2, epithelial cells and scales;

3, side view of epithelial cell pillars; 4, the same from above. (SEM micrographs by O. H. Volk.) Scale bar = 50 pm.

724 (1913); Arnell in Hepaticae of South Africa: 19,

20 (1963). Type: Kapland, 3320 (Montagu): Karoo

bei Matjiesfontein, auf sandigem Boden, zirka 900

m (-BA), Brunnthaler, s.n. XI, 1909 (G) Iconogr.

Hepat. (G).

Dioecious?, perennial, in crowded gregarious

patches. Gametophyte small, segments up to 8 mm

long, 2,5 mm broad and 1,5 mm thick; shape

narrowly oblong-ovate; single or furcate with

branches diverging at an acute angle (Fig. 4.1); when

wet dorsal surface whitish-green, apex bluntly

obtuse; in dry state upper surface white, plane to

concave; margins obtuse, flanks steep, often purple

to nearly black; ventral surface light brown to

purplish black, flat to slightly convex. Scales very

prominent, projecting beyond the margin, closely

imbricate (Fig. 5.1), laterally in one row, hyaline;

when dry, white to yellowish, up to 1,8 mm long and

1 mm wide at base which is purple (Fig. 4.3— 4.5);

shape triangular-acuminate, margins denticulate

apically, frequently ending with a narrow caducous

terminal cell; most cells oblong-hexagonal, larger at

the base and up to 110 pm long and 40 pm wide, with

straight thickened walls, cuticle striolate. Thallus in

transverse section 1-2 times broader than high (Fig.

4.2), epithelium about Vs to 2/s, assimilation tissue

Vs to 2/5, storage tissue Vs the height of the thallus

(Fig. 4.2); dorsal epithelium consists of colourless,

nearly cylindrical free cell pillars (Figs 4.6 &

5. 2-5. 4) 300 - 700 pm long, each pillar composed

of (4-) 5 (-6) fragile, tender cells which collapse

easily, (40—) 80—130 (-200) pm long, (20—) 30—50

(—70) pm wide, widening gradually towards base of

pillars, 2—3 times longer than wide (Table 2): below

epithelium a layer of chlorophyllose assimilation

tissue about 250—400 (-600) pm deep, consisting of

O. H. VOLK AND S. M. PEROLD

123

TABLE 2. — Measurements of the cells of the epithelium of R. villosa (in pm) (Perold 20)

rows of nearly isodiametric cells 25—60 [am in

diameter, with narrow air spaces between rows; at

the base a layer of parenchymatous tissue 300—400

pm thick, consisting of polygonous cells about 40 pm

in diameter, sometimes filled with oil droplets;

rhizoids colourless, both smooth and tuberculate, up

to 40 pm thick. Antheridia not seen. Archegonial

ostioles purple, about 60 pm long. Sporangia

develop July to September singly or close together

along median part of the thallus, causing slight

bulging of dorsal surface, blotched with purple

where it covers sporangia. Spores about 350 per

capsule, released by disintegration of surrounding

tissue, shape triangular-globular, polar, wingless,

(85—) 90—110 ( — 115) pm in diameter, colour brown

to very dark brown or black (Fig. 6. 1-6.6); distal

face convex, ornamentation papillose to vermicula-

te, in a whorl which spirals from the centre outwards

to the margin in 10-15 thick ridges (Figs 4.7 & 6.5);

some spores, probably riper, have flattened ridges;

proximal face with similar sculpturing, but ridges not

in obvious spirals (Fig. 6.1 — 6.3); apex blunt,

triradiate mark not prominent, each of the three

suture lines of the mark teoninate in a pore at the

margin.

Most plants are sterile. Of eighteen specimens

examined, only four were fertile, propagation

Fig. 6. — Riccia villosa. Spores. 1,

proximal face; 2, apex; 3, 2 facetsof

proximal face; 4, pore at margin; 5,

distal face; 6, side. (1, 2, 5, 6 from

Levyns s.n. July 1954, BOL; 3, 4

from Compton 5428, BOL. (SEM

micrographs by S. M. Perold.) Scale

bar - 50 pm.

124

STUDIES IN THE LIVERWORT GENUS RICCIA (MARCH ANTI ALES) FROM THE SOUTH-WEST CAPE

usually being adventitious. R. villosa produces

globular gemmae originating from the thallus and

later perforating the epithelium of the older parts.

This species is endemic to South Africa and

known from several localities in the drier winter

rainfall parts of the Cape like the Karoo, the

Knersvlakte and Namaqualand. It grows at altitudes

of 300-1 000 m in depressions, on sandy to gravelly

non-calcareous soils, fully exposed to the sun and in

association with small moss species, Desmatodon

convolutus (Brid.) Grout, Oedipodiella australis

(Wager & Dix.) Dix., with Mesembryanthemaceae

Ruschia spp., Dactylopsis digitata (Ait.) N.E.Br.,

Argyroderma spp., Conophytum spp., and with

Crassula spp.

R. villosa, with its free epithelial cell pillars, is a

member of the section Pilifer. It is recognized by the

large hyaline triangular scales. In drying out, the

thallus margins curve inwards and the large scales

meet and project upwards along the midline. This,

together with the collapse of the epithelial cells,

protects the living tissue of the thallus. It is a

distinctive plant. A drawing of R. villosa appears on

the cover of S. Arnell’s book on South African

hepatics.

Cape. — 2917 (Springbok): Springbok ( — DB), Vogel C791

(Mainz)+; Messelpad Pass, W Heights ( — DD), Schelpe 7814a

(BOL). 3118 (Vanrhynsdorp): nr Doringbaai, Kliphoek Farm

(-CD), Perold 20 (PRE). 3119 (Calvinia): S of Nieuwoudtville

on plateau (-AC), Leg. unknown 23885 (BOL). 3218

(Clanwilliam): N of Barrage Scheme (-BB), Stephens 24726

(BOL); Pakhuis Pass (-BB), Vogel C 5548 (Mainz)+. 3219

(Wuppertal): Bidouw Valley, on Mertenhof Farm ( — AB), Oliver

1467 (BOL); Cedarberg Range E side, Welbedacht (-AC),

Compton 5428 (BOL)* *. 3319 (Worcester): Inverdoorn (-BB),

Vogel C 1351 (Mainz)+; Worcester (— CB), Leg. unknown, 5430

(BOL). 3320 (Montagu): Matjiesfontein, 3 miles W "(-BA),

Pillans 10036 (BOL); Dobbelaarskloof (-CB), Levy ns s.n.

(BOL)*; Cogman’s Kloof (-CC), Arnell 803 (BOL); nr Montagu

Baths, on veld (-CC), Page s.n. (BOL). 3321 (Ladismith): Seven

Weeks Poort (-AD), Levyns 25272 (BOL)*; Ladismith (-AD),

Vogel C 11 (Mainz)+; Ladismith Kloof (-CA), Vogel C 1349

(Mainz)+; Olifants River (-DA), Vogel C 5449 (Mainz)+; locality

and collector unknown, 63 (BOL)*.

ACKNOWLEDGEMENTS

The authors wish to thank Prof. E. A. Schelpe for

the loan of the specimens from BOL and Dr habil T.

Bornefeld of the Department of Plant Physiology,

Gesamthochschule D-3500 Kassel, Germany, for

the chromosome counts and figures. Our thanks also

to Dr H. F. Glen for his kind assistance.

UITTREKSEL

'n Nuwe spesie van Riccia, R. parvo-areolata,

sowel as sporofiete en spore van R. villosa word

beskryf. Vroeer beskrywings van R. villosa is

gebaseer op steriele plante. Die unieke bou van die

dorsale epidermis wat bestaan uit los kolomme selle

in hierdie twee spesies (ook teenwoordig in verwante

spesies wat saam gegroepeer is in die seksie Pilifer)

word beskryf en geillustreer.

REFERENCES

Sim, T. R., 1926. The bryophyta of South Africa. Trans. R. Soc.

S. Afr. 15:8.

Volk, O. H., 1983. Vorschlag fur eine Neugliederung der

Gattung Riccia. Mitt. bot. StSamml., Munch. 19 : 453—465.

Volk, O. H., 1984. Beitrage zur Kenntnis der Marchantiales in

SWA. IV. Zur Biologie einiger Hepaticae mit besonderer

Berucksichtigung der Gattung Riccia. Nova Hedwigia. In

press.

+ Vogel’s collection is housed at D-6500 Mainz, Herbarium

Botanisches Institut der Universitiit.

* fertile specimen.

Bothalia 15, 1 & 2: 125-129 (1984)

Register of names and types : a comparison between Mesembryanthe

maceae and Poaceae

G. E. GIBBS RUSSELL* and H. F. GLEN*

Keywords: computerization of taxonomic literature, Mesembryanthemaceae, Poaceae, scientific names, type specimens

ABSTRACT

A register of names and types for southern Africa would make possible a more rapid completion of the Flora of

Southern Africa. Registers for Mesembryanthemaceae and Poaceae, families which are similar in number of

species and names, but different in distribution and importance, are compared, to give guidelines for an efficient

approach to extending the register to include other families.

INTRODUCTION

An essential part of the taxonomic work for the

Flora of Southern Africa is indexing all names that

must be evaluated, finding the original description of

each name and locating the type specimens for

examination (Leenhouts, 1968). This is an extremely

time-consuming task and, if the traditional procedu-

res could be streamlined, the production of the Flora

would be more rapid.

Registers for two other regions have been started

in the past. A register of names for North America

was begun for the Flora North America project

(Shetler & Skog, 1978). However, even though

computer encoding forms for this Register made

provision for a citation for each name in the

checklist, and for synonyms with citations, these do

not appear in the published checklist. The data-

capture form published with the checklist was

examined, but was felt to be too complex for the

purpose at hand. In Australia a computerized

register has been under development for about a

decade (Burbidge, pers. comm.; Chapman, 1979).

The Australian compilers have also recorded an

extensive amount of data for each name, and have

checked each name in its primary source.

The purpose of this paper is to compare existing

registers of names in two angiosperm families of

comparable size but widely different distribution,

importance and history of taxonomic study in

southern Africa, in order to determine guidelines for

extending the register to other families in the most

efficient way.

METHODS

Registers of names and types based on or applied

to southern African plants have been prepared for

two large families now being studied for the Flora,

Mesembryanthemaceae and Poaceae. The register

for Poaceae was computerized, so that listings sorted

by name, reference or type could be obtained, as

described by Gibbs Russell (1983).

The register for Mesembryanthemaceae was

Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

prepared by hand. The names listed by Jacobsen

(1974) as valid, together with those listed in a card

index held by one of us (HFG), formed the starting

point. Literature references for these were found,

and the list typed when complete. Synonyms listed

by Jacobsen were then noted, their sources found

and the results intercalated in the appropriate parts

of the typed list. The extended list was then re-typed

and any further additions were made. Index

Kewensis was the primary source of references, but

each one was verified in the library at Kew.

Repertorium Plantarum Succulentarum and the Kew

Record were important minor sources of names and

citations. In this way, a few dozen names not in

Index Kewensis but found in Jacobsen (1974) were

noted and the relevant data handed in for the next

Supplement to the Index. Only names published

before the end of 1974 were considered for inclusion

in the Register of Names in Mesembryanthemaceae.

The numbers of names and taxa and the

references for both families are compared to show

differences between the two groups. These differen-

ces are related to differences in history of taxonomic

study.

RESULTS

Mesembryanthemaceae and Poaceae have fairly

similar numbers of taxa and names applied to these

'taxa, but the number of references in which names

of Poaceae have been published is far greater than

the number of references in which names for

Mesembryanthemaceae have been published (Table

1). On average, more than three times as many

names occur per reference in Mesembryanthema-

ceae than in Poaceae. Figures 1 and 2 show the

numbers and percentages of names found in each of

the references contributing 1% or more of the total.

In Mesembryanthemaceae, over a third of the names

are derived from a single reference, and over half

the names are found in the three most important

references. Furthermore, the minor references, with

fewer than 1% of the total, account for only 16% of

the names. In Poaceae, fewer than one eighth of the

names are derived from a single reference, whereas

the minor references contain more than 40% of the

names. The generalization can be made that the

names for Poaceae are more scattered in the

literature than for Mesembryanthemaceae.

126 REGISTER OF NAMES AND TYPES: A COMPARISON BETWEEN MESEMBRYANTHEMACEAE AND

POACEAE

Fig. 1. — Mesembryanthemaceae

major references, which each

contribute 1% or more of the

total references. The number

of names is indicated across

the top, and the percentage

contribution of each refe-

rence is given at the end of its

bar.

TABLE 1. — Comparison between Mesembryanthemaceae and

Poaceae for numbers of taxa, names and references

Table 2 shows the 31 references common to both

families. Only one reference listed occurs among the

major references for both families. Five of the major

references for Mesembryanthemaceae and eight of

the major references for Poaceae occur on the joint

list. However, the other 17 references common to

the two families are minor references in both. The

low value (11.7%) for the Sorenson (1948)

coefficient of similarity comparing the occurrence of

references common to the two families emphasizes

the differing sources of names in each. Only one

sixth of the names in Mesembryanthemaceae are

derived from common references, but over a third of

the names for Poaceae are found in these sources.

The generalization can be made that there are few

common references of importance for both families,

but that the common references are more important

as sources of names in Poaceae than in Mesembry-

anthemaceae.

DISCUSSION

The differences in occurrence of names in the

literature between the two families is the result of

their differing history of taxonomic treatment, which

has occurred because of differences in their

distribution and economic and ecological impor-

tance. Although the two families have similar

numbers of taxa, they differ in their distribution and

importance.

Worldwide, Mesembryanthemaceae is centred on

southern Africa, with a few weedy species occurring

G. E. GIBBS RUSSELL AND H. F. GLEN

127

in the Mediterranean area, Australia and the warm

parts of America, whereas Poaceae occurs on all the

continents and forms about 20% of the world’s

vegetation (Heywood, 1978). In Africa, Mesembry-

anthemaceae ranks as fourth largest family in South

West Africa/Namibia and as ninth largest family in

the Cape Peninsula. It is not among the largest

families in any other region. Poaceae is one of the

three largest families (with Asteraceae and Faba-

ceae) in all completed African Floras (Gibbs

Russell, 1975). In Mesembryanthemaceae, 115 of

the 120 genera are endemic in the Flora of Southern

Africa area, whereas in Poaceae only 12 of the 204

genera are endemic.

Mesembryanthemaceae is economically important

for species cultivated as garden ornamentals, and

one species grown as a roadside sand binder,

whereas Poaceae comprises the world’s major grain

and fodder crops, as well as the most important

pasture plants (Lawrence, 1951). A measure of the

comparative ecological importance of these two

families is indicated by Acocks (1975), as shown in

Table 3. In his study, Poaceae has more taxa, which

are listed more frequently, and occur in more veld

types than Mesembryanthemaceae.

This difference in distribution and importance

explains the difference in history of taxonomic

treatment. Poaceae has wide-spread distribution and

high economic and ecological importance, so the

family has been studied by many botanists over the

years. Also, work done elsewhere on widespread

taxa is essential for studies in southern Africa.

Therefore Poaceae has an extensive literature, and

many of the references are general accounts of the

flora, or are treatments from distant places.

Mesembryanthemaceae has a restricted distribution,

and is of lower economic and ecological importance.

The family therefore has not been studied as

extensively as Poaceae and, because so few taxa are

widespread, not as much work has been done

elsewhere. Therefore, Mesembryanthemaceae has

fewer references in the literature, most of which are

of a specialist nature. These differences explain the

small number of references important for both

families.

Names

— ruu)-e>.uicn-\Jcoc£i(s — ro co cn cn

QQQQG!GH3(SQQ(S(3QQQQ

(SQQQQQQQSQQQQQQQQ

[ HI I 11 I I I I I I I I I I

-\J oo to

QG)(S

CD Q CD

11.8V.

140.9 V.

I

Fig. 2. — Poaceae major referen-

ces, which each contribute

1% or more of the total

references. The number of

names is indicated across the

top, and the percentage con-

tribution of each reference is

given at the end of its bar.

128 REGISTER OF NAMES AND TYPES: A COMPARISON BETWEEN MESEMBRYANTHEMACEAE AND

POACEAE

TABLE 2. References common to both Mesembryanthemaceae

and Poaceae

* Major reference, contributing 1% or more of the total names

for the family.

TABLE 3. — Ecological importance of the two families indicated

by their occurrence in the lists of Acocks (1975), for his seventy

veld types

CONCLUSION

If Poaceae, a widespread important family, and

Mesembryanthemaceae, a restricted family of

relatively lower economic and ecological impor-

tance, are regarded as ends of a continuum, other

families can be compared to them to determine how

much they would benefit from a complete register of

names. The more widespread and important a family

is, the greater the number of references it may be

expected to have, but a substantial proportion of

these may be found in sources it shares with other

families. The less widespread and important a family

is, the fewer references it may be expected to have,

but a smaller proportion of its names may be found

in common sources.

Because the majority of references in Poaceae and

Mesembryanthemaceae are not drawn from com-

mon sources, a complete register for all families may

be expected to have a high proportion of references

useful for only a single family. If computerized, such

a register would be difficult to manage because of its

large size, even though most of the references would

not be of use in more than one family.

Therefore, it is suggested that a complete register

of all names and types should be compiled in two

steps. In the first step, a preliminary register should

be drawn from sources which are of widespread

coverage or historical importance, such as Flora

Capensis, Flora of Tropical Africa, Thunberg’s

Prodromus, etc., as well as from journals known to

be of importance for southern African plants, such

as Kew Bulletin, Botanische Jahrbucher and Botha-

lia. This work would be performed once, and would

apply to a number of families, being most useful in

widespread and important groups. In the second

step, when a particular family is under study, the

references from these major sources will be readily

available as a starting point, and the references of

value only to that family can be traced and added to

the Register. The complete register of all names

would therefore grow as more families are studied

for the Flora.

ACKNOWLEDGEMENTS

Part of the work reported here was done while one

of us (HFG) was South African Liaison Botanist at

Kew. The Directors of the Botanical Research

Institute, Pretoria and the Royal Botanic Gardens,

Kew, are thanked for this opportunity and for

facilities made available.

Our thanks are due to those whose labours made

this study possible. Notable among these are the

ladies of the Kew typing pool, for soldiering on

through an unending morass of semi-legible ma-

nuscript, and Mrs D. Scott, Miss J. White and other

members of the Library staff at Kew, and Miss T. A.

Bence of Index Kewensis for their unfailing help in

tracing obscure names and their places of publica-

tion.

At the Botanical Research Institute, Mrs W. Roux

and Mrs C. Fourie have encoded names and

corrected computer printout. Mr S. Makgakga

counted names and references, and the comparisons

are based on his counts.

UITTREKSEL

’n Register van name en tipes vir Suider-Afrika sou

die voltooiing van die Flora van Suider-Afrika

bespoedig. Registers vir Mesembryanthemaceae en

Poaceae, families wat soortgelyk in getal spesies en

G. E. GIBBS RUSSELL AND H. F. GLEN

129

name is, maar wat verskil in verspreiding en

belangrikheid, word vergelyk om riglyne vir ’n

doelgerigte benadering aan die uitbreiding van die

register te gee, om ander families in te sluit.

REFERENCES

Acocks, J. P. H., 1975. Veld Types of South Africa, 2nd edn

Mem. bot. Surv. S. Afr. 40: 1-128.

Chapman, A. D., 1979. Australian Plant Name Index Aba-

rema- Acacia. Canberra: Australian Biological Resources

Study.

Gibbs Russell, G. E., 1975. Comparison of the size of various

African floras. Kirkia 10: ,123- 130.

Gibbs Russell, G. E., 1983. Register of names and types in

Poaceae: a computerized index for southern Africa. Bothalia

14: 943-944.

Heywood, V. H., 1978. Flowering plants of the World. Oxford:

Oxford University Press.

Jacobsen, H., 1974. Lexicon of succulent plants. London:

Blandford.

Lawrence, G. H. M., 1951. Taxonomy of vascular plants. New

York: Macmillan.

Leenhouts, P. W., 1968. A guide to the practice of Herbarium

taxonomy. Utrecht: International Bureau for Plant Taxono-

my.

Shetler, S. G. & Skog, L. E., 1978. A provisional checklist of

species for Flora North America. FI. N. Amer. Rep 84'

i-xix, 1-199.

Bothalia 15, 1 & 2: 131-138 (1984)

Intensity of plant collecting in southern Africa

G. E. GIBBS RUSSELL*, E. RETIEF* and L. SMOOK*

Keywords: check-lists, distribution maps, gamma diversity, herbarium, phytogeography, species diversity

ABSTRACT

The intensity of plant collecting in southern Africa is mapped using records from the Pretoria National

Herbarium Computerized Information System (PRECIS). For the entire area, over 85% of the quarter degree grid

squares have fewer than 100 specimens recorded. Collecting intensities are compared for different countries,

biomes and climatic zones. Future field work from the National Herbarium will be concentrated in areas most

seriously under-collected.

INTRODUCTION

Until now there has been no source of precise

information about the intensity of plant collecting in

southern Africa. Previous indications of collecting

intensity, such as the AETFAT maps showing the

extent of floristic exploration of sub-Saharan Africa

(Leonard, 1965; Brenan, 1965; Hepper, 1979 and

Leonard, 1979), were compiled from literature and

from the personal knowledge of experienced

botanists. Now, PRECIS (Pretoria National Herba-

rium Computerized Information System) can provi-

de quantitative information about collecting intensi-

ty based on the number of specimens from each

quarter degree grid square (Edwards & Leistner,

1971) in the entire Flora of Southern Africa region.

Why is it important to know the collecting

intensity of a region?

1. Knowledge of areas that are under-collected is

necessary when planning collecting trips, so that

additional localities as well as additional specimens

and taxa may be added to the Herbarium. In this

way, each taxon in the Herbarium will be more

completely represented both for distribution records

and for morphological characters over its range. The

time and money spent on collecting trips will thus be

used to the greatest possible effect.

2. Knowledge of areas that are well-collected is a

guide to using PRECIS to provide additional

information. For instance, the list of species for a

well-collected quarter degree square can be used as a

preliminary check-list for under-collected areas

nearby or in the same veld type.

3. Knowledge of collecting intensities is required

for phytogeographical studies because of the

interrelation between collecting intensity and species

diversity (gamma diversity, Whittaker, 1972).

Knowledge of collecting intensity will also help in

assessing the completeness of species lists.

METHODS

The PRECIS data base is explained in detail by

Magill et al. (1983) and Gibbs Russell & Gonsalves

* Botanical Research Pnstitute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

(in press). The system contains label information for

about 600 000 plant specimens in the National

Herbarium (PRE), of which about 55% (±325 000

specimens) have their locality expressed as a quarter

degree grid reference. In this form, the localities are

easily sorted by the computer to produce either

distribution maps of the specimens in any species, or

check-lists of the species in any quarter degree

square. The program that produces the check-lists

also reports the number of specimens held in each

quarter degree square. These numbers were mapped

as given by PRECIS, then broken into size classes as

shown in Fig. 1, for easier interpretation. The

numbers of quarter degree squares in different

regions were counted so that the percentages of

different size classes in each region could be

compared.

The data shown here, although based on part of

the specimens in one Herbarium, are thought to

reflect accurately the intensity of collecting for the

entire Flora area for the following reasons:

1. The National Herbarium, although located in

the Transvaal, has a high percentage of specimens

from other regions of the Flora area. Morris &

Manders (1981) report that 71,7% of the PRE

specimens from the Flora area were collected

outside the Transvaal.

2. When a portion of the data for Fig. 1, for which

all possible grids were known, is plotted separately,

the resulting map is extremely similar to Fig. 1.

Grids were determined for about 40 000 of the ±

50 000 specimens of Poaceae in PRE (the other

10 000 had poor locality records that could not be

expressed as a grid). Although based on only 12% of

the specimens represented in Fig. 1, the collecting

intensities shown for Poaceae are similar for

countries, biomes and rainfall regions to the

intensities shown by all specimens in Fig. 1.

3. When a sample of specimens from other

herbaria is plotted for collecting intensity, the results

agree with Fig. 1 for well collected and under

collected quarter degree squares. Collecting intensi-

ties for about 3 500 specimens of Ruschia and

Ruschiinae (Mesembryanthemaceae) from B, BM,

BOL, G, K, L, M, PRE, S and Z were plotted by Dr

H. F. Glen. This is important independent

confirmation of Fig. 1, because ± 90% of these

132

INTENSITY OF PLANT COLLECTING IN SOUTHERN AFRICA

Fig. l. Collecting intensity for the Flora of Southern Africa area, based on numbers of specimens per quarter degree square, as reported by PRECIS.

G. E. GIBBS RUSSELL, E. RETIEF AND L. SMOOK

133

specimens were from herbaria other than PRE, and

the total sample size was only 1% of the number of

specimens represented in Fig. 1.

4. The ± 325 000 specimens represented in Fig. 1

constitute more than 16% of the total number of

specimens housed in herbaria in the Flora area.

Index Herbariorum edn 7 (Stafleu, 1981) reports

2 028 400 specimens held in Flora area Herbaria

(Gibbs Russell, 1983). This percentage is higher

than that of Poaceae or Ruschiinae in Fig. 1,

discussed above. Furthermore, if the specimens

which are duplicates held in different herbaria,

specimens which were collected outside the Flora

area, and specimens for which it is impossible to

determine a grid reference were subtracted from the

total number of specimens, the true percentage

cover of the specimens reported in Fig. 1 would be

much higher.

In order to test the value of collecting in grid

squares shown by PRECIS to be under-collected, a

trip was made in January 1983 to the grid square

2728 in the eastern Orange Free State. The numbers

of specimens and taxa collected were compared with

the previous holdings of the Herbarium to determine

whether the trip had added materially to the

distribution records both for the taxa and for the

area.

RESULTS AND DISCUSSION

Information about collecting intensity can be

drawn from the map (Fig. 1) in several different

ways, either for the whole Flora area (Fig. 4.C), or

for divisions such as countries (Figs 3 & 4), biomes

(Fig. 5) and climatic zones (Fig. 6). • Regions

compared in this paper are shown in Fig. 2.

The first visual impression of the map (Fig. 1) is

that plant collecting, as reflected at PRE, has been

most concentrated in the central and eastern

Transvaal and the south-western and southern Cape,

with lesser centres in Natal, Swaziland, Lesotho and

the eastern Cape. South Africa, together with

Swaziland and Lesotho, is better collected than

Botswana or South West Africa/Namibia. However,

within South Africa, the central and northern Cape

is poorly collected. Closer examination shows that

even in reasonably well-collected areas there are

quarter degree squares with no specimens recorded,

and conversely, in under-collected areas particular

quarter degree squares with higher numbers of

specimens show the location of towns, roads or

research centres. Every whole degree square shows

at least a few specimens recorded.

Precise information about collecting intensities in

the different countries and in the entire Flora area is

given in Figs 3 & 4. Considered as a whole (Fig. 4C),

the Flora area is critically undercollected. Over 85%

of the area has fewer than 100 specimens recorded

per quarter degree square, over 25% of the quarter

degree squares have never been sampled at all, and a

further 33% has fewer than 10 specimens. The

individual countries, however, vary greatly in their

coverage. Best collected is Swaziland (Fig. 3A), with

no uncollected quarter degree squares, and with the

largest class having 201-500 specimens per square.

Lesotho (Fig. 3B) is also reasonably well-collected,

with a low percentage of uncollected squares, but

the largest class has only 11—50 specimens per

quarter degree square. Botswana (Fig. 3C) is the

most under-collected country, with nearly half the

quarter degree squares lacking specimens entirely,

and nearly 98% of the squares showing fewer than

100 specimens. It is the only country in which the

size class of 0 specimens per quarter degree square is

the largest.

Fig. 2. — Areas compared in Figs 3

(countries), 4 (countries), 5

(Karroid vegetation type)

and 6 (rainfall regions).

INTENSITY OF PLANT COLLECTING IN SOUTHERN AFRICA

134

Fig 3. — Percentages of collecting intensity size classes for the smaller countries in the Flora area. A,

Swaziland; 64% of quarter degree squares have more than 100 specimens. B, Lesotho; 33,2% of quarter

degree squares have more than 100 specimens. C, Botswana; 2,1% of quarter degree squares have more

than 100 specimens.

G. E. GIBBS RUSSELL, E. RETIEF AND L. SMOOK

135

oj

Fig. 4. — Percentages of collecting intensity size classes for the larger countries and total of the Flora area. A,

South West Africa/Namibia; 3,6% of quarter degree squares have more than 100 specimens. B, South

Africa; 25,6% of quarter degree squares have more than 100 specimens. C, total Flora area; 14,5% of

quarter degree squares have more than 100 specimens.

136

INTENSITY OF PLANT COLLECTING IN SOUTHERN AFRICA

The two larger countries, South West Africa/Na-

mibia (Fig. 4A), and South Africa (Fig. 4B), show

similar patterns, with the largest size class being

1 — 10 specimens per quarter degree square. How-

ever, South Africa has more quarter degree squares

with high numbers of specimens recorded than

South West Africa/Namibia, where only four

quarter degree squares have over 500 specimens

recorded. The entire Flora area (Fig. 4C) shows a

similar pattern to the two largest countries, because

their greater area overrides the pattern of the

smaller, better collected countries.

The Karoo was chosen as an example of

determining collecting intensity in a biome (Fig. 5).

The outlines of the Karoo biome were taken from

the map of Edwards & Scheepers (in Scheepers,

1983) and Fig. 5A shows the percentages of

specimens reported in each size class. In a more

restricted sample of the Karoo, the collecting

intensity in Acocks’s (1975) veld type 26, Karroid

Broken Veld, are shown in Fig. 5B. Although there

is a higher percentage of quarter degree squares with

no specimens reported, there are also higher

percentages of quarter degree squares with greater

numbers of specimens for the veld type than for the

biome. Both are obviously under-collected, and

reference to Fig. 1 shows where collecting should be

done to achieve reasonable coverage for these

vegetation types. PRECIS species-per-grid lists, for

the few quarter degree squares with a large number

of specimens reported, can serve as preliminary

checklists while further studies are in progress.

The mesic and arid climatic zones show striking

differences in collecting intensity (Fig. 6). The

differences are demonstrated both when comparing

percentages of size classes and when comparing

absolute numbers of grid squares within each region.

In the mesic eastern and southern parts of the Flora

area, over 44% of the quarter degree squares have

more than 100 specimens, whereas in the arid central

and western parts only 5,9% of the quarter degree

squares have more than 100 specimens. Table 1

shows that even though the drier regions cover more

than three times the area of the wetter regions, there

are more than twice as many squares with over 100

specimens in the wet regions than in the dry regions.

Reasons for the marked differences in collecting

intensity between wet and dry regions include the

following:

1. The mesic areas include most of the major

population centres, and 33 of the 38 herbaria in the

subcontinent. The ease with which roads can be

o

0

3

o-

w

24-

23-

22-

21 -

20-

19 —

ia-

17—

16-

15 —

14 —

13--

12-

1 1 --

10-

9-

8-

?--

6-

5-

4-

3-

2-

1-

0

B. Karroid Broken Veld

35 .5K

<M

Fig. 5. — Percentages of collecting

intensity size classes for the

karroid vegetation type. A,

Karoo Biome; 3,8% of quar-

ter degree squares have more

than 100 specimens. B, Kar-

roid Broken Veld Type 26;

9,7% of quarter degree

squares have more than 100

specimens.

G. E. GIBBS RUSSELL, E. RETIEF AND L. SMOOK

137

in o q

— Q

OJ

Fig. 6. — Percentages of collecting

intensity size classes for the

mesic region (more than 100

days with over 0,25 mm

rainfall) and the arid region

(less than 100 days with over

0,25 mm rainfall). A, mesic

region; 44,5% of quarter

degree squares have more

than 100 specimens. B, arid

region; 5,9% of quarter de-

gree squares have more than

100 specimens.

traced in dry areas (for instance from the Reef to

Kimberley and from the south-western Cape to the

Orange River) and urban centres can be located in

mesic areas (for instance Pretoria-Johannesburg

and Cape Town) shows that accessible places have a

higher collecting intensity than places difficult to

reach. Van Dijk (1971) has shown similar effects in

distribution maps for frog species.

2. Arid areas not only have lower rainfall, but the

rainfall is also more unpredictable (Weather

Bureau, 1965). The longer dry season and more

frequent droughts in dry areas allow fewer occasions

when plants are in a condition to make good

Herbarium specimens, in comparison to wetter

areas.

3. Differences in species diversity could also

make the recorded differences in collecting intensity

more apparent than they really are, in terms of true

coverage of species present. Species diversities are

reliably known for few localities, but if there is a

higher species diversity in mesic than in arid areas,

then a lower collecting intensity for arid areas, as

shown by fewer specimens reported, may be partly a

result of there being fewer species there available for

collection.

The trip made to grid square 2728, to test the

value of collecting in grids shown by PRECIS to

have low numbers of specimens recorded, yielded

the results shown in Table 2. In total, the trip added

about eight times the original number of specimens

and six times the original number of species to the

Herbarium for the degree grid square. This large

increase shows that excellent use can be made of

PRECIS to indicate areas in which field work would

be especially profitable.

TABLE 1. — Comparison between numbers of quarter degree

grid squares in areas with more than 100 days of rain and

with less than 100 days of rain

a. More than b. Less than % (a/b)

100 days rain 100 days rain

Number of quarter

degree grid

138

INTENSITY OF PLANT COLLECTING IN SOUTHERN AFRICA

TABLE 2. — Specimens and species in PRE herbarium as a result

of collecting trip to degree grid square 2728, previously

shown by PRECIS to be under-collected

CONCLUSIONS

Records of specimens collected in quarter degree

grid squares in the Flora area show that some areas

are relatively well-collected, but that the greater part

of the sub-continent is seriously under-collected.

Future collecting trips should be planned in a

co-ordinated way, using the information presented

here, so that new locality records as well as

additional specimens and taxa are added to the

Herbarium. As a result, each species will be better

represented, showing its morphological variation

throughout its range.

Quarter degree squares shown to be well-collected

can be a valuable source of information for future

studies in nearby areas, because PRECIS can

provide preliminary check-lists through the species-

per-grid listings. PRECIS thus makes the link

between previous collectors and Herbarium workers

and the present field worker, so that new studies can

begin on the basis of work done in the past.

ACKNOWLEDGEMENTS

Dr B. de Winter, the Director of the Botanical

Research Institute, has supported PRECIS

throughout the course of its development, and he

has also given advice during the preparation of this

paper. Dr O. A. Leistner provided helpful criticism

of the manuscript. Mr S. Makgakga plotted the

collecting intensities from the PRECIS printout, and

Mrs W. Roux prepared the map for publication.

UITTREKSEL

Die intensiteit van plantversameling in suidelike

Afrika word op ’n landkaart afgebeeld deur gebruik

te maak van gegewens uit die gerekaniseerde

inligtingsisteem, PRECIS, van die Nasionale Herba-

rium, Pretoria. Daar is gevind dat, binne die

betrokke gebied, meer as 85% van die kwartgraad-

ruite minder as 100 eksemplare bevat. Die ver-

samelintensiteite van verskillende landstreke, biome

en klimaatsones word vergelyk. Toekomstige veld-

werk van die Nasionale Herbarium sal op gebiede wat

uiters swak versamel is, toegespits word.

REFERENCES

Acocks, J. P. H., 1975. Veld types of South Africa. 2nd edn

Mem. hot. Surv. S. Afr. 40: 1 — 128.

Brenan, J. P. M., 1965. Map of the extent of floristic exploration

in Africa south of the Sahara. Webbia 19: 911-914.

Edwards, D. & Leistner, O. A., 1971. A degree reference

system for citing biological records in southern Africa. Mitt,

bot. StSamml., Munch. 10: 501—509.

Gibbs Russell, G. E., 1983. Book review of Index Herbariorum

edn 7, ed. F. A. Stafleu. Bothalia 14: 322—323.

Gibbs Russell, G. E. & Gonsalves, P., (in press). PRECIS - a

curatorial and biogeographic system. In F. Bisby, Symposi-

um on databases in systematics. London: Academic Press.

Hepper. F. N., 1979. Second edition of the map showing the

extent of floristic exploration in Africa south of the Sahara,

published by AETFAT. In J. Kunkel, Taxonomic aspects of

African economic botany, Proceedings of the IX Plenary

Meeting of AETFAT, Las Palmas de Gran Canaria.

Leonard, J., 1965. Carte du degre d'exploration floristique de

l’Afrique au sud du Sahara. Webbia 19: 907—910.

Leonard, J., 1979. Statistiques des progres accomplis en 24 ans

dans la connaisance de la flore phanerogamique africaine et

Malgache (1953 a 1976). In J. Kunkel, Taxonomic aspects of

African economic botany, Proceedings of the IX Plenary

Meeting of AETFAT, Las Palmas de Gran Canaria.

Magill, R. E., Gibbs Russell, G. E., Morris, J. W. &

Gonsalves. P., 1983. PRECIS — the Botanical Research

Institute Herbarium data bank. Bothalia 14: 481-495.

Morris, J. W. & Manders, R., 1981. Information available

within the PRECIS data bank of the National Herbarium,

Pretoria. Bothalia 13: 473—485.

Scheepers. J. C., 1983. The present status of vegetation

conservation in South Africa. Bothalia 14: 991—995.

Weather Bureau, 1957. Climate of South Africa, part 4, rainfall

maps. Weather Bureau WB 22.

Weather Bureau, 1965. Climate of South Africa, part 8, general

sur\>ey. Weather Bureau WB 28.

Whittaker, R., 1972. Evolution and measurement of species

diversity. Taxon 21: 213—251.

Bothalia 15, 1 & 2: 139-152 (1984)

Notes on African plants

VARIOUS AUTHORS

CYPERACEAE

A NEW SPECIES OF CARPHA FROM THE NATAL DRAKENSBERG, SOUTH AFRICA

Carpha filifolia Reid & Arnold, sp. nov., C.

capitellatae Boeck. affinis, sed culmis uni-nodis

interdum sine nodis, vaginis foliorum fragilibus

aureo-brunneis (veteribus nigris), laminis foliorum

filiformibus teretibus, acheniis 4-5plo longioribus

latitudine differt.

Type. — Natal, 2929 (Underberg): ‘Storm

Heights’, Mpendhle District (-BC), Hilliard &

Burtt 11791 (PRE, holo.; E; NU).

Perennial herb, erect, tufted, glabrous, 400—600

mm tall. Rhizome abbreviated, c. 2—3 mm long, c.

2-3 mm diam., concealed by persistent leaf bases.

Roots c. 1,5—2 mm diam., with a distinct outer layer

of aerenchymatous tissue. Culms overtopping

leaves, terete, c. 1 mm diam., single-noded (rarely

nodeless), each node bearing a leaf-like bract

110-190 mm long. Leaf sheaths with bases shiny,

dark brown, with raised nerves, usually turning

black and splitting vertically with age; upper parts

papery, golden brown, eventually opening and

becoming flattened, apex auriculate to truncate,

outer sheaths bladeless. Leaf blades spirally arran-

ged, filiform, 0,5-0,75 mm diam., c. 280-330 mm

long, terete with adaxial surface shallowly to deeply

grooved, venation usually obvious, apex obtuse.

Inflorescence a single terminal head, occasionally

accompanied by 2-3-peduncled or sessile reduced

partial inflorescences, each originating in the axil of

a leaf-like bract. Partial inflorescence a compact,

narrow panicle composed of 5—6 units, each

consisting of 3—5 sessile or pedicelled spikelets and

subtended by a bract. Bracts several, lowest 1^—3

times length of the panicle, leaf-like with slightly

dilated bases, remaining bracts reduced in length

and becoming glume-like towards the apex. Spikelets

lanceolate, c. 8—10 mm long, composed of 4-5

spirally arranged glumes. Glumes lanceolate, c. 6

mm long, apex acute to acuminate, golden-brown,

green-keeled when young. Lowest 2—3 glumes

empty, the remainder subtending bisexual florets or

the uppermost male or empty. Style excluding base

c. 5 mm long, slender, brownish, the 3 papillate

branches equalling the undivided portion in length.

Style base cream-coloured, trigonous, scabrid,

0,9— 1,4 mm long, persistent, forming a beak at the

achene apex. Stamens 3, anthers Unear, c 4-4,4 mm

long, pale, mottled green, crest minute, white,

pyramidal, with the crest height equalling its breadth

at the base. Bristles 6, terete, white, slightly

exceeding the achene in length, scabrid distally,

setose basally, those adjacent to achene faces

occasionally plumose. Achene trigonous, 4—5x1

mm, c. §— f glume length, brown at maturity with

ribs cream-coloured, beak and rib shoulders scabrid;

surface with epidermal cells 5-6 sided, arranged in

vertical rows.

This species is restricted to the sandstone and

basalt formations of the upper Karoo System in the

Drakensberg mountains, at altitudes between 1 800

and 2 800 m. It grows along streambanks and in

marshy areas, where it is frequently the dominant

species. Towards the end of the growing season its

populations are easily identifiable by the character-

istic yellow-green colour of the foliage. Flowering

begins in November at lower altitudes and is later at

higher altitudes. Mature fruits are present from

December onwards and are dispersed by the

subsequent breaking up of the spikelets. Figs 1-4.

Transvaal. — 2730 (Vryheid): ‘Oshoek’, Wakkerstroom dis-

trict (-AD), Devenish 1067, 1821 (PRE).

O.F.S. — 2829 (Harrismith): Platberg (—AC), Venter 7093,

(PRE; BLFU).

Natal. — 2929 (Underberg): path from Loteni Nature Reserve

to Redi (—AD), Hilliard & Burtt 16098, (E; NU); ‘Storm Heights'

(-BC), Hilliard & Burtt 11791 (E; NU; PRE); Highmoor Forest

Reserve (-BC), Killick & Vahrmeijer 3595 (PRE), Hilliard &

Fig. 1. — Distribution of Carpha filifolia.

140

NOTES ON AFRICAN PLANTS

Fig. 2. — 1. Carpha filifolia: a, habit x 2/3; b, inflorescence x 2; c, lowest inflorescence bract x 3; d, glume x 3. e, achene x 15; f,

style and branches x 20; g & h, bristles x 20. 2, anther crests of Carpha spp. (schematic representation): a, C. filifolia; b, C.

capitellata; c, C. bracteosa; d, C. glomerata; e, C. schlechteri (all x 50).

VARIOUS AUTHORS

141

Burtt 16258 (E; NU); Mlambonja Valley, Garden Castle Forest

Reserve (-CA), Hilliard & Burtt 14946 (E; NU; PRE); Sani Pass

(— CB), Hilliard & Burtt 9788 (E; NU; PRE); Cobham Forest

Station (-CB), Hilliard & Burtt 12509 (E; NU; PRE), 16005,

16059 (E; NU); Garden Castle Forest Reserve, (— CB), Hilliard

& Burtt 13462 (E; NU; PRE).

Lesotho.— 2929 (Underberg): Sehlabathebe National Park

(-CC), Hoener 2138 (PRE).

Cape. — 3127 (Lady Frere): Bastervoetpad, Elliot — Maclear

district boundary (-BB), Hilliard & Burtt 16677 (E; NU).

Previous workers in Carpha relied on gross

morphological features to separate the taxa,

neglecting spikelet characters. The present investi-

gation has, however, highlighted the importance of

these characters as a useful means of differentiating

taxa. The most important spikelet characters are

therefore discussed below for all five southern

African members of the genus, together with

vegetative characters and distribution patterns. This

comparison provides an insight into affinities

between taxa, as well as indicating possible

evolutionary trends within the group.

The existing southern African species of Carpha

have a more southerly distribution and a more

robust habit than C. filifolia. C. glomerata (Thunb.)

Nees, occurring at low altitudes from Piketberg in

the western Cape Province, through the southern

Cape to southern Natal, is the largest species in the

Fig. 3. — Carpha filifolia Reid & Arnold, Hilliard & Burtt 11791

(PRE, holotype).

Fig. 4. — Achenes of Carpha spp.: a, C. glomerata; b, C.

schlechteri; c, C. capitellata; d, C. bracteosa; e, C.

filifolia (all X 9,5).

genus. It grows to 3 m tall, has broad leaves up to 28

mm wide, culms with 1—2 sterile and 4—6 fertile

nodes, and a large, many-branched complex

inflorescence.

Closely related to C. glomerata is C. schlechteri

C.B.C1. This species is confined to the Koue

Bokkeveld mountains of the Ceres District in the

south-western Cape. It is less robust than C.

glomerata , reaching c. 1 m in height, with much

narrower leaves (c. 5—8 mm wide) and a less

complexly branched inflorescence. The number of

sterile and fertile culm nodes is the same as in C.

glomerata.

The remaining species, C. capitellata Boeck. and

C. bracteosa C.B. Cl., have a similar distribution to

C. glomerata although present records indicate that

they do not occur in Transkei or Natal. They differ

from the two previous species in being smaller (up to

600 mm in height), their leaf width ranges between

2-4 mm, the culms have 1-2 sterile and 2—4 fertile

nodes, and the inflorescence is more reduced,

comprising fewer branches with less complex partial

inflorescences. These two species are closely related,

differing mainly in C. bracteosa having large,

papyraceous inflorescence bracts. In Flora of the

Cape Peninsula, Levyns (1950) treated them as

conspecific.

C. filifolia is slightly smaller in habit than these

four species, reaching a height of 400—600 mm. The

leaves are typically filiformous, showing an adapta-

tion to high altitude montane conditions. The

inflorescence has undergone the greatest degree of

reduction, with the culm single-noded and suppor-

ting fewer, smaller partial inflorescences than either

C. capitellata or C. bracteosa.

In C. glomerata and C. schlechteri the anther

crests are characteristically linear-acuminate in

142

NOTES ON AFRICAN PLANTS

shape, being 2—2,5 times as long as broad at the

base in C. glomerata and 3—4 times as long as broad

at the base in C. schlechteri. The style bases in these

two taxa are respectively glabrous or minutely

scabrid.

In C. capitellata and C. bracteosa the anther crests

are pyramidal to globular in shape with the crest

length equal to or shorter than the width at the crest

base. The style base in both species is distinctly more

scabrid than in either C. glomerata or C. schlechteri.

In C. filifolia the size and shape of the anther crests

and scabrid nature of the style base are the same as

described for C. capitellata and C. bracteosa. This

suggests a much closer relationship with these two

taxa than with C. glomerata or C. schlechteri. This

affinity is further indicated by the amount of

reduction undergone by the inflorescence coupled

with the reduction in the number of culm nodes.

Despite its close affinity with these two taxa and

especially with C. capitellata, C. filifolia has a

number of characteristics that clearly set it apart

from other southern African species: these include

the tufted growth habit, filiformous leaves and

single-noded culms which have already been

discussed. Other distinctive features include the

large achenes, these being 1,6-2 times longer than

those of other taxa, in which the achenes are all

similar in size, ranging between 2, 3-2, 6 mm in

length. Associated with the large achene is a long

style base which varies between 0,9-1, 4 mm,

compared with 0,4— 0,7 mm in other taxa. Another

unusual feature found only in this species, although

not well-developed in all the specimens examined, is

the occurrence of plumose bristles.

Apart from the southern African species, the only

other Carpha species found in Africa is C. eminii (K.

Schum.) C.B. Cl. A detailed comparison between C.

filifolia and this species has not been carried out,

nevertheless it is clear from existing descriptions that

these .two taxa are not conspecific. Two characteris-

tics of C. eminii that distinguish it from C. filifolia

are its broad, flattened leaves, and small achenes

which are only i the length of the glumes.

Key to the southern African species

Plants rhizomatous; leaves dorsiventrally flattened, keel-

ed, 2-28 mm wide; inflorescence axis many-noded

(2— 6 fertile, 1—2 sterile); achene length 2, 3— 2,6 mm:

Plants of large stature, 0,8 to 3 m tall; fertile

inflorescence nodes 4-6; anther crests linear-

acuminate:

Leaves 15—28 mm wide, margins scabrid; spikelets

6-7 mm long C. glomerata

Leaves 5—8 mm wide, margins entire; spikelets c. 3

mm long C. schlechteri

Plants of small stature, less than 0,6 m tall; fertile

inflorescence nodes 2-4; anther crests pyramidal to

globular:

Inflorescence bracts narrowly lanceolate, chartaceous;

achene epidermal cells isodiametric, outer periclinal

walls crateriform C. capitellata

Inflorescence bracts broadly ovate, papyraceous;

achene epidermal cells longitudinally elongated,

outer periclinal walls ± planar C. bracteosa

Plants tufted; leaves filiform, 0,5-0,75 mm diam.;

inflorescence axis single-noded or nodeless; achene

length 4-5 mm C. filifolia

C. Reid & T. H. Arnold*

* Both Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

LAMIACEAE

A NEW SPECIES OF PLECTRANTHUS

Plectranthus dolomiticus Codd, sp. nov., forma

corollae P. petiolaris E. Mey. ex Benth. simile, sed

caulibus tenuioribus decumbentibus, foliis et flori-

bus parvioribus differt.

Herba perennis, semisucculenta, ramosa; caules

decumbentes cinereo-tomentulosi, 100-150 mm

longi. Folia petiolata; lamina semisucculenta, ovata

vel late ovata, 20—30 x 18-30 mm, subglabra, apice

rotundato, basi truncata, margine manifeste crena-

to; petiolus 15-30 mm longus. Inflorescentia

terminalis, simplex vel interdum ramosa, 70 — 130

mm longa; rhachis minute puberula; bracteae

ellipticae, 1,5 mm longae, persistentes. V erticillastri

10-25 mm distantes, 2-6-flori; pedicelli 4—5 mm

longi, minute puberuli. Calyx campanulatus, demum

5-6 mm longus, subglaber; lobus posticus ovatus,

erectus, 2 mm longus; lobus anticus subaequaliter

4-dentatus; dentes deltoideo-lanceolati, 1,5—2 mm

longi. Corolla violacea, glabra, 9 — 10 mm longa;

tubus aliquantum sigmoideus, basi 0,75 mm diam.,

ad fauce 2 mm diam.; labium posticum erectum, 2

mm longum, obscure 4-lobatum; labium anticum

concavum, 2 mm longum. Stamina 4; filamenta 2

mm longa, libera. Stylus 2 mm fauce exsertus.

Type. — Transvaal, Farm Ostend, 7 km north-west

of Penge Mine, 10 December 1982, Van Jaarsveld

7052 (PRE, holotype).

Perennial, semi-succulent, branched herb, about

100 mm tall and 300 mm across; stems decumbent,

slender, greyish-tomentulose, 100—150 mm long.

Leaves petiolate; blade semi-succulent, ovate to

broadly ovate, 20—30 x 18—30 mm, subglabrous,

apex rounded, base truncate, margin distinctly

crenate; petiole 15—30 mm long. Inflorescence

terminal, simple or sometimes branched, 70—130

mm long; rhachis minutely puberulous; bracts

elliptical, 1,5 mm long, persistent. Verticils 2—6-

flowered, 10—25 mm apart; flowers in 1— 3-flowered

sessile cymes; pedicels 4—5 mm long, minutely

puberulous. Calyx campanulate, eventually 5-6 mm

long, subglabrous; upper lip ovate, erect, 2 mm

long; lower lip subequally 4-toothed; teeth

lanceolate-deltoid, 1,5-2 mm long. Corolla violet-

purple, glabrous, 9—10 mm long; tube somewhat

sigmoid, 0,75 mm deep at the base, increasing about

the middle to 2 mm deep at the throat; upper lip

erect, 2 mm long, obscurely 4-lobed; lower lip

concave, 2 mm long. Stamens 4; filaments 2 mm

VARIOUS AUTHORS

143

Fig. 5. — Plectranthus dolomiticus (Van Jaarsveld 7052, PRE,

holotype).

long, free to the base. Style exserted by 2 mm,

shortly bifid. Nutlets broadly ovoid, 1,5 mm long,

pale brown. Figs 5 & 6.

Transvaal.— 2430 (Pilgrims Rest): 7 km NW of

Penge Mine (—AD); Van Jaarsveld 7052.

This species was discovered independently in 1982

by Mr E. J. van Jaarsveld of the Kirstenbosch

National Botanic Garden and Mr S. Venter of the

Transvaal Nature Conservation Department on the

dolomite formation which crops up in and near the

Olifants River and Blyde River valleys in the eastern

Transvaal. Here it grows in shallow, humus-filled

rock crevices in arid woodland, usually in open

Fig. 6. — Plectranthus dolomiticus (Van Jaarsveld 7052, PRE,

holotype), approximately natural size.

sunny places, forming small, semi-succulent plants

about 100 mm tall and 300 mm across with slender,

greyish-tomentulose, decumbent stems.

In growth habit the plants superficially resemble

Aeollanthus parvifolius, but the calyx places it in

Plectranthus. The shape and colour of the corolla are

reminiscent of P. petiolaris, in which the tube is also

somewhat sigmoid in shape, i.e. narrow at the base

and at first ascending, then deflexed about the

middle and expanding to the throat, but here the

resemblence ends. P. dolomiticus is a smaller plant

with semi-succulent, greyish stems and smaller

leaves and corolla 9-10 mm long, whereas P.

petiolaris is a robust herb, 0,5 — 1 m tall of forest

margins with fairly thick, 4-angled, strigose stems,

large, thin-textured leaves 40-140 x 35-110 mm,

and corolla 12—15 mm long.

L. E. CODD*

* Botanical Research Institute, Department of Agricul-

ture, Private Bag X101, Pretoria 0001.

MELIANTHACEAE

TWO NEW COMBINATIONS IN MELIANTHUS

Melianthus pectinatus Harv. in FI. Cap. 1, Add.

sub praef 21 (1860); Coulston & Bailey in Bailey,

Cycl. Hort. 4: 2024 (1916); Phillips & Hofmeyer in

Bothalia 2: 353 (1927); Bean in Chittenden, R.

Hort. Soc. Diet. 3: 1281 (1951). Type: Cape,

Namaqualand, Wyley s.n. (TCD, holo.).

Melianthus trimenianus Hook. f. in Trimen Jl. Bot. N.S. 2: 353

(1873); Hook. f. in Curtis’s bot. Mag. 37: t.6557 (1881). Type:

Cape, Namaqualand, Barkley s.n. (?K, holo.).

Shrub. Leaves up to 140 mm long; stipules paired,

acuminate, 8 mm long, 1 mm wide; rachis unwinged

to wings as broad as terminal leaflet; leaflets 7—27,

narrowly linear to lanceolate, sinuate to dentate,

dorsal surface stellately hairy over sunken veins,

sometimes sparsely pilose over blade, ventral

surface with a dense mat of short stellate hairs with

large emergent stellate hairs. Inflorescence a

sub-terminal raceme, erect to pendulous, elongating

144

NOTES ON AFRICAN PLANTS

to 70—140 mm. Flowers in whorls of 2—4 per node;

bracts tapering, apex acute, 10 mm long, 4 mm wide;

pedicels elongate to 25 mm in fruit. Sepals green,

posterior lateral sepal 20 mm long, odd sepal ovate,

acute, gibbous, 10 mm long. Petals red or pink, 16

mm long. Nectary horseshoe shaped. Ovary with 2

or 4 ovules per carpel. Fruit sharply 4 winged,

membraneous to parchment-like, glabrous, veins

prominent, 10 — 15 mm long.

Found in the Cape Province, Namaqualand and

Namibia, Liideritz District, from 27° to 31° S at

altitudes from sea level to 900 m above sea level.

Harvey (1860) misnamed this species M. pectina-

ta. As Melianthus is masculine in its Latinized form,

the name becomes M. pectinatus (Stearn, 1966).

a. subsp. pectinatus

Leaflets 11-27, narrowly linear to lanceolate,

sinuate to dentate; 20—50 mm long, 1-6 mm wide,

stellate hairs on upper surface usually restricted to

canal over veins; inflorescence usually erect; ovules

4 per carpel; capsule parchment-like.

Found in the Cape Province, Namaqualand, from

sea level to 900 m above sea level (Fig. 7).

Cape. — 2816 (Oranjemund): Waterkloof at Doornpoort

(-DA), Pillans 5374. 2817 (Vioolsdrif): 16 km NNE of

Stinkfontein (-CD), Leistner 3387. 2917 (Springbok): Klipfon-

tein ( — BA), Bolus 9485; between Steinkopf and Springbok

(-BD), Verdoorn & Dyer 1817, 8 km S of Springbok, on farm

Voelklip (— DB), Eliovson 13. 2918 (Gamoep): summit of

Kopperberg (-CA), Pillans 5666. 3018 (Kamiesberg): 26 km

NNE of Garies (—AC), Acocks 16469.

Fig. 7. — The geographical distribution of Melianthus pectinatus

subsp. pectinatus (•) and Melianthus pectinatus subsp.

gariepinus (A).

The most satisfactory key character for distinguish-

ing the two subspecies is the number of ovules per

carpel. In subsp. pectinatus there are 4 ovules per

carpel, whereas there are only 2 per carpel in subsp.

gariepinus.

The leaflet margins and numbers can be used to

distinguish the subspecies but this is not as clearcut.

In subsp. gariepinus the leaflets are dentate and

there are 7-9 leaflets. In subsp. pectinatus the

leaflets are sinuate to serrate and there are 11-27

leaflets. With respect to distribution, subsp. pectina-

tus occurs south of the Orange River, whereas

subsp. gariepinus occurs north of the Orange River.

b. subsp. gariepinus (Merxm. & Roessler)

Tansley, comb. nov. et stat. nov. Type: Namibia,

Luderitz-Sud, Numais-Bank, farm Spitskop, Merx-

muller & Giess 3402 (M, holo.; K; PRE!; WIND.,

iso.).

Melianthus gariepinus Merxm. & Roessler in Mitt. bot.

StSamml., Miinch. 7: 1 (1968); Merxm. in F.S.W.A. 76: 2 (1968).

Type: as for subsp. gariepinus.

Leaflets 7—9, lanceolate to oblanceolate, dentate;

50—60 mm long, 5—30 mm wide, stellate hairs on

upper surface not obviously restricted to veins;

inflorescence usually pendulous; ovules 2 per carpel;

capsule membraneous.

Found in Namibia, Luderitz District (Fig. 7).

S.W.A. — 2716 (Witpiitz): Udabib Mountains (-BB)., Muller

809; Luderitz — Slid on farm Spitskop III (-DC), Merxmuller &

Giess 2874; farm Namuskluft (-DD), Giess 12942.

Merxmiiller & Roessler (1968) distinguished M.

pectinatus from their new species, M. gariepinus, on

the strength of M. pectinatus having an erect raceme,

entire leaflets, and more leaflets than M. gariepinus.

It has been seen that M. pectinatus does not have

entire leaflets, the raceme is not always erect, and

the number of leaflets in M. gariepinus merely

continued the range of leaflet numbers in M.

pectinatus from 27 — 11 (M. pectinatus subsp. pectina-

tus) to 9—7 (M. pectinatus subsp. gariepinus). There

is a geographical discontinuity between the subspe-

cies.

Melianthus dregeanus .Sond. in FI. Cap. 1:

(1860) Phillips & Hofmeyer in Bothalia 2: 352

(1927). Type: Eastern Cape Province, grassy place

between Kachu, Geelhout River and Zandplaat,

300-700 m, Drege 4437 (Herb. Sond., holo.)

Shrub. Leaves up to 180 mm long; stipules paired,

cordate, acute, 7—35 mm long, 1.3 mm wide; rachis

unwinged to slightly winged; leaflets 5 — 15, elliptic,

30—80 mm long, 9-30 wide, serrate to dentate,

dorsal surface with sparse to dense covering stellate

hairs, ventral surface a mat of small stellate hairs

with few large stellate hairs. Inflorescence subtermi-

nal raceme, usually pendulous, 40—130 mm long.

Flowers alternate; bracts cordate at base, apex

acute, 10-35 mm long, 3—15 mm wide; pedicels

elongate up to 35 mm. Sepals green with dark red

spot in region of petals, posterior lateral sepal 20-35

mm long, odd sepal linear acute, 15—25 mm long.

Petals reddish, 10—15 mm long. Nectary elongated

anteriorly. Ovary with 2 ovules per carpel. Fruit 4

rounded lobes, woody, pubescent, wider than it is

long, 5-20 mm long.

Found in the eastern Cape between 30° and 33° S

and the Natal Drakensberg between 27° and 29 ° S,

occurring between 900 and 1 800 m above sea level,

on forest margins.

Although Sonder (1860) named this species M.

dregeana, because of the masculine ending of

VARIOUS AUTHORS

145

Melianthus, the specific name should read M.

dregeanus (Stearn, 1966).

a. subsp. dregeanus

Leaves up to 130 mm long; stipules 7—14 mm

long, 1-2 mm wide; leaflets 30—55 mm long, 9—16

mm wide, dorsal surface with sparse covering of

hairs, ventral surface with dense mat of short stellate

hairs. Inflorescence 40-70 mm long. Flowers bracts

10-18 mm long; pedicels up to 25 mm long. Sepals

posterior lateral sepals 20 mm long, odd sepal 15 mm

long. Petals 10 mm long. Fruit up to 10 mm long.

Found in the eastern Cape between 300 and 1 400

m above sea level (Fig. 8).

Cape. — 3126 (Queenstown): Bongola Neck ( — DD), Galpin

7967. 3226 (Fort Beaufort): Katberg (— BC), Sole 2737. 3227

(Stutterheim): Stutterheim (— CB), Rogers 12741; Prospect Farm,

Komga, 700 m (— DB), Flanagan 288.

Fig. 8. — The geographical distribution of Melianthus dregeanus

subsp. dregeanus (A) and Melianthus dregeanus subsp.

insignis (•).

The best character to distinguish the two

subspecies is size. The subsp. insignis is much larger

in all respects. The fact that subsp. insignis has large

emergent stellate hairs on the ventral surface of the

leaflets, whereas subsp. dregeanus does not, can be

used to distinguish the two sub-species. The

distribution also shows a spatial separation.

b. subsp. insignis (Kuntze) Tansley, stat. nov.

Type: Natal, Charlestown, 1 800 m, Kuntze s.n.

(NYBG.; holo.; PRE, photo!).

Melianthus insignis Kuntze, Rev. Gen. 3,2: 43 (1898); Dyer in

Flower. PI. Afr. 33: 1310 (1959). Type: as for subsp. insignis.

Melianthus dregeanus var. insignis (Kuntze) Phill. & Hofmeyer

in Bothalia 2: 351 (1927). Type: as for subsp. insignis.

Leaves up to 180 mm long: stipules 20-35 mm

long, 2-3 mm wide; leaflets 40-80 mm long, 15-30

mm wide, dorsal surface with dense mat of hairs,

ventral surface with dense mat of short stellate hairs

and some large scattered stellate hairs. Inflorescence

60-130 mm long. Flowers bracts 35 mm long;

pedicels up to 35 mm long. Sepals posterior lateral

sepals 35 mm long, odd sepal 25 mm long. Petals 15

mm long. Fruit up to 20 mm long.

Found in the Northern Natal Escarpment between

27° and 29° S, occurring between 900 and 1 800 m

above sea level, usually with Leucosida sericea on

forest margins (Fig. 8).

Transvaal. — 2630 (Carolina): Goede Hoop Ermelo (—DC),

Pott 4938. 2730 (Vryheid): Honeymoon Kloof, Wakkerstroom

(-AC), Galpin 9883.

O.F.S. — 2828 (Bethlehem): Golden Gate {-BC), Jacobs 8554.

Natal. — 2729 (Volksrust): Ingolo, Newcastle ( — DB),

Schweickerdt 981. 2830 (Vryheid): 27 km N of Utrecht (-AD),

Codd & Dyer 6260.

Apart from the larger size, this subspecies is no

different from subsp. dregeanus and therefore has

been included. The distribution is disjunct from the

distribution of subsp. dregeanus.

AN EXCLUDED SPECIES IN MELIANTHUS

Melianthus sibiricus Pall, ex Georgi, J.G. 1775.

Bemerkungen einer Reise in Russischen in Jahre

1772 3 (4): 1114. St Petersburg. This is an error,

cited in Daydon Jackson, Index Kewensis 3: 197

(1894). According to E. Launert (pers. comm.), it

should have read Melianthemum not Melianthus.

REFERENCE

Stearn, W. T., 1966. Botanical Latin. London: Nelson.

S. A. Tansley* & E. A. C. L. E. Schelpe*

* Bolus Herbarium, University of Cape Town, Private Bag,

Rondebosch 7700.

POACEAE

A NEW SPECIES OF EHRHARTA

Ehrharta eburnea Gibbs Russell, sp. nov. E.

capensi Thunb. affinis habitu, sed statura minore,

parvioribus fusiformis eburneis bulbis et vaginis

foliorum basilarium atropurpureis. E. barbinodi

Nees ex Trin. affinis structura spiculae, sed glumis

lemmatibus tantum triplo breviore, et sterilibus

lemmatibus in lateribus sparsim puberulis.

Gramen caespitosum perenne 200—500 mm

altum, interdum rhizomatibus longis gracilibus.

Culmi aggregati non ramosi, internodis imis 1 vel 2

uterque bulbum fusiformem valde costatum politum

eburneum formantis. Folia laminis planis vel

involutis ad 100 x 5 mm, marginibus interdum

undulatis leviter incrassatis; ligulae membranaceae

pilis fimbriatae; vaginae foliorum basilarium atro-

purpureae tenues saepe scissae ita bulbum vel

bulbos praebentes. Inflorescentia panicula angusta

secunda ad 130 x 25 mm, 12-32 patentibus spiculis.

Spiculae 9—13 mm longae; glumis lemmatibus triplo

breviore, plerumque atropurpureis. Flosculi lemma-

tibus sterilibus similaribus sed inaequalibus, primis

146

NOTES ON AFRICAN PLANTS

Fig 9. — Ehrhurta eburnea. Habit. Drawn from Van Breda 4102

and 1409 (spikelets).

nonnihil brevioribus quam secundis, lateribus spar-

sim puberulis longistrorsum porcatis, marginibus

rectis a basi ad apicem, carinis pilis longis mollibus

fimbriatis, basibus barbatis, apicibus aristatis;

lemma fertilis plus minusve primum sterilem lemma

aequans, parum politum. Stamina 6. Caryopsis

obovata, 1,2 mm longa.

Tufted perennial 200—500 mm tall, sometimes

with long slender rhizomes. Culms crowded,

unbranched, with lowest 1 or 2 internodes each

forming a fusiform, strongly ribbed, polished,

ivory-coloured bulb. Leaves with blades flat or

rolled, to 100 x 5 mm, margins sometimes undulate,

somewhat thickened; ligules membranous, fringed

with hairs; sheaths of basal leaves dark purple, thin,

often split, thus revealing bulb. Inflorescence a

narrow secund panicle to 130 x 25 mm, with 12—32

spreading spikelets. Spikelets 9—13 mm long; glumes

3 length of lemmas, usually dark purple. Florets with

sterile lemmas similar but unequal, the first

somewhat shorter than the second, sides sparsely

puberulous, longitudinally ridged, margins straight

from base to tip, keels fringed by long soft hairs,

bases bearded, tips aristate; fertile lemma ±

equalling first sterile lemma, somewhat polished.

Stamens 6. Caryopsis obovate, 1,2 mm long. Figs 9

& 10.

Type. — Cape 3220 (Sutherland): 15 miles from

Sutherland on Calvinia road (-BA), 12", frequent,

mixed karoo and grass, 5 000 ft, Story 4441 (PRE,

holo.; K, iso.).

Known only from localized places in the Calvinia,

Sutherland and Montagu Districts, where it grows in

rhenosterveld, mixed karoo and grassland, often in

rocky places, from 670-1 600 m. Flowering Sep-

tember-November.

Cape. — 3119 (Calvinia): Tierhoek Reserve (-BC), Henrici

3474 (PRE, K); Ekerdam (-BC), L. E. Taylor 2803 (NBG);

along road from Niewoudtville to Oorlogskloof ( — BD), Van

Breda 1409 (PRE); Calvinia (-BD), Agricultural College

Grootfontein 5 (PRE). 3220 (Sutherland): 14 miles north of

Sutherland (-BA), Acocks 15129 (PRE, K). 3320 (Montagu):

Laingsburg Dist., Bantams Karoo (— BB), Compton 12149

(NBG); Farm ‘Soutrivier’ (-BC), Van Breda 4102 (PRE).

Fig 10. — Ehrharta eburnea. a, spikelet; b, first sterile lemma; c,

fertile lemma showing exserted stamens and stigma; d,

palea; e, second sterile lemma; all x 2; f, lodicule x 4; g,

immature caryopsis x 4. Drawn from Van Breda 1409.

VARIOUS AUTHORS

147

E. eburnea is similar in habit to E. capensis

Thunb., though smaller in overall size. Both species

have crowded unbranched culms bearing bulbs at

the lower internodes, and flat leaf blades with

undulate somewhat thickened margins. However,

the bulbs of E. capensis are larger, either cylindrical

or spherical and bright orange or chalk-white,

depending on the variety. E. capensis never has dark

purple basal sheaths. The spikelets of E. eburnea are

quite different from those of E. capensis, but are

nearly identical to those of E. barbinodis Nees ex

Trin. in shape, size, ornamentation and colouring.

However, E. barbinodis has glumes the length

of the lemmas and long hairs along the margins of

the sterile lemmas. The similarity of E. eburnea to

two species, one bulb-forming and the other not,

illustrates the reticulate nature of speciation in

Ehrharta. It is not uncommon for species in groups

formed on the basis of spikelet structure to be very

similar vegetatively to species with different spikelet

types.

A recent hybrid origin for E. eburnea is ruled out

on the basis of the known distribution of the three

species. The range of E. capensis just touches the

range of E. eburnea at Calvinia and Montagu, but E.

barbinodis is a Namaqualand species with a range

separate from that of E. eburnea. Furthermore, E.

capensis and E. barbinodis are sympatric in the

Vanrhynsdorp and Kamiesberg degree squares, but

E. eburnea does not occur there.

The specific epithet eburnea refers to the bulb at

the culm base, which resembles an ivory bead.

G. E. Gibbs Russell*

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

NAME CHANGES AND ADDITIONAL SPECIES OF SOUTHERN AFRICAN POACEAE

A list of 227 name changes and 179 additional

species of southern African grasses, and accepted by

the National Herbarium, Pretoria, since 1955 was

published by Gibbs Russell & Smook (1980). Since

then, 98 name changes and 15 additional taxa have

been accepted by the National Herbarium. The

currently accepted names are listed alphabetically,

with the relevant reference indicated by a numeral

after the name. The additional taxa are listed

alphabetically.

Name Changes

Alloteropsis semialata (R.Br.) Hitchc. subsp. ecklo-

niana (Nees) Gibbs Russell 6

(= A semialata (R.Br.) Hitchc. var. ecklonii

(Stapf) Stapf)

Alloteropsis semialata (R.Br.) Hitch, subsp. semi-

alata6

(= A. semialata (R.Br.) Hitchc. var. semialata)

Andropogon amethystinus Steud ,4

(= A. abyssinicus sensu Chippind. non Fresen)

(= A. pilosellus Stapf)

Andropogon chinensis (Nees) Merr.4

(= A. schinzii Hack.)

Andropogon mannii Hook, f.4

(= A. platybasis J. G. Anders.)

Andropogon schirensis A. Rich.4

( = A. schirensis A. Rich. var. angustifolius Stapf)

Aristida adscensionis L.8

(= Aristida curvata (Nees) Trin. & Rupr.)

Arthraxon lanceolatus (Roxb.) Hochst. var. lanceo-

latus12

(= A. prionodes (Steud.) Dandy)

Axonopus affinis Chase 4

(—A. compressus sensu Chippind., non (Swartz)

Beauv.)

Bothriochloa bladhii (Retz.) S. T. Blake 4

(= B. glabra (Roxb.) A. Camus)

(= B. insculpta (A. Rich.) A. Camus var. vege-

tior (Hack.) C. E. Hubb.)

Brachiaria arrecta (Dur. & Schinz) Stent 4

(= B. latifolia Stapf)

Brachiaria chusqueoides (Hack.) Clayton 1

(= Panicum chusqueoides Hack.)

Brachiaria malacodes (Mez & K. Schum.) Scholz 10

(= B. poaeoides Stapf)

Brachiaria serrata (Thunb.) Stapf 4

(= B. serrata (Thunb.) Stapf var. serrata)

(= B. serrata (Thunb.) Stapf var. gossypinum (A.

Rich.) Stapf)

Brachychloa schiemanniana (Schweick.) S. Phill. 9

(= Heterocarpha schiemanniana Schweick.)

Chrysopogon serrulatus Trin. 4

(= C. montanus Trin. var. tremulus (Hack.)

Stapf)

Cladoraphis cyperoides (Thunb.) S. Phill. 9

(= Eragrostis cyperoides (Thunb.) Beauv.)

Cladoraphis spinosa ( L.f '.) 5. Phill. 9

(= Eragrostis spinosa (L.f.) Trin.)

Coelachyrum yemenicum (Schweinf.) S. Phill. 9

(= Cypholepis yemenica (Schweinf.) Chiov.)

Dichanthium annulatum (Forssk.) Stapf var. papil-

losum (A. Rich.) De Wet & Harlan 4

(= D. papillosum (Hochst.) Stapf)

Digitaria abyssinica (A. Rich.) Stapf 4

(= D. scalarum (Schweinf.) Chiov.)

Digitaria maitlandii Stapf & C. E. Hubb. 4

(= D. apiculata Stent)

D. thouaresiana (Fluegge) Camus 4

(= D. melanochila Stapf)

(= D. tricostulata (Hack.) Henr.)

Digitaria velutina (Forssk.) Beauv. 4

(= D. zeyheri (Nees) Henr.)

Eragrostis hierniana Rendle 3

(= E. uniglumis Hack.)

Eriochrysis brachypogon (Stapf) Stapf 4

(= E. brachypogon (Stapf) Stapf subsp. australis

J. G. Anders.)

Enteropogon prieurii (Kunth) Clayton.2

(= Chloris prieurii Kunth)

Eulalia aurea (Bory) Kunth.4

(= E. geniculata Stapf)

Hordeum stenostachys Godr.u

(= H. compressum Griseb.)

148

NOTES OF AFRICAN PLANTS

Ischaemum fasciculatum Brongn ,4

(= I. arcuatum (Nees) Stapf)

Melica decumbens Thunb.1

(= M. neesii Stapf)

Melica racemosa Thunb.1

(= M. ovalis Nees)

(= M. bolusii Stapf)

(= M. pumila Stapf)

(= M. brevifolia Stapf)

(— M. decumbens sensu Gordon-Gray in Ross FI.

Natal 96 (1972) non Thunb.)

Microstegium nudum (Trin.) A. Camus 4

(= M. capense (Hochst.) J. A. Camus)

Miscanthus Anderss.4

(= Miscanthidium Stapf)

Note: Four other species require new combina-

tions in Miscanthus.

Miscanthus capensis (Nees) Anderss. var. capensis

(= Miscanthidium capense (Nees) Stapf var.

capense )

Miscanthus sorghum (Nees) Pilg.4

(= Miscanthidium sorghum (Nees) Stapf)

Panicum fluviicola Steud.4

(= P. aphanoneurwyi Stapf)

Panicum hymeniochihim- Nees 4

(= P. filiculme Schinz)

(= P. hymeniochilum Nees var. glandulosum

Nees)

(= P. hymeniochilum I^ees var. hymeniochilum )

Paspalidium obtusifolium (Del.) Simpson 4

(= P. platyrrhachis C. E. Hubb.)

Paspalum scrobiculatum L.4

(— P. polystachyum R.Br.)

Pennisetum sphacelatum (Nees) Dur. & Schinz 4

(— S. sphacelatum (Nees) Dur. & Schinz var.

sphacelatum)

(= S. sphacelatum (Nees) Dur. & Schinz var.

tenui folium (Hack.) Stapf)

Pennisetum glaucum (L.) R.Br.4

(= P. americanum (L.) Leeke subsp. america-

num)

Rhynchelytrum grandiflorum Hochst.4

(= R. brevipilum (Hack.) Chiov.)

Rhynchelytrum longisetum (A. Rich.) Stapf & C. E.

Hubb.4

(= R. minutiflorum (Rendle) Stapf & C. E. Hubb.

var. melinioides (Stent) Stapf& C. E. Hubb.)

Rhynchelytrum nerviglume (Franch.) Chiov.

(= R. nyassanum (Mez) Stapf & C. E. Hubb.)

(= R. setifolium (Stapf) Chiov.)

(= R. ramosum Stapf & C. E. Hubb.)

Rhynchelytrum repens (Willd.) C. E. Hubb.4

(= R. villosum (Pari.) Chiov.)

Rhynchelytrum subglabrum (Mez) Stapf & C. E.

Hubb. 4

(— R. suberostratum Stapf & C. E. Hubb.)

Schizachyrium exile (Hochst.) Pilg.4

(=5. inclusion Stent)

Setaria incrassata (Hochst.) Hack.4

(=5. eylesii Stapf & C. E. Hubb.)

(— S. gerrardii Stapf)

(- S. holstii Herr.)

(=5. pabularis Stapf)

(=5. palustris Stapf)

(=5. perberbis De Wit)

(= S. phragmitoides Stapf)

(=5. porphyrantha Stapf)

(=5. rudifolia Stapf)

(=5. xvoodii Hack, subsp. bechuanica De Wit)

(= S. woodii Hack. var. fonssalutis De Wit)

(= S. woodii Hack. var. woodii)

Setaria pseudaristata (Peter) Pilg.4

(= S. tenuiseta De Wit)

Setaria sphacelata (Schumach.) Moss var. torta

( Stapf) Clayton 4

(= S. flabellata Stapf subsp. natalensis De Wit)

(= S. homblei De Willd.)

(= S. torta Stapf)

Setaria sphacelata (Schumach.) Moss. var. splendida

(Stapf) Clayton 4

(= S. splendida Stapf)

Setaria sphacelata (Schumach.) Moss var. sphace-

lata 4

(= S. decipiens De Wit)

(= S. flabellata Stapf subsp. flabellata)

(= S. neglecta De Wit)

(= S. perennis Hack.)

(= S. sphacelata (Schumach.) Moss subsp. aqua-

montana De Wit)

(= S. sphacelata (Schumach.) Moss var. stolonife-

ra De Wit)

(= S. stenantha Stapf)

Setaria sphacelata (Schumach.) Moss var. sericea

(Stapf) Clayton 4

(=5. almaspicata De Wit)

(= S. anceps Stapf)

(=5. cana De Wit)

(=5. flabelliformis De Wit)

(= S. sphacelata (Schumach.) Moss subsp. nodosa

De Wit)

(- S. sphacelata (Schumach.) Moss subsp.

pyropea De Wit)

Sorghastrum stipoides (Kunth) Nash4

(=5. rigidifolium (Stapf) Chippind.)

Sorghum bicolor (L.) Moench. subsp. arundinaceum

(Desv.) De Wet & Harlan s

(— S. verticilliflorum (Steud.) Stapf)

Sorghum bicolor (L.) Moench. subsp. drummondii

(Steud.) De Wet 5

(=5. sudanense (Piper) Stapf)

S. halepense (L.) Pers.5

(=5. almum Parodi)

Thamnocalamus tessellatus (Nees) Soderstrom &

Ellis 11

(= Arundinaria tessellata (Nees) Munro)

Tristachya lualabaensis (De Wild.) Phipps 3

(= T. hitchcockii (C. E. Hubb.) Conert)

Tristachya nodiglumis K. Schum 3

(= T. eylesii Stent & Rattray)

Urelytrum agropyroides (Hack.) Hack.14

(= U. squarrosum Hack.)

Urochloa oligotricha (Fig. & De Not.) Henr.4

(= Urochloa bolbodes (Steud.) Stapf)

Urochloa mosambicensis (Hack.) Dandy 4

(- U. rhodesiensis Stent)

Additional Taxa

Brachychloa fragilis S. Phill.

Bromus rubens L.

Chloris truncata R.Br.

Digitaria comifera Pilg.

VARIOUS AUTHORS

149

D. nuda Schumach.

Echinochloa jubata Stapf

Hordeum marinum Huds. subsp. gussoneanum

(Pari.) Thell.

H. murinum L. subsp. glaucum (Steud.) Tzvel.

H. murinum L. subsp. leporinum (Link) Arcangeli

Leersia tisserantii (A. Chev.) Launert

Melinis ambigua Hack.

Oplismenus undulatifolius (Ard.) Roem. & Schult.

Oxytenanthera abyssinica (A. Rich.) Munro

Panicum comorense Mez

Thelepogon elegans Roem. & Schult.

REFERENCES

1 Clayton, W. D., 1979. Some new African grasses. Kew Bull.

34: 557-560.

2 Clayton, W. D., 1982. Notes on the subfamily Chloridoideae

(Gramineae). Kew Bull. 37: 417-420.

3 Clayton. W. D., Phillips, S. M. & Renvoize, S. A., 1972.

Gramineae (Part II). In R. M. Polhill, Flora of tropical East

Africa. London: Crown Agents.

4 Clayton, W. D. & Renvoize, S. A., 1982. Gramineae (Part

III). In R. M. Polhill, Flora of tropical East Africa. London:

Crown Agents.

5 De Wet, J. M. J., 1978. Systematics and evolution of

Sorghum sect. Sorghum. (Gramineae). Am. J. Bot. 65,4:

477-484.

6 Gibbs Russell, G. E., 1983. The taxonomic position of Cj

and C4 Alloteropsis semialata (Poaceae) in southern Africa.

Bothalia 14: 205—213.

7 Gibbs Russell, G. E. & Ellis, R. P., 1982. The genus Melica

L. (Poaceae) in southern Africa. Bothalia 14: 37-44.

Gibbs Russell, G. E. & Smook, L., 1980. Names of southern

African grasses: name changes and additional species

reported since 1956. Proc. GrassldSoc. sth. Afr. 15: 89-97.

8 Melderis, A., 1971. Aristideae. In A. Fernandes et al., Flora

Zambesiaca 10,1. London: Crown Agents.

9 Phillips, S. M., 1982. A numerical analysis of the

Eragrostideae (Gramineae). Kew Bull. 37: 133-162.

10 Scholz, H., 1978. Bermerkungen iiber Gramineen aus dem

Berliner Herbar: Brachiaria und Megalachne. Willdenowia

8: 383-387.

11 Soderstrom, T. R. & Ellis, R. P., 1982. Taxonomic status of

the endemic South African bamboo, Thamnocalamus

tessellatus. Bothalia 14: 53-67.

12 Van Welzen, P. C., 1981. A taxonomic revision of the genus

Arthraxon Beauv. (Gramineae). Blumea 27: 255—300.

13 Von Bothmer. R., Jacobsen, N. & Nicora, E., 1980.

Revision of Hordeum sect. Anisolepsis Nevski. Bot. Notiser

13: 539-544.

L. Smook* & G. E. Gibbs Russell*

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

NOTES ON SPECIES OF EHRHARTA WITH A SHORT FIRST STERILE LEMMA

Like most other oryzoid grasses, Ehrharta has

spikelets composed of a single fertile floret with two

sterile florets, represented by empty lemmas, below

it. The African species of Ehrharta fall into two

natural groups: those with the first sterile lemma

similar to the second, and the fertile lemma different

from both (about 18 species); and those with the first

sterile lemma reduced and scale-like, and the fertile

lemma similar to the second sterile lemma (2

species).

The specimens with a short first sterile lemma

were classified by Stapf [FI. Cap. 7: 668—670 (1900)]

as five species and one variety. Chippindall (in

Meredith, Grasses & Past. S. Afr., 1955) gave a brief

account of the five species, but did not mention the

variety. Since these studies, many more specimens

of these closely-related taxa have become available,

most of which are the collections of E. Esterhuysen.

It is therefore possible to assess the relationships

within the group more effectively.

In general, Ehrharta species display a reticulate

pattern of relationships, with parallel or convergent

trends in vegetative parts occurring in different

species groups distinguished on the basis of spikelet

similarities. One of the most reliable spikelet

characters is the relative length of glumes to lemmas.

This character, insignificant though it may seem,

helps to identify a number of species and species

groups. Relative glume length is used here to

separate the taxa with short first sterile lemmas into

two species: E. rupestris, with the glumes about 3 the

length of the lemmas, and E. setacea, with the

glumes more than § the length of the lemmas.

Within each of these species, there are several

subspecies connected by intermediates, but there are

no known intermediates between the species. In

both, there are parallel trends for reduced size and

spikelet number, for rolled, folded or flat, erect or

spreading leaf blades, and for suffrutescence.

However, even though superficially similar vegeta-

tive forms occur in both species, each subspecies can

be differentiated on characters besides glume length.

The following key allows identification of the

taxa:

1 Glumes about 3 length of longest lemma 2

2 Plants delicate, herbaceous, less than 250 mm tall;

inflorescence of 1—4 spikelets, barely overtopping

leaves; spikelets 4,5-5 mm long

lc. E. rupestris subsp. dodii

2 Plants ± robust, herbaceous to suffrutescent, 200-450

mm tall; inflorescences of 4-9 spikelets, consider-

ably overtopping leaves; spikelets 4,5— 6,3 mm long 3

3 Leaf blades rolled and appearing setaceous, held

erect, or flat and held ± spreading, tips not

hooded; spikelets to 2 mm across, outline oblong

to linear lb. E. rupestris subsp. tricostata

3 Leaf blades folded, somewhat thickened, held at 45°

to culm, tips hooded; spikelets to 2,5 mm across,

outline oblong to nearly square

la. E. rupestris subsp. rupestris

1 Glumes longer than f (very rarely only 1) length of

longest lemma 4

4 Plants erect, 250—400 mm tall; inflorescences of 5—15

spikelets; glumes appressed to florets at maturity 5

5 Leaf blades tightly rolled, appearing setaceous,

rigid, erect or curved slightly outward from the

middle, texture smooth; spikelets 5,5— 6,5 mm

long 2a. E. setacea subsp. setacea

5 Leaf blades flat, to 6 mm across at base, rolled near

tip, bent 45° outward from culm at sheath, texture

scabrous; spikelets (6,5 — ) 7 — 8 mm long

2b. E. setacea subsp. scabra

150

NOTES OF AFRICAN PFANTS

4 Plants sprawling, or if erect then less than 250 mm tall;

inflorescences of 1—4 spikelets; glumes gaping

widely (more than 45°) at maturity 6

6 Plants sprawling or trailing, culms herbaceous,

lowest nodes bearing leaves with blades;

spikelets 4,5— 6,5 mm long, glumes usually a

little longer than lemmas

2c. E. setacea subsp. uniflora

6 Plants erect, culms suffrutescent below, lowest

nodes usually leafless; spikelets 4—5 mm long;

glumes slightly shorter than lemmas

2d. E. setacea subsp. disticha

1. Ehrharta rupestris Nees ex Trin., Phalaridea,

Zap. Imp. Akad. Nauk., ser. vi., 5(3): 25 (1839).

Type: Caledon Dist., Gnadenthal, Drege (K,

isosyn.; SAM, isosyn.; photo, and fragment in

PRE).

Rhizomatous perennial of widely varying size and

habit, from delicate plants 100 mm tall to robust

suffrutescent plants 450 mm tall. Culms erect or

decumbent, branched and creeping at base, bare of

leaves below. Leaves with blades erect or spreading,

inrolled, folded or flat; ligule a membrane fringed

with hairs; sheaths overlapping. Inflorescence an

erect raceme held above leaves. Spikelets 1—9,

erect, often distichous. Glumes about 3 length of

longest lemma, lower truncate or rounded, upper

acute. Florets with sterile lemmas dissimilar, the first

a thin scale with 3-5 raised nerves, 3—5 length of

second, often appearing to be a third glume; the

second hard and thickened, with 7 minutely

tubercled nerves, tip canoe-shaped; fertile lemma

similar to second sterile lemma, but slightly shorter,

broader and with more acute tip. Stamens 6.

(a) subsp. rupestris. Stapf in FI. Cap. 7: 668

(1900); Chippind. in Meredith, Grasses & Past. S.

Afr. 37 (1955).

An uncommon subspecies represented by scatter-

ed specimens from Worcester and Caledon east-

ward along the Swartberg to Prince Albert. It is most

abundant and appears in its typical form in the

Riviersonderend mountains of the Caledon and

Bredasdorp grids. Along the Swartberg there are

intermediates to subsp. tricostata and, as in that

subspecies, the smallest plants occur at the extreme

east of the range. The folded leaf blades with

hooded tips and the broad spikelets separate this

subspecies from all other taxa in the group. It grows

on mountain slopes among rocks at altitudes of 910

to 1 970 m.

One particularly robust specimen was collected on

coastal sand in the Vanrhynsdorp area (Van Breda

4436), north-west of the range of the rest of the

species. This may prove to be a separate taxon, but

more material is needed.

(b) subsp. tricostata (Stapf) Gibbs Russell,

comb, et stat. nov.

Ehrharta tricostata Stapf in FI. Cap. 7: 669 (1900); Chippind. in

Meredith, Grasses & Past. S. Afr. 35 (1955). Type: French Hoek,

2400 feet, Schlechter 9292 (K, holo.; PRE, iso.).

This is the commonest and most widespread of the

three subspecies of E. rupestris, extending from the

Cape Peninsula north to the Hex River mountains

and western end of the Swartberg, and extending

east along the coastal ranges nearly to Hermanus. It

is variable in habit: plants from the Peninsula and

Caledon are most robust and erect, with either

reduced, setaceous or flat leaf blades; plants from

the Hex River mountains are often decumbent,

herbaceous and with flat leaf blades; and plants from

the easternmost part of the range are small and fine,

although with suffrutescent culm bases. All these

forms grade gradually into each other. Furthermore,

numerous intermediates link this subspecies with the

other two, which may be considered extreme forms.

In habit it is most like E. setacea subsp. setacea, but

that taxon has scale-like overlapping bladeless

sheaths on the lower part of the culms, whereas this

subspecies has blade-bearing leaves nearly to the

culm base. It grows in wet places on mountain slopes

and at the base of cliffs, at altitudes of 300-2 030 m.

(c) subsp. dodii (Stapf) Gibbs Russell, comb, et

stat. nov.

Ehrharta dodii Stapf in FI. Cap. 7: 670 (1900); Chippind. in

Meredith Grasses & Past. S. Afr. 35 (1955). Type: Cape, rocks on

Constantia Berg, Wolley Dod 1961 (K, holo.; BOF, iso.; photo,

and fragment in PRE).

This subspecies occurs only immediately around

False Bay and northwards to the Hex River

mountains. It differs from subsp. tricostata mainly in

its smaller size and fewer spikelets, and is connected

to that subspecies by intermediates of slightly larger

stature and more spikelets. These are found within

the range of typical subsp. dodii as well as to the east

as far as Montagu. The Hex River mountain form of

subsp. tricostata should perhaps be counted among

these intermediates.

Because it is reduced, E. rupestris subsp. dodii is

difficult to distinguish from the two small subspecies

of E. setacea, subsp. uniflora and subsp. disticha,

except on spikelet characteristics. In general, E.

rupestris subsp. dodii is upright with erect rolled leaf

blades, E. setacea subsp. uniflora is sprawling with

flat leaf blades, and E. setacea subsp. disticha is

upright with spreading rolled or folded leaf blades.

However, the habit of each is variable to some

extent. This subspecies grows in wet places among

rocks and at the bases of cliffs, and at altitudes of

660 — 1 660 m.

2. Ehrharta setacea Nees, FI. Afr. Austr. 228

(1841). Type: Cape, in monte tabulari, alt. 3 000 ft,

Drege (PRE, isosyn., fragment).

Erect or trailing tufted perennial with creeping

rhizomes. Culms decumbent or prostrate, suffrute-

scent to herbaceous, often bare of leaves below.

Leaves with blades erect, recurved or spreading,

inrolled, folded or flat; ligule a membrane fringed

with hairs. Inflorescence an erect raceme barely to

considerably overtopping leaves. Spikelets 1-15, at

first erect but spreading at anthesis. Glumes § (rarely

only 5) as long to longer than lemmas, both acute.

Florets as in E. rupestris.

(a) subsp. setacea Stapf in FI. Cap. 7: 668

(1900); Chippind. in Meredith, Grasses & Past. S.

Afr. 37 (1955).

This is the most often collected and the most

widespread of the four subspecies of E. setacea,

occurring on the Cape Peninsula and as far

north-west as Sneeuwkop near Wellington and as far

VARIOUS AUTHORS

151

east as Bredasdorp. In the Klein River Mountains of

the Caledon District it is linked to subsp. uniflora

through intermediates. It grows in damp, peaty or

marshy places, in seepage areas in shale or in wet

sand from Table Mountain sandstone, at altitudes of

660 — 1 515 m.

(b) subsp. scabra (Stapf) Gibbs Russell, stat.

nov.

Ehrharta setacea Nees var. scabra Stapf in FI. Cap. 7: 669

(1900). Type: Cape, in a mountain peak near Swellendam,

Swellendam Div., Burchell 7312 (L, holo.; photo, in PRE).

This subspecies occurs along the Langeberg from

the Clock Peaks above Swellendam to Garcia’s Pass,

east of the range of the other subspecies. In the

Caledon quarter degree square a number of

intermediates link it to subsp. uniflora, which is also

linked in the same area to subsp. setacea. The

subspecies grows mostly in disturbed places on

mountainsides, such as beside paths and in burned

clearings, but is has occasionally been collected

among rocks and in seepage areas, at altitudes of

350-1 212 m.

(c) subsp. uniflora (Burch, ex Stapf) Gibbs

Russell, comb, et stat. nov.

Ehrharta uniflora Burch ex Stapf in FI. Cap. 7: 670 (1900);

Chippind. in Meredith, Grasses & Past. S. Afr. 37 (1955). Type:

Cape Div., Cape Flats near Rondebosch, Burchell 182 (K, holo.;

photo, in PR).

Known only from the Cape Peninsula and

Caledon Districts, this subspecies grades into both

subsp. setacea and subsp. scabra in the Caledon

District, thus linking the two most widespread and

disparate elements in the species. It grows in

seepage areas, marshy places and along water-

courses, and also at forest margins. It occurs at the

lowest altitudes of any taxon in this species group,

from 10—500 m.

(d) subsp. disticha Gibbs Russell, subsp. nov. a

subspecie typica parviore statura ad 250 mm ahum,

1-2 spiculis 4—5 mm longis, glumis hiantibus ad

maturitatem differt. A subspecie uniflora habitu

erecto, culmis basaliter suffruticosis destitutis folios,

glumis lemmatibus paulo brevioribus differt.

Plant perennial, erect, cushion forming, to 250

mm tall. Culms suffrutescent, branched near base,

bare of leaves below. Leaves profuse, with blades

rolled or folded, distichous, usually held at 45° to

culm, uncommonly erect, to 30 mm long; ligules

membranous, fringed with hairs; sheaths strongly

overlapping, with a tuft of hairs at sheath mouth and

auricles. Inflorescence of 1 or 2 spikelets, 5-10 mm

long, barely overtopping leaves. Spikelets 4-5 mm

long, glumes slightly shorter than lemmas, gaping

more than 45° at maturity. Florets with sterile

lemmas dissimilar, the first a thin scale with 3-5

raised nerves, 5— \ length of second, often appearing

to be a third glume, the second sterile lemma hard

and thickened, with 7 minutely tubercled nerves, tip

canoe-shaped; fertile lemma similar to second sterile

lemma, but slightly shorter, broader and with more

acute tip. Stamens 6. Mature caryopses not seen.

TYPE. — Cape 3419 (Caledon): Maanschynkop,

Rocklands Peak (—AD), rocky places on upper

slopes and along ridge, N. aspect, ± 2 500 feet,

Esterhuysen 31735 (PRE, holo.; BOL, iso.)

Known only from quarter degree grid square

3419 AD, where it grows in dry rocky places on

mountain slopes from 580 — 1 225 m. Flowering

October— December.

Cape. — 3419 (Caledon): Babylon’s Tower ( — AD), Esterhuysen

32319 (BOL, PRE), 34755 (BOL, PRE); Maanschynkop, Klein

River Mts (-AD), Esterhuysen 33647 (BOL, PRE); Vogelgat,

south of the Sheiling (—AD), Esterhuysen 35539a (BOL),

Williams 3086 (NBG, PRE); Fernkloof Nature Reserve (-AD),

Williams 2887 (NBG).

This subspecies does not appear to be linked by

intermediates to the other subspecies, even though it

occurs in the only quarter degree square where

intermediates are found between subsp. uniflora and

both subsp. setacea and subsp. scabra. However, in

subsp. disticha the position of the leaf blades is

variable even on the same plant. Normally they are

held at 45° to the culm as in subsp. scabra, but

occasionally they are held erect as in subsp. setacea.

G. E. Gibbs Russell*

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

SIMAROUBACEAE

FLOWERING IN KIRKIA W1LMSII ENGL.

Some plants, such as cycads, are known to

sometimes change sex under stress, e.g. when

transplanted or damaged, or when growing under

unfavourable conditions. Others, e.g. Cannabis

sativa, will change sex ratios gradually over the

growing season, the sex expression probably being

controlled by daylength. However, no record has

been found in the literature of a species in which the

two sexes alternate a number of times in one

flowering season, such as has been recently observed

in Kirkia wilmsii.

The genus Kirkia is a tropical African genus of

five species, recently revised by B. Stannard [ Kew

Bull. 35,4 : 829 (1981)]. It is usually placed in the

Simaroubaceae, but some consider it to constitute a

family on its own, the Kirkiaceae (Engl.) Tahkt. K.

wilmsii Engl, is endemic to the Transvaal lowveld,

where it is a common tree on rocky hillslopes and in

kloofs, occurring on both granitic and dolomitic

soils. The leaves are deciduous, and at anthesis the

new leaves are still immature. Flowers are produced

in axillary panicles in spring.

During 1981 and 1982 observations were made on

three trees of this species growing on a rocky,

north-facing slope in the Pretoria National Botanical

Garden. These three trees were growing outside the

natural range of the species, though only by a matter

of some kilometres. The nearest naturally-growing

152

NOTES OF AFRICAN PLANTS

Fig. 11. — Flowers of Kirkia wilmsii, x 6: 1, male; 2, female.

population I know of is at the Pyramids, near

Onderstepoort, about 15 km north of the Botanical

Garden.

Flowering season

The date of commencement of the flowering

season varied slightly, and was probably affected by

the weather. In 1981 the first flowers opened in the

1st and 2nd week of November but, in 1982, which

had a milder and less frosty winter, anthesis

occurred in the 3rd and 4th week of October. This

effect may be enhanced by the fact that the trees are

growing outside the natural range of the species. The

average duration of flowering during both years was

35 -(37) -39 days, with one specimen flowering for

an abnormal 14 days in 1982. Again, this may differ

in warmer areas.

Whether or not a tree flowers, appears to depend

on its size, but even large trees may not flower for a

season, or the inflorescences may die off soon after

being produced. It was also observed a number of

times that a tree would produce no flowers for as

long as eight days, then continue with the flowering

cycle as before.

Flower sex

When specimens of K. wilmsii are seen in the

herbarium, the species appears dioecious, with only

male or female flowers apparently present on a given

specimen. However, if nearly mature buds are

opened, it can be seen that they are always of the

opposite sex to the open flowers. It was decided to

observe the tres themselves and record the sex of

open flowers daily during the flowering season.

The flowers themselves were always unisexual,

and I have so far not seen one which could be

1981 (15 = 10 - 25:12) 1 cm = 10 days

considered hermaphrodite. The male flowers have

large anthers, long filaments, and a gynoecium

reduced to no more than a small pyramid of tissue in

the centre of a well-developed, nectiferous disc.

Female flowers, on the other hand, have a

well-developed gynoecium with four long, fused

styles and four fused, globular stigmas, and the

anthers are reduced and sterile with short filaments.

Both filaments and anthers in female flowers are

about 3 of the size of those in male flowers (Fig. 11).

Every tree produced flowers of both sexes, but

usually flowers of only one sex were open at any one

time on a given tree. When buds were examined, it

was found that the older buds were of the opposite

sex to the open flowers, and the younger buds were

of the opposite sex to the older buds. These open in

successive flushes, so that the tree, which is

morphologically monoecious, is functionally dioeci-

ous. It would be most interesting to follow up the

physiology and see what changes trigger the

development of flowers into either male or female.

Some overlap may occasionally occur, but usually

a tree which appeared completely male one day will

change to female overnight. Often where it seemed

that flowers of different sexes were overlapping in

time, it was found that either the female flowers had

blackened, probably non-receptive stigmas, or that

the male flowers had shed their pollen.

In this way, outcrossing is ensured, but each

individual tree still produces seed. The length of

time each flush of flowers remains open is very

irregular, thus staggering the presence of different

sexes even in a very small population.

No tendency was observed, either within the

species as a whole or on one tree, for either male or

female flowers to open first in the season. The length

of time flowers of a sex remained open varied

greatly, with each flush of female flowers lasting

3— (6) — 12 days, and of male 4— (13)— 17 days.

(Number in brackets is the mean.) The flush of male

flowers tended to last longer than the female. This

change of sex occurred 3—6 times in each season

(Fig. 12).

K. L. IMMELMAN*

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

1982 (15:10 -25=12)

^ 1 :

r ® — m

?

o'

2.

¥

Fig 12.— Duration of flowering of either sex in Kirkia wilmsii.

Bothalia 15, 1 & 2: 153-159 (1984)

Leaf anatomy of the South African Danthonieae (Poaceae). IX

Asthenatherum glaucum

R. P. ELLIS*

Keywords: Asthenatherum glaucum , C4, Danthonieae, Kranz, leaf anatomy

ABSTRACT

The leaf blade anatomy of Asthenatherum glaucum (Nees) Nevski was studied. Detailed descriptions of the leaf

section and abaxial epidermis are given and illustrated by means of photomicrographs. The leaf anatomy of all

specimens is undoubtedly Kranz with radiate chlorenchyma and specialized parenchyma bundle sheath

chloroplasts. This observation is the most important justification for separating Asthenatherum from Danthonia,

which has non-Kranz anatomy.

Significant epidermal variation was observed, particularly in the form and occurrence of macro-hairs and

prickles. These epidermal differences appear to correlate with geographical distribution and growth form of the

plants and may be of infraspecific taxonomic significance. The epidermal variation is continuous, but three more or

less distinct groups can be recognized: annual plants with unique macro-hairs with corrugated walls; annual or

perennial plants without macro-hairs but with large straight prickles without swollen bases; and perennial plants

without these large prickles. These groups appear to represent an intergrading cline along an increasing moisture

gradient eastwards from the Namib Desert in South West Africa/Namibia.

INTRODUCTION

Asthenatherum glaucum (Nees) Nevski is a

densely tufted, coarse perennial with culms much

branched from a stout, woody rhizome (Chippindall,

1955; Conert, 1962; Hubbard, 1970; Launert, 1970).

The base of the plant is clothed in cataphylls which

may be glabrous and papery or hairy.

All the above authors stress the strong perennial

nature of A. glaucum, yet in the National Herbarium

(PRE) a significant proportion of the collection of

this species consists of what are undoubtedly annual

plants. These specimens all display inflorescences,

although obviously only in their first year of growth

and it appears to be quite common for this perennial

grass to flower in its first year. In common with

many other grasses from arid environments, these

neotonous plants apparently behave exactly as

short-lived annuals unless sufficient moisture is

available to allow the establishment of a plant of

sufficient critical mass to overcome the regular

periods of drought, which frequently occur in the

desert regions it inhabits. This phenomenon has

resulted in a certain degree of taxonomic confusion

and also led to problems with identification within

this genus.

A. glaucum is a hardy species which is obviously

well-adapted to climatic extremes of moisture,

temperature and insolation. It inhabits the desert

and semi-desert areas of the Namib, Namaqualand,

the Orange River Broken Veld and the Kalahari

Thornveld. This grass favours a loose, sandy

substrate and occurs almost exclusively on sand

dunes of coarse to fine windblown sand.

Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

In the Namib Desert proper, A. glaucum is

common on the lower parts of the sand dunes. The

annual plants that have been collected all come from

the Namib Desert and there appears to be a certain

degree of niche separation between plants displaying

the annual habit and the perennial specimens.

Therefore, most collections of annual specimens

have been made from the fine gravel plains in the

interdune areas. This observation may be relevant in

future taxonomic considerations.

Away from the Namib Desert, A. glaucum

appears to be restricted to areas of deep, loose, red

sand, such as the Kalahari sand dunes. Only plants

with a perennial growth form have been recorded

from these areas.

Asthenatherum was originally described as be-

longing to the genus Danthonia DC. In 1934 Nevski

separated it from Danthonia, a decision which has

now been generally accepted (Conert, 1962;

Launert, 1970; Hubbard, 1970), although Hubbard

(1937) originally did not adopt Nevski’s genus due to

practical considerations. However, he did consider

the species that had been transferred to Asthenathe-

rum as forming a natural group. Chippindall (1955)

also retained the genus Danthonia for the species

presently assigned to Asthenatherum.

Asthenatherum can easily be distinguished from

typical Danthonia by the presence of a long, pungent

callus (Conert, 1971) as well as by the many-nerved

glumes separated by a distinct internode and the

transverse line of hairs on the lemma (Hubbard,

1970). The most meaningful diagnostic character,

however, is the presence of radiate chlorenchyma

which no other Danthonia species possesses. This

characteristic of Asthenatherum has been known for

almost fifty years (Hubbard, 1937) and constitutes

the most important justification for separating

Asthenatherum from Danthonia.

154 LEAF ANATOMY OF THE SOUTH AFRICAN D ANTHONIE AE (POACEAE). IX. ASTHENATHERUM

GLAUCUM

The generic status of Asthenatherum* is, therefo-

re, now generally accepted but considerable

confusion still surrounds the identity of the species

assigned to this genus — this in spite of a detailed

study of the whole genus by Conert in 1962. Conert

(1962) recognized five taxa in the genus (4 species

and 1 variety) of which four were recorded as

occurring in southern Africa. Launert (1970) and the

National Herbarium (PRE) (L. Smook, pers.

comm.) recognize only A. glaucum and A.

mossamedense (Rendle) Conert from South West

Africa/Namibia and the Flora of Southern Africa

region. The variety, A. glaucum var. lasiophyllum

(Pilg.) Conert, has not been accepted. A. forskalii

(Vahl) Nevski, a species distributed throughout

North Africa and the Middle East, but also recorded

from southern Angola (Conert, 1962, 1971) is not

recognized in the Flora of South West Africa

(Launert, 1970), although it very possibly occurs in

this territory.

Disjunct distributions between many arid area

species, which occur in the northern as well as in the

southern desert areas of Africa, are well documen-

ted (Monod, 1971; De Winter, 1971). A. forskalii is

considered to be an example of a taxon displaying

this distributional phenomenon by the above authors

as well as Conert (1971). However, Hubbard (1937)

comments that the specimens of A. forskalii from

Angola possess more deeply-lobed lemmas than

those from North Africa and one specimen ( Pearson

2174) has fulvous hairs on the lemmas. The identity

of the Angolan specimens requires verification in the

light of this statement and because specimens

collected in the extreme north of South West

Africa/Namibia, in almost identical habitats to those

of Mossamedes further north in Angola, have been

determined as being A. glaucum. Chippindall (1955)

also records A. forskalii from South West Africa.

Conert (1962, 1971) considered A. glaucum as

being endemic to South Africa and this view has

been confirmed by Monod (1971) and De Winter

(1971). However, Hubbard (1970) records A.

glaucum from the Turkana Desert in northern

Kenya. This observation implies that A. glaucum

also has a widely disjunct distribution. In addition, it

also raises the probability that A. glaucum and A.

forskalii are conspecific. It is remarkable that such

uncertainty still exists in a genus which has received

considerable taxonomic attention relatively recently.

Two other species are included in the genus: A.

fragile (Guinet & Sauvage) Monod is another

species from North Africa and Arabia and A.

mossamedense (Rendle) Conert is endemic to South

West Africa and Angola (Conert, 1971).

In the present study freshly fixed material of only

A. glaucum was available for study. Leaf anatomy

and epidermal characteristics were determined and

compared with the anatomy of several preparations

* Subsequent to going to press, the name Asthenatherum Nevski

has been replaced by Centropodia Reichb, and the new

combinations C. glauca (Nees) T. A. Cope and C. mossamedensis

(Rendle) T.A. Cope have been published by T.A. Cope in Kew

Bull. 37,4: 657-659 (1983).

prepared from herbarium material. These specimens

had been variously identified as A. forskalii, A.

glaucum var. lasiophyllum and A. mossamedense.

However, the A. forskalii specimens were later

determined as being var. lasiophyllum by Conert in

1973 but cited as A. glaucum by Launert (1970). All

the specimens previously quoted as being var.

lasiophyllum are now considered to be A. glaucum

by the National Herbarium. This study was

consequently restricted to A. glaucum with a few

comparisons being made with material prepared

from herbarium material of A. mossamedense. In

the anatomical descriptions which follow, the

terminology of Ellis (1976, 1979) will be followed

and the following abbreviations will be used:

vb/s — vascular bundle/s

l’vb/s — first order vascular bundle/s

3’vb/s — third order vascular bundle/s

ibs — inner bundle sheath; mestome sheath

obs — outer bundle sheath; parenchyma sheath

ANATOMICAL DESCRIPTION OF ASTHENATHERUM

GLAUCUM (NEES) NEVSKI

Leaf in transverse section

Leaf outline: broadly U-shaped to expanded and

flat; relatively thick (0,30 mm— 0,50 mm). Ribs and

furrows: very slight (Figs 2 & 4) to medium (Figs 1 &

5) adaxial ribs and furrows; furrows between all vbs;

adaxial ribs rounded; similar ribs over all vbs.

Abaxial ribs and furrows more pronounced than

adaxial ones; abaxial ribs somewhat flattened (Figs 1

& 5) and furrows shallow to medium depth. Median

bundle: not distinguishable structurally from other

l’vbs. Vascular bundle arrangement: 7, 9 or 11 l’vbs

in section; 3 3’vbs between consecutive l’vbs (Figs

1-4); seldom 2 (Fig. 5) or 4; no 2’vbs; all vbs

centrally located in blade. Vascular bundle structure:

3’vbs elliptical to tall and narrow (Figs 3 & 6); well

developed xylem and phloem and ibs. l’vbs elliptical

(Figs 3 & 6); phloem adjoins ibs; metaxylem vessels

narrow and circular; tend to have smaller diameter

than even the ibs cells. Vascular bundle sheaths:

elliptical to almost rounded; double; both sheaths

entire (Figs 1-6); no extensions; parenchyma sheath

cells numerous (>15), uniform in size and shape,

fan-shaped with straight radial walls and inflated

tangential walls (Figs 3 & 6); specialized, centripetal

chloroplasts conspicuous (Figs 3 & 6); ibs complete;

relatively large, unthickened cells particularly

noticeable in 3’vbs (Fig. 3). Sclerenchyma: small

adaxial strands associated with all vbs (Figs 3 & 6);

well developed abaxial strands associated with all

vbs; arched, follow shape of ribs (Figs 3 & 6);

sometimes with minute contact with obs cells (Fig.

6) . Fibres thick walled; cellulose, not lignin,

thickening. No sclerenchyma between bundles.

Margin with small sclerenchyma cap. Mesophyll:

radiate chlorenchyma (Figs 3 & 6); tabular cells

surround bundles completely (except where girders

in contact with bundle sheaths); successive chloren-

chyma groups almost completely separated by

bulliform cells. No colourless cells associated with

bulliform cells. Adaxial epidermis: fan-shaped

bulliform cell groups with central cell shield-shaped

(Fig. 6) or even diamond-shaped (Fig. 3); occupy

R. P. ELLIS

155

about half leaf thickness. Epidermal cell walls

unthickened; macro-hairs variable; absent (Figs

4-6) or present (Figs 1—3); slender and elongated;

costal; could be classified as prickles; bases

superficial (Fig. 3); small hooks present when

macro-hairs absent (Fig. 5); no papillae. Abaxial

epidermis: small fan-shaped bulliform cell groups at

bases of furrows; macro-hairs or prickles either

absent (Figs 4—6) or common (Figs 1-3); hairs

shorter and thicker than adaxial prickles and tend to

be interlocking; hooks present on specimens without

prickles; no papillae.

intercostal zones of specimens lacking prickles (Figs

10-12); unbarbed asperites may be present in costal

zones (Fig. 12). Micro-hairs: not observed on any

preparation examined. Macro-hairs: extremely

elongated, unicellular macro-hairs present on some

specimens (Figs 13-16); few specialized epidermal

cells associated with base (Fig. 16); base of hair

constricted and somewhat superficial (Fig. 14);

unique unrecorded type of hair with corrugated cell

walls; can be abundant or absent altogether; always

associated with specimens with prickles that resem-

ble short macro-hairs; never with specimens with

Figs 1—6. — Asthenatherum glaucum: transverse sections of the leaf blade. 1—3, form with well developed epidermal prickles. 1,

Ellis 2141, x 100; 2, Ellis 2178, x 100; 3, Ellis 2178, X 250. 4—6, form without elongated epidermal prickles. 4, Ellis 872, x

100; 5, Ellis 3605, x 160; 6, Ellis 3605, x 400.

Abaxial epidermis in surface view

Intercostal long cells: elongated, side walls parallel;

slightly undulating particularly in areas adjoining

costal zones (Figs 11 & 12); prickles or hooks

between long cells (Figs 8 & 9, 11 & 12). Stomata:

low to tall dome-shaped (Figs 9, 11 & 12) stomata

present in all files of intercostal cells; interstomatal

cells elongated; 1 or 2 between successive stomata in

a file. Intercostal short cells: absent except those

associated with hooks on edges of zone (Fig. 11).

Papillae: no papillae present. Prickles: large,

straight, sharp-pointed without swollen bases;

resemble short macro-hairs; interlocking over

intercostal zones (Figs 7-9); not found on all

specimens; both costal and along the margins of the

costal zones. Hooks: small, sharp-pointed hooks in

hooks only. Silica bodies: vertically elongated (Figs 9

& 12) to cuboid (Figs 9 & 11) and even dumb-bell

shaped (Figs 11 & 16); occur throughout costal

zones; sometimes associated with cork cells; often

not.

Specimens examined:

A. glaucum

S.W.A. — 2314 (Sandwich Harbour): Gobabeb Research

Station, Namib (-BD), Koch s.n.; S of Kuiseb River at

Gobabeb, Jensen 163*; Strey 2593*. 2419 (Aranos): Farm Bethel,

E of Aranos (-BD), Van Vuureti & Giess 1124 +. 2615 (Luderitz):

* Specimens personally determined by H. J. Conert in 1973 as

being A. glaucum var. lasiophyllum.

+ Specimens with reduced prickles on both epidermides.

156 LEAF ANATOMY OF THE SOUTH AFRICAN D ANTHONIE AE (POACEAE). IX. ASTHENATHERUM

GLAUCUM

% W

... ST!

'■m w

TS . JC

ca O:

Figs 7-12. — Asthenatherum glaucum: abaxial epidermis in surface view. 7-9, form with well developed epidermal prickles. 7, Ellis

2139, x 100; 8, Ellis 2178, x 250; 9, Ellis 2177 , x 250. 10—12, A. glaucum form with small epidermal hooks. 10, Ellis 872, x

100; 11, Ellis 872, x 250; 12, Van Vuuren & Giess 1124, x 250.

120 km W of Aus on road to Luderitz (-DB), De Winter & Hardy

7890*.

Cape. — 2816 (Oranjemund): 2 km W of Beesbank on road to

Alexander Bay (-DA), Ellis 2177, 2178, 2179. 2822 (Glen Lyon):

Matsap (-DB), Ellis 3605+. 2919 (Pofadder): 21 km E of

Pofadder on road to Kakamas ( — BA), Ellis 2139, 2140, 2141,

2142. 2921 (Kenhardt): 12 km from Kakamas turn-off on

Kenhardt-Upington road (-AC), Ellis 872+\ 15 km from

Kenhardt on Kakamas road, Botha & Panagos 21+.

A. mossamedense

S.W.A. — 1812 (Sanitatas): Orupembe waterhole ( — BA), De

Winter & Leistner 5725. 2016 (Otjiwarongo): 112 km W of

Welwitschia on road to Torra Bay (-AA); De Winter & Hardy

* Specimens personally determined by H. J. Conert in 1973 as

being A. glaucum var. lasiophyllum.

+ Specimens with reduced prickles on both epidermides.

8152. 2115 (Karibib): 300 km W of Windhoek at Khomas

Hochland-Swakopmund junction ( — DD), De Winter & Hardy

8021.

DISCUSSION AND CONCLUSIONS

Taxonomically the most significant features of the

leaf anatomy of A. glaucum are the radially arranged

chlorenchyma and the specialized chloroplasts in the

outer bundle sheath cells. This characteristic Kranz

anatomy implies that A. glaucum possesses the C4

photosynthetic pathway. This has been confirmed by

the determination of 12C/13C ratios of representative

specimens (6 = — 15,l%o, Ward 161; 6 = — 14,l%o,

Leistner 2006). These ratios are typical of C4 grasses

and corroborate the anatomical indications.

A. mossamedense also has Kranz anatomy (Fig.

R. P. ELLIS

157

Figs 13—16. — Abaxial epidermis of Asthenatherum glaucum var. lasiophyllum specimens with elongated macro-hairs. 13 — 14,

Jensen 163; 13, x 160; 14, X 250. 15 — 16, Strey 2593; 15, x 160; 16, X 250.

18 234

irh

Figs 17—18. — Anatomy and epidermal structure of Asthenatherum mossamedense, De Winter & Leistner 5725. 17, x 250, leaf in

transverse section; 18, x 250, abaxial epidermis.

17) and a C4 12C/13C ratio (6 = — 12,6%o, De Winter

& Hardy 8021). The anatomy of A. forskalii is

similar to that of A. glaucum (Conert, 1962) and,

therefore, it can be concluded that Asthenatherum is

a Kranz genus which exhibits the C4 pathway.

This observation is important because, in the

Danthonieae, virtually all genera are non-Kranz

(Brown, 1977; Renvoize, 1981) and the C4 pathway

is primarily associated with all the genera of the

Eragrostoideae and most of the Panicoideae. Of the

African representatives of the Danthonieae, De Wet

(1956) considered both Asthenatherum and Alloeo-

chaete to have panicoid anatomy and, consequently,

to be C4. Brown (1977) followed De Wet but did not

actually examine any material of Alloeochaete.

When this was done by Renvoize (1981), this genus

was found to have non-Kranz anatomy. Asthenathe-

rum is, therefore, the only South African genus of

the Danthonieae which is Kranz and, consequently,

is of particular phylogenetic significance.

On a worldwide basis, Asthenatherum , together

with Pheidochloa of Australia, are the only C4

genera of the Danthonieae recognized by Brown

(1977). Renvoize (1981) widened the concept of this

tribe (which he included in the Arundineae) and

incorporated several other Kranz genera such as

158 LEAF ANATOMY OF THE SOUTH AFRICAN D ANTHONIE AE (POACEAE). IX ASTHENATHERUM

GLAUCUM

Triraphis, Eriachne and Neyraudia. In the sub-

family Arundinoideae, Aristida and Stipagrostis are

additional C4 genera placed in the tribe Aristideae

(Renvoize, 1981).

The classification of Asthenatherum is, therefore,

critical to a better understanding of the derivation of

the Kranz syndrome and the taxonomy of this

interesting genus requires careful consideration. The

Arundinoideae, with the predominance of the C3

photosynthetic pathway, their rather limited ecolo-

gical success and their unspecialized spikelet

morphology appear to be a relatively primitive

group (Renvoize, 1981) and may have given rise to

some of the more modern sub-families. They

probably evolved from the ancestral grass type

before the other major sub-families became estab-

lished and from these pioneer Arundinoideae the

grasses radiated into the three most successful

groups extant today (Renvoize, 1981).

The leaf anatomy of Asthenatherum, with a Kranz

outer, parenchymatous sheath with centripetally

arranged chloroplasts resembles many of the

Eragrostoideae, some Paniceae and Stipagrostis

(Brown, 1977; Ellis, 1977). In the Eragrostoideae

these Kranz parenchyma sheath cells are, in

paradermal view, short and radially wide, whereas in

Asthenatherum they are longer than they are wide

(Brown, 1974). In addition, the silica cells of the

Eragrostoideae differ considerably from those of

Asthenatherum and any affinities with this sub-

family appear remote. Panicum species with

centripetal chloroplasts also have short and radially

wide Kranz cells and differ from Asthenatherum in

this respect. Stipagrostis, however, has elongate

Kranz sheath cells which resemble those of

Asthenatherum. This observation supports the

placement of both these genera in the same

sub-family.

Micro-hairs are uniform in shape and consistent in

occurrence and provide useful characters in grass

classification (Renvoize, 1981). De Wet (1954)

considered bicellular micro-hairs to be absent in A.

glaucum and A. forskalii but noted their presence on

A. mossamedense. Unfortunately no micro-hairs

were observed on the epidermis of any of the 18

specimens of Asthenatherum examined in this study

(including 3 of A. mossamedense). They, therefore,

appear to be of very infrequent occurrence.

Fortunately Palmer & Tucker (1981) give very clear

scanning electron photomicrographs of micro-hairs

on both the abaxial and adaxial epidermis of A.

glaucum ( Rains <£ Yalala 22, Botswana). These

micro-hairs are finger-like with a noticeably longer

basal cell. The apical cell is much shorter, usually

deflated and with a rounded apex.

The ratio of distal to basal cell length, together

with the rounded apex, are similar to the condition

in the danthonoid grasses studied by Tateoka et al.

(1959) and appears to lend weight to the classifica-

tion of Asthenatherum in the Arundinoideae. This

micro-hair shape is definitely not of the eggshape

type characteristic of the Eragrostoideae and the

rounded apex makes panicoid relationships unlikely.

E Stipagrostis also has a more tapering distal cell than

illustrated for A. glaucum.

Leaf anatomy, therefore, does not appear to

suggest alternative affinities for A. glaucum and it

seems to be best placed in the Arundinoideae

together with the danthonoid grasses. However, the

presence of the Kranz syndrome in Asthenatherum

implies that, in the Danthonieae, the Kranz

syndrome must have evolved at least once and

probably on several independent occasions to

produce Pheidochloa and the other genera currently

classified with the danthonoid group. Eragrostoi-

deae, on the other hand, probably all derive from a

single evolution of the syndrome.

Within the Danthonieae, Asthenatherum appears

to occupy an isolated and peripheral position. Its

closest relative is apparently Dregeochloa, with

which it shares certain anatomical characters (Ellis,

1977a). Dregeochloa, is a non- Kranz genus, how-

ever, but does share similar desert habitats, in the

Namib and Namaqualand, with Asthenatherum.

Considerable variation in epidermal structure was

observed within the sample of A. glaucum specimens

studied. Structures which contributed to this

variation were the unique corrugated macro-hairs

which were either present or absent and the length

of the interlocking prickle hairs. These were

sometimes reduced to unbarbed appendages or even

typical hooks. The quantitative nature of the

presence of macro-hairs represents a continuum with

the hairs being either abundant (Figs 13 & 14),

infrequent (Figs 15 & 16) or absent (Figs 7-12).

Similarly, variation in prickle hair length is

continuous and well-developed barbs (Figs 7—9)

intergrade with typical hooks (Figs 10-12).

Due to the continuous nature of the variation of

these characters, there appears to be no justification

for recognition of infraspecific taxa within A.

glaucum. Nevertheless, it seems that the epidermal

variation observed represents a definite ecological

trend associated with a moisture gradient from the

very arid Namib eastwards into the Kalahari. This

clinal variation may not be taxonomically meaning-

ful, but undoubtedly represents ecotypic adaptation

to differing environmental conditions. The genetic

basis of these differences is, however, unknown.

Three groups within A. glaucum can be recogni-

zed on the basis of this epidermal structure. The first

group possesses long macro-hairs with corrugated

walls as well as interlocking prickles (Figs 13 — 16). It

must be noted, however, that the epidermal

preparations examined in this study were prepared

from the mid-lamina region of the leaf blade and

only the abaxial surface was studied. Nevertheless,

the specimens that possessed these unique macro-

hairs had all been determined by Conert (1962) as

being var. lasiophyllum using these hairs as a

diagnostic criterion. It must be pointed out,

however, that these hairs occur on the leaf sheaths of

many of the A. glaucum specimens examined in this

study but are absent on the leaf blades. However,

those specimens with cushion hairs on the blades are

all genuine annual plants and occur only in the true

R. P. ELLIS

159

Namib Desert where they are found in the fine

gravel soils of the dune strates.

A. glaucum specimens with interlocking prickles

but lacking cushion hairs (Figs 7—9) are usually

perennial plants but may flower in their first year of

growth and thus appear annual. They also occur in

the Namib Desert but seem to favour the lower parts

of sand dunes in a sandy substrate. This form also

has a wider distribution than the first, occurring

southwards as far as Namaqualand and the Orange

River Broken Veld and extending eastwards into the

pre-Namib areas. Away from the Namib Desert

proper, only perennial plants have been collected.

Further to the east, in the Kalahari and northern

Cape, A. glaucum is restricted to Kalahari dunes on

deep, loose, red sand. All these plants are densely

tufted perennials. Specimens collected from these

areas show a considerable reduction in the length of

the interlocking prickles which, in these cases, can

only be termed intercostal hooks and unbarbed

costal asperites (Figs 10—12).

Superficially the epidermal structure of these

three groups is distinct but it must be remembered

that they gradually intergrade into one another and

the variation is not disjunct. Recognition of

infraspecific taxa does, therefore, not seem justified

but cognizance of this variation must be taken for

purposes of microscopic identification for applied

purposes.

A. mossamedense was not studied in detail during

this study because no freshly fixed material was

available. However, the leaf anatomy appears to be

similar to that of A. glaucum (Figs. 17) with the

epidermis exhibiting a further reduction in the

presence of prickles until they are totally absent (Fig.

18). Some specimens have numerous unbarbed

costal prickles. In this respect, A. mossamedense

may merely represent a continuation of the

reduction trend described in A. glaucum. It is

distinct morphologically, however, (Conert, 1962),

and the silica bodies are much more regularly

dumb-bell shaped than in A. glaucum and,

consequently, its specific status is not questioned.

ACKNOWLEDGEMENTS

The author is grateful to Miss H. Botha for

technical assistance, to Mrs A. J. Romanowski for

the photography and Mrs S. M. Thiart for typing the

manuscript. Dr J. C. Vogel of the Natural Isotopes

Division, NPRL, CSIR, kindly did the 12C/13C

determinations.

UITTREKSEL

Die anatomiese struktuur van Asthenatherum

glaucum (Nees) Nevski blare is bestudeer. Beskry-

wings van die blaaranatomie in dwarssnee en van die

abaksiale epidermis word gegee en geillustreer deur

middel van fotomikrograwe. Alle monsters wat

ondersoek is, toon duidelike Kranz anatomie met

chlorenchiem radiaal gerangskik en gespesialiseerde

chloroplaste in die parenchiem bondelskede. Hierdie

waarneming regverdig die skeiding van Asthenathe-

rum vanaf Danthonia wat nie- Kranz anatomie toon.

REFERENCES

Brown, W. V., 1974. Another cytological difference among the

Kranz subfamilies of the Gramineae. Bull. Torrey bot. Club

101: 120-124.

Brown, W. V., 1977. The Kranz syndrome and its subtypes in

grass systematics. Mem. Torrey bot. Club 26: 1-97.

Chippindall, L. K. A., 1955. In Meredith, The grasses and

pastures of South Africa. Johannesburg: CNA.

Conert, H. J., 1962. Uber die Gramineen-Gattung Asthenathe-

rum Nevski. Senckenberg. biol. 43: 239—266.

Conert, H. J., 1971. The genus Danthonia in Africa. Mitt. bot.

StSamml. , Munch. 10: 299-308.

De Wet, J. M. J., 1954. The genus Danthonia in grass

phytogeny. Am. J. Bot. 41: 204-211.

De Wet, J. M. J., 1956. Leaf anatomy and phytogeny in the tribe

Danthonieae. Am. J. Bot. 43: 175-182.

De Winter, B., 1971. Floristic relationships between the

northern and southern arid areas in Africa. Mitt. bot. St-

Samml., Miinch. 10: 424—437.

Ellis, R. P., 1976. A procedure for standardizing comparative

leaf anatomy in the Poaceae. I. The leaf blade as viewed in

transverse section. Bothalia 12: 65—109.

Ellis, R. P., 1977. Distribution of the Kranz syndrome in the

Southern African Eragrostoideae and Panicoideae according

to bundle sheath anatomy and cytology. Agroplantae 9:

73-110.

Ellis, R. P., 1977a. Leaf anatomy of the South African

Danthonieae (Poaceae). I. The genus Dregeochloa. Bothalia

12: 209-213.

Ellis, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in

surface view. Bothalia 12: 641-672.

Hubbard, C. E., 1937. In A. W. Hill, Flora of Tropical Africa,

Vol. 10. London: Crown Agents.

Hubbard. C. E., 1970. In E. Milne-Redhead & R. M. Polhill,

Flora of Tropical East Africa. Gramineae Part 1. London:

Crown Agents.

Launert, E., 1970. In H. Merxmiiller, Prodromus einer Flora

von Siidwestafrika. Germany: Cramer.

Monod, T., 1971. Remarques sur les symetrics floristiques des

zones seches Nord et Sud en Afrique. Mitt. bot. StSamml.,

Miinch. 10: 375-423.

Palmer, P. G. & Tucker, A. E., 1981. A scanning electron

microscope survey of the epidermis of East African grasses,

I Smithson. Contr. Bot. 49: 1 — 84.

Renvoize, S. A., 1981. The sub-family Arundinoideae and its

position in relation to a general classification of the

Gramineae. Kew Bull. 36: 85-102.

Tateoka, T., Inoue, S. & Kawano, S., 1959. Notes on some

grasses. IX. Systematic significance of bicellar microhairs of

leaf epidermis. Bot. Gaz. 121: 80-91.

Bothalia 15, 1 & 2: 161-166 (1984)

Embryo sac development in some South African Lantana species

(Verbenaceae)

J. J. SPIES*

Keywords: embryo sac, Lantana, Verbenaceae

ABSTRACT

Evidence that the South African Lantana camara L. complex only produces sexual embryo sacs is provided. It is

shown that the archesporium occasionally divides mitotically and that both archesporia form tetrads. The chalazal

megaspore of one tetrad and the micropylar megaspore of the second tetrad develop into Polygonum type embryo

sacs.

L. rugosa Thunb. also forms Polygonum type embryo sacs. The L. rugosa embryo sac has a much more densely

packed cytoplasm, smaller vacuole and the position of the polar nuclei differs from that of the L. camara embryo

sac. It is possible to distinguish between L. camara and L. rugosa on their embryo sac morphology alone.

INTRODUCTION

The genus Latana is represented in South Africa

by L. camara L., L. mearnsii Moldenke, L.

montevidensis (Spreng.) Briq., L. rugosa Thunb.

and L. trifolia L. L. camara is an introduced plant

that has escaped from cultivation and has become an

aggressive invader of the warmer subtropical regions

of South Africa. Of the indigenous Lantana species,

L. rugosa is the most abundant and widespread.

The occurrence of apomixis in L. camara has been

inferred from morphological studies of Fi plants

(Raghavan & Arora, 1960; Khoshoo & Mahal,

1967), although no evidence of apomixis was

obtained by embryological studies (Junell, 1934;

Patermann, 1935; Tatachar, 1940; Padmanabhan,

1959; Khaleel & Nalini, 1972).

Recently, Spies & Stirton (1982c) have described

the occurrence of two embryo sacs per ovule in L.

camara. One of these embryo sacs was found to be

sexual (resulting from a meiotic division of the

archesporium), whereas the origin of the other

embryo sac was, at that time, not determined. The

aim of this study is to examine the mode of origin of

the second embryo sac and to compare embryo sac

development in the exotic L. camara and the

indigenous L. rugosa.

MATERIALS AND METHODS

The plants used during this study were naturalized

plants collected in the field throughout South Africa

and transplanted in the Pretoria National Botanical

Garden. The study sample includes all material used

in the previous study (Spies & Stirton, 1982c) and, in

addition, Spies 750, 751, 752, 765, 767, 797, 834,

836, 851, 854, 857 and 887, and Stirton 8928 and 9834

were examined. L. rugosa was represented by Spies

1056 and 1060. Herbarium voucher specimens are

kept in the National Herbarium, Pretoria (PRE).

Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

Young inflorescences were fixed in Navashin

fixative (Stockholm modification — Maheshwari,

1939) at 4°C for at least 24 hours. The material was

dehydrated in an ethyl alcohol and tertiary butyl

alcohol series and subsequently embedded in Tissue

Prep (T565). Serial sections (7-10 p) were stained in

Safranin (Johansen, 1940) and Fast Green (Sass,

1951).

RESULTS AND DISCUSSION

(a) Lantana camara

Lantana camara has a bilocular ovary with a single

anatropous ovule per locule. A single massive

integument surrounds the uniseriate nucellus. The

archesporium is hypodermal and either functions

directly as a megaspore mother cell (Fig. la) or

divides mitotically (Fig. li) and subsequently acts as

two megaspore mother cells. The further develop-

ment of the single archesporium (Figs la — g, & 2a,

c, d) has been described elsewhere (Spies & Stirton,

1982c).

In those instances where two archesporia are

present, both cells divide meiotically to form two

parallel linear tetrads of megaspores (Figs lj & 2b).

The chalazal megaspore is always the functional

megaspore when only one tetrad is present (Spies &

Stirton, 1982c), but the chalazal megaspore of one

tetrad and the micropylar megaspore of the second

tetrad are functional when two tetrads are present

(Figs lk & 2b). As further development of the

megaspores into embryo sacs progresses, a differ-

ence in developmental stage between the two

embryo sacs is distinguishable. The embryo sac

originating from the chalazal megaspore is always

developmentally ahead of the embryo sac that

originated from the micropylar megaspore (Fig. 11 -

n). At maturity both embryo sacs eventually

resemble the sexual Polygonum type embryo sacs

(Maheshwari, 1950) with identical polarization (Fig.

2f, g).

During the early stages of embryo sac develop-

ment no differentiation of nuclei is visible (Fig. 2c,

162 EMBRYO SAC DEVELOPMENT IN SOME SOUTH AFRICAN LANTANA SPECIES (VERBENACEAE)

Fig. 1. — Schematic representation of embryo sac development in Lantana camara. a, archesporium; b, tetrad; c, tetrad with

chalazal megaspore development, d, 2-nucleate embryo sac; e, 2-nucleate embryo sac where the formation of a vacuole pushed

the nuclei to the opposite poles; f, 4-nucleate embryo sac; g, mature Polygonum type embryo sac; h-n, same as a-g, except

mitotic division of the archesporium occurred first and two embryo sacs are formed. Abbreviations: A, antipodal cell; AR,

archesporium; D, degenerating cell; E, egg cell; I, integument; K, embryo sac; N, nucellus; M, megaspore; P, polar nuclei; S,

synergid; V, vacuole.

d). All nuclei are round to oval with one large

nucleolus. The chromatin stains slightly darker

towards the membrane. The nuclear morphology

remains unchanged until after the 4-nucleate stage.

Differentiation of the nuclei occurs during the

8-nucleate stage.

The antipodal nuclei resemble the undifferen-

tiated nuclei, except for the number of nucleoli

sometimes formed. These nuclei often have more

than one small nucleolus. Whereas the antipodal

nuclei may be slightly elongated or elliptical, the

polar nuclei are usually round (Fig. 2f, g). Since the

cytoplasm is not as dense in the central region of the

embryo sac as in the polar regions, the polar nuclei

appear clearer and better defined than the

micropylar or antipodal nuclei. Several nucleoli of

different sizes can occasionally be seen in each polar

nucleus.

Various numbers of nucleoli might be a result of

different polyploid levels. Various polyploid levels

for Lantana camara in South Africa have been

described (Spies & Stirton, 1982a & b; Spies, 1984)

and as a result of polyploidy there may be a greater

number of nucleoli (Garber, 1950). Therefore, the

correlation between the number of nucleoli per

nucleus and the polyploid level of the plant might

provide an interesting study.

An egg cell surrounded by two synergids formed

the egg apparatus (Fig. 2g). The egg cell is oval to

rounded with a single large nucleolus and a vacuole

on the micropylar side of the nucleolus. The

synergids are more elongated and surround the egg

cell on three sides and each synergid has a prominent

hook on the side facing away from the egg cell.

Usually only one large nucleolus per synergid can be

seen. This nucleolus is separated from the chalazal

membrance of the synergid by a vacuole. The

cytoplasm of the synergids stains somewhat darker

than the cytoplasm of the egg cell.

The cytoplasm of the embryo sac is denser at the

poles than in the centre, with the micropylar pole

more dense than the chalazal one. The greater part

of the embryo sac consists of a very large vacuole.

The polar nuclei are usually suspended in cytoplasm

in the central part of the embryo sac next to the cell

membrane (Fig. 2e-g).

(b) Lantana rugosa

Lantana rugosa also has a bilocular ovary with a

single anatropous ovule per locule. A single massive

integument surrounds the uniseriate nucellus. The

archesporium is hypodermal and functions directly

as a megaspore mother cell (Figs 3a & 4a). Embryo

sac development (Fig. 3a — g) corresponds with that

of L. camara except that no case of the development

of more than one tetrad or embryo sac was seen in

either plant studied.

Although the development of the embryo sac is

similar in L. camara and L. rugosa, the L. rugosa

embryo sac can be distinguished morphologically

from that of L. camara at all stages. The main

distinction lies in the density and staining of the

cytoplasm. L. rugosa has a much denser and darkly

stained cytoplasm than L. camara at all develop-

mental stages of the embryo sac (compare Figs 2 &

4).

Undifferentiated nuclei in L. rugosa are round

with a single, large nucleolus. The density of the

chromatin increases towards the nuclear membrane

(Fig. 4b, c)’. This nuclear morphology persists till

after the 4-nucleate stage.

J. J. SPIES

163

Fig. 2. — Photomicrographs of embryo sac development in Lantana camara. a. archesporium; b. twin tetrads, initialization of the

development of two embryo sacs (M, developing megaspore; T, nonfunctional megaspores); c, 2-nucIeate embryo sac; d,

4-nucleate stage; e, 2 young 8-nucleate embryo sacs; f & g, 2 mature embryo sacs per ovule (breakage of embryo sac membrane

during preparation makes interpretation difficult), x 420. See Fig. 1 for legend to abbreviations.

During the 8-nucleate stage the antipodal cells

start to degenerate and integrate with the integu-

ment to such an extent that these cells cannot be

recognized as antipodal cells (Fig. 4e-g). This also

results in the embryo sac being fused to the

integument. When this integration process starts,

each antipodal cell has a large nucleus with a single

small nucleolus and scattered chromatin regions

throughout the nucleus. The antipodal cells vary

from round to elliptic, elongated and conical.

The polar nuclei are round with a single large

nucleolus in each (Fig. 4f, g). Although the polar

nuclei resemble the undifferentiated nuclei, the

nucleolus of the polar nucleus is darker stained than

in the undifferentiated nuclei. The polar nuclei are

situated in the proximity of the egg apparatus and

not in the central part of the embryo sac as in L.

camara.

The egg apparatus contains the egg cell and two

synergids. The egg cell is broad pear-shaped and a

single large nucleolus can be seen (Fig. 4f). In

cross-section the synergids have a more or less round

body with a long beak-shaped protrusion (Fig. 4f).

The nucleus is situated on the micropylar side of a

vacuole in the chalazal side of the synergids. Large

numbers of vacuoles dispersed throughout the

synergid are occasionally found (Fig. 4f).

The cytoplasm is stained much darker and is

denser than in L. camara. The large vacuole

between the polar nuclei and the antipodal cells is

surrounded by a dense cytoplasmic layer adjacent to

the cell membrane. Occasionally one or two small

164

EMBRYO SAC DEVELOPMENT IN SOME SOUTH AFRICAN LANTANA SPECIES (VERBENACEAE)

Fig. 3.— Schematic representation of embryo sac development in Lantana rugosa. a. archesporium; b, tetrad; c. tetrad with chalazal

megaspore development; d, young 2-nucleate embryo sac; e, old 2-nucleate embryo sac; f, 4-nucleate embryo sac; g, mature

embryo sac. See Fig. 1 for legend to abbreviations.

vacuoles occur between the polar nuclei and the egg

apparatus (Fig. 4d). When these vacuoles are

present, they are surrounded by a very thick

cytoplasmic layer. In some embryo sacs the polar

nuclei and egg apparatus are adjacent with no

vacuole present between them. As a result of the

denser cytoplasm the whole embryo sac appeared

more solid and stable than that of the flimsy L.

camara embryo sac which usually broke somewhere

during the cutting or staining process.

(c) Comparison with other V erbenaceae

All members of the Verbenaceae, that have been

examined embryologically, have a Polygonum type

of embryo sac development (for a list of contributors

— see Davis, 1966). A single exception is Avicennia

officinalis, which has an Allium type embryo sac

development (Karsten, 1891). Karsten’s report has

not been confirmed to date.

The occurrence of more than one archesporium

per ovule has been described in Pityrodia bartlingii

(Junell, 1934), Premna integrifolia (Patermann,

1935), Lippia nodiflora (Pal, 1951), Lantana camara

(reported as L. aculeata — Khaleel & Nalini, 1972)

and is now confirmed for L. camara. Khaleel &

Nalini (1972) distinguished up to three archesporia

in L. camara. Although only one usually developed

while the rest degenerated, two tetrads were

occasionally seen in their material. The present

study indicated a much higher frequency (up to

100% in some plants) of twin tetrads in the South

African L. camara complex.

Frequently more than one megaspore per tetrad

developed in Lantana camara, L. involucrata,

Bouchea incrassata, Petrea volubilis (Junell, 1934),

Citharexylum ilicifolium (Patermann, 1935) and

seldom in Lantana indica (Tatachar, 1940). In

addition to this phenomenon, Patermann (1935)

observed that, although only one megaspore

developed in Avicennia officinalis, the rest did not

degenerate. The same phenomenon was observed

during this study in Lantana camara.

It will be interesting to examine earlier stages of

embryonic development in plants that developed

more than one megaspore. Such a study would

indicate whether all developing megaspores origin-

ated from the same tetrad or whether different

archesporia formed different tetrads. In addition.

such a study would show whether each tetrad has

only one functional megaspore.

Very little information exists in the literature bn

the morphology of Verbenaceae embryo sacs. The

embryo sac appears to be usually elongated elliptical

with the micropylar region broader than the chalazal

end as seen during this study.

Schnarf (1931) studied embryo sac development in

the Angiospermae and concluded that the three

small antipodal cells degenerate at an early stage.

Junell (1934) demonstrated that the antipodal cells

in the Verbenaceae were different and often divided

to form many more cells. This phenomenon was also

observed in Clerodendron phlomidis (Misra, 1937)

and Lantana indica (Tatachar, 1940). Contrary to

this, both Patermann (1935) and Pal (1951)

confirmed Schnarf s (1931) observation in other

representatives of the Verbenaceae.

During the present study L. camara was observed

to have three small antipodal cells which persisted

until at least the completion of the formation of the

egg apparatus. The early degeneration of the

antipodals was observed in L. rugosa. The

degeneration or persistence of antipodal cells is,

therefore, not a constant characteristic of the

Verbenaceae.

The egg cell in the Verbenaceae is oval to

flask-shaped and a large vacuole in the micropylar

region was first observed by Misra (1939). In

contrast the pear-shaped synergids have a large

vacuole in the chalazal part (Misra, 1939). The

synergids are prominently hooked and a beak-

shaped apex was described in Clerodendron phlo-

midis, Caryopteris wallichiana (Misra, 1939), Lippia

nodi flora (Pal, 1951) and Lantana rugosa (present

study). The present study confirmed hooked

synergids in Lantana camara, but the beak-shaped

apex was not as prominent as in L. rugosa.

This study has, therefore, indicated that, although

similarities in embryo sac development exist in the

Verbenaceae, differences can even be found at

specific level.

CONCLUSIONS

The feproduction studies of Raghavan & Arora

(1960) and Khoshoo & Mahal (1967) indicated

J. J. SPIES

165

166

EMBRYO SAC DEVELOPMENT IN SOME SOUTH AFRICAN LANTANA SPECIES (VERBENACEAE)

matroclinous progeny in open pollinated L. camara.

Their assumption that the matroclinous progeny

represented at least facultative apomixis is not

supported by any cytological study. The only study

that might have supported an apomictic develop-

ment was the one by Spies & Stirton (1982c), where

a second embryo sac of unknown origin was

described. The current study proves that the second

embryo sac also has a sexual origin. Therefore, no

cytological evidence for apomixis in any Lantana

species has yet been presented.

Monosporic 8-nucleate embryo sacs, known as the

Polygonum type embryo sac, occur in at least 70%

of studied Angiospermae (Maheshwari, 1950). The

formation process is similar in all these plants but the

final product may differ significantly. Davis (1966)

discussed the differences in embryo sac development

and embryo sac components as a means to help in

the taxonomic treatment of taxa. However, she did

not indicate at what taxonomic level this criterion

could be used.

The morphological differences between the

embryo sac components of L. camara and L. rugosa

indicate that embryo sac studies may contribute to

taxonomic separation at the specific level. However,

these morphological differences may be an indica-

tion of greater phylogenetic differences than

presently assumed and more studies of closely

related species are necessary to test the taxonomic

effectiveness of this approach.

UITTREKSEL

Bewyse word gelewer dat die Suid-Afrikaanse

Lantana camara L. kompleks slegs geslagtelike

kiemsakke vorm. Daar word aangetoon dat die

archespoor soms mitoties deel en dat beide arches-

pore tetrades vorm. Die chalazale megaspoor van die

een tetrade en die mikropilere megaspoor van die

ander tetrade ontwikkel in Polygonum tipe kiemsak-

ke.

L. rugosa Thunb. vorm ook ’n Polygonum tipe

kiemsak. Die sitoplasma van die L. rugosa kiemsak

is baie digter saamgedruk, die vakuole is kleiner en

die posisie van die poolkerne verskil van die van L.

camara. Dit is moontlik om L. camara van L. rugosa

te onderskei slegs op grond van kiemsakmorfologie.

REFERENCES

Davis. G. L., 1966. Systematic embryology of the Angiosperms.

London: John Wiley.

Garber. E. D., 1950. Cytotaxonomic studies in the genus

Sorghum. Univ. Calif. Pubis Bot. 23: 283-361.

Johansen. D. A., 1940. Plant microtechnique. New York:

McGraw-Hill.

Junell. S.. 1934. Zur Gynaceummorphologie und Systematik der

Verbenaceen und Labiaten. Symb. bot. upsal. 4: 178-180.

Karsten, G., 1891. Uber die mangrove vegetation in Malayis-

chen Archipelago. Biblthca bot. 22.

Khaleel. T. F. & Nalini, A. S., 1972. Embryology of Lantana

aculeata L. var. nivea Bailey. Curr. Sci. 41: 491-494.

Khoshoo. T. N. & Mahal, C., 1967. Versatile reproduction in

Lantana camara. Curr. Sci. 37: 201-203.

Maheshwari, P., 1939. Recent advances in microtechnique. II

The paraffin method. Cytologia 10: 251-281.

Maheshwari. P., 1950. An introduction to the embryology of

Angiosperms. New York: McGraw-Hill.

Misra, K. C., 1937. The antipodals of Verbenaceae. Curr. Sci. 6:

98-99.

Misra. K. C., 1939. A contribution to the embryology of the

Verbenaceae. Proc. Indian Acad. Sci. B9: 49-56.

Padmanabhan. D., 1959. The development of endosperm in

Lantana camara L. Proc. Indian Acad. Sci. 49: 420—427.

Pal, N. P., 1951. Studies in the embryology of some

Verbenaceae. J. Indian bot. Soc. 30: 59-74.

Patermann, H., 1935. Beitrage zur Zytologie der Verbenaceen.

Diss. Berlin.

Raghavan, R. S. & Arora, C. M., 1960. Morphological and

cytological studies in the genus Lantana L. Bull. bot. Surv.

India 2: 299-303.

Sass, J. E., 1951. Botanical microtechnique. 2nd edn Iowa: Iowa

State Coll. Press.

Schnarf, K., 1931. Vergleichende Embryologie der Angiosper-

men. Diss. Berlin.

Spies. J. J., 1984. Determination of genome homology in

polyploids. 5. Afr. J. Sci. 80 : 44-46.

Spies, J. J., & Stirton, C. H., 1982a. Chromosome numbers of

South African plants. Jl S. Afr. Bot. 48: 21-22.

Spies. J. J. & Stirton, C. H., 1982b. Meiotic studies of some

South African cultivars of Lantana camara. Bothalia 14:

101-111.

Spies, J. J. & Stirton, C. H., 1982c. Embryo sac development in

some South African cultivars of Lantana camara. Bothalia

14: 113- 117.

Tatachar. T., 1940. The development of the embryo sac and

formation of haustoria in Lantana indica Roxb. and

Stachyturpheta indica Vahl. J. Indian bot. Soc. 19: 45—52.

Bothalia 15, 1 & 2: 167-174 (1984)

Classification of embryo sacs in the Eragrostis curvula Complex

T. B. VORSTER* and H. LIEBENBERG+

Keywords: apomixis, diplospory, embryo sac, Eragrostis curvula

ABSTRACT

At each of 17 collecting points between Johannesburg and Brits in the Transvaal, three plants which belong to

the Eragrostis curvula Complex were collected and studied. A total of 3 902 embryo sacs was examined in this

sample. Of the embryo sacs examined, 3 306 were apomictic by means of diplospory, whereas 99 were sexual

monosporic Polygonum- type embryo sacs. One hundred and nineteen embryo sacs were abnormal or divergent,

and 378 were degenerated. There are indications that seasonal climatic fluctuations may be responsible for embryo

sacs developing abnormally or degenerating. Simple and multiple correlations confirmed that sexual embryo sacs

usually do not develop abnormally or degenerate during the later developmental stages. This finding lends

credence to both the system of classification of individual embryo sacs and to the validity of the estimate of the

proportion of sexuality of the plants sampled at each sampling point.

INTRODUCTION

The Eragrostis curvula Complex produces two

main embryo sac types, namely apomictic 4-nucleate

diplosporic embryo sacs (Gustafsson, 1946) and

8-nucleate sexual monosporic Polygonum- type

embryo sacs (Liebenberg, 1961; Brix, 1974, Voigt &

Bashaw, 1972; Vorster & Liebenberg, 1977). Apart

from these, abnormal and degenerated embryo sacs

may also be found.

In the literature there seems to be some doubt

about the validity of allocating individual embryo

sacs to any one of the four types (sexual, diplosporic,

degenerate or abnormal). There is also uncertainty

about the factors governing the development of

sexual embryo sacs, which has in turn resulted in

doubts about the validity of estimates of sexuality in

individual plants. A study was therefore undertaken

to evaluate the variation in embryo sac development

of plants growing in a restricted area under natural

conditions (Vorster & Liebenberg, 1977).

MATERIAL AND METHODS

The plant material examined in this study

corresponds with that discussed by Vorster &

Liebenberg (1977), and material was collected from

different individuals occurring in 17 natural popula-

tions in a restricted area in the Transvaal (South

Africa). The inflorescences were fixed in Navashin’s

fixative (Stockholm modification — Maheshwari,

1939). Embedding and staining techniques outlined

in Johansen (1940) were followed, using Heiden-

hain’s Haematoxylin as the nuclear stain and Orange

G as the outer stain. A minimum of 50 embryo sacs

from every plant was investigated.

* Formerly: Botanical Research Institute, Department of

Agriculture, Pretoria. Presently: Anglo American, Centre for

Casava Research, Mtunzini 3867.

+ Department of Genetics, University of Pretoria, Pretoria 0002.

RESULTS

A total of 3 902 embryo sacs was examined the

results being summarized in Table 1. Of these, 3 306

embryo sacs were apomictic by means of diplospory,

whereas 99 were sexual monosporic Polygonum-

type (Maheshwari, 1950) embryo sacs. One hundred

and nineteen embryo sacs were abnormal or

divergent, and 378 were degenerated (Table 1).

1 Sexual embryo sac development

The archesporial mother cell differentiates from a

hypodermal cell of the nucellar tissue (Fig. 1) and

enlarges somewhat to undergo a transverse first

meiotic division. The second meiotic division follows

directly, with both the chalazal and the micropylar

cells dividing transversally to form a linear tetrad, or

the micropylar cell dividing perpendicularly to the

chalazal cell to form a T-shaped tetrad. The three

micropylar megaspores degenerate soon after the

formation of the tetrad and the chalazal megaspore

cell forms the embryo sac in the majority of cases

(Fig. 2).

This process proceeds as follows:

(a) the chalazal megaspore enlarges and under-

goes the first mitotic division. A vacuole develops

between the two nuclei and enlarges so that the

nuclei are forced to the opposite poles of the embryo

sac (Fig. 3). The enlargement of the vacuole

eventually also causes the embryo sac to be forced

against the nucellus so that the remnants of the

degenerated megaspores eventually disappear (Fig.

4);

(b) the second mitotic division only follows after

the embryo sac reaches this size. With the continued

enlargement of the vacuole the embryo sac is forced

deeper into the nucellus (Fig. 5);

(c) the third mitotic division gives rise to an

embryo sac with four nuclei at the micropylar pole

and four nuclei at the chalazal pole. Cell walls then

form around three of the micropylar nuclei to form

the egg cell and two synergids. One chalazal nucleus

migrates to the micropylar pole to form, together

168

CLASSIFICATION OF EMBRYO SACS IN THE ERAGROSTIS CURVULA COMPLEX

TABLE 1. — Classification of the embryo sac types

Di, diad; Te, tetrad; OD, old diplosporic mother cell; YD, young diplosporic mother cell; IN, one-nucleate stage; 2N, two-nucleate stage; 4N, four-nucleate

stage; 8N, eight-nucleate stage; M, mature embryo sac.

Collection dates: = 22.12.71; Z2 = 23.12.71; Z3 = 24.12.73; Y = 3.1.73; X, = 2.3.73; X2 = 3.3.73; = 11.10.73; W2 = 22.10.73; W3 = 6.11.73;

V= 1.12.73.

with the remaining micropylar nucleus, the two

polar nuclei. The two polar nuclei do not fuse before

fertilization (Fig. 6) and

(d) cell walls are formed around the remaining

three chalazal nuclei to form the antipodal cells.

These cells then divide until an average of 15 cells

are formed. Degeneration now takes place so that

only degenerated remnants can be seen in the

mature embryo sac, but these degenerated cells

always remain visible in sexual embryo sacs. (Fig. 6).

2 Diplosporic embryo sac development

As in sexual embryo sac development, the

archesporial mother cell also differentiates from a

hypodermal cell of the nucellar tissue (Fig. 1), but

does not undergo a meiotic division. It immediately

begins to enlarge to form the diplosporic mother

cell.

Vacuoles develop on either side of the nucleus and

slowly enlarge. The chalazal vacuole enlarges faster

and becomes dominant (Fig. 7) so that the nucleus is

forced to the micropylar pole. The micropylar

vacuole eventually disappears. The enlargement of

the chalazal vacuole continues to a stage where the

size of the embryo sac roughly equals that of a two-

to four-nucleate sexual embryo sac (Fig. 8).

The first apparent mitotic division then occurs

(Fig. 9) forming the 2-nucleate stage with both

nuclei at the micropylar pole (Fig. 10). The second

mitotic division follows, with the vacuole still

enlarging, to form the 4-nucleate stage (Fig. 11).

The 4-nucleate stage is short-lived, because cell walls

are almost immediately formed around three of the

nuclei to give rise to an egg cell and two synergids.

The fourth nucleus remains free in the cytoplasm

and functions as the polar nucleus (Fig. 12).

T. B. VORSTER AND H. LIEBENBERG

169

Figs 1-6.— Sexual monosporic Polygonum- type embryo sac development, original magnification ca. x 600. 1, archesporial mother

cell, 2, degenerated micropylar megaspores with the functional chalazal megaspore; 3, young 2-nucleate stage; 4, old

2-nucleate stage; 5, 4-nucleate stage — one of the micropylar nuclei appears on another section; 6, mature embryo sac; a,

antipodal cells; e, egg cell; m, micropylum; p, polar nuclei; s, degenerated synergids.

170

CLASSIFICATION OF EMBRYO SACS IN THE ERAGROSTIS CURVULA COMPLEX

& *

Figs 7- 12.— Diplosporic embryo sac development, original magnification ca. x 650. 7, young diplosponc mother cell; 8, old

diplosporic mother cell; 9, the first apparent mitotjc division; 10, 2-nucleate stage; 1, 4-nucleate stage, 12, mature

embryo sac; e, egg cell; m, micropylum; p, polar nucleus, s, synergids.

T. B. VORSTER AND H. LIEBENBERG

171

From the preceding, it is clear that it is only during

the archesporial mother cell stage that the two

embryo sac development types are not distin-

guishable from one another. Directly following the

archesporial mother cell stage the nucleus either

undergoes meiosis, or else vacuoles develop on

either side of the nucleus, to initiate diplosporic

embryo sac development.

DISCUSSION

1 Degeneration of young sexual embryo sac stages

A study of the literature leads to the conclusion

that doubt still exists as to what extent the

percentage of sexuality, found in a specific plant, is

influenced by degeneration. It is therefore possible

that due to the degeneration of young sexual embryo

TABLE 2. — Percentage sexual embryo sacs in young and old ovules

2N, two-nucleate stage; 4N, four-nucleate stage; M, mature embryo sac

172

CLASSIFICATION OF EMBRYO SACS IN THE ERAGROSTIS CURVULA COMPLEX

sacs the calculation of the percentage sexuality at

earlier stages could be subject to a certain amount of

error. For this reason, the collected data were

examined to determine whether such a discrepancy

could be detected.

This examination was based on the following

consideration: it is important to know what the

percentage of young sexual embryo sacs are during

the early stages, in relation to the percentage of

sexual embryo sacs present during the older stages.

It would then be possible to determine if young

sexual stages could be studied to give a reliable

estimate of the percentage of functional sexual

embryo sacs formed.

Table 2 was drawn up to determine these

proportions. The end of the 2-nucleate sexual and

diplosporic stages approximates the intermediate

point in both types of embryo sac development

(Liebenberg, 1961), and was therefore used as a

dividing point in this Table.

Comparisons could not be drawn within single

collections, because often either young or only

advanced stages were found (Table 2). Average

percentages of 3,12% and 2,29% (Table 2) were

found for young sexual and mature sexual stages,

respectively, all the collections examined having

been taken into account.

If the young sexual ovules developed abnormally

or degenerated before reaching maturity, then the

average percentage sexuality in young ovules would

be expected to be higher than the average

percentage sexuality in mature ovules. Table 2

shows that these percentages are very much the

same, one can therefore conclude that sexual

embryo sacs are not inclined to degenerate or to

develop abnormally, after the 2-nucleate stage.

2 Classification of embryo sacs

A further problem, similar to the one discussed

above, was that young stages, classified as sexual,

could just as well have been divergent diplosporic or

abnormal embryo sacs.

At first, considerable difficulty was experienced

with the classification of the 2- and 4-nucleate stages

in sexual embryo sacs. These difficulties have also

been reported by other researchers. Remnants of

the three degenerated megaspores still showed in the

two- as well as in some of the 4-nucleate sexual

stages of diploid Eragrostis curvula var. conferta

(Voigt & Bashaw, 1972) and of E. plana (Lieben-

berg, 1961), whereas Streetman (1963) recorded

remnants only in 2-nucleate stages of E. superba.

Of 99 sexual ovules investigated in this study, 29

were 2-nucleate stages (Table 1). Only one of these

showed remnants of the three degenerated micropy-

lar megaspores and even in this instance these had

nearly disappeared. (Fig. 3). None of the ten

4-nucleate stages examined in this study showed any

such remnants.

If all the abovementioned 2-nucleate sexual stages

had been classified as abnormal or divergent, this t

would have resulted in the average percentage of

young sexual stages being lower than the percentage

of mature sexual stages. This would imply that a

larger number of mature sexual and diplosporic

stages were observed in ovules of plants where

sexuality was found, whereas younger stages were

observed mainly in plants with a very low percentage

sexuality or without sexuality at all. The results show

that the opposite is in fact true (Table 1). The sexual

2- and 4-nucleate stages were furthermore usually

accompanied by younger and/or older sexual stages

which were no problem to classify (Table 1).

The classification of 2- and 4-nucleate stages as

being sexual, even when remnants of the degenera-

ted megaspores are no longer visible, therefore

seems to be justified.

3 Abnormal embryo sacs

Only 3,5% of all the embryo sacs investigated,

were abnormal or divergent. Taking into considera-

tion that sexual embryo sacs show no tendency to

develop abnormally after the 2-nucleate stage, the

question now arises as to what causes the

degeneration and abnormal development of embryo

sacs in the earlier stages of development of the

sexual embryo sacs and of the non-sexual embryo

sacs.

The collections of this study were made during

three different growing seasons. The number of

diplosporic, sexual, abnormal and degenerated

embryo sacs were grouped according to the seasons

in which they had been collected so that comparison

between the different embryo sac types and the

collecting periods was possible (Table 3).

From Table 3 it is obvious that the number of

abnormal and degenerated embryo sacs fluctuated

seasonally, especially when correlated against the

amount of rainfall received. The rainfall for

December 1971 was characteristic for that month.

Table 3 shows that average percentages of abnormal

and degenerated embryo sacs, from plants collected

during that month to be much lower than those

collected during the earlier months of 1973, when a

much lower than average rainfall was measured. On

the other hand, the end of 1973 formed part of a

particularly good rainy season with the result that

the percentages of abnormal and degenerated

embryo sacs are lower (Table 3).

The average percentages of sexual embryo sacs

also show this tendency but it is not as clearly

defined as in the case of the other two embryo sac

types.

It seems as if climatological changes may,

therefore, be a factor responsible for inducing

abnormal embryo sac development as well as

embryo sac degeneration in the plants studied.

4 Correlations amongst embryo sac types

If the above statement is erroneous and it is in fact

sexual embryo sacs which developed abnormally or

degenerate, then one would expect a positive

correlation between the number of sexual embryo

sacs and the number of abnormal or degenerated

embryo sacs investigated for a specific collection.

With this in mind, simple and multiple correlations

were made between the four different embryo sac

T. B. VORSTER AND H. LIEBENBERG

173

TABLE 3. — Comparative table between the period of collection and the different embryo sac types

types according to the standard method and the

results are summarized in Tables 4 and 5.

The percentages of diplosporic embryo sacs are in

all cases significantly negatively correlated with the

other embryo sacs predominate in all collections.

The percentages of sexual, abnormal and degene-

rated embryo sacs when compared show no

significant correlation (Table 4).

In order to find any possible, more complex

multiple relationships (correlations), a multiple

TABLE 4. — Simple correlations between the different

embryo sac types

sacs

* Correlation (becomes positive in the case of the

multiple correlations)

• R-squared

regression analysis was done. No further relation-

ships could be found (Table 5), so that it seems as if

no significant connections exist amongst the diffe-

rent embryo sac types. The percentage of abnormal

and degenerated embryo sacs do not, therefore,

seem to depend on the percentage of sexual embryo

sacs.

CONCLUSIONS

Seen as a whole, one has to conclude from this

study, that the sexual embryo sacs showed no

tendency to degenerate or to develop abnormally

during the later stages of their development. This

finding lends credence to both the system of

classification used for individual embryo sacs, and to

the estimation of the proportion of sexuality in each

collection.

It seems, furthermore, that abnormal develop-

ment and degeneration of embryo sacs may be

caused by climatological changes.

UITTREKSEL

Drie plante versamel by elk van 17 versamelpunte

tussen Johannesburg en Brits in die Transvaal,

behorende tot die Eragrostis curvula Kompleks is

bestudeer. Uit 'n totaal van 3 902 kiemsakke wat

bestudeer is, was 3 306 apomikties dear middel van

diplosporie terwyl 99 geslagtelike monosporiese

Polygonum-hpe kiemsakke was. Honderd en ne-

gentien kiemsakke was abnormaal of afwykend

terwyl 378 kiemsakke gedegenereerd was. Aandui-

dings is gevind dat klimaatskommelinge hoofsaaklik

daarvoor verantwoordelik was dat kiemsakke neig

om abnormaal te ontwikkel of te degenereer.

Enkelvoudige en meervoudige korrelasies het verder

bevestig dat geslagtelike kiemsakke nie geneig het om

abnormaal te raak of later te degenereer nie. Dit

impliseer nie net dat die bepaling van die persentasies

vir geslagtelike kiemsakke wat by die verskillende

versamelings verkry is, betroubaar is vir daardie

spesifieke versameling nie, maar ook dat die

beslissing of n kiemsak geslagtelik is of nie,

betroubaar was.

174

CLASSIFICATION OF EMBRYO SACS IN THE ERAGROSTIS CURVULA COMPLEX

REFERENCES

Brix, K., 1974. Sexual reproduction in Eragrostis curvula

(Schrad.) Nees. J. PI. Breed. 71: 25-32.

Gustafsson, A., 1946. Apomixis in higher plants. I. The

mechanism of apomixis. Lunds Univ. Arsskrift N.F. Avd. 2.

42: 1-66.

Johansen, D. A., 1940. Plant microtechnique. New York:

McGraw-Hill.

Liebenberg, H., 1961. Apomiksie by Eragrostis en Themeda.

M.Sc. verhandeling, Universiteit van Stellenbosch.

Maheshwari. P., 1939. Recent advances in microtechnique. II.

The paraffin method. Cytologia 10: 251-281.

Maheshwari, P., 1950. An introduction to the embryology of the

angiosperms. New York: McGrawe-Hill.

Streetman, L. J., 1963. Reproduction of the lovegrasses, the

genus Eragrostis, i.e. E. chloromelas Steud., E. curvula

(Schrad.) Nees. E. lehmanniana Nees and E. superba Peyr.

Wrightia 3: 41 - 51.

Voigt, P. W. & Bashaw, E. C., 1972. Apomixis and sexuality in

Eragrostis curvula. Crop Sci. 12: 843—847.

Vorster, T. B. & Liebenberg, H., 1977. Cytogenetic studies in

the Eragrostis curvula Complex. Bothalia 12: 215—221.

Bothalia 15, 1 & 2: 175-227 (1984)

A syntaxonomic and synecological study in the Humansdorp region of

the Fynbos Biome

R. M. COWLING*

Keywords: Fynbos Biome, mediterranean-type shrublands, south-eastern Cape, syntaxonomy, vegetation characterization, vegetation

dynamics

ABSTRACT

A hierarchical syntaxonomic scheme of vegetation in the eastern border of the Fynbos Biome (Humansdorp

region) is presented as a second approximation after the earlier work by Acocks (1953) in the area. Details on the

physiography, geology, climate, soils, historical features and present management of the area are given to provide

the setting for this and other papers dealing with community characterization and ecological relationships,

dynamics, structure and biogeography. A community classification was generated using a hierarchical numerical

classificatory technique (TWINSPAN) which produced ordered two-way phytosociological tables. Tabular

comparisons and final sorting of tables is according to the methods of the Ziirich-Montpellier School. Higher

syntaxonomic ranks (classes and orders) are subjectively defined. Four classes, seven orders and 22 communities

are recognized in the study area. The classes are Cape Fynbos Shrublands (3 orders, 10 communities). Cape

Transitional Small-leaved Shrublands (1 order, 4 communities) Subtropical Transitional Thicket (2 orders, 6

communities) and Afromontane Forest (1 order, 2 communities). Discussion of the scheme is focused on the level

of the order (roughly equivalent to a veld type). In addition to diagnostic floristic elements, syntaxa are further

characterized using biogeographic, structural and habitat criteria. The role of historical land use on vegetation

dynamics and interrelationships is briefly discussed. As far as possible, syntaxonomic concepts are extrapolated to

the entire Fynbos Biome as well as biomes adjacent to its eastern boundary. Certain syntaxa are examined relative

to the shrubland/heathland concepts developed for mediterranean-type ecosystems.

CONTENTS

Introduction 175

1 Study area 176

1.1 Physiography 177

1.2 Geology 178

1.3 Climate 179

1.4 Soils 182

1.5 Historical land use 186

2 Methods 187

2.1 Data collection 187

2.2 Data analysis 188

2.3 Community dynamics 190

3 The syntaxa 191

3.1 Cape Fynbos Shrublands 191

3.1.1 South-eastern Mountain Fynbos 193

3.1.2 Grassy Fynbos 194

3.1.3 South Coast Dune Fynbos 201

3.2 Cape Transitional Small-

leaved Shrublands 203

3.2.1 South Coast Renosterveld 205

3.3 Subtropical Transitional Thicket 209

3.3.1 Kaffrarian Thicket 213

3.3.2 Kaffrarian Succulent Thicket 216

3.4 Afromontane Forest 217

3.4.1 Knysna Afromontane

Forest 218

* Botany Department, University of Cape Town, Private Bag,

Rondebosch 7700.

4 Discussion 219

4.1 Soil nutrients and

shrubland types 219

4.2 The role of soil moisture 221

4.3 Tension zones and phytochorological

complexity 221

4.4 Conservation 222

Acknowledgements 223

References 223

INTRODUCTION

The work by the late J. P. H. Acocks, ‘Veld types

of South Africa’ is a standard guide for most

vegetation studies in this country. Acocks’s (1953)

concept of a veld type is an agro-ecological unit of

vegetation ‘. . . whose range of variations is small

enough to permit the whole of it to have the same

farming potentialities.’ He is to be commended for

his brilliant and perceptive overview of vegetation

patterns in South Africa, and the production of a

vegetation map which is of great value to both

academic and applied ecologists. Unfortunately,

Acocks’s criteria for distinguishing veld types are

never fully defined and often incorporate vague and

untestable statements on history, utilization and

dynamics (see also Martin & Noel, 1960). He

completed his major work at a time when the

assumptions of Clementsian dynamics were largely

accepted. This allowed for the grouping of

structurally and floristically unrelated types into a

single veld type, based on assumed successional

relationships. Furthermore, Acocks’s groupings of

units above the level of the veld type are often

176 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

Fig. 1. — Block diagram showing location, topography and isohyets of the study area. Isohyets from 1: 250 000 Average Annual

Rainfall Series.

unsound: Knysna Forest (4)* is included as a Coastal

Tropical Forest Type (cf. White, 1978), Valley

Bushveld (23) as a Karroid Type (cf. Moll & White,

1978), whereas False Fynbos (70) and Fynbos (69)

are placed in different units due to the former’s

presumed derivation from mountain .grassland,

thicket and forest.

Clearly the time is long overdue for a critical

re-evaluation of certain veld type concepts. This is

especially true for the Fynbos Biome (Kruger, 1978)

where Acocks (1953) explicitly states that his system

is inadequate. In this paper I present a syntaxonomic

hierarchy of vegetation units focussing on the level

of the order, which is roughly equivalent to a veld

type (cf. Van der Meulen, 1979). I have used

floristic, biogeographic, structural, dynamic and

habitat criteria to formulate syntaxonomic concepts

and to make testable hypotheses regarding their

origins and interrelations.

This study was carried out in the Humansdorp

region, towards the eastern limit of the Fynbos

Biome. This area provides a microcosm of many

Biome vegetation types, as well as types representa-

tive of adjacent biomes. In Marloth’s (1923) words

‘. . . conditions of existence of associations of plants

(can) be studied nowhere better than the boundaries

of the areas where they encounter another

vegetation of different requirements.’

* In this paper bracketed numbers refer to the Acocks (1953)

veld type number.

In addition to clarifying syntaxonomic concepts

above the level of the community, I expand on the

physical and historical environment of the study area

and thus provide the setting for future papers on

community characterization and dynamics, phy-

tochorology and vegetation history (Cowling,

1983a), diversity relations (Cowling, 1983b), growth

form distribution (Cowling & Campbell, 1983a;

Cowling & Campbell, 1983b).

The scheme outlined below must, after Acocks’s

(1953) early work, be regarded as a second

approximation (cf. Poore, 1962) to provide a

hypothetical framework to be refined or rejected in

future years.

1 THE STUDY AREA

The study area is in the Humansdorp region of the

south-eastern Cape (Fig. 1). The eastern Cape

comprises a complex tension zone where elements of

the Cape, Karoo-Namib, Afromontane and

Tongaland-Pondoland phytochorological regions

meet and combine to form a bewildering variety of

vegetation types (Werger, 1978; Goldblatt, 1978;

Gibbs Russell & Robinson, 1981; White, 1983). A

combination of a transitional and variable climate,

complex topographic and geological patterns and

species assemblages of diverse phytochorological

affinities has resulted in communities highly vulner-

able to the destabilizing influences of settled

agriculture (Liebenberg, 1945; Trollope, 1980;

R. M. COWLING

177

Gibbs Russell & Robinson, 1981), resulting in the

invasion and thickening up of weedy species (for a

review see Trollope, 1970).

Few plant-ecological studies have been under-

taken in the eastern Cape in the last two decades,

and none in the study area. A check-list for the

Humansdorp District has been published (Fourcade,

1941). There are short descriptions of fynbos and

thicket communities near Port Elizabeth (Olivier,

1977) and thicket communities in the Sundays River

Valley (Archibald, 1955; Penzhorn & Olivier, 1974;

Olivier, 1981; Taylor & Morris, 1981). Acocks

(1953) recognized three veld types in the study area:

False Fynbos (70) is mapped as covering most of the

region with tongues of Valley Bushveld (23)

extending up the Gamtoos River Valley and

Alexandria Forest (2) occurring along the coast as

far as Jeffreys Bay (cf. Fig. 3).

1.1 Physiography

The study area has both physiographic elements of

the southern Cape coastal region: the rugged

quartzite mountains of the Cape Folded Belt and an

undulating Coastal Foreland (Wellington, 1955)

stretching from the mountain pediments (approxi-

mately 300 m) to the present shoreline (Fig. 1). The

wide alluvial valley and braided course of the

Gamtoos River (Fig. 1) is a feature more typical of

the south-eastern Cape.

The dominating feature of the landscape is

the Great Winterhoek range culminating in the

Cockscomb Peak (1 750 m). This range forms the

eastern end of the inland Swartberg-Baviaanskloof

axis. East of the Gamtoos River a subsidiary range,

with a more south-easterly trend, is known as the

Elandsberg. The coastal axis of the Folded Belt is

represented by the Kareedouw Mountains which

terminate in the study area as two planed parallel

anticlines which jut into the sea at Cape Seal and

Cape St Francis. The only other mountain area is the

Klipfontein Range (300 - 400 m) north-west of

Humansdorp, which is the eastern termination of the

Kouga Mountains.

The Coastal Foreland comprises a plain which cuts

across all geological formations (Fig. 2). In the

southern Cape, the coastal plain has been inter-

preted as an early Tertiary (Eocene) surface (pene-

plain) which slopes at a gradient of 1° from

approximately 1 000 m in the mountain foothills (Du

Toit, 1966; Haughton et al., 1937). In the study area

the plain forms gravel-covered terraces above the

deep and narrow v-shaped incised valleys. Away

from the mountains, the plain is more extensive and

in places, especially where the underlying material is

of the Uitenhage Group, is covered by a veneer of

limestone and marine gravels of the Alexandria

Formation (Fig. 2).

r r

25°

J I I —

Fig. 2. — Geological map (after Haughton et al., 1937).

178 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

King (1972) argues that, up to an elevation of

about 350 m, the coastal plain in the south-east Cape

is one of marine abrasion. Uplift due to monoclinal

tilting of the sub-Miocene surface along an east-west

axis resulted in a Pliocene marine transgression.

Marine sediments of the Alexandria and Bredasdorp

Formations were deposited in a transgressive phase

and redistributed in the regressive phase (King 1972,

1978). The plain is therefore interpreted as a

‘composite marine and continental bevelled surface’

(Heydorn & Tinley, 1980) and not simply a

subaerially formed peneplain.

The geomorphology of the coastal foreland is

dominated by planation forms related to a descend-

ing sequence of high sea-levels (Butzer & Helgren,

1972). The most striking surface occurs immediately

north of Humansdorp (Fig. 1) on the 200 m contour

— this is the Coastal Platform (largely gravel-capped

in the study area). Other surfaces and terraces can

be seen in places at 100 m, 60 m, and 30 m (Butzer &

Helgren, 1972).

The dominant feature of the coastline is the series

of half-heart bays (Silvester, 1960; Heydorn &

Tinley, 1980). They are usually formed as a result of

the seaward truncation of resistant quartzitic rocks

of the Table Mountain Group (thereafter referred to

as TMG) and the greater erosion of softer rocks to

the north. This results in asymmetric east-facing

bays in which the rock outcrops form headlands at

the apex of the short part of the curve (e.g. St

Francis Bay, Fig. 1). Recent deposits of siliceous

sands accumulate against the long curve of the

half-heart embayments (Heydorn & Tinley, 1980).

The most extensive dune-fields in the study area

occur between Oyster Bay and Cape St Francis and

also near the Gamtoos River mouth. In the former

region, fixed dunes have been re-activated, possibly

in the recent past (Keet, 1936). Transverse dunes,

aligned at right angles to the prevailing wind

direction, predominate in the re-activated dune-

fields. Fixed dunes are aligned parallel to the

direction of the prevailing (SW) winds and are

termed linear hairpin dunes (Heydorn & Tinley,

1980). These are essentially parabolic dunes, the

noses of which have been blown out, and must have

been formed under a wind regime considerably

stronger than present (Tinley, pers. comm.).

The Gamtoos River has its source in the

mountains of the Great Karoo. After breaching the

Groot Winterhoek Range it enters the study area in

a deep meandering valley and flows in a general

south-easterly direction to the sea (Fig. 1). The

other major rivers — the Kromme and the

Kabeljous — flow in strike valleys which are deeply

incised into the coastal plain.

1.2 Geology

The oldest rocks in the region belong to the late

Pre-Cambrian Cango Formation of the Malmesbury

Supergroup. They occur in a faulted block

truncating the southern limbs of the Elandsberg

Range with down-faulted Uitenhage beds to the

south (Fig. 2). The lower beds consist of phyllites

with two thick horizons of limestone; the upper beds

are principally arenaceous consisting of felspathic

grits, small pebbled conglomerates and quartzites

(Haughton et al., 1937). In the field, it is difficult to

distinguish the last-mentioned from TMG quart-

zites. Rocks of the Cape Supergroup predominate in

the study area. These were laid down in a basin

depository on a planed surface of Malmesbury-

Cango rocks in a largely marine environment in late

Silurian and Devonian times* (Rust, 1967). Folding

was initiated 200—300 m.y. ago during the

Cape— Karoo orogeny.

Table Mountain quartzites and sandstones com-

prise the mountain ranges in the area but also

underlie considerable portions of the coastal plain

(Fig. 2). The beds are highly folded, usually

asymmetrically with steeper dips to the north, and

overfolding is common. They are made up of

well- jointed, massive quartzitic sandstones; there

are no conglomerates or basal shales. Shale bands

sometimes occur near the top of the group

(Haughton et al., 1937).

Bokkeveld beds are found in a narrow wedge on

the coastal plain (Fig. 2). They are argillaceous

throughout and composed of soft, yellowish to

greenish grey shales (Haughton et al., 1937).

In the late Jurassic times (140 m.y.), faulting

occurred in the southern Cape contemporaneous

with the separation of the African plate from the

South American plate (Sclater et al., 1977). This

faulting, which exposed pre-Cape rocks in the study

area, also created minor tectonic valleys into which

Cretaceous fluviatile, estuarine and shallow marine

sediments were laid down (Du Toit, 1966). These

are the sediments of the Uitenhage Group which

occur extensively in the Gamtoos River Valley,

where they are bounded in the north by the fault

plane and rest unconformably on Bokkeveld rocks in

the south (Fig. 2). Only Enon beds are present

although there is some evidence of Variegated Marls

near Loerie (Haughton et al., 1937). The variability

of these deposits has been stressed by Du Toit

(1966). In the lower Gamtoos Valley, coarse-grained

conglomerates and sandstones, deposited under

fluvial conditions, are the predominant rocks. Along

the north side of the river and towards the mouth the

beds become finer, passing into fine sandstones with

reddish marls and grey sandy clays (Haughton et al.,

1937). These are well displayed in the cliffs on both

sides of the Gamtoos mouth.

Tertiary deposits include the subaerial high-level

gravels and marine deposits of the Alexandria

Formation (Fig. 2). The former, which abut against

steep mountain slopes, have probably originated

mostly as talus material (Haughton et al., 1937)

resting on an early Tertiary surface (King, 1972).

Silicification took place subsequent to deposition,

probably during an arid climatic phase (Du Toit,

1966).

The marine deposits of the Alexandria Formation

are of Pliocene age (King, 1972). They occur on the

northern bank of the Gamtoos River, overlying the

Uitenhage beds (Fig. 2) and have probably been

removed from the remainder of the coastal plain by

Quaternary subaerial denudation (King, 1972). The

beds are usually covered by white tufaceous

limestone and sand (Haughton et al., 1937).

R. M. COWLING

179

S.E. Mountain Fynbos

Grassy Fynbos

fSiSSSl S. Coast Dune Fynbos - Kaffrarian

Thicket mosaic

□

S. Coast Renosterveld

Renoster veld -Grassy Fynbos

transition

Kaffrarian Succulent Thicket

Afromontane Forest

Gamtoosrivier

34-

ape St Francis

al Point

INDIAN

OCEAN

10 o

iFT I— I H I— I I— I I

Kilometre

25

I

Fig. 3. — Vegetation map. Units mapped are orders (see Table 3).

Recent and Quaternary deposits occur as alluvium

in the Kromme and Gamtoos Valleys and as dune

sands along the coast (Fig. 2). In the latter area,

deposits of calcrete crop out between the longitudin-

al dune ridges and also immediately behind the

dunes along the Jeffreys Bay coast.

1.3 Climate

A discussion of the climate is necessarily

hampered by a lack of climatic data, particularly

temperature data, which are recorded only at Cape

St Francis. These data were extrapolated to other

stations on the coastal plain (Oyster Bay, Humans-

dorp, Jeffreys Bay). At Otterford temperature data

from the nearby Van Stadens climate station (452 m)

were used and those of Uitenhage (108 m) for

Hankey. Corresponding stations are physiographic-

ally matched. Climate diagrams are shown in Fig. 4.

1.3.1 Seasonal climatic controls

The climate in the study area is warm temperate.

The climatic regime is dominated by an alternating

succession of east-moving cyclones budded off from

the circum-polar westerlies, and high pressure

anti-cyclones which ridge in behind the lows

(Jackson & Tyson, 1971; Schulze, 1972).

The frequency and intensity of cyclonic fronts is

greatest in winter as a result of the northward

migration of the pressure belts; the weather is cool

and much rain falls. In summer a strengthening and

southwards migration of the South Atlantic high

pressure cell effectively blocks westerly cyclones

reaching the south-western Cape resulting in a

markedly reduced precipitation (mediterranean-

type climate). This effect is not felt so strongly along

the southern Cape coast where a considerable

amount of summer rain falls. Summer precipitation

180 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

Fig 4. — Walter-Lieth climate diagrams. Data from Anon. (1942) and Weather Bureau (1954).

R. M. COWLING

181

is usually associated with the approach of cool post-

frontal air from the south-west, moving over a

relatively warm ocean, as an anti-cyclone progresses

north-eastwards along the coast, or as a result of

cut-off lows (Anon., 1942; Schulze, 1972).

1.3.2 Climatic classification

The study area has a climate transitional between

Koppen’s Cfb and Csb climates (cf. Schulze &

McGee, 1978). The temperature regime is warm

temperate and rain may fall at any time of the year

although the three summer months (Dec. - Feb.) are

always driest (Fig. 4). Three climatic types can be

recognized at the mesoclimatic level:

(i) Coastal plain subhumid climate: Occurs on the

coastal plain and lower slopes of the Elandsberg

where the annual rainfall ranges from 500 to 800 mm

(Fig. 1) (e.g. Oyster Bay, Humansdorp, Cape St

Francis). Along the coast it is cooler due to the

tempering effect of onshore winds. To the west of

the study area it merges with a Tsitsikamma humid

climate and in the east, near Jeffreys Bay, with the

valley climate (see below).

(ii) Semi-arid valley climate: Restricted to the

low-lying areas of the Gamtoos and Seekoei Rivers

where the annual rainfall is less than 500 mm (Fig.

1). As with other river valleys in the south-eastern

Cape, the climate is much drier and more variable

than conditions on the adjacent mountains and

interfluves (Anon., 1942; Liebenberg, 1945).

(iii) Humid coastal mountain climate: This is the

climate of the moist, sea-facing slopes of the

Elandsberg Mountains where annual precipitation is

in excess of 800 mm (Fig. 1). The climate is equable

with sufficient rain throughout the year; it is the

forest climate of the George-Knysna-Tsitsikamma

area (cf. Phillips, 1931).

Below, I discuss briefly some climatic features

with reference to the climatic types listed above:

1.3.3 Winds

The usual wind pattern during the path of a

cyclonic front is the backing from NE to NW (often

associated with an increase in temperature if ‘berg

winds’ persist) and then from W to SW as the front

passes (Anon., 1942). The latter condition is usually

accompanied by high winds, cooler weather and

rain. With the onset of anti-cyclonic conditions the

wind backs from S to E, the skies clear and the

weather warms up. At the equinoxes there are often

strong southerly winds and flood conditions resulting

from ‘cut-off low pressure cells in the interior

advecting cool, moist air from anti-cyclones off the

coast (Heydorn & Tinley, 1980).

Wind roses for Port Elizabeth are shown in Fig. 5.

The area is liable to strong winds and occasional

gales at any time of the year. Specht & Moll (1983)

observe that wind regimes on the Cape coast are

much stronger than climatically analogous areas in

Australia and that this has implications for plant

growth in the former region due to increased

evapotranspiration. Along the southern and south-

eastern Cape coast, the calmest period is in the

autumn months (March— May), whereas spring

(Sept. -Nov.) is the windiest (Anon., 1942). The

prevailing wind direction is from W to SW and there

is a significant increase in E and SE winds in summer

(Fig. 5). Hot, dry and turbulent berg winds, which

are composed of subsidiary air masses draining

seawards from the interior, have their highest

frequency in the winter months (Tyson, 1964; Fig.

5). They are always accompanied by a sharp

increase in temperature, giving rise to the anoma-

lous situation where some of the highest tempera-

tures (up to 38°C) are recorded in winter (Anon.,

1942; Tyson, 1964; Louw, 1976).

The wind regime shown in Fig. 5 is representative

of the coastal plain climate. The valley climate is less

susceptible to the cooling effect of summer sea

breezes (E to SE) (Louw, 1976; pers. obs.). It is

likely that the wind regime in the mountains differs

in many aspects from that on the coast (H. T. Scharf,

pers. comm.).

1.3.4 Precipitation

Most precipitation falls in the form of rain. Snow,

associated with ‘cold snaps’ in winter, falls rarely on

the mountains and then persists for only a few days

(pers. obs.). Mists are important in the mountains

where drizzle from S and SE winds can result in

considerable precipitation (cf. Nagel, 1962). Sea

(advection) fog is uncommon (average of 20 days per

year at Port Elizabeth) and shows its greatest

Fig. 5. — Summer and winter wind roses for Port Elizabeth. Areas

represent 5% intervals. Percentage calms within the circle.

Data from Schulze (1972).

182 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

frequency in late summer (Anon., 1942; Heydorn &

Tinley, 1980). There are no data for land (radiation)

fog. Indications are that it is restricted almost

entirely to the Gamtoos Valley and occurs mainly in

winter (May— June) (Anon., 1942; pers. obs.).

Mean annual isohyets from 1: 250 000 Average

Annual Rainfall Series are shown in Fig. 1.

Precipitation decreases in a north-easterly direction

from Oyster Bay to the Gamtoos Valley and then

increases sharply up the Elandsberg Range. Mean

annual rainfall for eight stations ranging in elevation

from 61 m on the floor of the Gamtoos Valley to 535

m in the Elandsberg, showed a linear increase of

rainfall with elevation (mean annual rainfall = 228, 4

+ 0,8031. altitude; r = 0,93; p < 0,001).

Rainfall distribution varies from sub-

mediterranean on the coast (Dec. -Feb. dry period)

to bimodal in the valley climate (additional

June -Aug. dry period). In the Elandsberg, pro-

longed dry periods are probably rare. Using the

difference between Thornthwaite’s index of poten-

tial evapotranspiration and rainfall, Bond (1980a;

1981) has shown that semi-arid and subhumid

climates in the southern Cape constant rainfall

region have an effective moisture regime that is

unequivocally mediterranean as typified by winter

moisture surpluses.

Low intensity orographic rain predominates

throughout the year although pre-frontal thunder-

storms are occasional in summer. Falls of up to 120

mm per day have been recorded during cut-off

conditions while low pressure systems prevail (see

above) (Anon., 1942). Heaviest downpours are

associated with post-frontal SSW and S winds. The

valley area receives little frontal rain (W to SW

winds) (pers. obs.).

1.3.5 Temperature

Mean annual temperatures are shown in the

climate diagrams (Fig. 4). Generally the tempera-

ture regime for the coastal plain and mountains is

equable, whereas in the Gamtoos region diurnal and

annual variations are greater. Frost is uncommon.

Cold snaps occur in winter when strongly developed

cold fronts are followed by an anticyclone which

advects cold polar air landwards (Schulze, 1972).

Hot spells are correlated with persistent berg winds

(usually in spring) and with warm summer anticy-

clones.

River valleys in the south-eastern Cape show great

annual and daily extremes in temperature (Anon.,

1942). At Uitenhage, in the Swartkops River Valley,

the highest mean daily maximum is recorded in

February (29,1°C) and the coldest in (July 5,9°C);

corresponding temperatures for Cape St Francis are

22,8°C and 10,1°C respectively (Anon., 1942). This

pattern is further illustrated in Fig. 6 which shows

the monthly mean maximum and minimum tem-

peratures at Port Elizabeth Airport (climatically

similar to Cape St Francis) and Uitenhage,

measured over a period of two years (Louw, 1976).

The minimum mean daily range at Uitenhage is

12,2°C (Nov.) and the maximum 15,5°C (June).

Corresponding figures for Cape St Francis are 4,9°C

(March) and 8,4°C (June). The months with the

highest and lowest mean absolute maxima at

Uitenhage are February (39,2°C) and July (-0,4°C)

respectively; corresponding data for Cape St Francis

are April (29,7°C) and August (5,0°C) (Anon.,

1942). There are no detailed temperature data for

the mountain climatic type. H. T. Scharf (pers.

comm.) has measured an environmental lapse rate

of 0,6°C per 100 m in the Great Winterhoek

Mountains near Uitenhage.

Fig. 6. — Mean monthly maximum and minimum temperatures at

Port Elizabeth Airport and Uitenhage. Data from Louw

(1976).

1.3.6 Solar radiation

Solar radiation has an important bearing on

ecological studies (Schulze & McGee, 1978) as the

effects of terrain slope and aspect on radiant energy

regimes lead to corresponding variation in soil

moisture status (Holland & Steyn, 1975; Holland et

al., 1977; Granger & Schulze, 1977) and pedogenesis

(Garland, 1979). Fig. 7 shows the influence of

topography on solar radiation at 34°S (data from

Schulze, 1975). During summer there is little

difference in potential radiation on all slopes and

aspects. At the equinoxes there is a trend for

increased radiation on steeper north slopes and a

more marked decrease in radiation on steep south

slopes. In the southern winter, radiation is strongly

affected by a lower sun azimuth and steep north

slopes receive markedly more radiation than steep

south slopes (Fig. 7).

1.4 Soils

The diversity of parent material and the complex

topography of the study area combine to produce

complex soil patterns. Furthermore, a sequence of

late-Cainozoic changing environmental factors in-

cluding tectonic deformation, eustatic fluctuations of

sea level and fundamental changes in climate and

vegetation have affected pedogenesis in a number of

R. M. COWLING

183

MIDSUMMER

N NE/NW E/W SE/SW S

_ l I I

25 I 1 1 1 1 L

EQUINOXES

SLOPE

0°

5°

10°

15°

20°

25°

Aspect

Fig. 7.— Influence of topography on solar radiation. Daily

incoming radiant flux densities on sloping terrain as a

function of slope, aspect and season, for cloudless days at 34°

S. Data from Schulze (1975).

ways (Butzer & Helgren, 1972). Information on soils

in the Fynbos Biome is given by Lamprechts (in

Boucher, 1978; 1979), Bond (1981) and in a

comprehensive review for the mountains of the

biome by Campbell (1983a). There are no published

data on the soils of the study area.

It is not within the scope of this study to produce a

comprehensive survey of soils. I restricted my

observations to augerings (where possible) in sample

plots and notes from roadside cuttings. Chemical

characteristics are generalized from the results of

101 analyses of soil samples collected at approxi-

mately 15 cm depth (see Methods). Results of these

analyses are shown in Fig. 8. These data are used for

comparisons among substrate and vegetation types

and do not adequately reflect nutrient regimes. The

term ‘soil fertility’ reflects the nutrient requirements

of common crop and pasture plants. Soils are

classified according to the South African Soil

Classification System (MacVicar et al., 1977). I

describe the soils of each parent material separately

and discuss briefly vegetation-soil patterns. A brief

characterization of the major vegetation types

referred to in this section, is presented in Table 3.

Examples of profiles of the soil forms and their

correlation with the FAO and USDA systems are

shown in MacVicar et al. (1977).

1.4.1 Cango Formation

I restrict my discussion to the soils derived from

limestones and phyllites near the Kleinfontein

quarry below Otterford (Fig. 2). The soils are deep

(1-3 m) with uniformly red, apedal subsoils (Hutton

Form). Topsoils are slightly acid to neutral loams,

rich in exchangeable bases (S value; 14-28 meq %)

especially calcium. Values for total nitrogen are

high, but available phosphorus is fairly low (4—10

ppm). Soils are well drained and support dense

thicket.

1.4.2 Table Mountain Group

Soils derived from sandstones and quartzites of

the TMG are mostly infertile and invariably support

fynbos vegetation (Fig. 8). Kruger (1979) has

described the fynbos soils of the quartzitic fold

ranges as generally greyish, acid, shallow sands to

sandy loams with low base saturation (see also

Lamprechts, 1979). He also stressed the importance

of podzolisation as a pedogenic process in fynbos

soils. Recent work by Bond (1981) and Campbell

(1983a) has shown that mountain soils are: much

deeper than usually assumed; that red and

yellow-brown profiles are common; and that true

podzols are rare. Campbell (1983a) has demon-

strated a west-east gradient of increasing clay

fraction and finer sand fraction, and increasing

exchangeable bases in the mountains of the Fynbos

Biome. In the south-eastern Cape, mountain soils

are generally more fertile than the southern and

western Cape.

In the mountains a combination of steep

topography and slow weathering of resistant rocks

usually results in lithosols on steep slopes and deeper

colluvial deposits in valley fills, pediments and talus

slopes (Lamprechts, 1979). Soil patterns in the

Elandsberg Mountains are complex and similar to

those on the seaward slopes of the Outeniqua

Mountains (Bond, 1981). On steep north-facing

slopes and ridges, soils are predominantly litholic

(Mispah and Glenrosa Forms) while Cartref Form

(Amabele Series Cf 20), with an eluvial (E) horizon

on a poorly developed illuvial B horizon, predomi-

nates on rounded anticlinal ridges and gentle slopes.

Deeper Clovelly and Hutton Forms occur at lower

altitudes on talus slopes and pediments; Oakleaf

(low clay content in B horizon) and Fernwood

Forms are less common. Mountain fynbos occurs on

these soils. Valley fills and minor colluvial-alluvial

basins have deep (> 1,5m) Oakleaf Form

(Koedoesvlei Series Oa37) soils with a high clay

content in the B horizon and pronounced organic

staining in the upper soil layer. These soils have

good moisture-retaining properties and support

forest. No true podzols were seen.

Mountain Fynbos soils are very strongly acid (pH

3,9— 4,2), loamy sands to sandy loams and have low

levels of all major nutrients (Fig. 8). Forest soils are

very strongly acid loams to sandy loams, rich in

oxidizable carbon. A strong correlation between

carbon and other nutrients (see Cowling, 1983b)

suggests that the relatively high fertility of these soils

(Fig. 8) depends largely on organic enrichment from

plant remains in the soil.

TMG soils on the coastal forelands differ from the

mountain soils in that they are mostly colluvial,

deeper and somewhat more fertile. Lateritic

hardpans of ferricrete, overlain by colluvial gravels

and sands are found on the Coastal Platform north

of Humansdorp and south of the Kromme River

near Oyster Bay. These may well correlate with

palaeosols described elsewhere in the southern Cape

by Helgren & Butzer (1977), although I did not

TABLE MOUNTAIN SANDSTONE BOKKEVELD SHALE ENON DEPOSITS QUATERNARY DUNE SAND

184 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

a S.E. Mountain Fynbos n-5 □ S. Coast Renosterveld n=17 o Grassy Fynbos n= 8 v S. Coast Dune Fynbos n=13

O Grassy Fynbos n*17 ■ Kaffrarian Succulent Thicket n- 3 □ S. Coast Renosterveld n= 3 • Kaffrarian Thicket n= 7

a Afromontane Forest n»2 • Kaffrarian Thicket n= 7 ■ Kaffrarian Succulent Thicket n= 5

• Kaffrarian Thicket n^4 • Succulent Thicket(Termitaria) n = 2

R. M. COWLING

185

excavate profiles below the ferricrete to confirm

this.

A dominant feature of the Coastal Platform to the

north of the Klipfontein Mountains is a capping of

sub-rounded to sub-angular colluvial gravel (1 — 1,5

m) with reworked ferricrete, and set in a reddish

sandy loam matrix. These soils are best described as

a dystrophic Hutton Form (stony phase), although

the rooting layer is mainly confined to greyish sandy

loam topsoil. On the upper slopes of anticlinal

remnants (Klipfontein Mountains) soils are of

Mispah and Glenrosa Forms, whereas deep Hutton

Form soils, supporting thicket vegetation, in

contrast to Grassy Fynbos elsewhere, occur on

stable, north facing screes and talus slopes. The

anticline to pediment catena on the southern flanks

of the hills is Mispah — » Glenrosa — > dry Fernwood

— » wet Fernwood and Longlands on the pediment.

The two last-mentioned soils are underlain by a

ferricrete hardpan, are seasonally waterlogged, and

support predominantly herbaceous vegetation (res-

tioid grassland). Weakly developed podzols are

occasionally found: Houwhoek Form on the lower

slopes and Lamotte Form in depressions on the

pediment.

A more or less similar catena exists on the

southern sides and valleys of the planed folds south

of the Kromme River (see Fig. 1). A schematic soil

and vegetation catena is shown in Fig. 21. Deep

sands with a strongly developed bleached (E)

horizon (Constantia Form) occur on north slopes

above the Kromme River.

Along the valley walls of the Kromme and other

rivers there are a number of slope breccias and

colluvial deposits usually on concave slopes. Soils

are often deep, sometimes comprising an angular

rock rubble in a matrix of greyish to brownish

organic enriched loam to sandy loam. Deep red

Hutton Form soils are also found. These soils are

associated with thicket vegetation.

Grassy fynbos soils are infertile, strongly acid (pH

4, 2-5,1) sands, loamy sands and sandy loams;

thicket soils are moderately acid (pH 4,8 -6,5) loams

to sandy loams whose improved fertility status is

probably due to organic matter enrichment (Fig. 8) .

1.4.3 Bokkeveld Group

Soils derived from the Bokkeveld shales can be

grouped into two classes. The first class comprises

the shallow duplex soils of the level interfluves

below the Coastal Platform (see Fig. 1). These soils

are at a relatively youthful stage of development as

evinced by the poorly developed B horizons.

Subsoils are clay-rich lithocutanic (sometimes pedo-

cutanic) horizons which have developed in situ;

topsoils (0,15-0,3 m) are sandier (loams to sandy

loams) and typically demarcated by a stone line,

indicating colluvial origin. The Glenrosa Form

(Williamson Series GS 16) is predominant although

the Swartland Form (Breidbach Series SW 12) is

commonly found. As a consequence of the duplex

structure these soils become waterlogged in winter,

especially where run-off is minimized by a low relief

and good vegetal cover. Under these conditions they

support grasslands, but in regions of steeper relief

and where overgrazing has resulted in prolonged soil

exposure and capping, and topsoil truncation,

renosterveld is the predominant vegetation.

The second group comprises deeper, well-drained

soils of the slopes and bottoms of the river valleys.

They are invariably thicket covered, have a super-

ficial enrichment of organic matter and a high water-

holding capacity. Common forms are Clovelly and

Hutton Form soils although Glenrosa Form soils

do occur where fractured bedrock crops out along

valley slopes. In the last-mentioned situation, it is

difficult to estimate effective soil depth since roots

can penetrate down soil-filled fissures in the

bedrock. Slope breccias are also common in the

valley topography. A schematic soil-vegetation

catena is shown in Fig. 29. Renosterveld soils are

strongly to moderately acid sandy loams to loams,

considerably more fertile than fynbos soils (Fig. 8).

Thicket soils are fairly fertile, moderately acid

loams, sandy loams and clay loams (Fig. 8).

1.4.4 Uitenhage Group

The variable nature of the Enon beds in terms of

the mode of deposition and resultant parent material

is fully expressed in the complexity of soil types

associated with the Uitenhage Group. A rough

distinction can be made between soils derived from

parent materials deposited under lacustrine and

estuarine conditions and those deposited under

fluviatile conditions (cf. Haughton et al., 1937; Du

Toit, 1966). The former are fine-grained rocks

(mudstones, shales, soft sandstones) which are

strongly weathered giving rise to deep uniformly

yellow-brown to red soils (Clovelly, Hutton and

Griffin Forms). These soils are well-drained and

support thicket.

The fluviatile Enon sediments are massive

coarse-grained conglomerates and sandstones; soil

structure and fertility status are broadly similar to

those derived from the sandstones and quartzites of

the TMG. Soils of the conglomerates of the coastal

plain and southern margin of the Gamtoos Valley

are of Oakleaf Form (stony phase). Topsoils are

shallow (0,1 -0,3 m) and the rooting zone does not

usually penetrate the underlying conglomerate.

Above Hankey and Loerie, shallow topsoils overlie

coarse false-bedded sandstones (Glenrosa Form);

deeper Cartref and Constantia Forms, with bleached

E horizons, are occasionally found on gently sloping

ground and depressions. The boundaries of the

thicket and fynbos communities (Fig. 3) in the area,

correlate with the distribution of the two above-

mentioned suites of soils.

Thicket soils are moderately acid, fairly fertile

sandy loams to sandy clay loams (Fig. 8). Fynbos

soils are texturally and chemically similar to TMG

fynbos soils, whereas renosterveld soils are of

intermediate fertility and similar to renosterveld

soils on Bokkeveld shale (Fig. 8). They differ in

being sandier and well-drained.

1.4.5 Recent dune sands

Soil patterns of the aeolian dune sands are

comparatively simple and the various forms are

strongly associated with dune topography. The

186 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

parallel dune ridges have well-drained Fernwood

Form (Langebaan Series FW 21). The climax

vegetation is thicket, although successional dune

fynbos communities are prevalent today. On

southern slopes of the dunes, soils are organic

stained to a depth of about 0,5 m. On drier, less

densely vegetated north slopes, the humus layer is

often absent suggesting gravitational transport of

soil surface material. Near Oyster Bay soil

development has resulted in iron oxide coatings on

the sand grains and these no longer qualify as regie

sands: they are classified as Clovelly Form (Oranje

Series Cv 41) or more rarely Hutton Form (Nyala

Hu 41) and are probably allied to the Red Berea

Sands (Heydorn & Tinley, 1980).

Soils of the dune valleys are mostly poorly drained

owing to the presence of a calcrete hardpan. Deep,

seasonally waterlogged sands with an abnormal

accumulation of organic matter predominate. These

are classified as Fernwood Form (Soetvlei Series FW

41). In some situations soft plinthic subsoils are

found within 1,2 m of the surface and the soils are

therefore classed as Longlands Form. Lamotte

Form, showing a vesicular hardening in the B

horizon, is occasionally found in dune hollows. The

abovementioned suite of soils support predomi-

nantly herbaceous vegetation.

Where the calcrete is exposed or covered by a thin

mantle of sand, the soil form is Mispah (Kalkbank

Series Ms 22) and the vegetation is dune fynbos. A

schematic soil-vegetation catena in the dune

topography near Cape St Francis is shown in Fig. 25 .

Dune soils are neutral to alkaline medium sands.

Fertility status is high but the availability of nutrients

such as phosphorus may be limited by high pH

values (cf. Brady, 1974) (See Fig. 8).

1.4.6 Alluvium

Stratified alluvium (Dundee Form) occurs in the

lower reaches of the Kromme and Gamtoos Rivers.

I sampled alluvial soils at one site in the Gamtoos

Valley near Hankey. They support thicket and are

deep (> 1,5 m), neutral loams and extremely fertile

(85 to 90 ppm available phosphorus).

1.5 Historical land use

The past two decades have seen much research on

the prehistory of the south and south-eastern Cape

(Klein, 1974; Deacon & Brooker, 1976). These

studies have made an important contribution to our

understanding of primitive man from Middle to Late

Stone Age cultures and also on environmental

changes during the time spanning these cultures

(about 100 000 years ago to historical times). In this

section I focus on patterns of land use associated

with indigenous peoples prior to their displacement

by European settlers and on settled agricultural

practices associated with the latter peoples. I

concentrate on those forms of land use which

contributed to recent gross vegetational changes in

the study area.

The aboriginal people encountered by the first

Europeans in the Humansdorp area were terminal

Later Stone Age ‘Bushmen’ and ‘Hottentots.’ The

former were largely hunter-gatherers, whereas the

latter were mainly herdsmen. Archaeological studies

are beginning to elucidate the rather complex way in

which they shared and utilized the landscape.

Domestic livestock (sheep) are known from the

southern Cape from at least 17 000 B.P. (Schweitzer

& Scott, 1973) and the proportions of bones in the

most recent layers of some cave deposits suggest that

the inhabitants were principally herders (Klein,

1977). Early travellers (Sparrman, 1785; Thunberg,

1796) indicate that primitive herders made judicious

use of fire and were astute veld managers who

moved their settlements as soon as veld conditions

began to deteriorate. The scenario before the advent

of the Europeans can be summarized briefly as

follows: a localized but intense grazing by fairly low

populations of domestic and indigenous ungulates

(‘pulse disturbance grazing’ of Noble & Slatyer,

1980). Grazing was most likely concentrated on the

more nutritive pastures on Bokkeveld shale, dune

sands and some Enon beds (cf. Stinder et al., 1965).

Veld on TMG quartzites would then, as now, have

had a lower cover of grasses which were predomi-

nantly ‘sour’ and supported a relatively low

herbivore biomass. Fires were started both by man

and other factors (lightning, rock falls) and the fire

regime was probably variable.

A reconstruction of vegetation at the beginning of

historic times can unfortunately only be inferred

from the vague accounts of early travellers and also

from present day vegetation relics. The course of the

eastern migration of the 18th Century ‘trekboers’

proceeded through the Langkloof and into the

Humansdorp region. As early as 1744 a ‘loan’ farm

had been registered at the Kabeljous River near

Jeffreys Bay and in 1770 the Gamtoos River was

declared the eastern colonial boundary (Botha,

1923). Many travellers (Sparrman, 1785; Paterson,

1790; Thunberg, 1796; Lichtenstein, 1812; Camp-

bell, 1815) commented favourably on the grassveld

between the Kromme and Seekoei Rivers. In

Lichtenstein’s (1812) words: ‘Large tracts of land

near Gamtoos are covered with wholesome nou-

rishing grass ... on which were grazing ... a great

variety of wild animals, particularly oribi’. Eland,

quagga, zebra and hartebees were also noted in the

surroundings (Paterson, 1790). Buffalo and elephant

were observed in the valley thicket and hippo in the

rivers (Sparrman, 1785; Paterson, 1790; Skead,

1980). From Seekoei River mouth, travellers headed

northwards across the Coastal Platform, towards the

Gamtoos Valley where the village of Hankey now

stands. They would therefore have had to traverse

fynbos about which Steedman (1835) had this to say:

‘The surface of the plain consisted of a sandy

gravelly soil, perfectly hard, with moor-like vegeta-

tion and scarcely a tree to enliven the dreariness of

the route’.

One of the first results of early settled agriculture

was the elimination of the aboriginal graziers and

indigenous ungulates, and the replacement of the

latter with the settler’s domestic livestock. Farms

were large and uncamped and stocking rates initially

low (Gerryts, 1949). There followed a period of light

selective grazing. Gradually livestock numbers

increased, farms were subdivided and stocking rates

R. M. COWLING

187

rose sharply. Veld was, and is, burnt as often as

possible (four-to-five-year cycle) in mid- to late

summer (Smith, 1967). A combination of continuous

overgrazing and the lack of post-fire rests resulted in

widespread veld deterioration (Smith, 1967; Lie-

benberg, 1945). These malpractices have persisted

until recent times. In a survey of the Humansdorp

region carried out by Smith (1967), 33% of the

sample farms were not divided into grazing camps

and only 14% of the farmers rested their veld for

more than six months after a fire.

Major changes in vegetation are probably

correlated with the transition from pulse- to

continuous-disturbance grazing regimes and from a

patchy to a rigid and uniform fire regime. Today, the

grasslands on the Bokkeveld-shale and Enon-

conglomerate coastal flats have been largely

replaced by renosterveld shrublands, although a few

grassland relics remain. The spread and thickening

up of Elytropappus rhinocerotis (renosterbos) in

southern and south-eastern Cape coastal grasslands

has been well documented (Sparrman, 1785;

Levyns, 1926; Hall, 1934; Bagshaw-Smith, 1937; Du

Toit & Du Toit, 1938; Smit, 1943; Levyns, 1956). I

propose a mechanism for the encroachment of

grassland by E. rhinocerotis later in this paper.

Changes in fynbos vegetation have probably been

less drastic, involving mainly the elimination of the

fire-sensitive seed-regenerating species. Undis-

turbed thicket is not fire-prone and its total extent is

unlikely to have contracted much in historical times.

The great fire of 1869, which raged uninterrupted

from the south-western Cape to Uitenhage, prob-

ably destroyed isolated pockets of thicket and forest

in the study area (cf. Gerryts, 1949).

2 METHODS

2.1 Data collection

2.1.1 Sampling strategy

The two major physiographic units in the study

area (coastal foreland, valley/mountains) each

required a different sampling strategy. In the former

region an investigation of community response to

disturbances such as burning, bush-cutting and

grazing formed a major part of the study. After

initial reconnaissance, I divided the vegetation into

units on the basis of geological substrate. As a result

of consultations with the local extension officer of

the Department of Agriculture I was able to choose,

as study sites on each parent material, a number of

farms which reflected a wide range of management

practices. Each farm formed a major site where

releves were selectively placed in homogeneous

vegetation stands (cf. Werger, 1974) with a known

history of utilization. To ensure an even sampling

spread I placed a number of releves outside these

major sites. This sampling approach is essentially

similar to the approach used by the Ziirich-

Montpellier School (Westhoff & Van der Maarel,

1973; Werger, 1974). Random sampling would have

resulted in unnecessary replication and the possibili-

ty of an unrepresentative sample with an imperfectly

known utilization history.

The disturbance regime in the Gamtoos River

Valley and Elandsberg Mountains is considerably

less intense than the coastal forelands. I used direct

gradient analysis (Whittaker, 1967) to compare

coenoclines (cf. Whittaker, 1973) in mature fynbos

and non-fynbos vegetation along parallel elevational

gradients (Cowling & Campbell, 1983b). On both

coenoclines sample sites were stratified by elevation

and releves were randomly located on each major

aspect (N, S, E and W) within the site. Further

releves were placed along the coenoclines to ensure

a representative sample for classificatory purposes.

2.1.2 Sampling intensity and plot size

Sampling intensity is largely governed by the aims

of a particular study and is usually limited by logistic

and time constraints. In this study, I was not

concerned with obtaining an accurate classification

at the level of the association ( sensn Westhoff & Van

der Maarel, 1973), but rather the identification of

communities which expressed and characterized the

floristic variation and integrity of higher syntaxono-

mic units. However, the degree of variation that can

be integrated into a meaningful expression of a

vegetation unit is still a matter of judgement

(Mueller-Dombois & Ellenberg, 1974). I sampled a

total of 194 releves, including 879 species. After the

classifications were constructed I located 51 ‘test’

plots in undersampled areas. With a few exceptions

it was possible to integrate these plots into the

established communities.

Werger (1972) defines optimal plot size as that

size nearest to the minimal area, giving the best

compromise between information obtained and

effort expended. Much research has been aimed at

clarifying the concept of minimal area but it appears

that an objective definition is impossible (Werger,

1972; Kershaw, 1973). In this study it was necessary

to have a fixed plot size to facilitate comparisons of

species richness (see Cowling, 1983b). I used the

approach of Werger (1972) that determines the

optimal plot size as having a 50-55% content of a

hectare information (number of species in a

hectare). Using the standard species-area relation-

ship and regarding one hectare as having an

information content of 100% optimal plot size can

then be determined on the basis of the resolution

desired for the study.

Nested quadrats containing plot sizes 1, 5, 10, 100

and 1 000 m2 (see Whittaker et al., 1979) were

sampled in a wide range of vegetation types in the

study area. With a few exceptions, 10 x 10 m plots

retrieved the desired level of information (>50%)

(Table 1) and this size was considered optimal for all

vegetation in the study area. Two of the communi-

ties on coastal dunes (Themeda— Stenotaphrum and

Restio—Maytenus ), apparently required larger plots.

Here communities form a complex mosaic of

successional states and it is often difficult to locate

homogeneous stands of 1 000 m2.

A plot size of 10 x 10 m has been used in fynbos

and Afromontane forest in the south-western Cape

(Werger et al., 1972; McKenzie et al., 1977;

McKenzie, 1978) but most fynbos studies use smaller

plots (5 x 10m) (e.g. Taylor, 1969; Bond, 1981).

188 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

TABLE 1. — Information content of 100 nU (10 x 10 m) plot size determined from species-area

extrapolations (1 ha = 100% info.)

* Structural characterization here and elsewhere in this paper according to Campbell et al. (1981)

2.1.3 Vegetation parameters and site variables

Within each releve I subjectively estimated

percentage projected canopy cover of each species.

Species were classed into growth forms and

additional structural data were recorded (for details

see Cowling & Campbell, 1983b). Nomenclature

follows the Albany Museum Herbarium (GRA,

Botanical Research Institute) in Grahamstown,

where species were identified. Environmental vari-

ables recorded in the releves are shown in Table 2.

2.2 Data analysis

2.2.1 Community classification

On the coastal plain I initially subdivided the

vegetation according to geological substrate because

parent material appeared to be a major factor in

determining the distribution of vegetation types (cf.

Taylor, 1978; Kruger, 1979; Boucher & Moll, 1980).

This assumption was largely corroborated by the

results of a polythetic agglomerative classification of

the whole data set (group-average sorting using

relativized Czekanowski coefficient (Campbell,

1978). An identical classification of the releves from

the Gamtoos/Elandsberg region showed that this

data set could be divided into two groups — fynbos/

renosterveld and thicket/forest communities.

I used two-way indicator species analysis (TWIN-

SPAN; Hill, 1979) to classify releves of each data

set. TWINSPAN is a recent improvement on

indicator species analysis (ISA: Hill et al., 1975).

ISA is a polythetic, divisive classificatory technique

which has proved successful in a variety of

vegetation types (Hill et al., 1975; Hall & Swaine,

1976; Basset, 1978; Daniels, 1978). TWINSPAN and

ISA produce a classification of stands by the

progressive splitting of ordinations (reciprocal

averaging; Hill, 1973) at their centres of gravity. At

each split indicator (diagnostic) species are chosen to

define the two groups of data; serious misclassifica-

tions are rare (Hill, 1979). TWINSPAN produces a

classification of species as well as stands and is

designed to construct an ordered two-way table

which approximates the tabular matrix arrangement

of the Zurich-Montpellier School and, as such, is a

highly promising technique (Gauch & Whittaker,

1981). TWINSPAN therefore approaches the much

desired integration of classical syntaxonomy with

numerical phytosociology (cf. Dale & Webb, 1975;

Komarkova, 1980). The technique has been used

successfully in southern Cape fynbos communities

by Bond (1981).

I compared the two-way phytosociological tables

produced in the classifications and extracted

synoptic tables for each order (see below). These

tables (Tables 4 to 10) summarize species constancy

to communities within an order and the diagnostic

value of species of that order. Constancy is rated on

a 1-5 scale: 1 = 1-20% frequency; 2 = 21-40%; 3

R. M. COWLING

189

TABLE 2. — Environmental variables recorded in all releves. Soil chemical data for the A horizon

only from a subset of 97 samples. Some details on methods and classes of variables are shown

= 41—60%; 4 = 61—80%; 5 = >81%. Only those

species with values greater than 1 in one or more

communities were included in the synoptic tables. I

determined diagnostic values for species in relation

to orders. Species marked * are diagnostic for a

particular order and include both character and

differential species (cf. Werger, 1974); species

marked 1 are also diagnostic species but are regional

and local endemics and are either local or general

character taxa (Werger & Van Gils, 1976). A more

detailed analysis of species diagnostic values must

await more extensive phytosociological studies in the

southern and eastern Cape.

2.2.2 Syntaxonomic ranking

At present there are inadequate plant community

data in the Fynbos Biome to facilitate a rigorous and

formalized syntaxonomic treatment. The ranks I

propose here are tentative and are not all formally

defined in terms of quantitative measures of plot

similarity. Communities were identified in the

TWINSPAN classifications and depicted in two-way

phytosociological tables and ordination diagrams

(Cowling, in prep.). Orders and classes were

established subjectively from a study of the tables,

field observations elsewhere in the southern

south-eastern Cape, and from a perusal of the

relevant literature. The hierarchical system I present

(Table 3) is an attempt to meet the guidelines

proposed by the Botanical Research Institute (n.d.).

In this study, sampling was not sufficiently

intensive to extract associations ( sensu Werger,

1974) as the final syntaxonomic unit. I have termed

the ultimate units ‘communities’; in some cases they

are equivalent to associations but could mostly be

ranked as alliances. The pivotal rank in this

treatment is the order (orders are mapped in Fig. 3).

Following Van der Meulen’s (1979) treatment of

western Transvaal bushveld, I regard the order as

being roughly equivalent to an Acocks (1953) veld

type. It is an agro-ecological unit comprising a range

of communities which are biogeographically and

structurally related and have broadly similar habitat

190 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

requirements. The order is a useful mapping unit as

a first stage in planning and development. It must be

stressed that most of Acocks’s veld types would

include many orders as defined here. I have called

the highest unit a class which comprises a group of

related orders.

These definitions are unfortunately vague and

clearly do not correspond to the more explicitly

stated definitions in classical phytosociology. A

clearer picture of the ranks emerges in later

discussions. In the Cape fynbos communities where

gamma and delta diversities are high (Kruger &

Taylor, 1979), there are likely to be large numbers of

syntaxa of all ranks. Many orders will have the same

agro-economic potential since, particularly in the

western Cape, composition may change drastically

with very little change in environment.

2.2.3 Community nomenclature

Varied approaches to the classification of the

plant communities in South Africa have resulted in a

chaos of syntaxonomic nomenclature. Recently the

Botanical Research Institute (n.d.) has proposed a

series of guidelines, based largely on the code laid

down by Barkman et al. (1976), for a standardized

system of syntaxonomic nomenclature for South

Africa. I have used these guidelines in the

community nomenclature presented below (see

Tables 4-10).

The nomenclature includes a binomial specific

connotation where the first species is a dominant and

the second that of a differential or character species.

This is followed by a locality-structural term, e.g.

Humansdorp Grassy Fynbos. At least one of the

species names of the binomial (preferably the first,

dominant one) is a species whose presence and

abundance is not grossly affected by the currently

applied fire and grazing regimes. Each community is

given a formal structural description according to the

system of Campbell et al. (1981).

2.2.4 Biogeographical analysis

I determined the phytochorological affinities of

each species in terms of established phytochoria

(Werger, 1978; White, 1983). The distributions of

taxa were established from locality records in the

Albany Museum Herbarium (GRA) and the Bolus

Herbarium (BOL) and from distribution maps in

recent revisions. Species were classified as:

(i) endemic to a particular phytochorion;

(ii) linking two (usually) adjacent phytochoria;

(iii) widely distributed, common in tropical and

subtropical phytochoria;

(iv) widely distributed, occurring in temperate and

tropical phytochoria and often extending

outside Africa.

Categories (ii)— (iv) are ecological and chorological

transgressor species of White (1971).

Regional and local endemics were classed as being

of Cape, Afromontane, karroid (Karoo-Namib) or

subtropical (chiefly Tongaland-Pondoland) affinity.

Cape endemics are restricted to Weimarck’s (1941)

South-Eastern Centre. Karroid and subtropical

endemics are restricted to the Kaffrarian Transition

Zone (Cowling, 1983a), a region extending from the

Humansdorp District eastwards to the Kei River and

bounded in the north by the Sneeuwberg-

Winterberg axis. I give details and examples of

distribution patterns elsewhere (Cowling, 1983a).

2.3 Community dynamics

An understanding of community dynamics is

essential for the full characterization of vegetation

subjected to recurrent disturbances. In managed

natural systems a knowledge of dynamics provides

the soundest basis for long-term management

(Slatyer, 1976). I studied the dynamics of fynbos and

renosterveld communities of the coastal forelands

and the results are fully discussed in Cowling (in

prep). As these results are referred to in this paper, I

undertake below to outline the approach and

methods used to study dynamics and to clarify the

assumptions implicit in these methods.

In South Africa, the study of vegetation dynamics,

particularly secondary succession, has been severely

limited by the Clementsian dynamics paradigm (e.g.

Phillips, 1931; Acocks, 1953 and many others). The

classical view of ecological succession and the climax

(Clements, 1916) is that, following a disturbance

there is a community replacement sequence which

culminates in the final community (climax) which is

in equilibrium with the prevailing environment.

Implicit in this view is the assumption that each

assemblage of species (community) modifies the site

conditions so that it becomes less suitable for its own

persistence and more suitable for its successor.

Furthermore, Clements assumed that, given time, a

climax vegetation of the same general type

(monoclimax) will be produced and stabilized in a

similar climatic region, irrespective of earlier site

differences. These concepts are still upheld in the

literature (e.g. Odum, 1969).

Clements’s concepts have proved to be unrealistic

and unworkable in the field and attempts to uphold

them have resulted in a confusion of terms to

accommodate situations which do not fit the ideal.

Two papers have effectively undermined the

assumptions of Clementsian ecology. Whittaker’s

(1953) incisive and exhaustive paper on climax

concepts liberated botanists from the monoclimax

paradigm in a manner more realistic than Tansley’s

(1935) earlier attempt. The essence of Whittaker’s

analysis is as follows: ‘There is no absolute climax

for any area, and climax composition has meaning

only relative to position along environmental

gradients . . . All climaxes are edaphic as well as

topographic and climatic climaxes . . . All are part of

the climax pattern.’ Drury & Nisbet (1973) present a

devastating criticism of the generalizations made by

classical succession theory. They refute the univer-

sality of the classical ‘relay floristics’ model of

succession (see also Egler, 1954) and stress the

importance of initial floristic composition of the site

in determining successional pathways. They also

note that modification of the environment by plants

acts generally to delay succession and not facilitate

the invasion of the site by new species.

The reviews cited above and later papers (e.g.

Horn, 1974; Pickett, 1976; Connel & Slatyer, 1977;

R. M. COWLING

191

Van Hulst, 1978; Noble & Slatyer, 1980; Glenn-

Lewin, 1980; Peet & Christensen, 1980) have

emphasized life history properties (colonizing abili-

ty, dispersal, growth rates, longevity) of individual

species (cf. Gleason, 1926) and not the emergent

properties of communities, in determining the

pattern of succession. Two useful developments are

Connel & Slatyer’s (1977) description of succession-

al pathways in terms of species facilitation, tolerance

and inhibitive properties, and Noble & Slatyer’s

(1980) use of species’ vital attributes to predict

post-disturbance successional pathways.

Methods to study community dynamics range

from inductive studies (observations and experi-

ments) to deductive studies using both simple

qualitative and complex quantitative models (Good-

all, 1977). For philosophical and practical reasons,

the deductive approach is preferred, although

long-term observations will always be required to

validate predictions.

The approach I used was to select sites of known

disturbance history, encompassing a wide range of

disturbance features (see 2.1 above). I used

TWINSPAN (see 2.4.1. above) to classify releves in

terms of floristic and structural attributes (Cowling,

in prep.). In this way samples were organized into a

manageable number of floristic and structural

community-types which were then subjected to

further investigation (cf. Slatyer, 1976; Austin,

1977). A comparison of the two classifications

showed that within each floristic community there

were a number (usually two) of structural communi-

ties or structural ‘cover states’ assumed to corres-

pond to different successional developments of that

community. The assumption that spatially separate

vegetation represents states of the same system at

different stages of development has been criticized

recently (Drury & Nisbett, 1973; Goodall, 1977). In

my study this assumption is supported by the high

floristic similarity of diverse structural types and the

existence of fenceline contrasts (Figs 30, 32) where

the disturbance regime was the only site variable

that differed across the fenceline.

I used the qualitative model employing vital

attributes of species (Noble & Slatyer, 1980) to

predict successional patterns under different distur-

bance regimes. These predictions were tested

against vegetation samples, of known disturbance

history, that were collected for the multivariate

analyses.

3 THE SYNTAX A

In this section the syntaxonomic concepts are

defined and named. Data summarizing structural

and ecological relationships of the classes and orders

are shown in Table 3. Tables 4—10 summarize

floristic data. Figs 9 & 35 show phytochorological

spectra and patterns of endemism, and Figs 10 & 36

show some structural data for selected communities.

3.1 Cape Fynbos Shrublands

I use the term ‘Cape Fynbos Shrublands’ to

describe a class of communities that includes the

fynbos vegetation of the Cape Floristic Region.

Specht (1979) and Specht & Moll (1983) include

Cape fynbos within a global heathland concept.

However, not all Cape Fynbos Shrublands are

heathlands sensu Specht (1979). In accordance with

the structural scheme proposed by Campbell et al.

(1981) I restrict the use of the term heathland to

those communities having a high cover of true

heaths (Ericales) (see also Bond, 1981; Campbell,

1983b). Therefore the more general term ‘shrub-

land’ which includes heathlands, is preferred.

A precise definition and delimitation of Cape

fynbos vegetation eludes botanists up to the present.

In a recent review of the Cape phytogeographical

region, Taylor (1978) defined fynbos as follows:

‘Floristically fynbos can be defined by . . . the lack of

single species dominance and/or the conspicuous

presence of members of the family Restionaceae.

Physiognomically fynbos is characterized by three

elements, restioid, ericoid and proteoid.’ Kruger

(1979), in another review, accepts Taylor’s defini-

tion and states that ‘. . . the only constant and

differential floristic element is the Restionaceae.’

Both authors give a list of typical genera.

These definitions are easily refuted. A general

lack of single species dominance in fynbos is a myth:

fynbos communities often have dominance concen-

trated in one or two species (Cowling, 1983b).

Restionaceae are prominent in the Southern

Variation of the Strandveld (34a) (Acocks, 1953;

Boucher & Jarman, 1977) and occasional in Coastal

Renosterbosveld (46) (Acocks, 1953; Boucher,

1983; Table 7), both non-fynbos types. Afromon-

tane fynbos (Killick, 1979), although often lacking

Restionaceae, has a similar structure and generic

composition to Cape fynbos (see Story, 1952; Phipps

& Goodier, 1962; Killick, 1963; Edwards, 1967; Van

der Schijff & Schoonraad, 1971). In these communi-

ties, typical fynbos genera as defined by Taylor

(1978) and Kruger (1979) include Protea, Erica,

Muraltia, Cliffortia, Passerina, Phylica, Metalasia,

Stoebe, Helichrysum, Merxmuellera and Pentaschis-

tis. Some workers regard the Cape as an extension of

the Afromontane region (Adamson, 1948; Tinley,

1975; Linder, 1983).

I define the class Cape Fynbos Shrubland in terms

of biogeographic, structural and ecological criteria

as follows:

(i) Sample floras show a phytochorological spec-

trum in which approximately 50% of the species are

restricted to the Cape phytochorion as delimited by

Werger (1978). The majority of the remaining

species are largely Cape linking taxa (Fig. 9).

(ii) A high incidence of regional endemism (cf.

Weimarck, 1941). Regional endemics are mostly of

Cape affinity. Some data are available for the study

area (Fig. 9) but more are needed for a quantitative

assessment.

(iii) Structural definition according to Campbell

(1983b) reflecting characteristics of the ‘heathland

syndrome’ (small leaves, proteoid isobilateral

leaves, sclerophylly, evergreen hemicryptophytes

(cyperoids and restioids) (Specht, 1979; Fig. 10)).

Campbell’s (1983b) structural definition of fynbos of

the Cape Fold Mountains and my biogeographic

definition show good agreement in the study area.

192 A SYNTAXONOMIC AND SYNECOLOGICAL

S.E. MOUNTAIN FYNBOS GRASSY FYNBOS

STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

S .COAST DUNE FYNBOS S.COAST RENOSTERVELD

Tetraria-Thamnus community Erica-Trachypogon

n=8(no.of relev6s) n=16

Restio-Agathosma

n=13

Elytropappus- Metalasia

18,0

sclerophyll

\orthophyll

fleshy

Leaf texture (°/o woody cover)

Fig. 9. — Phytochorological spectra and endemism for Cape Fynbos Shrublands and Cape Transitional Small-leaved Shrublands

( Elytropappus — Metalasia Community). C = Cape endemics; C-A = Cape-Afromontane linking species; C-TP =

Cape-Tongaland-Pondoland linking species; C-KN = Cape-Karoo-Namib linking species; TW = Tropical-subtropical wides;

W = wides; RES = residual (groups contributing less than 5% total species). Geographical affinity of endemics: C = Cape; T

= Subtropical; K = Karroid.

S.E. MOUNTAIN FYNBOS

Tetraria-Thamnus community

n = 8 (number of relevgs)

GRASSY FYNBOS

Erica -Trachypogon

n-16

SOUTH COAST DUNE FYNBOS

Restio - Agathosma

n=13

SOUTH COAST RENOSTERVELD

Elytropappus -Metalasia

n=12

III IS BOZO

C C A C-TP C KN TW W RES

Geographical affinity

Fig. 10. — Growth form composition of Cape Fynbos Shrublands and Cape Transitional Small-leaved Shrublands ( Elytropappus —

Metalasia Community). Definition of growth forms as in Campbell et al. (1981). Leaf texture categories defined in Cowling &

Campbell (1983a).

R. M. COWLING

193

TABLE 3. — Syntaxonomic and synecological relationships of higher vegetation units in the Mumansdorp region

1 Structural characterization according to Campbell et al. (1981).

2 Rainfall limits are approximate and do not take into account local variations of slope, aspect, soil drainage and texture, all of which affect soil moisture.

3 The term ‘Renosterveld' (Boucher, 1980) is used instead of Rhenosierbosveld (cf. Acocks, 1953).

(iv) Ecologically restricted to areas receiving a

substantial proportion of winter rainfall and having

sandy, infertile soils and alkaline calcareous sands

(Table 3).

Due to the high gamma diversity of Cape Fynbos

(Kruger & Taylor, 1979) future phytosociological

studies will most certainly result in a proliferation of

phytocoenoses at all syntaxonomic ranks. A

structural treatment of fynbos (Linder & Campbell,

1979; Bond, 1981; Campbell, 1983b) is probably

more pragmatic at this stage.

I recognized three Cape Fynbos Shrubland orders

in the study area.

3.1.1 South-eastern Mountain Fynbos

Mountain Fynbos is the term used by Taylor

(1978) and Kruger (1979) to replace Acocks’s (1953)

Macchia (69) and False Macchia (70). The term is an

unfortunate choice since Mountain Fynbos commu-

nities are frequently found near the coast (cf.

Coastal Fynbos). However, the term is well

entrenched in the literature and there is little point

in discarding it.

Mountain Fynbos comprises communities of the

Cape Folded Belt from the Cedarberg in the

north-west to the Groot Winterhoek Mountains in

the south-east. It occurs under a sub-humid to humid

rainfall regime in the western and southern Cape

(>400 mm yr1) and in moist upland regions

(>750mm yr1) in the south-eastern Cape. Mountain

Fynbos is invariably restricted to acid infertile and

sandy soils derived from quartzites and sandstones

of the TMG. Communities are almost entirely

composed of Cape endemics and regional endemism

is high (Fig. 9). Grasses are rare and when present,

are of the C3 type (e.g. Pentaschistis, Merxmuellera)-,

large-leaved dorsiventral shrubs and other subtropi-

cal growth forms are lacking (Fig. 10). Structurally

the communities are Mid-dense to Closed Shrub-

194 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

TABLl 4. — South-eastern Mountain Eynbos Communities

* Diagnostic species for South-eastern Mountain Eynbos, including:

+ South-eastern endemics (Wcimarck, 1941 ) largely restricted to South-eastern Mountain I ynbos

lands, Heathlands and Restiolands. Mountain

Fynbos communities have been described by Taylor

(1969), Werger et al. (1972), Boucher (1978),

McKenzie et al. (1978), Glyphis et al. (1978), Laidler

et al. (1978), Bond (1981) and Taylor & Van der

Meulen (1981).

South-Eastern Mountain Fynbos occurs in the

eastern Baviaanskloof and Kouga Mountains and

the Groot Winterhoek and Elandsberg Ranges

(pers. obs.) (see Table 3). Mountain Fynbos in the

study area was undersampled and I recognized only

two communities; floristic data are shown in Table 4.

The Tetraria—Thamnus Community occurs in

humid (800-1 000 mm yr1) upper (>400 m) regions

of the eastern Elandsberg Range (Fig. 11). Soils are

shallow, leached, acid and infertile loamy sands

(Fig. 8) of the Mispah, Glenrosa, Cartref and

occasionally Oakleaf Forms. On the drier lower

slopes is the Leucospermum — Tetraria Community

which is transitional to Grassy Fynbos (see below).

Soils are mainly Glenrosa and Mispah. Community

dynamics were not studied.

3.1.2 Grassy Fynbos

The concept of Grassy Fynbos presented here is

new. It includes the fynbos communities of the lower

and north slopes and planed forelands of the Cape

Folded Belt in the south eastern Cape (Table 3).

Fig. 11. — Mature South-eastern

Mountain Fynbos, Tetraria-

Thamnus proteoid Commu-

nity on shallow leached loa-

my sand (Cartref Form) in

the Elandsberg Mountains

(c. 500 m elevation). Domi-

nant species: Leucadendron

loreriense, Hypodiscus stria-

tus, H. synchroolepis, Erica

copiosa. Between Loerie and

Otterford forest stations.

R. M. COWLING

195

Both Taylor (1978) and Kruger (1979) mention the

increased grassiness of the south-eastern Cape

fynbos but include it with Mountain Fynbos.

3.1.2a Characterization

Biogeographically, Grassy Fynbos is characterized

by a high proportion of Cape-Afromontane linking

species and widely distributed tropical C4 grasses

( Themeda , Trachypogon, Heteropogon, Brachiaria,

Eragrostis) (Fig. 9). The order has good diagnostic

species including a number of regional and local

character taxa (Table 5). Structurally, Grassy

Fynbos communities are similar to Mountain Fynbos

except for the prominence of grasses in the

understorey at the expense of restioids (Fig. 10).

Soils are marginally more fertile than Mountain

Fynbos soils (Fig. 8).

3.1.2b Grassy Fynbos as a false fynbos

The grassiness and Afromontane links of these

communities probably led Acocks (1953) to his

conjecture that they are derived from grassland and

forest similar to the Afromontane communities of

the eastern Cape (Dohne Sourveld, 41). This is

Acocks’s (1953) False Macchia concept which is

clearly untenable for southern and south-eastern

Cape Mountain Fynbos where tropical grasses are

absent. I argue below that the notion of Grassy

Fynbos as a recently derived vegetation type is also

untenable.

Evidence refuting Acocks’s hypothesis is the large

number of regional and local endemics of fynbos

affinity in Grassy Fynbos. (Fig. 9; Martin, 1966;

Cowling, 1983a). Acocks (1979) recognized this as a

fundamental flaw in his theory: he therefore

introduced fynbos as a permanent feature of the

landscape by invoking a ‘rotating climax’ according

to the Clementsian replacement sequence: scrub

forest (thicket) — » (fire) — > grassland — > fynbos — »

thicket. My observations in the study area and

elsewhere provide little support for this hypothesis.

The distribution of thicket in Grassy Fynbos is

determined largely by the availability of special

edaphic sites (deep water-retaining soils of talus

slopes, slope breccias and termitaria) (Fig. 12) (see

also Martin, 1965). Some of the sites are fire-

protected which probably facilitated the initial

establishment of thicket seedlings. However, most

Grassy Fynbos soils are shallow and stony or poorly

drained and there is no evidence of thicket

development in mature communities. Established

thicket is extremely resistant to most fires: only

thicket margin species are partially destroyed and

they soon regenerate from epicormic buds (Martin,

1966). Furthermore, in the study area and elsewhere

(see Martin, 1966), a predominantly grassy state in

the post-fire succession of most Grassy Fynbos

communities is very short-lived (6-12 months) after

which post-fire resprouters start assuming domi-

nance. Some of these fire-resistant resprouting

species are regional endemics (e.g. Erica pectinifo-

lia, Gnidia coriacea, Protea tenax, Phylica gnidi-

oides ). Grassy Fynbos is not a derived vegetation

type nor does it form a transitory successional stage

between thicket and grassland. The implementation

of a uniform fire regime combined with continuous,

selective grazing has reduced the resilience (Holling,

1973; Walker et ai, 1981) of these systems to a

critical point where the ‘domains of attraction’

(Holling, 1973) of populations are transcended

resulting in shifts in patterns of dominance and local

extinctions. I investigate these ideas more fully

below.

3.1.2c Dynamics

At present, nearly all Grassy Fynbos in the study

area is burnt to a 4- to 5-year rotation. Vegetation is

grazed soon after the fire by domestic livestock

which select only young, growing grasses.

In prehistoric times, grazing was probably

irregular and localized and the fire regime more

patchy (see 1.5). I have chosen the Erica— Trachy-

pogon Community (Figs 12-15) to illustrate the

effect of different disturbance regimes on communi-

ty composition and structure. Noble & Slatyer’s

(1980) vital attributes model was used to predict

successional changes.

Fig. 12. — Erica- Trachypogon

Grassy Fynbos on colluvial

gravels north of Humans-

dorp. Kaffrarian Thicket on

termitarium. Fynbos species

include Phylica abietina, Eri-

ca pectinifolia, Leucaden-

dron salignum, Trachypogon

spicatus, Restio triticeus,

Themeda triandra, Merx-

muellera stricta and Bobartia

orientalis. Near Misgund (c.

220 m elevation).

196 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

TABLE 5.— Grassy Fynbos Communities

A: Themeda triandra-Passerina pendula Hankcy Dry Grassy I'ynbos

B: Erica pectinifolia— Trachypogon spicatus Humansdorp Grassy Fynbos

C: Protea neriifolia-Clutia alatemoides Humansdorp Grassy Fynbos

D: Thamnochortus glaber- Erica diaphana Tsitsikamma Grassy Fynbos

I Thamnochortus fru ticosus- Tristachya leucothrix Tsitsikamma Restioid Grassland

* Diagnostic species for Grassy Fynbos including:

* South-eastern endemics (Weimarck, 1941) largely restricted to Grassy Fynbos

R. M. COWLING

197

Fig. 13. — Erica— Trachypogon

Grassy Fynbos on shallow

stony soil derived from Enon

sandstones. Species include

Protea neriifolia, Erica de-

mis sa, Themeda triandra,

Trachypogon spicatus, Restio

triticeus and Muraltia squar-

rosa. Near Hankey on road

to Loerie (c. 150 m eleva-

tion).

Fig. 14. — Destruction by fire of a

relic stand of Protea neriifo-

lia in Erica— Trachypogon

Grassy Fynbos. This stand

included the only known

population in the study area

of the rare species Agathos-

ma unicarpellata. Near Mis-

gund (c. 200 m elevation).

Fig. 15. — Erica— Trachypogon

Grassy Fynbos on red collu-

vial gravels (Hutton Form).

Frequently burnt and grazed.

Species include Leucaden-

dron salignum, Erica pectini-

folia, Lanaria lanata , The-

meda triandra , Restio triti-

ceus, Leucospermum cu-

neiforme and Brachiaria ser-

rata. Klipfontein Mountains

north of Humansdorp on

road to Hankey (c. 300 m

elevation).

198 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

Species types in the Erica —Trachypogon Commu-

nity include those species which have the ability to

rootsprout or sprout from a lignotuber and in which

new recruitment only occurs immediately after a

disturbance (VI species of Noble & Slatyer, 1980).

Examples are Leucadendron salignum (cf. Williams,

1972), Leucospermum cuneiforme (cf. Rourke,

1972) and Erica pectini folia (pers. obs.). Cl species

(e.g. Protea neriifolia, P. repens ) are species having

a storage of relatively long-lived propagules in the

canopy of the individuals; new recruitment occurs

only after a fire. Seeds are stored in closed cones and

Cl Cape Proteaceae are therefore serotinous (Bond

et al., 1983). The Protea spp. take 4 to 5 years to

reach reproductive maturity and are not long-lived

(20-30 yrs) (Rourke, 1980).

DI species are well dispersed (D), intolerant (I)

species. Examples are Anthospermum aethiopicum

and Cliffortia linearifolia. These species store seed in

the soil and Cliffortia also resprouts from a rootstock

(Martin, 1966). They start declining in importance

after 8 to 10 years. The preponderance of intolerant

(I) species stresses the importance of initial floristic

composition (Connel & Slatyer, 1977) in determi-

ning patterns of secondary succession in Grassy

Fynbos communities.

Fig. 16 shows the replacement sequence in the

Erica -Trachypogon Community under two distur-

bance regimes. Of great significance is that under a

uniform fire regime, the model predicts that the Cl

species will be eliminated since second and later

disturbances occur before juveniles mature, and

propagules are poorly dispersed and short-lived (cf.

Bond, 1980b). VI and DI species types are able to

persist and become dominant under these condi-

tions. These predictions are supported by field

observations. On the Humansdorp coastal plain,

frequent burning has resulted in the elimination of

fire-vulnerable seed-regenerating species from the

Erica -Trachypogon Community (Figs 14 & 15).

Near Hankey this community is not intensively

grazed and fires are patchy and their frequency

variable. Cl species ( Protea neriifolia, P. repens,

Erica demissa (Martin, 1966) dominate (Fig. 13).

Bond et al. (1983) have recently shown that

seedling recruitment of Cape Cl Proteaceae in the

southern Cape differed significantly with fire season.

Best regeneration in reproductively mature stands

occurred after autumn and late summer fires,

whereas spring and winter fires led to very poor

seedling establishment. Bond (1983) demonstrated

that recruitment was dependent on the length of

time that dormant seed was exposed to predation by

rodents. He indicated the failure of simple

population models of succession (such as Noble &

Slatyer’s model) to predict recruitment of Cl

Proteaceae since these models did not take into

account the relationship between the plants and

rodent predators and other interactions. In the study

area, fire season is an unlikely factor in the

elimination of Cl Protea spp. since Grassy Fynbos is

burnt almost exclusively in late summer and early

autumn — Feb. -April (Smith, 1967). Noble &

Slatyer’s (1980) model predicts adequately the

elimination of Cl species under a fire interval too

short to allow for the accumulation of canopy-stored

seed reserves.

3.1. 2d Grassiness

The grassiness of the eastern fynbos communities

requires some discussion. Acocks (1953) and Kruger

(1979) attribute it to a higher proportion of summer

rain favouring the competitive growth of C4 tropical

grasses. An alternative hypothesis is that the growth

of predominantly deciduous C4 species is favoured

on the more fertile eastern fynbos soils (Heddle &

Specht, 1975; Specht et al., 1977; Bond, 1981). The

assumption implicit in this hypothesis, that C4

grasses are more competitive than, for example,

Restionaceae under more fertile conditions, needs

to be critically examined for local conditions.

A third hypothesis, related to the first mentioned,

is that C4 grasses are favoured in those regions where

temperatures are high during the growing season

(Vogel et al., 1977). This hypothesis explains the

high cover of C4 grasses on north slopes, even at high

altitudes, since incoming radiation (and hence

surface temperature) is highest on the north slopes

a Erica pectinifolia

VI Leucadendron salignum

Leucospermum cuneiforme

q| Cliffortia linearifolia

Anthospermum aethiopicum

£1 Protea neriifolia

Protea repens

b PATCHY FIRE REGIME

Life stage

0 4 10 20 40 OC

— m ?ie

- m 1

e

m

le

D/j+Vlj +Clj — DI +VI +Clj -^DJ +VI Cl — ^Dl + VI ♦ Cj | Q

;( io 3/

\ * Dlj tVlj DJ 4 VI +CI*

\

C UNIFORM FIRE *D| * V|j DI +VI D| + VI *0

REGIME —

4 - 5 yr rotation v '

Fig. 16. — Vegetation replacement

sequence for the Erica

- Trachypogon Grassy Fyn-

bos communities. Fig. 16a

summarizes the vital attribu-

te data required to derive the

replacement sequences un-

der a patchy fire regime (Fig.

16b) and a uniform fire

regime (Fig. 16c) (see text).

Life stage parameter charac-

teristics are: m, time to reach

reproductive maturity; /, the

longevity of the population;

and e, the longevity of the

propagule pool. Underlined

species types indicate that

component species are mori-

bund but not locally extinct.

* represents the limited dis-

persal of Cl propagules from

adjacent unburned vegeta-

tion (only possible under a

patchy fire regime). Sub-

script j indicates species in

juvenile (premature) stage.

R. M. COWLING

199

during the equinoxes (Fig. 7) which coincide with

rainfall peaks (Fig. 4). On southern slopes at high

altitudes temperatures during the growing season

are lower and Restionaceae predominate (Cowling

& Campbell, 1983). The hypothesis also explains the

decline in grass cover with increasing age of Grassy

Fynbos communities. Shade cast by the overstorey

creates a cooler microclimate at ground level

favouring species with low optimum temperatures

for growth (C3 grasses and Restionaceae); burning

destroys this effect resulting in higher soil tempera-

tures which favour C4 grasses (cf. Rice & Parenti,

1978). The assumption that dominant Restionaceae

(e.g. Restio triticeus) outcompete C4 grasses at lower

growing-season temperatures should be tested

experimentally. Bond (1980a) and Pierce & Cowling

(1984) have found that in the southern and

south-eastern Cape respectively R. triticeus grows in

the cooler, winter months.

Patterns in the relative coverage of C3 and C4

grasses are also explained by the growing season-

temperature hypothesis (Cowling, 1983c). In Grassy

Fynbos, C3 grasses comprise on average about one

third of the total grass cover (Cowling, 1983c). This

relative coverage increases with vegetation age and

on cool southerly slopes, indicating a microclimatic

lowering of growing season temperatures, favouring

the competitive growth of C3 species. In the study

area C3 grasses persist as a usually minor component

of mature Grassy Fynbos, long after most C4 species

have declined and virtually disappeared (Cowling,

1983c).

3.1.2e Synonymy

Phytocoena conforming to my concept of Grassy

Fynbos have been described from Albany Zuurberg

and quartzitic forelands by Dyer (1937 — Fynbos),

Martin & Noel (1960 — Erica demissa Mixed Heath

Alliance), Martin (1965 — Mature Erica demissa

Heath, Erica demissa — Phylica axillaris Heath,

Tetraria capillacea Mixed Dwarf Shrub Heath) and

Jessop & Jacot Guillarmod (1969 — Macchia- Grass-

veld). Olivier (1977) gives a brief description of

Grassy Fynbos near Port Elizabeth and some of

Phillips’s (1931) lithophilous macchia in the Knysna

District is Grassy Fynbos. At this stage it is uncertain

whether the Grassy Fynbos concept could include

the Protea nitida Communities (‘Waboomveld’)

described by Bond (1981) in the Outeniquas and

Taylor & Van der Meulen (1981) in the Rooiberg.

Campbell’s (1983b) Grassy Fynbos concept is

essentially a structural one; he includes types from

the south-western Cape dominated by C3 grasses and

excludes grassy communities with proteoid oversto-

reys.

3.1.2f Communities in the study area

I have described five Grassy Fynbos communities

from the study area (Table 5). The Themeda -Pas-

serina Community (Fig. 17) is transitional between

Grassy Fynbos and Coast Renosterveld. It is a

species-rich and highly variable community occur-

ring on stony Oakleaf soils derived from Enon

conglomerate, and is restricted to the southern wall

of the Gamtoos Valley where the rainfall is between

400 and 500 mm yr1. The Erica— Trachypogon

Community occurs on the TMG sandstones of the

coastal plain (Figs 12, 14 & 15) and on Enon

sandstones in the Gamtoos region between Hankey

and Loerie (Fig. 13). Soils are litholic, infertile

sandy loams (Glenrosa, Hutton (stony phase)

Forms). Rainfall is 500—600 mm yr'1. The communi-

ty occurs in two structural cover states: one

dominated by resprouting fire-resistant species (Fig.

15), the other dominated by seed-regenerating

species (Fig. 13) (see below). The Protea - Clutia

Community occupies the southern slopes of the low

quartzite hills northwest of Humansdorp which

receive a higher rainfall (600-700 mm yr1) than the

surrounding plains. Soils are shallow Mispah and

Glenrosa Forms on upper slopes and stony Hutton

Form on lower slopes. Structural cover states similar

to the Erica —Trachypogon Community were identi-

fied. The Thamnochortus — Erica (Fig. 18)

Community occurs on deep well-drained and highly

leached sands (Constantia Form) south of the

Kromme River. This community is entirely domina-

ted by resprouting species. Rainfall is 650-750 mm

Fig. 17. — Themeda —Passerina

Dry Grassy Fynbos on shal-

low loamy sands on Enon

conglomerate. Species inclu-

de Protea repens, Elytropap-

pus rhinocerotis, Leucaden-

dron salignum, Passerina

pendula, Themeda triandra,

Restio cuspidatus and Phylica

a bietina. Brandkoppen

above the Gamtoos River

(c. 150 m elevation).

200 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

Fig. 18. — Thamnochortus — Erica

Grassy Fynbos four years

after a fire. Gentle north

slope above Kromme River

(c. 60 m elevation). Soil is a

deep, well-drained sand

(Constantia Form) and is

very infertile. Dominant

shrubs all resprouting from

lignotubers. Erica diaphana,

E. pectinifolia , Leucaden-

dron salignum, Leucosper-

mum cuneiforme. Grami-

noids include Tetraria com-

pressa , Hypodiscus willdeno-

wia, Restio triticeus and Di-

heteropogon filifolius. Lin-

derhof, between Cape St

Francis and Oyster Bay.

Fig. 19. — Thamnochortus -Tris-

tachya Restioid Grassland on

deep, seasonally waterlogged

sands (Longlands Form).

Dominant species include

Thamnochortus fruticosus,

Heteropogon contortus, Tris-

tachya leucothrix and Elion-

urus mutica. Buffelsbos, be-

tween Cape St Francis and

Oyster Bay (c. 80 m eleva-

tion).

Fig. 20. — Restioid Grassland on

pediment below the Klipfon-

tein Mountains, north west

of Humansdorp. Soil is a

deep, seasonally waterlogged

sand overlying a ferricrete

hardpan. Dominant species

are Themeda triandra,

Thamnochortus fruticosus,

Elegia sp., Helictotrichon

hirtulum, Heteropogon con-

tortus and Helichrysum ano-

malum. Rondebosch (c. 220

m elevation).

R. M. COWLING

201

Fig. 21. — Schematic soil and vegetation catena on the coastal foreland (TMG) south of the Kromme River. Communities are fynbos

(Thamnochortus— Erica Community) and restioid grassland ( Thamnochortus— Tristachya Community).

yr1. The Thamnochortus —Tristachy a Community is

a restioid grassland on deep seasonally waterlogged

sands of pediments and valley fills in the Oyster

Bay— Cape St Francis region (Fig. 19) and imme-

diately north and west of Humansdorp (Fig. 20).

Fig. 21 shows a schematic soil and vegetation catena

including the two last-mentioned communities.

3.1.3 South Coast Dune Fynbos

The current concept of Coastal Fynbos (47)

includes a heterogeneous group of vegetation types.

Acocks (1979) subdivided Coastal Fynbos into west

and south coast blocks but acknowledges that

further subdivision is needed. Taylor (1978) and

Kruger (1979) have largely adopted Acocks’s

scheme in their treatment of the Coastal Fynbos.

There are at least four fundamental types of

Coastal Fynbos. Acocks (1953) has long recognized

that the dwarf fynbos of the Elim flats should be

considered a distinct veld type. The south coast

calcrete fynbos centred between Bredasdorp and

Mossel Bay is another well circumscribed type

(Acocks, 1953; Taylor, 1978). These communities

are restricted to calcareous, neutral to alkaline,

shallow sands overlying calcrete of the Bredasdorp

Formation, and are characterized by a great number

of endemics (see Taylor, 1978, for a partial list). Van

der Merwe (1977) has described communities from

the De Hoop Nature Reserve and Muir (1929) gives

a general account of calcrete fynbos in the

Riversdale district. Another group of Coastal

Fynbos communities occurs on' deep, acid, infertile

sands and is mainly confined to the west coast

(Milewski & Esterhuyzen, 1977; Boucher, 1983) but

also occurs locally on the south coast (e.g. sandy

flats vegetation on reddish sands in the Riversdale

district; Muir, 1929). These Coastal Fynbos types

are structurally similar to Mountain Fynbos. The

presence of tall tufted Restionaceae ( Thamnochor-

tus spp., Willdenowia striata ) and a greater

importance of geophytes and annuals in the calcrete

and acid sand fynbos and the occurrence of

subtropical thicket species (Rhus, Euclea) in the

former, could be regarded as diagnostic features.

A fourth type, defined below, is South Coast

Dune Fynbos. Communities belonging to this order

occur on Recent, deepish, calcareous sands from the

Cape Flats to Cape Recife, near Port Elizabeth.

They are characterized by good diagnostic species

including Euclea racemosa subsp. racemosa, Restio

eleocharis, R. leptoclados and Agathosma apiculata

(Table 6). South Coast Dune Fynbos is further

characterized by a strong component of non-fynbos

shrubs and trees which links it to the Tongaland-

Pondoland dune thicket (e.g. Rhus crenata, Olea

exasperata, Maytenus procumbens, Rhoicissus tri-

dentata, Sideroxylon inerme) (Fig. 9; Table 6; see

also Moll & White, 1978). Proteaceae are usually

lacking.

Structurally, South Coast Dune Fynbos is distin-

guished from other fynbos orders by the strong

component of large-leaved dorsiventral shrubs

(subtropical species mentioned above) and the lack

of proteoid shrubs (Fig. 10). Although grass cover

can be quite high, restioids and small-leaved shrubs

dominate in the herb and shrub strata respectively.

According to Campbell’s (1983b) structural criteria,

communities in the study area would be classified as

fynbos.

Taylor (1978) and Boucher & Moll (1980)

partially include my concept of South Coast Dune

Fynbos with strandveld, a large-leaved thicket of

subtropical affinity. They do this because of the

complex successional interplay between fynbos and

thicket communities. On deep well-drained sands,

secondary succession proceeds from a fynbos to a

thicket cover state with a mixed fynbos/thicket

mid-successional state. On shallower sands, overly-

ing calcrete, fynbos communities, identical to the

early fynbos state on deep sands, persist without any

invasion of thicket elements. While recognizing

these successional relationships, I have chosen to

place the thicket and fynbos states in separate

orders.

South Coast Dune Fynbos communities have been

described from the Cape Flats by Taylor

(1972 — Metalasia Coast Dune Fynbos), from Cape

Hangklip by Boucher (1978 — Ehrharta-Ficinia

202 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

TABLE 6. — South Coast Dune Fynbos Communities

* Diagnostic species for South Coast Dune Fynbos, including:

t South-eastern endemics (Weimarck 1941 ) largely restricted to Dune Fynbos (Tsitsikamma group)

Strand Pioneer Community), from Groenvlei by

Martin (1960 — Dune Heath) from Goukamma by

Van der Merwe (1976 — Helichrysum teretifolium —

Metalasia muricata Shrubland Community, Tetraria

cuspidata — Restio eleocharis Fynbos Community),

from Robberg by Taylor (190a —Phylica — Pas-

serina Microphyllous Evergreen Steppe Coast Fyn-

bos, Metalasia- Restio Gray Evergreen Microphyl-

lous Shrub Pseudo-Savanna) and from near Port

Elizabeth by Olivier (1979 — Coastal Fynbos inter-

mingled with scrub). Phillips (1931) gives a general

account of South Coast Dune Fynbos (Psammophi-

lous Macchia) from the Knysna region and Muir

(1929) describes some communities from the

Riversdale District. Possible outliers occur on the

west coast near Langebaan ( Metalasia — Myrica

Dune Dense Evergreen Ericoid Shrubland of

Boucher & Jarman, 1977) and east of Port Elizabeth

(Dyer, 1937 — Psammophilous Macchia; Martin &

Noel, 1960 — Maritime Heath).

Three South Coast Dune Fynbos Communities

occur in the study area (Table 6). The Themeda

—Stenotaphrum Community occurs on deep

seasonally-waterlogged sands in dune valleys (Fig.

22) and also on well-drained sands subjected to

frequent bush-cutting (Fig. 23). In the latter areas,

in the absence of bush-cutting, the community

develops from a closed grassland to small-leaved

shrubland and thicket; in the dune valleys the

establishment of shrubland is prevented by years of

an abnormally high watertable when fynbos initials

such as Passerina vulgaris, Metalasia muricata,

Anthospermum aethiopicum and Aspalathus spinosa

subsp. spinosa are killed (pers. obs.). Despite the

tropical affinity of the grass flora and some of the

forbs, the presence of many Dune Fynbos diagnostic

species (Table 6) and strong successional relation-

ships links the Themeda — Stenotaphrum to the

South Coast Dune Fynbos.

The Restio — Agathosma Community (Figs 24 &

37) is found on well-drained deep sands (Fernwood

Form) on dune ridges, where it is successional to

thicket, and on shallower well-drained sands

R. M. COWLING

203

Fig. 22. — Themeda -Stenotaph-

rum Dune Grassland (fore-

ground) on deep seasonally

waterlogged sands (Fern-

wood Form). Dune thicket

on well-drained deep sands

(dune ridge) in background.

Grassland species include

Themeda triandra, Restio

eleocharis, Imperata cylindri-

ca, Aspalathus spinosa, Ra-

panea gilliana and Tetraria

cuspidata. Cape St Francis

(c. 20 m elevation).

Fig. 23. — Dune Grassland, Fyn-

bos and Thicket complex on

deep well-drained sands.

Grassland maintained by fre-

quent bushcutting. Grassland

species include Stenotaphrum

secundatum, Themeda trian-

dra, Tristachya leucothrix,

Rhoicissus tridentata and

Thamnochortus glaber.

Brandewynkop near Oyster

Bay (c. 100 m elevation).

overlying calcrete (Mispah Form). The lack of

thicket development in the latter area is probably

due to the inability of thicket species to outcompete

shallow-rooted fynbos species where rooting depth is

limited to 50 cm or less, and that the shallow soils are

droughty. The Restio — Maytenus Community is a

Small-leaved Shrubland/Large-leaved Shrubland

(thicket) confined to deep well-drained sands on

dune ridges where it often replaces the Restio -

Agathosma Community in a successional sequence

to dune thicket. Fig. 25 shows a schematic soil and

vegetation catena on linear dunes near Cape St

Francis.

3.2 Cape Transitional Small-leaved Shrub lands

Within and adjacent to the Fynbos Biome are a

variety of non-succulent Small-leaved Shrublands

which, according to the concepts developed above,

are distinctly non-fynbos. Although some of these

shrublands have been the subject of a recent review

(Boucher & Moll, 1980), there are no published

phytosociological surveys and the formulation of

syntaxonomic concepts is therefore severely limited.

I tentatively define Cape Transitional Small-leaved

Shrublands as follows:

(i) Phytochorological spectra are dominated by

ecological and chorological transgressor species

linking the Cape region with adjacent phytochoria,

particularly the Karoo-Namib and Afromontane

regions. Cape endemics comprise about one third of

a given sample flora (Fig. 9).

(ii) Regional endemism is lower than for Cape

Fynbos Shrublands and many endemics are not

members of typical Cape genera (Fig. 9).

(iii) Structurally, the communities are Small-

leaved Shrublands but they lack most of the

‘heathland’ ( sensu Specht, 1979) features typical of

Cape Fynbos Shrublands. Restioid and proteoid

growth forms are almost entirely lacking and grasses

dominate in the field layer (Fig. 10). Deciduous

geophytes are prominent, if not in cover, then in

richness. A considerable proportion of the small-

leaved woody shrubs have fleshy (semi-succulent)

leaves (Fig. 10) and some show seasonal leaf

204 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

Fig. 24. — Restio — Agathosma

Dune Fynbos on well-

drained deep, yellow-brown

sands (Clovelly Form). Do-

minant ericoid shrub is Pas-

serina vulgaris. Other species

include Muraltia squarrosa,

Metalasia muricata, Agatho-

soma stenopetala , Restio ele-

ocharis, Stipagrostis zeyheri

and Ehrharta calycina.

Brandewynkop near Oyster

Bay (c. 80 m elevation).

dimorphism. Shrubs with large dorsiventral leaves

are occasional.

(iv) Ecologically, they are restricted to fine-

grained soils derived from Cretaceous mudstones

and conglomerates, Malmesbury and Cango phyl-

lites, Bokkeveld shales, Cape Granites and the

tillites and shales of the Karoo Supergroup (Table

3). The soils are generally more fertile than Cape

fynbos soils (Fig. 8). Cape Transitional Small-leaved

Shrublands are found in areas receiving at least 30%

winter rain where the annual average precipitation is

from 300-600 mm (Table 3).

Characteristic genera of Transitional Small-leaved

Shrublands are Relhania, Elytropappus, Aspalathus,

Euryops, Pteronia, Passerina, Eriocephalus, Selago,

Hermannia and Helichrysum.

The following Acocks’s (1953) veld types are

included here: Coastal Rhenosterbosveld (46),

Mountain Rhenosterbosveld (43) and parts of the

Karroid Merxmuellera Mountain Veld (60). The last

mentioned has strong links with Afromontane

grasslands of the north-eastern Cape (Acocks,

1953). Campbell’s (1983b) Renosterveld, Clan-

william Karroid Shrubland and much of his Grassy

Shrubland are Cape Transitional Small-leaved

Shrublands.

A hypothesis is outlined here to explain the origin

of Cape Transitional Small-leaved Shrublands (for

details see Cowling, 1983a). Implicit in this thesis is

the assumption that the junction between the

shale-derived soils of the coastal forelands and the

sandy, infertile soils of the TMG quartzites forms an

effective migration barrier under most climatic

conditions. I assume that under moister conditions

than present, there was limited movement of fynbos

taxa onto the shale-derived coastal foreland soils and

that under drier conditions fynbos on the mountains

was not greatly displaced by karroid shrublands. The

general tenet of this assumption is that this soil

fertility barrier served to isolate the mountains as a

series of ‘islands’ with respect to the adjacent

lowland flora. Extreme climatic differences between

the mountains and lowlands reinforces this effect.

Transitional Small-leaved Shrublands occupy

tension zones between adjacent phytochoria: not

only between the Cape and Karoo-Namib regions

but also between these and the Afromontane and

Tongaland-Pondoland regions (Fig. 26). They are

R. M. COWLING

205

Mountain Fynbos

Fig. 26. — Renosterveld on the south coast forelands as a tension zone vegetation. Renosterveld communities occupy a zone

between karoo and grassland/thicket. The location of this zone would change with climatic fluctuations, and disappear under

arid conditions when karoo reached the TMG/shale junction. Dashed lines define a sensitive zone vulnerable to

mismanagement and the establishment of ‘false’ shrublands.

often dominated by one or two generalist species

(DI species type of Noble & Slatyer, 1980) such as

Elytropappus rhinocerotis (renosterbos), Metalasia

muricata, Relhania spp. and Euryops spp. These

species are widely distributed, are able to grow

under a wide range of conditions and have ‘weedy’

characteristics (Levyns, 1926 & 1935a; Trollope,

1970 & 1980). They are mostly derived from Cape

taxa. Elytropappus is a Cape genus of six species:

five have fairly localized distributions in the

south-western Cape, whereas E. rhinocerotis is

widely distributed throughout the Cape Region, the

Great Escarpment of the south central Cape and

parts of the eastern Cape (Levyns, 1927). Similarly

Metalasia is a Cape genus of 33 species, most of

which have restricted distributions in the Cape

Region (Pillans, 1954). M. muricata is a wide-

ranging ‘weedy’ species common in certain Transi-

tional Small-leaved Shrublands.

The tension zone occupied by Transitional

Small-leaved Shrublands would have been unstable

during Late Cainozoic climatic fluctuations. During

the last glacial, when conditions were drier than at

present (Deacon, 1983), the expansion of karroid

shrublands on the coastal forelands would have

eliminated these transitional shrublands in places; in

other regions they would have expanded at the

expense of grassland and thicket/forest (Cowling,

1983a). The instability of this tension zone is likely

to have favoured the establishment of weedy species

recruited from adjacent phytochoria (Gibbs Russell

& Robinson, 1981).

In drier regions marginal to karroid shrublands,

Transitional Small-leaved Shrublands are regarded

as ‘natural’ communities (Acocks, 1953). Under

these conditions E. rhinocerotis and other shrubs

would occupy most niches, as species with more

mesic requirements were eliminated prior to the

invasion of karoo. In moister, bimodal rainfall

regions, where Transitional Small-leaved Shrub-

lands are marginal to grasslands of tropical affinity,

man-induced disturbances have telescoped these

events by creating environments which are effective-

ly more arid and by minimizing competition through

the selective utilization of species by domestic

livestock (Fig. 26). The recent spread of E.

rhinocerotis in South Coast Renosterveld is descri-

bed below.

The hypothesis presented above has several

testable predictions. Palynological studies such as

those being undertaken by Deacon and associates in

the southern Cape (e.g. Deacon, 1979) will

undoubtedly be of value in determining the

historical sequence of vegetation types. Demo-

graphic studies to determine the population strate-

gies of the major shrub species and other

autecological investigations may explain the compe-

titive ‘superiority’ of these species. Of interest, too,

is the possible role of allelopathy: Squires &

Trollope (1979) have clearly demonstrated the

allelopathic effects of Chrysocoma tenuifolia, the

principal invasive species of false karroid shrublands

and a component of drier Transitional Small-leaved

Shrublands.

However, most important at this stage are

detailed phytosociological studies to clarify the

biogeographical and ecological relationships of these

shrubland communities.

3.2.1 South Coast Renosterveld

Acocks (1953 & 1979) recognizes two coastal

renosterveld types: a west coast form and a south

coast form. This distinction has been upheld by

Taylor (1978) and Boucher & Moll (1980).

West Coast Renosterveld differs from the south

coast form in having a sparser grass cover composed

largely of C3 genera ( Ehrharta , Pentaschistis,

Merxmuellera , Lasiochloa), a higher diversity of

deciduous geophytes and annuals and the presence

of Relhania ericoides and Leyssera gnaphaloides as

206 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMAN5DORP REGION OF THE FYNBOS

BIOME

TABLE 7. — South Coast Renosterveld Communities

A: Themeda triandra-Cliffortia linearifolia Humansdorp False Fynbos

B: Elytropappus rhinocerotis-Metalasia muricata Humansdorp Coast Renosterveld

C: Elytropappus rhinocerotis-Eustachys paspaloides Hankey Coast Renosterveld

D: Elytropappus rhinocerotis— Relhania genistaefolia Humansdorp Coast Renosterveld

* Diagnostic species for South Coast Renosterveld including:

f South-eastern (Weimarck, 1941) and Kaffrarian Transition Zone (Cowling, 1983a) endemics largely restricted to South Coast

Renosterveld and related grasslands

characteristic subdominants with Elytropappus rhi-

nocerotis. West coast communities have a stronger

fynbos influence, especially on granitic soils.

Boucher (1983) has suggested that west coast

renosterveld is derived from Mountain Fynbos.

South Coast Renosterveld ranges from the

Caledon district to Humansdorp with outliers on the

Albany coastal forelands (Table 3). Unlike the west

coast form. South Coast Renosterveld receives a

considerable amount of summer rain. Typical

subdominants and diagnostic shrub species are

Relhania genistaefolia, R. cuneata, Helichrysum

anomalum, Indigofera denudata and Hermannia

flammea (Table 7). Tropical C4 grasses, which do

not reach the west coast, are particularly common

(Fig. 9, Table 7).

Acocks’s (1953) conjecture that the original

vegetation of the South Coast Renosterveld was a

scrub forest (thicket) (see Acocks, 1953; p. 8, Map

R. M. COWLING

207

Fig. 27. — Elytropappus —Metala-

sia Renosterveld on eroded

duplex soils (Glenrosa and

Swartland Forms). Thicket

along drainage line and on

termitaria in background.

Renosterveld species include

Elytropappus rhinocerotis,

Themeda triandra, Herman-

nia flammea and Metalasia

muricata. Thicket species are

Olea europaea, Tarchonan-

thus camphoratus, Rhus luci-

da and Pittosporum viridiflo-

rum. Above Seekoei River,

south of Humansdorp (c. 80

m elevation).

Fig. 28. — Elytropappus —Eustac-

hys Renosterveld (grassland

cover state) on shallow loa-

my sands on Enon conglo-

merate. Species include The-

meda triandra , Trachypogon

spicatus, Metalasia aurea,

Helichrysum anomalum,

Eustachys paspaloides and

Elytropappus rhinocerotis.

Thickets in background on

termitaria. Papiesfontein (c.

70 m elevation).

1, p. 86) is not corroborated by historical accounts

and the habitat requirements of present-day thicket

‘relics’. Judging from the earliest records it appears

that much of the South Coast Renosterveld was a

dense grassveld dominated by Themeda triandra.

The south coast forelands were once known as the

‘blougrasveld’, a reference to the blueish hue of the

once common Themeda (Smit, 1943). However, as

early as 1775, Sparrman (1785) noted the demise of

the grasslands and the spread of Elytropappus

rhinocerotis. He attributed this to the continuous

overgrazing of the grassveld by domestic livestock

(see also Hall, 1934; Levyns, 1935b; Du Toit & Du

Toit, 1938; Smit, 1943; Levyns, 1956).

A study of the present-day distribution of thicket

in the Coastal Renosterveld of the study area shows

that it is always restricted to deep, well-drained soils;

renosterveld and grassveld are generally confined to

shallow soils, often with poorly-developed clayey,

impermeable subsoils (Glenrosa and Swartland

Forms) (Figs 27-29). Only a few shallow-rooted

thicket species (Rhus glauca, R. lucida, R. incisa and

Euclea crispa) are occasionally associated with

renosterveld.

I suggest that South Coast Renosterveld is derived

from grassveld and that thicket was always

subordinate in the landscape where it was restricted

to special edaphic sites largely located along river

courses and on rock outcrops and termitaria of the

interfluves. The advent of settled agriculture shifted

the disturbance regime from pulse-disturbance

grazing and variable fire intervals to a fixed burning

cycle and continuous overgrazing (1.5). Continuous

immediate post-fire grazing, which was widely

practised, is particularly harmful to the grass sward

(Du Toit & Du Toit, 1938; Smit, 1943; Smith, 1967).

The increased predictability of the disturbance

regime led to a reduced resilience of the system (cf.

Walker et ah, 1981). As resilience is lowered, there

is usually a critical point where slight increases in

mortality can trigger population collapses, i.e.

populations are forced close to the boundary of their

‘domain of attraction’ and then ‘flip’ over it

(Holling, 1973). In this way the system shifted from

a grass-dominated state to a shrub-dominated one.

208 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

<y-rT rTTJ|-^f rrTi'TT'tl'T'

Termitarium

Clayey subsoil

GOOD MANAGEMENT

Duplex, seasonally waterlogged soils

(G lenrosa, Swart land ) ^crf>yyr\

sandy loam topsoil

1 'rrrrTT

Thicket

POOR MANAGEMENT

Eroded topsoil, capping, increased run-off

T-Ttt

Themeda grassland

Renosterveld

Bokkeveld shale

Well drained, deepish Hutton & Clovelly

Fig. 29. — Schematic soil and vegetation catena on the Bokkeveld shale flats, south of Humansdorp.

In the study area there are some fenceline

contrasts separating grassland and shrubland cover

states of the same community (Figs 30-32). Pure

grasslands are maintained by occasional heavy

grazing in a rotational system employing a large

number of small camps. This encourages a mat-like

growth of Themeda with a high basal cover and the

development of a Cynodon dactylon sward which

effectively binds the soil and reduces the area of bare

soil. Renosterveld communities on the Bokkeveld-

shale flats south of Humansdorp occur on duplex

soils which are seasonally waterlogged. Continuous

overgrazing of the grass sward causes soil capping

which results in increased run-off, and erosion, the

truncation of topsoils and the eventual disturbance

of water-table dynamics (Fig. 29). This process

favours the establishment of woody species (Tinley,

1977). Once established, E. rhinocerotis and other

species are difficult to eliminate because they have

deep roots (Scott & Van Breda, 1937) which are able

to utilize relatively reliable subsoil moisture. This

moisture is not normally available to shallow rooted

grasses in the droughty topsoils. Shrubs must be

removed by fire or mechanical means but are often

re-established within a few years (Levyns, 1935b;

Cowling, in prep.).

Communities belonging to the South Coast

Renosterveld have been described from the Swel-

lendam District by Grobler & Marais (1967 —

Renosterbos Community of the plains, Renoster-

bos —Themeda Community), from near Riversdale

by Taylor (1970b — Elytropappus Evergreen

Microphyll Shrub Steppe Savanna). General ac-

counts of the South Coast Renosterveld are given by

Muir (1929) for the Riversdale District and Jordaan

(1947) for the Caledon and Bredasdorp Districts.

The Acacia karroo -Themeda triandra Association

of Martin & Noel (1960) and the Acacia karroo

Savanna of Jessop & Jacot Guillarmod (1969) have

many diagnostic renosterveld species. These com-

munities occur on Bokkeveld shales on the Albany

forelands. Mismanagement often results in the

co-dominance of E. rhinocerotis and Acacia karroo

(pers. obs). I have sampled a renosterveld commu-

nity on Dwyka shales, north of Grahamstown

(unpublished data). Dominant species included

Elytropappus rhinocerotis, Felicia filifolia, Aspala-

thus lactea subsp. adelphea, Relhania genistaefolia

and Themeda triandra (Table 7).

I have described four South Coast Renosterveld

communities from the study area (Table 7). These

communities are difficult to characterize since

variation in type and intensity of recurrent

disturbances is reflected in variations in species

composition (Cowling, in prep.).

The Themeda — Cliffortia Community is confined

to the wetter (550-650 mm yr'1) parts of the

Bokkeveld shale flats, along the Kromme River.

The community consists of two structural cover

states: a Closed Grassland dominated by Themeda

and a derived Small-leaved (Grassy) Shrubland

dominated by Cliffortia linearifolia (Figs 30 & 31).

Soils are shallow, duplex types (Glenrosa and

Swartland Forms) which are waterlogged in winter.

Destruction of the grass sward and subsequent

topsoil erosion has adversely affected the water-

table dynamics (see above).

The strong Afromontane affinities of this commu-

nity (e.g. Helichrysum odoratissimum, Cliffortia

linearifolia, Festuca spp.. Gladiolus longicollis) link

it to Afromontane fynbos and grassland [see Story’s

(1952) Cliffortia linearifolia lowland macchia and

Killick’s (1963) Cave Sandstone Scrub]. Species of

Cape affinity include Leucadendron salignum, Erica

unilateralis, Tetraria cuspidata and Restio sieberi.

This community best fits the concept of false or

derived fynbos as formulated by Acocks (1953).

Structurally these communities would not be

classified as fynbos by Campbell (1983b). Because of

the presence of diagnostic renosterveld species in the

Themeda — Cliffortia Community (Table 7), I have

included it in South Coast Renosterveld.

I have observed allied communities with C.

linearifolia as a dominant on the Coastal Platform

between George and Groot Brak, in the Zuurberg

and the Amatole Mountains (cf. Story, 1952). In the

eastern Cape these communities do not indicate a

massive eastwards movement of fynbos elements in

historical times (cf. Acocks, 1953; Trollope, 1970,

1973) but rather a localized thickening up of species

R. M. COWLING

209

Fig. 30. — Overgrazed Cliffortia

—Themeda False Fynbos on

shallow sandy loam overlying

a poorly developed clayey

subsoil (Glenrosa Form) on

Bokkeveld shale. Beyond

fenceline in background is

high watertable Themeda

grassland. Dominant species

in foreground: Cliffortia line-

arifolia, Themeda triandra,

Sporobolus africanus, Gnap-

halium repens, Rhus incisa

and Restio sieberi. Osbosch

near Kromme River mouth

(c. 30 m elevation).

Fig. 31. — Grassland and shrub-

land cover states on the

Themeda — Cliffortia False

Fynbos Community. Grass-

land is pure and exceptional-

ly dense. Species include

Themeda triandra (domi-

nant), Helictotrichon hirtu-

lum, Pentaschistis augustifo-

lia, Sporobolus africanus,

Festuca caprina, Setaria fla-

bellata and Restio sieberi.

Grassmere, between Hu-

mansdorp and Cape St Fran-

cis (c. 50 m elevation).

present in the pre-disturbance grasslands, albeit in

restricted habitats.

The Elytropappus — Metalasia Community (Figs

27, 32 & 33) occurs in the drier regions (450—550

mm yr'1) of the Bokkeveld shale flats on shallow

duplex soils. Grassland and shrubland cover states

have been identified. The Elytropappus — Relhania

Community is restricted to the driest regions

(400-450 mm yr1) of the Coastal Platform above

the Gamtoos River Valley. The Hutton Form soils

are mostly deep (0,8— 1,0 m).

The Elytropappus —Eustachys Community (Figs

28 & 34) is exceptionally species rich and

chorologically complex. It is restricted to Enon

conglomerates on the coastal forelands. Soils are

stony well-drained loamy sands and are slightly less

fertile than renosterveld soils derived from Bokke-

veld shale (Fig. 8). Many releves from this

community are transitional to the Themeda -Passe-

rina Grassy Fynbos Community (3.1.2).

3.3 Subtropical Transitional Thicket

Closed Large-leaved Shrublands ( sensu Campbell

et al., 1981) of essentially tropical and subtropical

affinity penetrate into the Fynbos Biome from the

east and extend along the west coast as far as

Lamberts Bay. Physiognomically these shrublands

consist of an impenetrable tangle of shrubs and low

trees usually interwoven by woody climbers.

Communities of similar structure and generic

composition are found throughout tropical and

subtropical Africa (Okali et al., 1973; Tinley, 1975;

White, 1982) and are termed thicket. Tinley (in

Heydorn & Tinley, 1980) extended the thicket

concept to the Fynbos Biome. Boucher & Moll

(1980) call Fynbos Biome thicket communities

mediterranean shrublands, a clearly erroneous

interpretation, since affinities are strongly tropical

and there are virtually no thicket species endemic to

the Cape mediterranean-climate region (Cowling,

1983a).

210 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

Fig. 32. — Fenceline contrast

showing shrubland and

grassland cover states of the

Elytropappus —Metalasia

Renosterveld Community.

Duplex soils on Bokkeveld

shale (Glenrosa to Swartland

Forms). Note thicket deve-

lopment (Rhus glauca) along

fenceline which affords a

perch for frugivorous birds.

Near Aloe Ridge at Aston

Bay (c. 20 m elevation).

i 4%

Fig. 33.— Degraded Themeda-

dominated grassland being

invaded by Elytropappus rhi-

nocerotis (Elytropappus

—Metalasia Community).

Bokkeveld shale flats be-

tween the Kromme and See-

koei Rivers (c. 30 m eleva-

tion).

Fig. 34. — Elytropappus -Eustac-

hys Renosterveld (shrubland

cover state) on Enon conglo-

merate. Dominant species

include Elytropappus rhino-

cerotis, Passerina rubra , As-

palathus chortophila, The-

meda triandra, Eustachys

paspaloides and Pentaschistis

angustifola. Papiesfontein (c.

40 m elevation).

R. M. COWLING

211

Fig. 35. — Phytochorological spec-

tra and endemism for Sub-

tropical Transitional Thicket

and Afromontane Forest. A

= Afromontane endemics;

TP = Tongaland-Pondoland

endemics; KN = Karoo-

Namib endemics; A-TP =

Afromontane-Tongaland-

Pondoland linking species;

TP-Z = Tongaland-

Pondoland-Zambezian link-

ing species; TP-KN =

Tongaland-Pondoland-

Karoo-Namib linking spe-

cies; TW = Tropical-

subtropical wides; W = Wi-

dest RES = Residual (groups

containing less than 5% total

species). Geographical affi-

nity of endemics as in Fig. 9.

KAFFRARIAN SUCCULENT THICKET KAFFRARIAN THICKET

Sideroxylon -Euphorbia community Pterocelastrus -Euclea

AFROMONTANE FOREST

Rapanea-Ocotea

HZ! 0111] Q LZH [Z] " _!

KN A TP TP-2 TP-KN W TW RES

Geographical affinity

Subtropical Transitional Thicket occurs roughly

from the Kei River to the south-western Cape

(Table 3). It is defined as follows:

(i) Phytochorological spectra are dominated by

ecological and chorological transgressor species

linking the Tongaland-Pondoland Region with

Afromontane, Karoo-Namib and to a lesser extent,

Cape regions. Tongaland-Pondoland endemics

comprise about one-quarter of a given sample flora

(Fig. 35).

(ii) Regional endemics are few and usually of

non-tropical affinity. Karroid shrubs, particularly

succulents ( Euphorbia , Crassula, Delosperma, Aloe)

comprise most of the endemics (Fig. 35).

(iii) Structurally, the communities are dominated

by large-leaved evergreen sclerophyllous shrubs

many of which are stem-spinescent. Succulents are

conspicuous in dry areas and vines are generally

common (Fig. 36).

(iv) Ecologically, Subtropical Transitional

Thicket is restricted to deepish, well-drained fertile

soils (Fig. 8, Table 3). Transitional Thicket occurs in

areas which receive at least some winter rain.

Subtropical Transitional Thicket includes thicket

in Acocks’s (1953) Southern Form of the Eastern

Province Thornveld (7b), False Thornveld Eastern

Cape (21), Southern Variation of the Valley

Bushveld (23b), Fish River Scrub (23c), Addo Bush

(23d(i)), Sundays River Scrub (23d(ii)), Gouritz

River Scrub (23e), Noorsveld (24), as well as thicket

in Dense Strandveld Scrub (34a), Coastal Rhenos-

terbosveld (46) and Coastal Macchia (47). Also

included are certain scree and rock-outcrop commu-

nities in Mountain Fynbos (69 and 70).

KAFFRARIAN SUCCULENT

THICKET

Sideroxylon -Euphorbia community

n-15(no. of relev6s)

KAFFRARIAN THICKET

Pterocelastrus-Euclea

n = 18

AFROMONTANE FOREST

Rapanea-Ocotea

n=4

100

90.

i

D

a

K

40.

30.

20.

10.

Stemspine.- 27,6

Leaf spine: 4,6

Vines : 20,9

DDD

U

□

Large leaves

dorsiventral

Small leaves

Succulents

Herbs

sclerophyll orthophyll

fleshy

Leaf texture(% woody cover!

Fig. 36. — Growth form composition of Subtropical Transitional Thicket and Afromontane Forest.

212 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

Transitional Thicket has few character taxa. A list

of trees and tall shrubs includes Euclea racemosa

subsp. racemosa, E. polyandra, E. tomentosa,

Maytenus lucida, M. oleoides, Cassine maritima, C.

reticulata, Rhus longispina, R. crenata, R. schlechte-

ri, Cussonia gamtoosensis; C. thyrsiflora, Olea

exasperata, Rapanea gilliana, Heeria argentea,

Maurocenia frangularia, Brachylaena neriifolia,

Aloe africana and Euphorbia tetragona. Many of

these are derived from species more widely

distributed in the tropics and subtropics (e.g.

Robson, 1966 for Maytenus; F. White, pers. comm,

for Euclea). The drier succulent form of the

Transitional Thicket has a number of endemic

succulent shrubs of karroid affinity, belonging to the

genera Delosperma, Senecio, Euphorbia, Crassula,

Zygophyllum and Lampranthus.

The general trend within the Transitional Thicket

flora is a rapid depauperization westwards. Of the

tropical and sub-tropical shrub and tree species

reaching the Kei River, 49,5% do not extend beyond

the Kaffrarian Transition Zone (2.2.4) (Cowling,

1983b; see also Gibbs Russell & Robinson, 1981). In

the study area, which is on the western boundary of

the transition zone, there are 97 non-succulent

thicket shrubs and trees; on the south and west coast

forelands there are 43 and 24 species respectively

(data from Muir, 1929; Acocks, 1953, 1979; Boucher

& Jarman, 1977; pers. data).

Structurally, Transitional Thicket is readily distin-

guished from other African thicket types. There is a

predominance of evergreen sclerophyllous leaves

(Fig. 36) and a high cover of succulent shrubs of

karroid affinity. Thicket communities to the north

often have a strong component of orthophyllous

deciduous species (Wild, 1952; Edwards, 1967;

Fanshawe, 1968; Tinley, 1977). Winter decidu-

ousness is a predictable strategy of tropical

summer-rainfall climates (cf. Orians & Solbrig,

1977). However in the south-eastern Cape where

rainfall distribution is highly erratic (Gibbs Russell

& Robinson, 1981) and in the southern and

south-western Cape where a high proportion of the

rain falls in winter, plants must be capable of

utilizing soil moisture whenever its availability

coincides with other optimum growth conditions.

Clearly long-lived, sclerophyllous ‘high-cost— slow-

profit’ leaves (Orians & Solbrig, 1977) will be

favoured (Cowling & Campbell, 1983a).

The distribution of thicket communities is

determined by a complex of interrelated factors.

Fire is cited as a factor limiting the distribution of

thicket in the eastern Cape (Du Toit, 1972;

Trollope, 1974). Thicket is often restricted to what

are usually interpreted as fire-protected sites (screes,

rock outcrops, ravines, termitaria and river valleys).

Trollope (1974) emphasizes the historic role of fire

and browsing ungulates in restricting thicket

development.

Tinley (1977; Heydorn & Tinley, 1980) stressed

edaphic controls on the distribution of thicket.

Thicket development is often restricted to deep well

drained soils. In a study of thicket clump formation

on the Accra Plains, Ghana, Okali et al., (1973)

conclude that thicket is restricted to deep, uniform,

fine-grained soils; surrounding grassland occurs on a

shallow duplex soil (sandy topsoil) which is

seasonally waterlogged. In my study area, the

densest thicket occurs on deep apedal sandy loams

to sandy clay loams (Hutton and Clovelly Forms)

derived from Enon shales and mudstones. The soils

are well drained and probably capable of retaining

much moisture at the rooting depth of the thicket

species. Thicket is also found on dune sands,

termitaria, stable talus and scree slopes, rock

outcrops and river valleys. All these sites are

characterized by deep, well drained soils. These

profiles probably offer the rooting depth and

water-retaining capacity necessary for thicket for-

mation (cf. Okali et al., 1973; Tinley, 1977).

It has been shown that the soils of thicket clumps,

especially when they are associated with termitaria,

are more fertile than the soils of surrounding

vegetation (Watson, 1967; Fanshawe, 1968; Okali et

al., 1973; Fig. 8). The association of thicket and

termite mounds is a constant feature throughout

sub-Saharan Africa (e.g. Wild, 1952; Fanshawe,

1968). It has been suggested that termitaria favour

thicket formation by providing fire protection for

woody plants and by improving soil nutrient,

drainage and water storage conditions (Lee &

Wood, 1971; Trapnell et al. , 1976; Tinley, 1977). On

non-termitarium sites, it is possible that the higher

fertility status of the soils is due to plant-induced soil

changes resulting from increased litter accumulation

and decomposition (cf. Paulsen, 1953; Charley &

West, 1976; Aweto, 1981). Correlative studies of soil

variables show strongly significant positive relation-

ships between organic carbon and major nutrients in

thicket soils suggesting plant-induced enrichment

(Cowling, in prep.). There are strong indications

that increased organic matter is a prerequisite for

thicket formation.

Most thicket species have fleshy fruits which are

mainly bird dispersed (Bews, 1917; Tinley, 1977;

pers. data). Avian dispersal undoubtedly plays an

important role in the establishment of thicket

particularly in the vicinity of perches (rock outcrops,

termitaria, established thicket) (Bews, 1917; Tinley,

1977). Many of these perch sites also fulfil the

edaphic requirement for thicket initiation.

Fire, edaphic factors and dispersal agents all

contribute to the dynamics of thicket formation. Fire

protection is of critical importance in the early stages

of thicket initiation but becomes less important as

the thicket matures. The roles of soil depth,

drainage and organic matter, and dispersal agents

have not been adequately stressed as factors

contributing to the development of Transitional

Thicket. A more fruitful approach to the study of the

structure and dynamics of Fynbos Biome thicket

communities is to look to the work on subtropical

and tropical African thicket and not to studies of

mediterranean shrublands on other continents.

Post-disturbance dynamics of Transitional Thicket

communities in the study area were not studied.

Thicket communities are stable and have low

resilience. They regenerate slowly after occasional

R. M. COWLING

213

TABLE 8. — Kaffrarian Thicket Communities

* Diagnostic species for Kaffrarian Thicket including:

f Kaffrarian Transition Zone endemics (Cowling, 1983a);

• Southern and south-western Cape endemics

catastrophic disturbances such as fire or clearing,

and are eliminated if these disturbances occur at

intervals of less than a few decades. The thicket

communities are not fire prone but are vulnerable to

overstocking of domestic livestock, particularly

goats (Aucamp & Barnard, 1980).

3.3.1 Kaffrarian Thicket

Kaffrarian Thicket consists of the non-succulent

Subtropical Transitional Thicket communities. They

have their maximum expression in the Kaffrarian

Transition Zone and depauperate outliers extend to

the south-western Cape.

Kaffrarian Thicket has strong affinities with the

Afromontane Forest flora although Tongaland-

Pondoland endemics and linking species dominate

the phytochorological spectrum; endemism is low

(Fig. 35). Structurally, the thicket is a closed

shrubland to low forest dominated by evergreen,

sclerophyllous trees and shrubs with a high cover of

stemspines and vines (Fig. 36). Some diagnostic

species are listed in Table 8.

Included in Kaffrarian Thicket are the thickets in

Acocks’s (1953), Eastern Province Thornveld (7b),

Alexandria Forest (2), False Thornveld of Eastern

Cape (21), Coastal Rhenosterbosveld (46) and

Coastal Macchia (47). Available data do not indicate

how far westwards the Kaffrarian Thicket extends. It

is possible that the scree thickets in Mountain

Fynbos, with character species such as Maytenus

oleoides, Heeria argentea, Euclea polyandra and

Maurocenia frangularia should be placed in a

separate order within the Subtropical Transitional

Thicket.

Kaffrarian Thicket occurs on a wide range of

parent materials, wherever conditions are suitable.

At a later stage, it may be possible to subdivide it

into alliances representing dune thicket, clay thicket

and mountain thicket. Kaffrarian Thicket on dunes

ranges from the Algoa Bay coast to the Cape

Peninsula and is characterized by a number of

species largely restricted to deep, calcareous, coastal

dune sands (e.g. Olea exasperata, Euclea racemosa

subsp. racemosa, Rhus crenata, Maytenus procum-

214 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

bens and Cassine maritima). It must be distinguished

from the more tropical Mimusops caff ra— Br achy -

laena discolor Dune Thicket of the Natal and

Transkei coasts (Acocks, 1953; Moll & White,

1978). Dune thicket communities are described from

the Algoa Bay coast by Taylor & Morris (1981 —

Olea exasperata Bush, Pterocelastrus tricuspidatus

Bushclumps, Dune Woodland), from Goukamma by

Van der Merwe (1976 — Sideroxylon inerme

Community), from Robberg by Taylor (1970a —

Mesophyllous Evergreen Broad Sclerophyll Mixed

Coastal Scrub), from De Hoop Nature Reserve by

Van der Merwe (1977 — Sideroxylon inerme Forest

and shrubland community), from near Stanford by

Taylor [1961 — Calvaria (= Sideroxylon) — Euclea

Short Forest], from Betty’s Bay by Boucher (1978 —

Colpoon — Rhus Scrub, Sideroxylon Scrub), from

the Cape Flats by Taylor (1972 — Pterocelastrus

Dune Scrub), and from the Cape of Good Hope

Nature Reserve by Taylor (1969 — Sideroxylon

Scrub Association). Taylor (1978) gives a general

account of coast scrub (dune thicket) for the Cape

region and Phillips (1931) and Muir (1929)

respectively describe dune thickets from Knysna and

Riversdale Districts.

Kaffrarian Thicket communities on clayey sub-

strates are described from the Keiskammahoek

District by Story (1952 — Acacia Scrub, Fort Cox

Scrub, Nqhumeya Scrub, Zanyokwe Bush), from

the Albany and I3athurst Districts by Martin & Noel

(1960 — warm temperate forest scrub. Acacia

karroo - Cussonia spicata Alliance, particularly

bush-clump savanna type), from near Grahamstown

by Jessop & Jacot Guillarmod (1969 — forest and

thicket areas D, F, G and H), and from the Addo

National Park by Archibald (1955 — Bontveld).

Muir (1929) gives a general account of clay thicket in

the Riversdale District and Taylor (1970b) describes

a single community near Riversdale.

The mountain thicket form of Kaffrarian Thicket

is restricted to arenaceous substrates of the TMG

and has strong Afromontane links. Phillips (1931)

and Von Breitenbach (1974) describe mountain and

clay thickets from the Knysna Forest enclave.

Four Kaffrarian Thicket communities are distin-

guished in the study area (Table 8). The Cassine -

Cussonia (Figs 23 & 37) and Cassine - Schotia

Communities are both dune thickets. The former is

synonymous with Van der Merwe’s (1976) Sider-

oxylon inerme Forest Community and occurs on

deep humic sands (Fernwood Form) of the coastal

dunes. It is closely related to the dune thicket of the

southern and south-western Cape, but has a richer

component of subtropical shrubs. In the study area

the Cassine — Cussonia Community is confined to

the wetter parts of the coast west of the Kromme

River.

The Cassine —Schotia Community is a drier dune

thicket with a higher proportion of succulents, spiny

shrubs and vines. It also has a number of Kaffrarian

Succulent Thicket diagnostic species indicating

strong links with this type. It resembles the Coastal

Bush described by Archibald (1955) at Addo, and

the Sundays River Scrub described by Taylor &

Morris (1981) on the Algoa Bay coast. The Cassine

— Schotia Community occurs east of the Kromme

River on deepish dune sands.

The Pterocelastrus — Euclea Community occurs

on well drained clayey soils in the valleys incised into

the coastal plain and also on termitaria on the

interfluves (Figs 28 & 38). Another block occurs in

the foothills of the Elandsberg above the drier

succulent thicket (Fig. 39). The Pterocelastrus —

Gonioma Community (Fig. 40) is restricted to TMG

sandstone and quartzites and occurs on stable scree

and talus slopes, slope breccias and other rocky

colluvial sites where soils are deep and stony. Under

identical rainfall regimes, this community is more

mesic and has stronger Afromontane links than the

Pterocelastrus — Euclea Community. This is

probably due to better infiltration afforded by the

sandier soils of the former community.

Fig. 37. — Cassine -Cussonia Du-

ne Thicket (right) on deep

well-drained sand (Fernwood

Form) and Restio-Agathos-

ma Dune Fynbos (left) on

shallow sand overlying cal-

crete (Mispah Form). Cape

St Francis (c. 20 m eleva-

1 - tion).

R. M. COWLING

215

Fig. 38 — Pterocelastrus —Euclea

Thicket in the Kromme Ri-

ver Valley on well-drained

Hutton soils (Bokkeveld

shale). Dominant species are

Pterocelastrus tricuspidatus,

Cassine aethiopica, Euclea

undulata, Aloe pluridens and

Sideroxylon inerme. ( c . 30 m

elevation).

Fig. 39. — Kaffrarian Thicket,

Euclea —Pterocelastrus Com-

munity, on deep, reddish

loams (Hutton Form) deri-

ved from Cango limestone,

Elandsberg foothills (c. 300

m elevation). Dominant spe-

cies Polygala myrtifolia,

Euclea undulata, Olea euro-

paea subsp. africana, Ochna

serrulata, Schotia latifolia.

Kleinfontein limestone quar-

ry near Loerie.

Fig. 40. — Pterocelastrus— Goni-

oma Knysna Thicket on

TMG Sandstone on colluvial

talus and breccia soils.

Krommellenboog between

Cape St Francis and Oyster

Bay (c. 60 m elevation).

216 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

TAB LI 9. — Kaffrarian Succulent Thicket Communities

* Diagnostic species for Kaffrarian Succulent Thicket including:

t Species endemic to the Kaffrarian Transition Zone (Cowling, 1983a)

3.3.2 Kaffrarian Succulent Thicket

Kaffrarian Succulent Thicket consists of the

Transitional Thicket communities where succulents

contribute 20-30% relative cover (Fig. 36). They

occur in the hot, dry river valleys from the Kei to the

Gouritz (Table 3). Like the Kaffrarian Thicket they

have their maximum expression in the Kaffrarian

Transition Zone, particularly the Fish and Sundays

River valleys. Succulents are well adapted to grow in

the valley climates (1.3) where soil moisture is

limiting for extended periods (Cowling & Campbell,

1983a).

Diagnostic species include a number of regional

and local character species, many of which are

succulent shrubs of karroid affinity (Table 9).

Phytochorological affinities are complex but

Tongaland-Pondoland endemics and linking spe-

cies predominate; a Karoo-Namib influence is strong

(Fig. 35). Endemism is more pronounced than in

Kaffrarian Thicket owing to a strong representation

of regional endemics of karroid affinity (Fig. 35).

The Kaffrarian Transition Zone is a centre of

endemism of succulent Euphorbia species (Croizat,

1965), where 14,5% of southern African species are

confined (data from White et al., 1941), most of

which are found in Kaffrarian Succulent Thicket.

Aloe, Crassula and Delosperma have numerous

species in these communities and many are endemic.

There are very few subtropical shrub and tree

endemics (Fig. 35).

A great diversity of growth forms are found in

Kaffrarian Succulent Thicket. These include leaf and

stem succulent shrubs, trees and vines, arborescent

rosette succulents, succulent herbs, large- and

small-leaved sclerophyllous and orthophyllous

shrubs, low trees and vines, grasses, forbs, annuals

and geophytes (Fig. 41). Bews (1925) believes that

the high incidence of succulents, spines and vines

(Fig. 36) in eastern Cape succulent thicket reflects

an extreme specialization of the subtropical flora.

My concept of Kaffrarian Succulent Thicket

encompasses much of Acocks’s (1953) Valley

Bushveld west of the Kei River [Southern Variation

of the Valley Bushveld (23b), Fish River Scrub

(23c), Addo Bush, Sundays River Scrub (23d) and

Gouritz River Scrub (23e)]. It may be possible to

incorporate in this concept the thicket of the

Noorsveld (24) and Spekboomveld (25). The latter

two veld types have a lower and more open

R. M. COWLING

217

Fig. 41. — Kaffrarian Succulent

Thicket near the mouth of

the Kabeljous River (c. 10 m

elevation). Dominant species

include arborescent succu-

lents: Euphorbia triangularis,

E. grandidens, Aloe africana;

sclerophyll shrubs: Euclea

undulata, Sideroxylon iner-

me, Azima tetracantha, Rhus

longispina; and vines:

Rhoicissus digitata, Capparis

sepiaria and Euphorbia

mauritanica.

physiognomy, a higher cover of succulents and

stronger Karoo-Namib links than Kaffrarian Succu-

lent Thicket. They could be possibly grouped, along

with the thicket of the Little Karoo (26b) as karroid

succulent thicket, a separate order within the

Subtropical Transitional Thicket.

There are no published detailed phytosociological

surveys of Kaffrarian Succulent Thicket. Archibald

(1955), in a survey of the Addo Park, recognized

‘spekboomveld’ which covered 90% of the park.

Martin & Noel (1960) describe the Euphorbia

grandidens — E. triangularis Alliance and the

Euphorbia bothae —Portulacaria afra Alliance of the

Succulent Woodland Formation in the river valleys

of the Albany and Bathurst Districts. Checklists for

the Addo Park and Springs Nature Reserve in the

Sundays River basin have been made by Penzhorn &

Olivier (1974) and Olivier (1981) respectively.

I have recognized two Kaffrarian Succulent

Thicket communities from the Gamtoos Valley in

the study area (Table 9). I call both communities

Sundays River Thicket since they probably belong to

a single alliance centred in the Sundays River basin.

The Euclea — Brachylaena Community occurs on

Hutton soils derived from Bokkeveld shale on the

southern margin of the Gamtoos valley. It merges

with the Pterocelastrus — Euclea Community in the

Kabeljous valley. Rainfall is between 400 and 450

mm yr'1.

The Sideroxylon — Euphorbia Community is an

exceptionally dense, impenetrable thicket on deep

soils (Hutton and Clovelly Forms) derived from

Enon shales and mudstones. Relics are found on

deep alluvial, bottomland soils that have not been

cleared for agriculture. The Schotia — Crassula

Sub-community (Fig. 42) is typical of the drier,

north-facing sites whereas the Capparis - Crassula

Sub-community occurs in moister areas. Rainfall is

between 450 and 500 mm yr"1.

3.4 Afromontane Forest

Afromontane forests have been the subject of an

excellent review by White (1978) where the

Fig. 42. — Sideroxylon —Euphorbia

Succulent Thicket on deep,

sandy clay loams (Clovelly

Form). Succulents include

Aloe ferox, Euphorbia gran-

didens, Senecio pyramidatus

and Portulacaria afra. Other

species are Carissa haemato-

carpa, Euclea undulata, Cus-

sonia gamtoosensis and Si-

deroxylon inerme. Near

Hankey, on road to Loerie

(c. 150 m elevation).

218 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

TABLE 10. — Knysna Afromontane Forest Communities

* Diagnostic Afromontane I'orest species (after White, 1978)

composition, affinities and structure of the forests

are explicitly defined. They occur in an archipelago

of mountain ‘islands’ distributed on the African

continent from Somalia, along the east African

uplands to the Cape Peninsula, and westwards to

Sierra Leone (Fig. 1 in White, 1978). In southern

Africa, where increasing latitude compensates for

altitude, Afromontane elements are found at low

altitudes, especially west of the Kei River (Tinley,

1967; White, 1978; Moll & White, 1978). The

Alexandria high forests, where Podocarpus falcatus

(an Afromontane endemic) and Erythrina caffra (a

Tongaland-Pondoland endemic) are canopy codo-

minants, clearly demonstrate the transitional nature

of eastern Cape coastal forests. The Knysna forests

are typically Afromontane (White, 1978).

Afromontane Forests are fairly well circumscribed

with numerous endemic species but few local or

regional endemics, and a low degree of familial and

generic endemism (White, 1978; 1981; Fig. 35).

A list of diagnostic Afromontane Forest trees found

in southern African forests is shown in White (1978,

pp. 472-473). Some of these are shown in Table 10.

Afromontane Forests are dominated by species of

Afromontane affinity (Fig. 35). In South Africa,

there is often a strong component of Afromontane-

Tongaland-Pondoland linking species (e.g. Apodytes

dimidiata, Teclea natalensis, Calodendron capense,

Cassine tetragona, Canthium inerme, Diospyros

villosa, Cussonia spicata, Ochna arborea, Trimeria

grandiflora and Hippobromus pauciflorus ) (see also

Moll & White, 1978). These species indicate a strong

Afromontane influence in the Tongaland-Pondoland

Region, since north of Natal they are mainly

confined to Afromontane forests (cf. Tinley, 1967).

Typical Afromontane Forest in South Africa

includes forest vegetation in the following Acocks’s

(1953) veld types (White, 1978): Knysna Forest (4),

North-Eastern Mountain Sourveld (8), Highland

and Dohne Sourveld (44a and b), Natal Mist Belt

Ngongoni Veld (45) and forest patches in the

Mountain Fynbos (69 and 70). Transitional

Afromontane-Tongaland-Pondoland Forests include

Alexandria Forest (2), Pondoland Coastal Plateau

Sourveld (3) and Ngongoni Veld (5).

Afromontane Forests comprise the only true

forest vegetation in the Fynbos Biome. They are

restricted to the southern coastal forelands and

mountains, on sites where soil moisture is available

throughout the year (White, 1978; McKenzie, 1978;

Fig. 43). Compared to allied forests of Transkei,

Natal and Transvaal, the Cape Afromontane Forests

are depauperate (Von Breitenbach, 1974; White,

1978).

Afromontane Forest communities have been

studied in the Grahamstown District by Dyer (1937)

and Martin (1965), in the Knysna enclave by Phillips

(1931) and Von Breitenbach (1974), in the

Riversdale District by Muir (1929), at Grootva-

dersbosch by Taylor (1955), and at Cape Hangklip

by Boucher (1978). McKenzie (1978) undertook a

detailed phytosociological study of Afromontane

Forest communities between George and the Cape

Pensinsula. Pensinsula forest communities have

been described by Campbell & Moll (1977) and

McKenzie et al. (1977).

3.4.1 Knysna Afromontane Forest

The Afromontane Forest vegetation in South

Africa is very uniform and probably consists of only

a few orders within the hierarchy proposed in this

R. M. COWLING

219

Fig. 43. — Afromontane Forest on

the Elandsberg on a convex

slope (colluvium) at the line

of contact between TMG

quartzite and Cango phyllite

and quartzite. Road between

Otterford and Kleinfontein

(c. 400 m elevation).

Fig. 44. — Afromontane Forest,

Rapanea —Ocotea Communi-

ty, on deep colluvial soil

(Oakleaf Form) in the

Elandsberg. Dominant spe-

cies: Ocotea bullata, Curtisia

dentata, Alsophila capensis,

Trichocladus crinitus, Rapa-

nea melanophloeos. Loerie

‘Stinkwood’ Forest (c. 500 m

elevation).

paper. White (1978) recognized five local systems,

three of which occur in the Cape Province. These are

the Cape Province east of Knysna, the Knysna

forests, and the forests west of Knysna.

The first system is centred in the southernmost

spur of the Drakensberg (Winterberg and Amatole

Ranges) and has been studied in the Keiskamma-

hoek District by Story (1952). Outliers occur along

the Zuurberg Range (Martin & Noel, 1960; Martin,

1965). These forests have a number of species not

found further westwards (see White, 1978 p. 503)

and are characterized by a high cover of Canthium

ciliatum in the understorey (cf. Story, 1952; Acocks,

1953; pers. obs.).

Forests in the Knysna region have been studied

extensively in the past (Phillips, 1931; Von

Breitenbach, 1974). I include them with the forests

of the south-western Cape as a separate order

(Knysna Afromontane Forest). Knysna forests

usually have a great deal of Gonioma kamasii in the

understorey, whereas in south-western Cape forests,

Hartogiella ( =Hartogia) schinoides is a dominant

understorey species (see Campbell & Moll, 1977;

McKenzie et al., 1977; Boucher, 1978; McKenzie,

1978).

I recognized two Knysna Afromontane Forest

Communities in the study area: Rapanea -

Canthium Community and the Rapanea - Ocotea

Community (Table 10). Both occur in colluvial

basins or valley fills in the Elandsberg mountains

(Figs 43 & 44). The former is found in isolated

pockets at the junction of the TMG and Cango

quartzite where the rainfall is between 700 and 900

mm yr1. The latter is restricted to a small basin

above Loerie Forest Station which receives a rainfall

in excess of 1 000 mm yr"1. Soils are deep with loamy

topsoils which grade into sandy clay loam subsoils

(Oakleaf Form). They receive runoff throughout the

year and have a good water retaining capacity.

4 DISCUSSION

4.1 Soil nutrients and shrubland types

Specht & Moll (1983) include the Humansdorp

region in the mediterranean climate zone of South

220 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

TABFF 1 1. —Correspondence of vegetation types (shrublands only) in the Humansdorp

region to base-rich, nutrient-poor and calcium-rich soil suites defined by Specht

(pers. comm.; Specht & Moll, 1983)

1 T.M.G. sandstone and quartzites

2 Bokkeveld shale

3 Enon deposits

4 Recent sands

Africa (see also Di Castri, 1980; 1.3). The dominant

vegetation classes in the study area (Cape Fynbos

Shrublands and Cape Transitional Small-leaved

Shrublands) are typical of the south-western Cape

which does enjoy a true mediterranean climate. The

shrublands of the Humansdorp study area are

discussed here in terms of current concepts of

mediterranean shrublands (Specht, 1979; Di Castri,

1980; Specht & Moll, 1983).

Specht & Moll (1983) make a fundamental

distinction between South African mediterranean

shrublands (with open-scrub overstorey of evergreen

sclerophyllous shrubs over an understorey of

seasonal grasses and herbs) on base-rich soils, and

heathlands (with evergreen sclerophyllous strata of

Proteaceae, Ericaceae and Restionaceae) on

nutrient-poor soils. A third category of shrublands is

confined to calcium-rich soils where a stunted

overstorey of evergreen sclerophyllous trees or

shrubs occurs over a ground stratum showing a

gradation from seasonal grasses and herbs to

evergreen hemicryptophytes (restioids and cyper-

oids) (Specht & Moll, 1983).

Table 11 shows the relations of Humansdorp

shrublands to the soil sites defined by Specht (in

Specht & Moll, 1983; Campbell, 1983a).

Thicket grows on base-rich and calcium-rich soils.

The structure and composition of the thicket on the

different soil suites is very similar (Table 8). They

deviate from the structural characterization given for

South African mediterranean shrublands on base-

rich soils by Specht & Moll (1983; Table 5) in that

they are closed (never savannoid) and have a sparse

understorey of shade-tolerant grasses and herbs.

South Coast Renosterveld occurs on soils which are

marginally base-rich or nutrient-poor (Table 11, Fig.

8). Specht & Moll (1983) include renosterveld as a

mediterranean shrubland on base-rich soils which

has been derived recently from ‘an open-scrub

formation, dominated by Olea africana and Sider-

oxylon inerme’. The latter is regarded as the

‘true-mediterranean’ shrubland analogous to mator-

ral, maquis, chaparral and mallee of other lands with

mediterranean-type climate (Boucher & Moll, 1980;

Di Castri, 1980). It is argued (3.2.1) that South

Coast Renosterveld is derived from a grassland (or

open grassy small-leaved shrubland) and that the

‘ Olea - Sideroxylon open-scrub’ is restricted to

special edaphic sites and represents a westward

penetration of subtropical thicket. South Coast

Renosterveld does have the understorey characte-

ristics of shrublands ( sensu Specht & Moll, 1983) but

does not have any clear analogue in other

mediterranean lands. Axelrod (1978) argues a

post-glacial origin for California coastal sage and a

recent spread induced by man’s disturbances to the

landscape (see 2.3.1). Westman (1981; pers. comm.)

sees very little structural similarity between renos-

terveld and coastal sage scrub.

TABLI 12. —Correlation between soil variables and dominant phytochorological groups in shrubland types in the Humansdorp

study area. Data from 97 releves. Sample distribution as in Fig. 8 and including two samples from alluvium (Kaffrarian Succulent

Thicket) and two samples from Cango limestone and phyllite (Kaffrarian Thicket)

•+ p < ,05

••++ p < ,02

•••+++ p < ,01

••••++++ ; p < ,001

R. M. COWLING

221

Shrublands on nutrient-poor soils or heathlands

(sensu Specht, 1979; Specht & Moll, 1983) include

South-eastern Mountain Fynbos and Grassy Fynbos.

Bond (1981) and Campbell et al. (1981) restrict the

use of heathland to communities dominated by

ericaceous shrubs whereas Specht’s concept has a

much wider applicability. Grassy Fynbos is anomo-

lous since it has an understorey that is often

dominated by seasonal grasses (Fig. 10) even when

the over-storey is composed of ericaceous shrubs.

Grassy Fynbos soils are slightly more fertile than

Mountain Fynbos soils (Fig. 8; 3.1.2).

Dune Fynbos on calcium-rich soils shows both

heathland and shrubland characteristics (Fig. 10).

These transitional features are not so much reflected

in the structure of the understorey (cf. Specht &

Moll, 1983) but in intermingling and mosaic patterns

of shrubland (large-leaved subtropical shrubs) and

heathland (small-leaved fynbos shrubs) overstoreys

on deep, well drained sands. Dune Fynbos on

shallow sands overlying calcrete conforms to

Specht’s (1979) heathland concept.

Table 12 shows the correlations between some soil

variables and phytochorological groups characteris-

tic of shrubland types in the study area. Cape

endemics are characteristic of Mountain and Grassy

Fynbos (Cowling, 1983a; Fig. 9) and show a highly

significant negative relationship with soil depth,

litter, organic carbon and all major nutrients; they

show significant positive relationships with percen-

tage sand and rock cover. In general, a high

incidence of fynbos elements is associated with

shallow, rocky, infertile, sandy soils.

Cape-Afromontane linking species are typical of

moist renosterveld ( Themeda - Cliff ortia Commu-

nity) and to a lesser extent. Grassy Fynbos

(Cowling, 1983a; Fig. 9). Correlations with soil

variables are similar to Cape endemics except that

there are no significant correlations with percentage

sand, pH and rock cover. Cape-Karoo-Namib

linking species are characteristic of drier renoster-

veld communities ( Elytropappus — Eustachys and

Elytropappus — Relhania ) (Cowling, 1983a). They

show very weak relationships with soil nutrients

indicating the intermediate fertility status of

renosterveld soils in relation to the full spectrum of

soils sampled. The highly significant positive

correlation between the percentage of Cape-Karoo-

Namib linking species and rock cover is a result of

the concentration of these species in the Elytropap-

pus — Eustachys Community on stony soils derived

from Enon conglomerate.

The Tongaland-Pondoland endemics typify

thicket communities (Cowling, 1983a; Fig. 35) and

show relationships with soil variables that are exactly

the inverse of the Cape endemics (Table 12). Soils

are deep base-rich, sandy loams to sandy clays.

Cape-Tongaland-Pondoland linking species are

commonly found in Dune Fynbos but also occur in

renosterveld (Cowling, in press; Fig. 9). They show

highly significant positive correlations with available

calcium and pH (calcium-rich soils sensu Specht &

Moll (1983) and weaker positive relationships with

percentage sand, S-value and available phosphorus

(Table 12).

Soil nutrients are of value in differentiating

amongst classes and orders of shrublands in the

study area. There is a fertility gradient ranging from

Mountain Fynbos soils (least fertile), through

Grassy Fynbos and South Coast Renosterveld, to

thicket soils. Dune Fynbos occurs on calcareous

sands with markedly different chemical characteris-

tics to other fynbos soils. The only soil factor

common to all fynbos shrublands in the study area is

a sandy texture.

Specht & Moll’s (1983) plant formations (shrub-

land types) characteristic of base-rich, nutrient-poor

and calcium-rich soils are of limited value when

applied to shrublands in the Humansdorp study

area. Specifically, the marginal status of renoster-

veld soils is obscured. Moreover, many shrublands

on true base-rich soils are not Cape mediterranean

communities but subtropical thicket.

4.2 The role of soil moisture

Soil moisture plays a critical role in the

distribution of phytocoena at all levels of the

hierarchy. Most thicket communities and forest are

confined to deep, well drained soils with good

water-holding capacity. Succulent thicket, thicket

and forest can be ranged along a gradient of

decreasing soil-moisture deficit (Cowling & Camp-

bell, 1983b).

Within the Cape Fynbos Shrublands and Transi-

tional Small-leaved Shrublands, there are edaphic

grasslands and restioid grasslands on hardpan soils

with seasonally high watertables. Well drained

shallow or stony soils support small-leaved or

proteoid shrublands on all substrate types. On deep,

excessively drained dune sands, fynbos is succes-

sional to thicket whereas similarly-structured sand

derived from TMG sandstone support fynbos as a

climax.

Soil structure, moisture and nutrient status from a

complex group of interrelated factors that determine

vegetation composition and structure. These factors

are explored more fully in Cowling (in prep.).

4.3 Tension zones and phytochorological complexity

The Cape mediterranean climate zone is not

isolated from the summer-rainfall region but grades

into it along the southern and south-eastern coastal

forelands, thus facilitating the penetration of

subtropical elements into the Fynbos Biome. The

hot, dry valleys of the Fish, Sundays, Gamtoos and

Gouritz Rivers provide a coastwards migratory

pathway for karroid elements from the dry

intermontane valleys and upland basins to the north.

Furthermore, in the south-eastern and southern

Cape, where there is a warm-temperate climate at

the coast, Afromontane species are widespread at

low altitudes.

The chorological complexity of the south-eastern

Cape is due to the transitional nature of the climate

and complex topographical, geological and soil

patterns (Gibbs Russell & Robinson, 1981). The

whole region comprises a huge tension zone where

major phytochoria converge. In the south-eastern

and southern Cape this complexity is best expressed

on the relatively fertile soils of the coastal forelands.

222 A SYNTAXONOMIC .AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

Subtropical thickets of the river valleys have

admixtures of Karoo-Namib taxa; in more mesic

areas there is a strong Afromontane influence.

South Coast Renosterveld has species contributed

from the Cape, Karoo-Namib, Afromontane and

Tongaland-Pondoland phytochoria. On the lower

slopes and planed surfaces of the Cape Fold

Mountains, there is Grassy Fynbos which links

Mountain Fynbos and the Afromontane grasslands

of the eastern Cape [Dohne Sourveld (41)]. There

are patterns of intermingling of elements (e.g.

sub-tropical grasses in Grassy Fynbos and South

Coast Renosterveld) and interdigitation of commu-

nities (e.g. thicket on special edaphic sites in Grassy

Fynbos and South Coast Renosterveld).

Mountain Fynbos, of wholly Cape affinity, occurs

on mountain ‘islands’ within the ‘sea’ of chorological

complexity that comprises the vegetation on the

coastal forelands in the south-eastern and southern

Cape. Towards the east, this Mountain Fynbos

island is pinched into a narrow peninsula on the

upper slopes of the Elandsberg and Groot Winter-

hoek Mountains.

A study of the chorological complexity and

endemism of these vegetation types, together with a

knowledge of the ecological relations of the modern

taxa that contribute to them, provides the basis for

hypotheses on the historical biogeography of the

area. This is explored elsewhere (Cowling, 1983a).

4.4 Conservation

A fundamental aim of the Fynbos Biome Project

is the stimulation of research focusing on the

management and conservation of biome communi-

ties (Kruger, 1978). It is not my intention to digress

at length on the conservation status of each

community recognized in this study. Rather, I wish

to highlight situations where unique components will

be lost unless immediate action is taken.

Much attention has been given to the conservation

of Mountain Fynbos and Kruger (1977) has

proposed a series of mountain reserves within

recognized biogeographical regions. Much of the

area of Mountain Fynbos is controlled by the

Department of Environment Affairs and, although

some of this land could be afforested in the future,

theoretically the situation is good as the department

explicity provides for conservation in these areas

(Kruger, 1977).

In direct contrast, the conservation status of the

lowland regions in the Fynbos Biome area is critical

(Taylor, 1978; Boucher & Moll, 1980). Coastal

Rhenosterbosveld (46) now covers only 9% of its

former extent, having been largely replaced by

agricultural crops (Taylor, 1978). About 1% of its

former total area is conserved (Edwards, 1974). The

conservation status of Coastal Fynbos (47) is equally

dismal with only 2,1% of its former total extent

conserved (Edwards, 1974).

Most of the study area is on the lowlands where

land is privately owned and not subject to State

control. Grassy Fynbos is maintained in a semi-

natural state for veldgrazing. Since this form of

land-use is likely to persist over much of its area, the

conservation status of these communities is not

regarded as critical. The current fire regime in

Grassy Fynbos has resulted in the local elimination

of a seed regenerating species (Fig. 14). Grassy

Fynbos in the south eastern Cape is conserved in the

Kouga, Groendal and Suurberg Wilderness Areas

(Kruger, 1977).

South Coast Renosterveld communities in the

study area are severely threatened because they

occur on arable soils. Their conservation invokes the

problem of the withdrawal of high potential land

from agriculture. The present trend of ploughing up

natural veld in order to establish artifical pastures

and cereal crops is proceeding at an alarming rate.

There are great economic incentives to replace

degraded renosterveld with more productive agri-

cultural crops. Renosterveld communities are

scientifically important as they reflect to the fullest

the biogeographical complexity of the south-eastern

Cape vegetation. Immediate steps should be taken

to ensure their adequate conservation. A reserve

should be established to incorporate the hilly

country and parts of the adjacent coastal plain on the

southern side of the Gamtoos Valley. This area

would include Elytropappus - Relhania Renoster-

veld and Euclea — Brachylaena Succulent Thicket

on Bokkeveld shale and Elytropappus — Eustachys

Renosterveld and Themeda — Passerina Grassy

Fynbos on Enon conglomerate. The topography is

such that most of the area is unsuitable for crop

husbandry.

Attention has already been drawn to the critical

state of dune fynbos and thicket communities in the

study area (Cowling, 1980). Dune coasts are

dynamic, fragile ecosystems, extremely vulnerable

to disturbance and quickly destroyed by human

irresponsibility (Heydorn & Tinley, 1980). The dune

communities are certainly the most threatened in the

region. A number of factors, including indiscrimi-

nate and poorly planned resort development and

poor agricultural management have contributed to

the destruction of natural vegetation, the thickening

up and spread of alien Acacia spp. and the

re-activation of stable dunes.

The only two provincial nature reserves in the

study area are located on the coast. However,

together they cover less than 100 ha and provide

inadequate conservation for the region’s flora and

fauna (Cowling, 1980). As an absolute minimum to

meet the conservation requirements of the dune

communities, the present reserve at Cape St Francis

should be extended to include at least 150 ha of

indigenous vegetation (cf. Heydorn & Tinley, 1980).

Succulent thicket and thicket occurs mainly along

water courses and in the steeply dissected country of

the Gamtoos Valley. Some attempts are being made

to clear the thicket in the latter area but most of the

land is non-arable. The immediate purchase of land

for conservation is not a priority. Afromontane

Forest occurs on state-controlled land and is

conserved by the Department of Environment

Affairs.

R. M. COWLING

223

ACKNOWLEDGEMENTS

I thank S. Pierce, B. M. Campbell, H. P. Linder,

K. L. Tinley, E. J. Moll, H. Deacon, A. Scholtz, C.

Boucher, H. C. Taylor, W. J. Bond, F. J. Kruger,

R. L. Specht and W. Westman for useful discussion

during the course of this study. B. M. Campbell, E.

J. Moll, J. C. Scheepers, M. J. A. Werger, W.

Westman and P. Zedler commented on an earlier

draft of this paper. I am grateful to E. Brink, G. V.

Britten, H. R. H. Smith and G. Thompson of the

Department of Agriculture for their help and

co-operation. This study forms part of the Fynbos

Biome Project and was funded by the CSIR. Many

thanks to Mrs M. L. Jarman of CSP:CSIR for

liaison and other supporting facilities.

UITTREKSEL

’n Hierargiese sintaksonomiese skema van vegeta-

sie aan die oostelike grens van die Fynbosbioom

(Humansdorpstreek) word as ’n tweede benadering

aangebied, na die vroeere werk van Acocks (1953) in

die gebied. Besonderhede oor die fisiografie,

geologie, klimaat, gronde, historiese kenmerke en

huidige bestuur van die gebied word verskaf as

agtergrond vir hierdie en ander artikels wat handel

oor gemeenskapskarakterisering en ekologiese ver-

houdings, dinamika, struktuur en biogeografie. 'n

Gemeenskapsklassifisering is ontwikkel deur middel

van 'n hierargiese, numeriese klassifiseringstegniek

(TW1NSPAN) wat geordende tweerigting fitososiolo-

giese tabelle produseer. Die vergelykings en uiteinde-

like sortering van tabelle is volgens die metodes van

die Zurich-Montpellier-skooI. Hoer sintaksonomiese

range (klasse en ordes) word subjektief gedefinieer.

Vier klasse, sewe ordes en 22 gemeenskappe word in

die studiegebied herken. Die klasse is Kaapse

Fynbosstruikveld (3 ordes, 10 gemeenskappe),

Kaapse Oorgangs-kleinblarigestruikveld (1 orde, 4

gemeenskappe) , Subtropiese Oorgangsruigteveld (2

ordes, 6 gemeenskappe) en Afromontane — Woud (1

orde, 2 gemeenskappe). By die bespreking van die

skema word gefokus op die vlak van die orde

(ongeveer ekwivalent aan 'n veldtipe). Ter aanvulling

van die diagnostiese floristiese elemente, word

sintaksa verder gekarakteriseer deur middel van

biogeografiese, strukturele en habitatkriteria. Die rol

van historiese bodemb emitting op vegetasiedinamika

en onderlinge verwantskappe word kortliks bespreek.

Sintaksonomiese konsepte word, sover moontlik, op

die hele Fynbosbioom, sowel as biome aangrensend

aan sy oostelike grens, ge-ekstrapoleer. Sekere

sintaksa word relatief tot die struikveld/heideveld

konsepte wat vir meditereense-tipe ekosisteme ont-

wikkel is, ondersoek.

REFERENCES

Acocks, J. P. H., 1953. Veld types of South Africa. Mem. bot.

Surv. S. Afr. 28.

Acocks, J. P. H., 1979. The flora that matched the fauna.

Bothalia 12: 673-709.

Adamson, R. S., 1948. Some geographical aspects of the Cape

flora. Trans. R. Soc. S. Afr. 31: 437-464.

Anon., 1942. Weather on the coasts of southern Africa. Part 3.

Union of South Africa from Mossel Bay to East London.

Pretoria: Government Printer.

Archibald, E. E. A., 1955. An ecological survey of the Addo

Elephant National Park. J1 S. Afr. Bot. 20: 137-154.

Aucamp, A. J. & Barnard, H. H. , 1980. Die ontplooiing van die

veekundpotensiaal van die droe gras-bosgemeenskappe van

die Oos-Kaap. Proc. Grassld Soc. sth. Afr. 15: 137-140.

Austin, M. P., 1977. Use of ordination and other multivariate

descriptive methods to study succession. Vegetatio 35-

165-175.

Aweto, A. O., 1981. Organic matter build-up in fallow soil in

part of south-western Nigeria and its effects on soil

properties. J. Biogeogr. 8: 67-74.

Axelrod, D. I. , 1978. Origin of coastal sage vegetation, Alta and

Baja California. Am. J. Bot. 65: 1117-1131.

Bagshawe-Smith, L., 1937. Rhenosterbush in the district of

Albany. 5. Afr. J. Sci. 33: 335.

Barkman. J. J., Moravec, J. & Rauschert, S., 1976. Code of

phytosociological nomenclature. Vegetatio 32: 131-185.

Basset, P. A., 1978. The vegetation of a Camargue pasture. J.

Ecol. 66: 803-827.

Bews, J. W., 1917. The plant succession in the thorn veld. S. Afr.

J. Sci. 14: 153-172.

Bews, J. W., 1925. Plant forms and their evolution in South

Africa. London: Longmans.

Bond, W. J., 1980a. Periodicity in fynbos of the non-seasonal

rainfall belt. Jl S. Afr. Bot. 46: 343-354.

Bond. W. J., 1980b. Fire and senescent fynbos in the Swartberg,

Southern Cape. S. Afr. For. J. 114: 68-71.

Bond. W. J., 1981. Vegetation gradients in the southern Cape

mountains. M.Sc. thesis, University of Cape Town.

Bond., W. J., 1984. Seed predators, fire and Cape Proteaceae —

limits of the population approach to succession. Vegetatio. In

press.

Bond, W. J., Vlok, J. & Viviers. M., 1984. Variation in seedling

recruitment of Cape Proteaceae after fire. J. Ecol. 72:

209-221.

Botanical Research Institute (BRI), no date. The Standardi-

zation of syntaxonomic nomenclature. Pretoria: BRI.

Unpublished.

Botha, C. G., 1923. The dispersion of the stock farmer in the

Cape Colony in the eighteenth century. S. Afr. J. Sci. 20:

574-580.

Boucher, C., 1978. Cape Plangklip area. 11. The vegetation.

Bothalia 12:455-497.

Boucher. C., 1980. Notes on the use of the term ‘Renosterveld’.

Bothalia 13:237.

Boucher, C., 1983. Floristic and structural features of the coastal

foreland vegetation south of the Berg river, western Cape

Province, South Africa. Proceedings of the 10th AETFAT

Congress, Pretoria. Bothalia 14:669-674.

Boucher, C. & Jarman, M. L., 1977. The vegetation of the

Langebaan area, South Africa. Trans. R. Soc. S. Afr.

42:241-272.

Boucher, C. & Moll, E. J., 1980. South African mediterranean

shrublands. In F. di Castri, D. W. Goodall & R. L. Specht,

Mediterranean-type shrublands. Amsterdam: Elsevier.

Brady, N. C., 1974. The nature and property of soils, 8th edn.

New York: Macmillan.

Butzer, K. W. & Helgren, D. M., 1972. Late Cainozoic

evolution of the Cape coast between Knysna and Cape St

Francis, South Africa. Quat. Res. 2:143-169.

Campbell, B. M., 1978. Similarity coefficients for classifying

releves. Vegetatio 37:101-109.

Campbell, B. M., 1983a. Montane plant environments in the

Fynbos Biome. Bothalia 14:283—298.

Campbell, B. M., 1983b. Vegetation classification in a

floristically complex zone: the Cape Floristic Region. In

prep.

Campbell, B. M. & Moll, E. J., 1977. The forest communities of

Table Mountain, South Africa. Vegetatio 34:105 — 115.

Campbell, B. M., Cowling R. M., Bond, W. J. & Kruger, F.

J., 1981. Structural characterization of vegetation in the

Fynbos Biome. South African National Scientific Program-

mes Report No. 53. Pretoria: CSIR.

Campbell, J., 1815. Travels in South Africa. London: Black &

Parry.

Charley, J. C. & West, N. E., 1975. Plant induced chemical

patterns in some shrub dominated semi-desert ecosystems of

Utah. J. Ecol. 63:945-963.

Clements, F. E., 1916. Plant succession. Carnegie Inst. Publ. No.

242, Washington.

224 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

Connel, J. H. Slatyer. R. O., 1977. Mechanism of succession

in natural communities and their role in community stability

and organization. Am. Nat. Ill: 1119-1144.

Cowling, R. M., 1980. The coastal dune ecosystems of the

Humansdorp District — a plea for their conservation. East.

Cape Natur. 70: 25—28.

Cowling, R. M., 1983a. Phytochorology and vegetation history

in the south-eastern Cape, South Africa. /. Biogeogr. 10:

393-419.

Cowling. R. M., 1983b. Diversity relations in Cape shrublands

and other vegetation in the south-eastern Cape, South

Africa. Vegetatio 54: 103—127.

Cowling, R. M., 1983c. The occurrence of Q and C4 grasses in

fynbos and allied shrublands in the south-eastern Cape,

South Africa. Oecologica 58: 121 — 127.

Cowling, R. M. & Campbell. B. M., 1983a. The definition of

leaf consistence categories in the Fynbos Biome and their

distribution along an altitudinal gradient in the south-eastern

Cape. Jl S. Afr. Bot. 49: 87-101.

Cowling, R. M. & Campbell, B. M., 1983b. A comparison of

fynbos and non-fynbos coenoclines in the lower Gamtoos

River valley, south-eastern Cape, South Africa. Vegetatio

53: 161-178.

Croizat, L., 1965. An introduction to the subgeneric

classification of Euphorbia L. with stress on the South

African and Malagasy species. Webbia 20: 573—706.

Dale, M. B. & Webb, L. J., 1975. Numerical methods for the

establishment of associations. Vegetatio 30: 77-87.

Daniels, R. E., 1978. Floristic analysis of British Mires and Mire

communities. J. Ecol. 66: 773-803.

Deacon, H. J., 1979. Excavations at Boomplaas Cave: a

sequence through the Upper Pleistocene and Holocene in

South Africa. World Archaeol. 10: 241 — 257.

Deacon, H. J., 1983. The comparative evolution of

mediterranean-type ecosystems: a southern perspective. In

F. J. Kruger, D. T. Mitchell & J.U.M. Jarvis,

Mediterranean-type ecosystems. The role of nutrients.

Berlin— Heidelberg— New York: Springer.

Deacon, H. J. & Brooker, M., 1976. The Holocene and Upper

Pleistocene sequence in the southern Cape. Ann. S. Afr.

Mus. 71: 203-214.

Di Castri, F. , 1980. Mediterranean-type shrublands of the world.

In F. di Castri, D. W. Goodall & R. L. Specht,

Mediterranean-type shrublands. Amsterdam: Elsevier.

Drury, W. H. & Nisbet, I. C. T., 1973. Succession. J. Arnold

Arbor. 54: 331—368.

Du Toit, A. L., 1966. The geology of South Africa. 3rd edn.

Edinburgh & London: Oliver & Boyd.

Du Toit, P. F., 1972. Acacia karroo intrusion: the effect of

burning and sparing. Proc. Grassld Soc. sth. Afr. 7:

429-430.

Du Toit, R. & Du Toit, E. , 1938. Rhenosterbush as a weed. Fmg

S. Afr. 13: 429-430.

Dyer, R. A., 1937. The vegetation of the divisions of Albany and

Bathurst. Mem. bot. Surv. S. Afr. 17.

Edwards, D., 1967. A plant ecology survey of the Tugela River

Basin. Mem. bot. Surv. S. Afr. 36.

Edwards, D. , 1974. Survey to determine the adequacy of existing

conserved areas in relation to vegetation types. A

preliminary report. Koedoe 17: 3—38.

Egler, F. E., 1954. Vegetation science concepts. 1. Initial

floristic composition — a factor in old field development.

Vegetatio 4: 412—417.

Fanshawe, D. B., 1968. The vegetation of Zambian termitaria.

Kirkia 6: 169—179.

Fourcade, H. G., 1941. Check-list of the flowering plants of the

divisions of George, Knysna, Humansdorp and Uniondale.

Mem. bot. Surv. S. Afr. 20.

Garland, G. C., 1979. Assymetry in Drakensberg valleys: an

alternative point of view. 5. Afr. J. Sci. 75: 138-139.

Gauch, H. G. & Whittaker, R. H., 1981. Hierarchical

classification of community data. J. Ecol. 69: 537-557.

Gerryts, E. J., 1949. Humansdorp se groei en bloei.

Humansdorp: Biblioteek Pers.

Gibbs Russell, G. E. & Robinson, E. R., 1981. Phytogeography

and speciation in the vegetation of the eastern Cape.

Bothalia 13: 467-472.

Gleason, H. A., 1926. The individualistic concept of the plant

association. Bull. Torrey bot. Club. 53: 7—26.

Glenn Lewin, D. C., 1980. The individualistic nature of plant

community development. Vegetatio 43: 141 — 146.

Glyphis, J. P., Moll, E. J. & Campbell, B. M., 1978.

Phytosociological studies on Table Mountain, South Africa:

1. The Back Table. Jl S. Afr. Bot. 44: 281-289.

Goldblatt, P., 1978. An analysis of the flora of southern Africa:

its characteristics, relationships and origins. Ann. Mo. bot.

Gdn 65: 369-436.

Goodall, D. W., 1977. Dynamic changes in ecosystems and their

study: the roles of induction and deduction. J. Environ.

Manage. 5: 309-317.

Granger, J. E. & Schulze, R. E., 1977. Incoming solar

radiation patterns and vegetation response: examples from

the Natal Drakensberg. Vegetatio 35: 47—54.

Grobler, P. J. & Marais, J., 1967. Die plantgroei van die

Nasionale Bontebokpark, Swellendam. Koedoe 10:

132-148.

Hall, T. D., 1934. South African pastures: retrospective and

prospective. S. Afr. J. Sci. 31: 59-97.

Hall. J. B. & Swaine, M. D., 1976. Classification and ecology of

closed-canopy forest in Ghana. J. Ecol. 64: 913-915.

Haughton. S. H., Frommurze, H. F. & Visser, D. J. L., 1937.

The geology of portions of the coastal belt near the Gamtoos

Valley, Cape Province. Geological Survey Report No. 151.

Pretoria: Government Printer.

Heddle, E. M. & Specht, R. L., 1975. Dark Island Heath

(Ninety^VIile Plain, South Australia). VIII. The effect of

fertilizers on composition and growth, 1950—1972. Aust. J.

Bot. 23: 151-164.

Helgren, D. M. & Butzer, K. W., 1977. Palaeosols of the

southern Cape coast. South Africa. Implications for laterite

definition, genesis and age. Geogrl Rev. 67: 430—445.

Heydorn, A. E. F. & Tinley, K. L., 1980. Estuaries of the Cape.

Part 1. Synopsis of the Cape coast: natural features, dynamics

and utilization. C.S.I.R. Research Report 380. Pretoria:

CSIR.

Hill, M. O., 1973. Reciprocal averaging: an eigenvector method

of ordination. J. Ecol. 61: 237—249.

Hill, M. O. 1979. TWINSP AN — A FORTRAN programme for

arranging multivariate data in ordered two-way tables by

classification of individuals and attributes. Ithaca, New York:

Section of Ecology and Systematics, Cornell University.

Hill, M. O., Bunce, R. G. H. & Shaw, M. W., 1975. Indicator

species analysis, a divisive polythetic method of classification

and its application to a survey of native pinewoods in

Scotland. J. Ecol. 63: 597-613.

Holland, P. G. & Steyn, D. G., 1975. Vegetational responses to

latitudinal variations in slope angle and aspect. J. Biogeogr.

2: 179-183.

Holland, P. G., Steyn, D. G. & Fuggle, R. F., 1977. Habitat

occupation by Aloe ferox Mill. (Liliaceae) in relation to

topographic variations in direct beam solar radiation

income. J. Biogeogr. 4: 61-72.

Holling, C. S., 1973. Resilience and stability of ecological

systems. A. Rev. Ecol. Syst. 4: 1-23.

Horn, H. S., 1974. The ecology of secondary succession. A. Rev.

Ecol. Syst. 5: 25—37.

Jackson, S. P. & Tyson, P. D., 1971. Aspects of weather and

climate over southern Africa. Environmental Studies

Occasional Papers 6: 1-11.

Jessop, J. P. & JacotGuillarmod, A., 1967. The vegetation of

the Thomas Baines Nature Reserve. Jl S. Afr. Bot. 35:

367-392.

Jordaan, P. G., 1964. Plantegroei van die distrikte Bredasdorp

en Caledon. Tydskr. Wet. Kuns, Nuwe Reeks 6: 47—58.

Keet., D. J. M., 1936. Report on driftsands in South Africa.

Forestry Department Bulletin 172.

Kershaw, K. A., 1973. Quantitative and dynamic ecology, 2nd

edn. London: Arnold.

Killick, D. J. B., 1963. An account of the plant ecology of the

Cathedral Peak area of the Natal Drakensberg. Mem. bot. j

Surv. S. Afr. 34.

Killick, D. J. B., 1979. African mountain heathlands. In R. L.

Specht, Heathlands of the world. A. Descriptive studies.

Amsterdam: Elsevier.

King. L. C., 1972. The coastal plain of southern Africa. Its form,

deposits and development. Z. Geomorph. N. F. 16:

239-251.

King, L. C., 1978. The geomorphology of central and southern

Africa. In M.J.A. Werger, Biogeography and ecology of \

southern Africa. The Hague: Junk.

Klein. R. G., 1974. Environment and subsistence of prehistoric

R. M. COWLING

225

man in the southern Cape Province, South Africa. World

Archaeol. 5: 249-284.

Klein, R. G., 1977. The ecology of early man in southern Africa.

Science 197: 115-126.

Komarkova, V., 1980. Classification and ordination in the

Indian Peaks area, Colorado Rocky Mountains. Vegetatio

42: 149-163.

Kruger, F. J., 1974. The physiography and plant communities of

the Jakkalyt ivier Catchment. M.Sc. thesis. University of

Stellenbosch.

Kruger, F. J., 1977. Ecological reserves in the Cape Fynbos:

towards a strategy for conservation. S. Afr. J. Sci. 73:

81-85.

Kruger, F. J., 1978. A description of the Fynbos Biome Project.

South African National Scientific Programmes Report No.

28. Pretoria: CSIR.

Kruger, F. J., 1979. South African Heathlands. In R. L. Specht,

Heathlands of the world. A. Descriptive studies. Amsterdam:

Elsevier.

Kruger. F. J., & Taylor, FI. C., 1979. Plant species diversity in

the Cape fynbos: gamma and delta diversity. Vegetatio 41:

85-93.

Laidler, D., Moll, E. J., Campbell, B. M. & Glyphis, J., 1978.

Phytosociological studies on Table Mountain, South Africa.

2. The Front Table. Jl S. Afr. Bot. 44: 291-295.

Lamprechts, J. J. N., 1979. Geology, geomorphology and soils.

In J. Day, W. R. Siegfried, G. N. Louw & M. L. Jarman,

Fynbos ecology: a preliminary synthesis. South African

National Programmes Report No. 40. Pretoria: CSIR.

Lee, K. E. & Wood, T. G., 1971. Termites and soils. London:

Academic Press.

Levyns, M. R., 1926. A preliminary note on the rhenosterbush

and the germination of its seed. Trans. R. Soc. S. Afr. 14:

383-388.

Levyns, M. R., 1927. A revision of Elytropappus. Jl S. Afr. Bot.

1: 89-103.

Levyns, M. R., 1935a. Germination in some South African seeds.

Jl S. Afr. Bot. 1: 161-170.

Levyns, M. R., 1935b. Veld burning experiments at Oakdale,

Riversdale. Trans. R. Soc. S. Afr. 23: 231-243.

Levyns, M. R. , 1956. Notes on the biology and distribution of the

rhenoster bush. S. Afr. J. Sci. 52: 141-143.

Lichtenstein, FL, 1812. Travels in southern Africa. Vol. 1.

London: Colburn.

Liebenberg, L. B. D., 1945. Veld management and soil

utilization in the south east Cape Province. Fmg S. Afr. 20:

617-628.

Linder, H. P., 1983. The historical phytogeography of the

Disinae (Orchidaceae). Proceedings of the 10th AETFAT

Congress, Pretoria 1982. Bothalia 14: 565—570.

Linder. H. P. & Campbell, B. M., 1979. Towards a

structural-functional classification of fynbos: a comparison

of methods. Bothalia 12: 723-729.

Louw, W. J., 1976. Mesoclimate of the Port Elizabeth-Uitenhage

metropolitan area. Pretoria Weather Bureau Technical

Paper No. 4. Pretoria: Government Printer.

MacVicar, C. N., Loxton, R. F., Lamprechts. J. J. N., Le

Roux, J., De Villiers, J. J., Verster, E., Merrywea-

ther, F. R., Van Rooyen, T. H. & Von Harmse, H. J.,

1977. Soil classification. A binomial system for South Africa.

Pretoria: Department of Agricultural Technical Services.

Marloth, R., 1923. Observations on the Cape Flora: its

distribution on the line of contact between the south-western

districts and the Karoo. S. Afr. J. nat. Hist. 4: 335-344.

Martin, A. R. H., 1960. The ecology of Groenvlei, a South

African fen. 1. The primary communities. J. Ecol. 48:

55-71.

Martin, A. R. H., 1965. Plant ecology of the Grahamstown

Nature Reserve: 1. Primary communities and plant

succession. Jl S. Afr. Bot. 31: 1—54.

Martin, A. R. FL, 1966. The plant ecology of the Grahamstown

Nature Reserve: 2. Some effects of burning. Jl S. Afr. Bot.

32: 1-39.

Martin, A. R. H. & Noel, A. R. A., 1960. The flora of Albany

and Bathurst. Grahamstown: Rhodes University.

McKenzie, B., 1978. A quantitative and qualitative study of the

indigenous forests of the south-western Cape. M.Sc thesis,

University of Cape Town.

McKenzie, B., Moll. E. J. & Campbell, B. M., 1977. A

phytosociological study of Orange Kloof, Table Mountain,

South Africa. Vegetatio 34: 41-53.

Milewski, A. V. & Esterhuysen, E., 1977. Habitat of

Restionaceae endemic to western Cape coastal flats. J! S.

Afr. Bot. 44: 55—65.

Moll, E. J. & White, F., 1978. The Indian Ocean Coastal Belt.

In M. J. A. Werger, Biogeography and ecology of southern

Africa. The Hague: Junk.

Mueller-Dombois, D. & Ellenberg, H., 1974. Aims and

methods of vegetation ecology. New York: Wiley.

Muir, J., 1929. The vegetation of the Riversdale area. Mem. bot.

Surv. S. Afr. 13.

Nagel, J. F., 1962. Fog precipitation measurements on Africa’s

south-west coast. Notos 11: 51-60.

Noble, I. R. & Slatyer, R. O., 1980. The use of vital attributes

to predict successional changes in plant communities subject

to recurrent disturbance. Vegetatio 43: 5—21.

Odum, E. P., 1969. The strategy of ecosystem development.

Science 164: 267—270.

Okali, D. U. U., Hall., J. B. & Lawson, G. W. 1973. Root

distribution under a thicket clump on the Accra Plains,

Ghana: its relevance to clump localization and water

relations. J. Ecol. 61: 439-454.

Olivier, M. C., 1977. A systematic check list of the

spermatophyta of the Baakens River Valley, Port Elizabeth,

Jl S. Afr. Bot. 43: 145-159.

Olivier, M. C., 1979. Notes on the vegetation of the Sardinia

Bay Area. East. Cape Natur. 67: 13—14.

Olivier, M. C., 1981. An annotated systematic check list of the

spermatophyta of the Springs Reserve, Uitenhage. Jl S. Afr.

Bot. 47: 813-828.

Orians.G. H. & Solbrig, O. T., 1977. A cost-income model of

leaves and roots with special reference to arid and semi-arid

areas. Am. Nat. Ill: 677-690.

Paterson, W., 1790. A narrative of four journeys into the country

of the Hottentots and Caffraria. London: Johnson.

Paulson, H. A., 1953. A comparison of surface soil properties

under mesquite and perennial grass. Ecology 34: 727-732.

Peet, R. K. & Christensen, N. L., 1980. Succession: a

population process. Vegetatio 43: 131-140.

Penzhorn, B. L. & Olivier, M. R., 1974. A systematic check list

of flowering plants of the Addo Elephant National Park.

Koedoe 17:121-136.

Phillips, J. F. V., 1931. Forest succession and ecology in the

Knysna region. Mem. bot. Surv. S. Afr. 14.

Phipps, J. P. & Goodier, R., 1962. A preliminary account of the

plant ecology of the Chimanimani mountains. J. Ecol. 50:

391-419.

Pickett, S.T.A., 1976. Succession: an evolutionary interpreta-

tion. Am. Nat. 110: 107-119.

Pierce, S. M. & Cowling, R. M., 1984. The phenology of

fynbos, renosterveld and subtropical thicket in the

south-eastern Cape. S. Afr. J. Bot. 4: 1-17.

Pillans, N. S., 1954. A revision of Metalasia. Jl S. Afr. Bot. 20:

47-90.

Poore, M. E. D., 1962. The method of successive approximation

in descriptive ecology. Adv. ecol. Res. 1: 35-68.

Rice, E. L. & Parenti, R. L., 1978. Causes of decreases in

productivity in undisturbed tall grass prairie. Am. J. Bot. 65:

1091-1097.

Roberts, B. R. & Opperman, D. P. J., 1974. Veld management

recommendations — a reassessment of key species and

proper use of factors. Proc. Grassld Soc. sth. Afr. 9:

133-140.

Roberts, B. R., Anderson, E. R. & Fourie, J. H., 1975.

Evaluation of natural pastures: quantitative criteria for

assessing conditions in the Themeda veld of the Orange Free

State. Proc. Grassld Soc. sth. Afr. 10: 133 — 140.

Robson, N. K. B., 1966. Celastraceae. In A. W. Exell, A.

Fernandes & H. Wild, Flora Zambeziaca 2, 2. London:

Flora Zambeziaca Managing Committee.

Rourke, J. P., 1972. Taxonomic studies on Leucospermum R.

Br. Jl S. Afr. Bot. Suppl. 8.

Rourke, J. P. , 1980. The Proteas of southern Africa. Cape Town:

Purnell.

Rust, I. C., 1967. On the sedimentation of the Table Mountain

Group in the western Cape Province. D.Sc. thesis, University

of Stellenbosch.

Schulze, B. R., 1972. South Africa. In J. H. Griffiths, World

Survey of climatology. Vol. 10. Climates of Africa.

Amsterdam: Elsevier.

Schulze, R. E., 1975. Incoming radiation on sloping terrain: a

226 A SYNTAXONOMIC AND SYNECOLOGICAL STUDY IN THE HUMANSDORP REGION OF THE FYNBOS

BIOME

general model for use in southern Africa. Agrochemophysi-

ka 7: 55-61.

Schulze, R. E. & McGee, O. S., 1978. Climatic indices and

classification in relation to the biogeography of southern

Africa. In M. J. A. Werger, Biogeography and ecology of

southern Africa. The Hague: Junk.

Schweitzer. F. R. & Scott, K., 1973. Early occurrence of

domestic sheep in sub-Saharan Africa. Nature 241: 547.

Sclater. J. C., Hellinger. S. & Tapscott, C., 1977. The

palaeobathymetry of the Atlantic ocean from the Jurassic to

the present. J. Geol. 85: 509-552.

Scott, J. D. & Van Breda, N. G., 1937. Preliminary studies on

the root system of the rhenosterbos (Elytropappus

rhinocerotis) on the Worcester Veld Reserve. S. Afr. J. Sci.

33: 560-569.

Silvester, R., 1960. Stabilization of sedimentary coastlines.

Nature 188: 467-469.

Skead, C. J., 1980. Historical mammal incidence in the Cape

Province. Vol. 1. The western and northern Cape. Cape

Town: Cape Dept. Nature and Environmental Conserva-

tion.

Slatyer, R. O., ed., 1976. Dynamic changes in terrestrial

ecosystems: patterns of change, techniques for study and

applications to management. M. A. B. Tech. Notes No. 4

Paris: UNESCO.

Smit, N. L., 1943. Reclamation of rhenosterbushveld. Fmg S.

Afr. 18: 287-289.

Smith, H. J., 1967. Ondersoek na die bestuur en benutting van die

skynfynbosveldtipe in die Humansdorp grondbewaring-

distrik. Eastern Cape Region: Department of Agricultural

Technical Services.

Sparrman, A., 1785. Voyage to the Cape of Good Hope. Vol. 2.

London.

Specht, R. L., 1979. Heathlands and related shrublands of the

world. In R. L. Specht, Heathlands of the world. A.

Descriptive studies. Amsterdam: Elsevier.

Specht, R. L. & Moll, E. J., 1983. Heathlands and

sclerophyllous shrublands — an overview. In F. J. Kruger,

D. T. Mitchell & J. U. M. Jarvis, Mediterranean-type

ecosystems. The role of nutrients. Berlin, Heidelberg, New

York: Springer.

Specht, R. L., Conner, D. J. & Clifford, H. T., 1977. The

heath-savanna problem; the effect of fertilizers on

sand-heath vegetation of North Stradbroke Island,

Queensland. Aust. J. Ecol. 2: 179-186.

Squires, V. R. & Trollope, W. S. W., 1979. Allelopathy in the

Karoo shrub, Chrysocoma tenuifolia. S. Afr. J. Sci. 75:

88-89.

Steedman, A. , 1835. Wanderings and adventures in the interior of

southern Africa. London: Longmans.

Stinder, H. W., Cilliers, P. G., Joubert, J. G. & Perold, I.

S., 1965. The nutritive value of natural pastures in the

Humansdorp area of the south-east Cape Province of the

Republic of South Africa. Technical Communications 45.

Pretoria: Department of Agricultural Technical Services.

Story, R., 1952. A botanical survey of the Keiskammahoek

district. Mem. bot. Surv. S. Afr. 27.

Tansley, A. G., 1935. The use and abuse of vegetational

concepts and terms. Ecology 16: 284—307.

Taylor, H. C., 1955. Forest types and floral composition of

Grootvadersbosch. Jl S. Afr. For. Ass. 26: 33—46.

Taylor, H. C., 1961. Ecological account of a remnant coastal

forest near Stanford, Cape Province. Jl S. Afr. Bot. 23:

153-165.

Taylor, H. C., 1969. A vegetation survey of the Cape of Good

Hope Nature Reserve. M.Sc. Thesis, University of Cape

Town.

Taylor, H. C., 1970a. l.B.P. checksheet for the Robberg Nature

Reserve. Stellenbosch: Botanical Research Institute.

Taylor, H. C., 1970b. l.B.P. checksheet for the Riversdale

Nature Reserve. Stellenbosch: Botanical Research Institute.

Taylor, H. C., 1972. Notes on the vegetation of the Cape Flats.

Bothalia 10: 637-646.

Taylor, H. C., 1978. Capensis. In M. J. A. Werger,

Biogeography and ecology of southern Africa. The Hague:

Junk.

Taylor. H. C. & Morris,.!. W., 1981. A brief account of coastal

vegetation near Port Elizabeth. Bothalia 13: 519-525.

Taylor, H. C. & Van der Meulen, F., 1981. Structural and

floristic classification of Cape Mountain Fynbos on

Rooibcrg, southern Cape. Bothalia 13: 557—567.

Thunberg, K. P., 1796. Travels in Europe , Africa and Asia. Vol.

1. Performed between the years 1770 and 1779. London:

Richardson, Cornhill & Egerton.

Tinley, K. L., 1967. The moist evergreen forest— tropical dry

semi-deciduous forest tension zone in north-eastern Zulu-

land and hypotheses on past temperate/montane rain forest

connections. In E. M. van Zinderen Bakker, Palaeoecology

of Africa 2. Cape Town: Balkema.

Tinley, K. L., 1975. Habitat physiognomy, structure and

relationships. In Die Soogdier Navorsingsinstituut

1966—1975. Symposium Proceedings. Univ. Pretoria Publ.

New Series No. 97: 61—11.

Tinley, K. L., 1977. Framework of the Gorongosa ecosystem.

D.Sc. thesis, University of Pretoria. Published by K. L.

Tinley.

Trapnell, C. G., Friend, M. T., Chamberlain, G. T. & Birch,

H. F., 1976. The effect of fire and termites on Zambian

woodland soil. J. Ecol. 64: 577—588.

Trollope, W.’S. W., 1970. A consideration of macchia (fynbos)

encroachment in South Africa and an investigation into

methods of macchia eradication in the Amatole Mountains.

M.Sc. thesis, University of Natal, Pietermaritzburg.

Trollope, W. S. W., 1973. Fire as a method of controlling

macchia (fynbos) vegetation in the Amatole Mountains of

the eastern Cape. Proc. Grassld Soc. sth. Afr. 8: 35—41.

Trollope, W. S. W., 1974. Role of fire in preventing bush

encroachment in the eastern Cape. Proc. Grassld Soc. sth.

Afr. 9: 67-72.

Trollope, W. S. W., 1980. Encroachment and control of

indigenous undesirable plants in the eastern Cape. Alice:

Agricultural Development Research Institute, University of

Fort Hare.

Tyson, P. D., 1964. Berg winds of South Africa. Weather 19:

7-11.

Van der Merwe, C. V., 1976. Die plantekologiese aspekte en

bestuursprobleme van die Goukamma-Natuurreservaat.

M.Sc. thesis, University of Stellenbosch.

Van der Merwe, C. V., 1977. 'n Plantegroeiopname van die De

Hoop Natuurreservaat. Bontebok 1: 1—29.

Van der Meulen, F., 1979. Plant sociology of the western

Transvaal bushveld. South Africa. A syntaxonomic and

synecological study. Dissertationes Botanicae. Band 49.

Vaduz: Cramer.

Van der Schiyff, E. & Schoonraad, E., 1971. The flora of the

Mariepskop complex. Bothalia 10: 461-500.

Van Hulst, R., 1978. On the dynamics of vegetation: patterns of

environmental and vegetational change. Vegetatio 38:

65-76.

Vogel, J. C., Fuls, & Ellis, R. P., 1978. The geographical

distribution of Kranz grasses in South Africa. S. Afr. J. Sci.

74: 209-215.

Von Breitenbach. F., 1974. Southern Cape forests and trees.

Pretoria: Government Printer.

Walker, B. R., Ludwig. D., Holling, C. S. & Peterman, R.

M., 1981. Stability of semi-arid grazing systems. J. Ecol. 69:

473-498.

Watson, J. P., 1967. A termite mound in an iron age burial

ground in Rhodesia. J. Ecol. 55: 663-669.

Weather Bureau, 1954. Climate of South Africa. Part 2.

Rainfall statistics. Pretoria: Weather Bureau, WB 20.

Weimarck, H., 1941. Phytogeographical groups, centres and

intervals within the Cape Flora. Lunds Univ. Arsskrift N.F.

Adv. Z Bd. 37, Nr 5: 1-143.

Wellington, J. H., 1955. South Africa. A geographical study.

Vol. 1. Cambridge: Cambridge University Press.

Werger. M. J. A., 1972. Species area relationships and plot size:

with some examples from South African vegetation.

Bothalia 10: 583—594.

Werger, M. J. A., 1974. On concepts and techniques applied in

the Zurich -Montpellier method of vegetation survey.

Bothalia 11: 309-323.

Werger, M. J. A., 1978. Biogeographical division of southern

Africa. In M. J. A. Werger, Bio geography and ecology of

southern Africa. The Hague: Junk.

Werger, M. J. A., Kruger, F. J. & Taylor, H. C., 1972. A

phytosociological study of the Cape fynbos and other

vegetation at Jonkershoek, Stellenbosch. Bothalia

10:599-614.

Werger, M. J. A. & Van Gils, H., 1978. Phytosociological

classification in chorological borderline areas. J. Biogeogr.

3: 49-54.

R. M. COWLING

227

Westhoff.V. & Van der Maarel, E., 1973. The Braun-

Blanquet approach. In R. H. Whittaker, Ordination and

classification of communities. The Hague: Junk.

Westman, W. E., 1981. Diversity relations and succession in

Californian coastal sage scrub. Ecology 62: 170-184.

White, F., 1971. The taxonomic and ecological basis of

chorology. Mitt. bot. StSamml., Munch. 10: 91 — 112.

White. F., 1978. The Afromontane region. In M. J. A. Werger,

Biogeography and ecology of southern Africa. The Hague:

Junk.

White, F., 1981. The history of the Afromontane archipelago and

the scientific need for its conservation. Afr. J. Ecol. 19:

33-54.

White, F. ed., 1983. Vegetation of Africa. Paris: UNESCO/-

AETFAT/UNSO.

White, A., Dyer, R. A. & Sloane, B. L., 1941. Succulent

Euphorbiaieae of southern Africa. Vol. 1 & 2 Pasadena:

Abbey Garden Press.

Whittaker, R. H., 1953. A consideration of climax theory; the

climax as population and pattern. Ecol. Monogr. 23:41-78.

Whittaker, R. H., 1967. Gradient analysis of vegetation. Biol.

Rev. 42: 207-264.

Whittaker, R. H., 1973. Direct gradient analysis: techniques

and results. In R. H. Whittaker, Ordination and classifica-

tion of communities. The Hague: Junk.

Whittaker, R. H., Niering, W. A. & Crisp, M. D., 1979.

Structure pattern and diversity of a mallee community in

New South Wales. Vegetatio 39: 65—76.

Wild, H., 1952. The vegetation of Southern Rhodesia termitaria.

Rhodesia agric. J. 49: 280—292.

Williams, I. J. M., 1972. A revision of the genus Leucadendron.

Contr. Bolus Herb. 3: 1-425.

Bothalia 15, 1 & 2: 229-239 (1984)

Distinguishing features of forest species on nutrient-poor soils in the

Southern Cape

J. C. VAN DAALEN*

Keywords: evergreenness, forest, fynbos, mast fruiting, nutrition, phenolic compounds, roots, sclerophylly, soils

ABSTRACT

Soils of the indigenous forest-fynbos interface in the Southern Cape were sampled for chemical and physical

analyses and compared by means of anlyses of variance. Correlations among soil variables were investigated by

subjecting the correlation matrices to cluster analysis. Soil data were compared with that of fynbos and tropical

forest areas.

Morphological and physiological features of the forest vegetation, such as evergreenness, sclerophylly, phenolic

compounds in the leaves, mast fruiting (i.e. gregarious fruiting) and root mat, were correlated with the soil

nutritional status.

INTRODUCTION

The Afromontane forests of the Southern Cape

are surrounded by fynbos. Schimper (1903) postula-

ted that the distribution of the forest is controlled by

soil moisture. Neethling (1970) showed that forest

occurs on soil types as shallow, highly leached and

infertile as those on which fynbos occurs. Kruger

(1979) mentioned that seasonally severe soil

moisture deficit, together with periodic fires might,

by and large, preclude the growth of native trees.

Van Daalen (1980) postulated that the poor nutrient

status of the forest soils might prevent the

regeneration of trees when the forest has been

destroyed.

Jordan & Herrera (1981) proposed two types of

nutrient cycling strategies for tropical forests, i.e. an

oligotrophic strategy, which occurs on nutrient-poor

soils, and an eutrophic strategy occurring on

nutrient-rich soils. The majority of the tropical

forests are, however, oligotropic ecosystems (Jan-

zen, 1974; Me Key et al., 1978), susceptible to

leaching loss of nutrients as the result of rapid

decomposition of litter and heavy, frequent rains

(Jordan & Herrera, 1981).

Are the Southern Cape forest soils richer in

nutrients than the surrounding fynbos soils? Or do

the forests grow bn nutrient poor soils, where an

oligotrophic strategy applies? If so, what are the

distinguishing features of the forest species on these

poor soils?

In an attempt to answer these questions, available

data on the morphology and physiology of the forest

species are correlated with results obtained from soil

analyses from the forest-fynbos interface.

METHODS

Soils within and outside the forest were described

and sampled for chemical and physical soil

analyses. Five study areas were sampled: parts of

Kleineiland (33°58'S, 23°13'E) and Grooteiland

(33°57'S, 23°13'E) at Kaffirkop State Forest, Dirk se

* Saasveld Forestry Research Centre, Department of Environ-

ment Affairs, Private Bag X6531, George 6530.

Eiland (33°56'S, 23°13'E) at Diepwalle State Forest,

and Forest Creek Concession area (33°54'S,

22°52'E) and Ratelbos Island (33°53'S, 22°53'E) at

Goudveld State Forest (Fig. 1). The former three

sites are on the 220 m coastal plateau, whereas the

latter two are in the foothills of the Outeniqua

Mountains.

The following chemical and physical soil analyses

were made (see Appendix for the description of

analytical methods):

(a) pH in FLO and CaCL.

(b) Particle size analyses.

(c) Organic carbon of the A! horizon.

(d) Exchangeable cations (K, Ca, Mg, Na).

(e) Exchangeable acidity and aluminium.

(f) Available phosphorus.

(g) Total phosphorus.

(h) Total nitrogen.

Accuracy of determinations were checked by

including laboratory standard samples for each

element and calculating coefficients of variation of

the results.

The chemical and physical features of soils under

forest and fynbos, and of soils on different study

sites were compared by means of analyses of

variance for each element and ratios of different

elements separately (Table 1). Homogeneity of

variance was assumed.

The correlations among the different chemical

elements and particle sizes were investigated by

calculating Pearson’s correlation coefficient for the

variables of the Ax and B2i horizons separately,

except for C/P (perchloric acid), C/K, C/Ca and

percentage base saturation. The two correlation

matrices were subjected to cluster analyses (Orloci,

1975; Campbell, 1978 and Webster, 1979) using a

cluster procedure of group average sorting to

construct the dendrogram (Pritchard & Anderson,

1971) (Figs 2 & 3).

Elemental concentrations of the Aj horizon were

compared with available elemental information from

selected tropical and fynbos surface soils (Table 2).

Specific leaf mass (dry mass/unit area) was

230 DISTINGUISHING FEATURES OF FOREST SPECIES ON NUTRIENT-POOR SOILS IN THE SOUTHERN CAPE

Fig. 1. — Map of study areas.

determined for seventeen selected forest species

(Table 3). Leaf area of adult leaves (| to 1 year old)

was measured and the samples were then dried and

the mass determined.

RESULTS

Analyses of variance

No major differences between the morphology of

forest and fynbos soils were noticed on any single

site. Where differences did occur, they did not

coincide with the forest edge, indicating that the

forest-fynbos edge was artificially induced in these

cases* (Van Daalen, 1980). Differences between

sites were found; soils of the mountain foothills were

without any E (A2) horizons, whereas the plateau

soils had pronounced E horizons and heavy subsoils.

Only mean pH, and potassium, calcium and

aluminium concentrations showed significant dif-

ferences between forest and fynbos. Mean pH (in

H:0 and CaCh) of the B2i horizons of the forest was

lower than that of the fynbos B21 horizons. pH was

negatively correlated (p <0,01) with aluminium.

The forest B21 horizons had 65% higher aluminium

compared to the fynbos B2i horizons, thus

decreasing the pH through increased hydrolysis

associated with an increase in aluminium

(Etherington, 1975). There was a 55% decrease in

aluminium of the A] horizon from fynbos to forest

with no noticeable effect on pH.

Although aluminium was positive correlated with

percentage cla in the A, and B21 horizons (see

* The present study was limited to the coastal plateau and the

southern mountain foothills. On the more northern foothills on

the southern side of the Outeniqua Mountains, i.e. on the

northern boundary of the main forests, the distribution of forest

seems to be controlled by soil moisture to a greater extent than

that of the main plateau forest.

below), no significant difference in clay content

could be found between forest and fynbos sites in

both A and B horizons.

The pH of fynbos B21 horizons was significantly

higher than that of fynbos Ax horizons, whereas the

pH of both the forest and fynbos Ax horizons of the

plateau sites was significantly higher than that of the

mountain foothill sites.

Potassium of both A and B horizons decreased by

50% from forest to fynbos, and also from the Ax to

the B21 horizons in both forest and fynbos.

Calcium decreased from forest to fynbos only in

the Aj horizon.

When standardized per unit clay, all the cations,

except sodium, decreased significantly from the Aj

to the B21 horizon (Table 1). With the exception of

two values, all the S-values/100 g clay for the B21

horizons were in the mesotrophic range. The

S-values for the Ax horizons varied widely,

especially where the clay content was low.

Percentage base saturation decreased significantly

from the Ax to the B21 horizons in the forest. Due to

large variation in the values for fynbos the decrease

from fynbos Ax to B21 horizons, and the decrease

from forest to fynbos Ax horizons, were statistically

insignificant (Table 1).

Organic carbon was consistently higher in the

forest than in the fynbos, although not statistically

significantly so. A larger sample will be needed to

confirm this result.

No significant differences in nitrogen, and total or

available phosphorus between forest and fynbos

sites could be found (Table 1). The nitrogen content

of the B21 horizon was significantly lower than that

of the A! horizon, both in the forest and fynbos. In

the B2i horizons of the mountain foothill sites it was

about three times higher than in the dense clayey B21

TABLF J . — F-values for analyses of variance (only significant ones listed) of chemical soil analyses

<N

(N

U

X

a

\o \o OO h

o_fn ^o^oo o o \o «no N n

C'l t o” -i m ri -h

CC Cl M3 OO

-cocoooo-it^'Or't^'O'trot

qq i cm c- t--

co i/", M3r--i~~'r^sOMOOcocico('---

o o in co h in ov — i r— cc cc cc M3

,orv'-iici'tOcoitin(NNci^t

o-l rtrtTttnooi-Hor-i

MO SO 00 t~~

O h o- o in in co i — O' o in oi <jv

tq<qscqOM3ccooor'-cocoOM3

h o o o h o h" o’ o" of cc" —< cf

sO o "tf CC MD OJ ON cq om of O 00 cq

cq^cococo'cfoicococo'd-co^f

qq'Ooooo'H'tMOM'C'f

corococo'sfcocococococococo

—i o co in O'

'D'OXt^oO"tfncoifcorirf

NcciON'f^-OO'OCf

OfOnOOiOcor^cO

O CD CO cf Clf ci o'' o' of CO MO

MO O O

h OO C

q q o cf o in x Tf >h -Gc of oo

cq cf —i cf cf CO 00 00 — i — q cq -q •— I

cfcfcfcfcfcfc<3c<3cfcfcfcfcf

< CQ

CM <C

=0

Q a£

o _c

^03 0

so c~ oo m3

O O O co

O' t' o in

CO CM CO —I

o o o o

CM OO — 1 o

co co cn if

oo oo mo o

CO fO co V

M MO 't O

om r~~ mo mo

MO CM of 00

00 CM MO cf

CO ^ CO ^f

cq cq c*3 Om

Tf Cf Cf Cf

■a

o

<C QQ < £Q

<1> t-j

as <c3

& =$3

s 3-c

o

OO MO ■'t O

oo t' f-

■ef c- mo r-

in CM cf CM

MO cq

of of

rq O O

M3 cf Cf^

>n q oo

M3 cf Cf

< cq

CM <C

m

r-H r-1 1— ( ’ ’ ' N

Cl Cl CM C3

.03 XCQ ®

cl f' OO 't

—i o cq cq

— f o' — r mo

O 43 O 43

S a

S3 44

<l> O

=® '33

< CQ < 33

Source of variation

00 OO MO O

sO OO CO O

oo c- oi q

MO oo in OM

o M oo Cl

in 3 h O'

't in oo oo

CO o OM mo

sO MO C~~ mo

o —i r- h

cm cf cq cf

cm" d'oo

C l c~ CM mO

O' r — Om O

cf co cq cq

of o' cf cm"

00 00 cf Cl

O O OM MO

cl

£ O

2 ■c

— I CMcc

—< CM 33 -—

<CQ^-

C sO O' O'

I/O ’ — I C' o

cq Cq O. Cq

o’ cf o’ o

cl oo r~ mo

O cf CO Om

H Cl Cl H

Cl Cl CC — l

mo cf to om

— I CC) CM <— I

SO MO OM O'

Tt SO MQ O

>n 't co c

OO so 00 — i

M3 O ci 00

O r~ c*o cf

c- Cl 00 Cl

't -H SO O'

C — O’ C>

O O O cf

o

< CQ <C 33

I — so co O

Cl. Cl. aa

CM O' O O'

SO H Cl H

Cq cq ©. cq

o o" o' c[

OM O O' CM

mg ci >n so

— q Cq — q O,

o' o" o" o'

OM CC OM MO

— 1 M3 SO_ MG.

in O fl 'f

SO O' Cl O'

so O' MO oo

oo o oo ci.

r~- mo

so co

• ^ C

t-< c 43

O 43 _

43 < 03

i—t CM 20

CM pQ

sO M3 OM OM

Cl O ■'* O'

m3 cq sq Mq

— " o" cf r~

so

I'OOOM

_I o

cq °, °„

o" o" o o_

so O 't Mf

OM CM M3 — 1

o o co o

o o o o

so MO OM Om

cc C- Cl M3

ci oo cq oq

cc c~ so •— i

O0 —I £O.Cl

q' c- oo — i

|S rH q q

cio MO O'

S5 g-

“ S3 44

o ScS

c £ ^

>, tJ ^

VH O M O

C 43 O 43

43 ^43 ^

-M CM —l cl

<' CQ <C SO

OO r- M3 O

■^f T-H OM O

ci o. oo cq

cf Tf~ CC rf

m3 Mi" cc ^

cq °q cq cq

MO CC MO Cl

SO CC MO M3

00 C- M3 O

o o c- t

00 C- M3 O

o C- C~- Om

cf — q M3 Cq

so c m" MO

OM —I

00 cc

CM —I

. II o

M o 43

w o 43 _

o

4 dm

cm 03 ==■

< ®

CM

C- MO Om OM

00 MO CM M3

so om c— ■ oq

00 C M3 oo

so sc in Mf

C~ so O' OM

00 OM MO CM

cq rq cq Mjq

in c c tn

om om oo

c O' in oo

C Mf Cl SO

CM O OM OO

>. "O

feL s

£

a s

o o

P- u

Pm Pm

>

T3

^ o

qj CJ

o ^

S3 44 tn

> O

15 q u.

CD

CQ cm

M O M O

O 43 o 43

43 ^

H(S4d

< 33 < 03

tu

u

u-

C— SO cc O

-q cq

cc so

M3 M3 Mi" M3

C Mf cc If

q cq O cq

o' o" o" cq

O ^f o M3

— M CC »— I C-

CM O CM 1-H

o" cf o' o'

Tf i cc

OM Cq cq M3

00 rf ^ oo

OM CC CM CC

—1 00 M3 00

cq Cq O cq

o' cf o o_

Tj- OM Tf O'

MO Tf —I

oo o oo cq

o" o" o" o"

so M3 OM OM

O Mi" M3 CM

cq co ■— q — q

o" oo" — r o

M3 M3 MO -4

M3 — I

Cl O^ CM 0_

o o" 0^0^

M3 — i Om cc

SOM Tfr M3 'Cf

—i o^ <q O

o' o" o" o"

SO M3 OM OM

r — cf c~ gc

M3 Tf OM O

oq o__ —q —q

o" o" o_ o_

OM SO MO C-

cn m3 't c

so O cq cq

o’ o" o’ o

O Q 2 4!

(43

CQ < S3

p < 0,05

Thl fust figure (1 ) is the degrees of freedom for the treatments and the second figure is the error degrees of freedom

Kaffirkop includes Kleineiland, Grooteiland and Dirk se Inland.

Goudveld includes Forest Creek Concession Area and Ratelbos Island.

% Base saturation was calculated as S-value -f (S-value + exchangeable aci lty).

o

J. C. VAN DAALEN

233

horizons of the plateau sites. Root penetration into

the more sandy B21 horizons of the mountain foothill

sites was much better than in the plateau B21

horizons. Nitrogen could therefore be added to

these sandy horizons much more efficiently.

No significant differences in the carbon/

potassium, carbon/calcium and carbon/total phos-

phorus ratios between forest and fynbos sites were

found. The Ar horizons of the mountain foothills

fynbos sites had significantly lower (more favourab-

le) carbon/calcium and carbon/total phosphorus

ratios than that of the plateau fynbos sites.

Cluster analyses

Fig. 2 shows the strong positive correlations

among the standardized S-value, nitrogen, cation

(except sodium) and carbon values for the A,

horizon. For the B21 horizon these relations were

much weaker (Fig. 3), possibly as a result of a lower

carbon content.

Sand content (especially in the A, horizon), silt

content and pH were relatively independent. Strong

positive correlations between clay content and

aluminium, both in the Ax and B21 horizons were

observed.

Comparison of elemental concentrations

1 Phosphorus and carbon/ phosphorus ratio

Available phosphorus values for this study, and

those of tropical forest and fynbos areas, were in the

QJ

Fig. 2.— Dendogram for A! hori-

zon variables using Pearson’s

correlation coefficients as si-

milarity indices.

0,3

(U

Fig. 3. — Dendogram for B2! hori-

zon variables using Pearson’s

correlation coefficient as si-

milarity indices.

TABLE 2. —Elemental concentrations (ppm) in surface soils (Aj horizons) of selected tropical forest and fynbos areas

234

DISTINGUISHING FEATURES OF FOREST SPECIES ON NUTRIENT-POOR SOILS IN THE SOUTHERN CAPE

<D

c

,CD

~o3

OC

-h ro

rO rf

O S GO

<D , Q

O 1

) os os as

so so so

r as as os

as

CD ^ T3 ^ K a T3

QJ CJ >— i

to GO (j)

as

c ~ —< c

- i 3 xf I S3

2 oo 5 3^.5 1 -a

OSioSlIHHZwhW co wO

i , | ' *

I I I i/) U ro

*

u

GO

o

sO

O

GO

— < go as

OO os GO I

Tj- ro r*o

I

c*o oo

00 os

so so

I I I I I

go CD

go r-

I I I I I I I I I

as o a\ <n

A OO o

u O os oo

I I so as so as

GO 00 GO l"-

I I 00 I I GO go I OO I | GO |

■ rf | | |

u o- o o n as

— < ^ o o — < r-

ri -H so sO GO rn

OO SO C"*

h M as u

1 ro ro (N SO

roO^TOOOsOOOOOOO

P'OOroO-'Ol't'trN'O't^

OO OO go rf U' -H o: " “

Tf U' -H Tt SO -H

M (N O ri

^ x, x, O CO u u

r*0 OO no | -rf 00 -H I |

CO N r)1- o

r-so^iroasoor-Gosoas— ir^o

MOOOOOOioMOSCOOUO

go so • — i ro ■ — i -h (N n

ci so as go

oo as go as

— i — i so o

O ror^ooooo

I | | I | I | | |^- | COSOM-Ot O

1 1 1 1 1 1 ' 1 'Os Osor^O-sO'^t— i

Tt o o o

i i i i i i ij if5°S1' i i i

ro ro

i so 't Os ro

' — r CO © go I I I I I I I I I ’t CO I

cl cl

c^J

u u

II*

£ £

QZ C£ E C

d,fl,Uo.cuUUi

CD

a & a> <d c. CL

O C3 ^ (J U WJc^)

.SMrt3300cs-.(L)t)

o r 0 -t rrt i_, CJ "♦“* O.

-S'flBssgg j j? g g *

o co co o O

J^OOqOO^O^

o c_> .55

Cl cl qj

Ul U-

<d <u

.5 .S

CD CD

0) Q

o S

U- U-

C c

a> --a

CD c^

- &

£ o

c

c fc .5 .S

HHHQSeSoSfivSa

S jb:<mIuSS

Exchangeable concentration. All other values are total concentrations.

J. C. VAN DAALEN

235

same range. Total phosphorus and carbon/total

phosphorus values were, however, in the range of

the fynbos site values (Table 2).

2 Nitrogen, cations and carbon! nitrogen,

carbon! potassium and carbon! calcium ratios

On the whole, these values were higher (lower in

the case of the ratios) than those of the fynbos sites

and lower (higher in the case of the ratios) than

those of tropical forest sites (Table 2).

DISCUSSION

Differences between forest and adjacent fynbos sites

The higher aluminium content of the forest B21

horizons compared to forest Ax horizons might

reflect a more effective podzolisation under forest

than under fynbos because of a different organic

composition. Usually the aluminium-pH relation in

the A horizon is complicated by litterfall, nutrient

circulation, decomposition rate and the type of

vegetation (D. C. Grey, personal communication,

1980) . This can mask any aluminium-pH correlation.

As mentioned before, all the study sites had

artificially induced forest-fynbos edges; therefore

the differences in nutrients are probably vegetation

induced. A possible explanation for the decrease in

potassium from forest to fynbos is that most of it is

locked up in the vegetative biomass. The forest, with

a higher biomass will keep more potassium than the

fynbos. Alternatively, the fynbos might take up less

potassium per biomass unit. When a forest is

destroyed, more potassium is available than can be

utilized by the invaders, in this case fynbos, with a

consequent leaching of this mobile element

(Etherington, 1975). This theory is confirmed by the

strong positive correlation between potassium and

organic carbon (Fig. 2) and the results of Brasell,

Unwin & Stocker (1980). They found a higher

potassium content in the litter of the more fertile

rain forest and Araucaria cunninghamii plantation

sites in tropical Australia than on the less fertile

forest and plantation sites.

The same arguments as for potassium can be

applied to explain the decrease in calcium from

forest to fynbos Ax horizons. Again, bio-cycling of

calcium is important in restricting leaching loss by

returning calcium to the soil surface (Etherington,

1975). As with potassium, Brasell etal. (1980) found

a higher calcium content in the litter of the more

fertile sites. Although calcium is not immobilized in

organic matter, it does show chelation with a

number of organic acids, for example citric and

gluconic (Stevenson, 1967 ex Etherington, 1975).

(Also see the strong positive correlation between

calcium and organic carbon in Fig. 2.) This can

restrict leaching of calcium out of the soil profile.

The higher cation content of the Ax horizon

compared to that of the B21 horizon can be

attributed to the greater number of exchange sites

associated with a higher carbon content in the Ax

horizon than in the B21 horizon (see Fig. 2 for

positive correlation between standardized cations

and standardized carbon content).

Correlations between elements

The positive influence of standardization for clay

on the correlations among different elements,

including organic carbon (Fig. 2), emphasizes the

importance of the organic carbon in maintaining the

nutritional status of the Ax horizon.

Comparison with fynbos and tropical forest areas

Unlike Brasell et al. (1980), who found

significantly higher concentrations of total

phosphorus and nitrogen in the litter of tropical rain

forest in Australia than in that of adjoining

Araucaria cunninghamii plantation, the differences

in nitrogen and total or available phosphorus

between forest and fynbos sites were very small and

insignificant. On the whole, the elemental

concentrations of the forest horizons were in the

range of concentrations of fynbos and heathland

sites in South Africa and Australia (Table 2). Total

phosphorus, in particular, was much lower than that

of savanna and tropical forest areas.

Evers (1967) used carbon-based ratios to

characterize the nutritional status of spruce (Picea

excelsa) sites. Large ratios implied unfavourable

nutritional conditions. He suggested the following

limiting values for these forests:

C/N 24-26

C/P 350-450

C/K 400-500

C/Ca — uncertain, but preferably less than 100.

He found the C/P ratio a particularly good indicator

of the general nutritional status, in contrast to the

C/N ratio.

C/P ratios of this study were very high (Table 2).

This implies poor nutritional status. Generally, the

southern Cape indigenous forests are growing on

fynbos soils. As a result of this, certain disting-

uishing features of the forest species can be

observed.

Distinguishing features of forest species

1 Evergreenness and sclerophylly

Evergreen sclerophylls could be related to low

nutrient availability. These leaves are longer lived,

have relatively thick cuticles and wax cover and may

be more resistent to nutrient loss by parasites and

herbivores than mesophytic leaves (Grubb, 1977).

Loveless (1961 & 1962) regarded the sclerophyllous

leaf as the expression of a metabolism found in

plants that can tolerate low levels of phosphate. He

suggested that a phosphorus content of leaves below

about 0,3% resulted in sclerophylly. Cowling &

Campbell (1980) demonstrated that the higher

degree of evergreenness and sclerophylly in the

South African fynbos compared to shrublands in

Chile and California is related to low soil nutrient

levels, rather than winter rainfall and summer

droughts, as suggested by Schimper (1903).

In Table 3 some forest species are listed in a

decreasing order of specific leaf area of adult leaves.

By comparing these values to those of Sobrado &

Medina (1980) for the scleromorphic low-tree forest

on sandy podzolised soils of Amazonas (which vary

236 DISTINGUISHING FEATURES OF FOREST SPECIES ON NUTRIENT-POOR SOILS IN THE SOUTHERN CAPE

between 455 for old Mouriri uncitheca leaves to 89

for young Catostemma sp. leaves) and by using

Schimper’s (1903) and Bond’s (1981) guidelines for

the field recognition of scleromorphic leaves (i.e.

non-succulent leaves which break when folded, they

are leathery and tend to maintain their shape on

drying), the species with a specific leaf mass of 140

and more can be regarded sclerophyllous. These

species form 66% of the total forest tree composition

(Geldenhuys, 1975; Geldenhuys, 1980). Leaf

analyses for some of these species yielded

phosphorus levels of 0,08% and lower (Table 3).

TABLt 3. — Specific leaf mass, and percentage P and total

phenol content in leaves of some forest species

1 Specific leaf mass determined on fully grown leaves (\ 1

year old.

2 Average of all seasons (Koen, J.H., 1981).

3 Averajte of determinations made by J.H. Koen (personal

communication, 1981 (and A.H.W. Seydack (1981).

Although Trichocladus crinitus (Thunb.) Pers.,

the dominant ground layer shrub in the forest, did

not fit the pattern of sclerophylly associated with low

leaf phosphorus levels (Table 3), in general

sclerophylly seems to be related to low soil

phosphate levels in the southern Cape.

The following benefits of evergreenness on

nutrient poor soils have been suggested:

(i) Evergreens have a higher nutrient use

efficiency than deciduous plants, i.e. higher carbon

gain per unit of nutrient used in leaf construction

(Small, 1972).

(ii) If the evergreen leaves are retained for more

than one year, they have a lower annual nutrient

requirement and loss (Monk, 1971; Chapin et al.,

1980).

(iii) Chapin et al. (1980) suggested that evergreens

have a reduced annual cost of translocation of

nutrients from senescent leaves, and of synthesis and

breakdown of nutrient storage compounds. How-

ever, when leaves are shedded, large proportions of

especially P, N and K are retranslocated to the twig

before abscission (Small, 1972).

(iv) Evergreens have a year-round abscission and

production of new leaves so that nutrients are

retained in leaves rather than in soil and thus fewer

nutrients are lost to the system by leaching (Monk,

1966; Thomas & Grigal, 1976).

Goldberg (1982) proposed that evergreen plants

would grow and survive at least as well on fertile as

on infertile soils in isolation. As a result of lower

photosynthetic rates and larger amounts of

structural carbon per leaf (Grime & Hunt, 1975),

and a higher investment in anti-herbivore

compounds (see below), they have a lower

maximum growth rate and therefore a lower

competitive ability than deciduous plants on fertile

soils.

Deciduous tree species present in the Southern

Cape forests occur mainly on the few shale bands

and other more fertile sites. Very few occur on the

nutrient poor sandstone derived soils (C. J.

Geldenhuys, personal communication, 1981).

2 Phenolic compounds in leaves

Janzen (1974) postulated that the vegetation on

tropical poor, leached soils is protected against

herbivory by exceptionally high levels of toxic

phenolics and other secondary compounds. In a

habitat with extremely low primary productivity, yet

a climate favourable to animals year round, there

should be strong selection for plants that are

exceptionally rich in chemical defenses. This is so for

the following reasons:

(a) In infertile habitats it is costly to replace

nutrients of consumed and damaged vegetative

parts.

(b) In a habitat with low productivity and

evergreen plants, proportionally more of the plant’s

resources are expected to be spent on defenses to

prevent herbivore damage before the leaf must be

replaced for internal economic reasons.

Determinations of total phenols in leaves of some

of the sclerophyllous species in Table 3 yielded

values of more than 50 mg/g dry leaf material, except

for Olea capensis L. subsp. macrocarpa (C. H. Wr.)

Verdoorn, which had a phenol content of 44,8 mg/g.

These values are in the same order as those for some

African rain forests (McKey et al., 1978). Some

values for the mesophytic species in Table 3 were

above 50 mg/g too, but the average of the

sclerophyllous species was 59,6 mg/g compared to

38,9 mg/g for the mesophytic species.

3 Mast fruiting

Mast fruiting (i.e. gregarious fruiting) in

Dipterocarpaceae has been observed in South-East

Asia, where trees have population- and

community-level fruiting that is synchronized at

intervals greater than one year. The adaptive

significance of synchronous fruiting on habitats with

low primary productivity is the advantage of

saturating seed predators, thereby increasing the

J. C. VAN DAALEN

237

probability of reproductive success. Furthermore,

large population buildups of seed predators are

prevented due to the lack of food between fruiting

years (Janzen, 1974).

Fruiting intervals of more than one year are

well-known for many forest species in the Southern

Cape, for example Olea capensis subsp. macrocarpa,

Podocarpus latifolius (Thunb.) R.Br. ex Mirb., P.

falcatus (Thunb.) R.Br. ex Mirb., Rapanea

melanophloeos (L.) Mez, Apodytes dimidiata E.

Mey. ex Am. and Olinia ventosa (L.) Cufod.

(Phillips, 1926; F. von Breitenbach, 1965). The

release of mature seed of all these species seems to

be synchronized to a certain extent, although it can

take several months before all seed of a species has

been dropped (Phillips, 1926).

4 Root mat

According to Jordan & Herrera (1981) the root

mat on the soil surface is one of the most important

mechanisms for direct nutrient cycling and nutrient

conservation in the Amazonian rain forest. They

found that the gradient of decreasing thickness of

the root mat parallelled the gradient of increasing

soil fertility. Stark & Jordan (1978) showed that

99,9% of all 45Ca and 32P sprinkled on these root

mats was immediately absorbed and only 0,1%

leached through the mat. Rapid growth of small

roots is another nutrient conserving mechanism

(Jordan & Escalante, 1980), and Herrera et al.

(1978) demonstrated the rapid colonization of a

fallen leaf by small roots and its rapid decomposi-

tion.

Dense root mats on and in the surface soil layer

occur in the Southern Cape indigenous forests (Van

Daalen, 1980). Small roots that have colonized

decomposing logs above the root mat and humus

layer have been observed in the forest (Fig. 4).

Plant litter falling on the floor, is decomposed and

nutrients that become available are utilized

immediately. Disturbance of the root and litter

layers will have significant negative effects on forest

nutrition, and implicitly will effect forest

regeneration and development.

CONCLUSIONS

Apart from some higher elemental concentrations

in forest than in fynbos A] horizons due to a higher

carbon content, forest soils in the Southern Cape are

not richer in nutrients than the surrounding fynbos

soils. Distinguishing features of the forest species on

these soils are evergreenness and sclerophylly, high

levels of phenolics and other secondary compounds,

fruiting intervals of more than one year and dense

root mats on and in the surface soil layer.

This has several implications for the management

of these forests:

(a) During exploitation the root mat and soil

surface must be left intact as far as possible. The

more this is disturbed, the more difficult it will be to

restore the closed nutrient cycle of the forest.

(b) The smaller the gap created in the canopy

when felling trees, the more likely it will be that

plant litter reaches the opened forest floor and that

the nutrient conserving mechanisms can be

maintained.

(c) Assuming a positive relation between forest

growth and soil nutrient status, organic carbon can

be an indicator of the potential of a site for forest

growth (Ojeniyi & Agbede, 1980). Exploitation

should then be limited to sites with high organic

carbon in the A horizon.

(d) Forest edges should be left undisturbed. The

better nutrient conserving mechanisms and nutrient

cycles can be maintained, and the healthier and

more vigorous the ecotone vegetation is, the better

Fig. 4. — Roots colonizing a de-

composing log above the root

mat and humus layer in the

southern Cape indigenous

forest.

238 DISTINGUISHING FEATURES OF FOREST SPECIES ON NUTRIENT-POOR SOILS IN THE SOUTHERN CAPE

these edges can tolerate negative external in-

fluences, like disturbances through fire or exploita-

tion of adjoining plantation.

(e) When re-establishing forest, the

re-establishment of the soil nutrient status is of great

importance. Re-establishment should be done with

species that can tolerate low nutrient levels, i.e. the

most sclerophyllous species.

Further studies are needed concerning the relation

between nutrient status and forest regeneration,

growth and distribution of species, on mast fruiting,

phenolic compounds in leaves, sclerophylly and

nutrient cycling in the forest.

ACKNOWLEDGEMENTS

This report is based partly on an M.Sc. study. The

research forms part of the conservation forestry

research programme of the Directorate of Forestry,

Department of Environment Affairs. Thanks are

due to Messrs C. J. Geldenhuys, W. J. Bond and D.

C. Grey of the Saasveld Forestry Research Station,

Mr J. J. N. Lambrechts of the University of

Stellenbosch and Mr F. J. Kruger of the South

African Forestry Research Institute in Pretoria for

constructive discussions and criticizing the draft.

UITTREKSEL

Gronde van die inheemse woud-fynbos-

skeidingsvlak in die Suid-Kaap is bemonster vir

chemiese en fisiese grondontledings en deur middel

van variansie-analise vergelyk. Korrelasies tussen

grondveranderlikes is ondersoek deur die

korrelasiematryse aan klont- ( = ‘cluster’) analise te

onderwerp. Die gronddata is vergelyk met die van

fynbos en tropiese woudareas.

Morfologiese en fisiologiese eienskappe van die

woudplantegroei, soos bladhoudendheid, sklerofillie ,

fenoliese bestanddele in blare, gesamentlike

saadlewering en die wortelmat, is met die

voedingstatus van die grond gekorreleer.

REFERENCES

Allison, L. E., 1965. Organic carbon. In C. A. Black, Methods

of soil analysis. Amer. Soc. Agron. Monog. 9: 1367-1378.

Bond. W. J., 1981. Vegetation gradients in Southern Cape

mountains. M.Sc. thesis, University of Cape Town.

Brasell, H. M., Unwin. G. L. & Stocker. G. C., 1980. The

quantity, temporal distribution and mineral-element content

of litterfall in two forest types at two sites in tropical

Australia. J. Ecol. 68: 123 — 129.

Campbell, B. M., 1978. Similarity coefficients for classifying

releves. Vegetatio 37: 101 — 109.

Chapin, F. S., Johnson, D. A. & McKendrick, J. D., 1980.

Seasonal movements of nutrients in plants of differing

growth form in an Alaskan tundra ecosystem: implications

for herbivory. J. Ecol. 68: 189-210.

Cowling, R. M. & Campbell, B. M., 1980. Convergence in

vegetation structure in the mediterranean communities of

California, Chile and South Africa. Vegetatio 43: 193—197.

Day, P., 1965. Particle fractionation and particle size analysis. In

C. A. Black, Methods of soil analysis. Amer. Soc. Agron.

Monog. 9: 545—566.

Etherington, J. R., 1975. Environment and plant ecology.

London: John Wiley.

Evers, F. H., 1967. Kohlenstoffbezogene Nahrelementverhalt-

nisse (C/N, C/P, C/K, C/Ca) zur characterisierung der

Ernahrungsituation in Waldboden. Mitt. Ver. forstl. Stan-

dortskart. 17: 69—76.

Geldenhuys, C. J., 1975. Die autekologie van Podocarpus

falcatus. M.Sc. thesis, University of Stellenbosch.

Geldenhuys, C. J., 1980. The effect of management for timber

production on floristics and growing stock in the southern

Cape indigenous forests. S. Afr. For. J. 113: 6—15.

Goldberg, D. E., 1982. The distribution of evergreen and

deciduous tree relative to soil type: an example from the

Sierra Madre, Mexico, and a general model. Ecology 63:

942-951.

Golley, F. B., McGinnes, J. T., Clements, R. G., Child, G. I.

& Duever, M. J., 1975. Mineral cycling in a tropical moist

forest ecosystem. Athens: University of Georgia Press.

Grime, J. P. & Hunt, R., 1975. Relative growth rate: its range

and adaptive significance in a local flora. J. Ecol. 63:

393-422.

Groves, R. H., 1980. Heathland soils and their fertility status. In

R. L. Specht, Ecosystems of the world: heathlands and

related shrublands , 143-150. Amsterdam: Elsevier.

Grubb, P. J., 1977. Control of forest growth and distribution on

wet tropical mountains. A. Rev. Ecol. Syst. 8: 83-107.

Herrera, R., Jordan, C. F., Klinge. H & Medina, E., 1978.

Amazon ecosystems: their structure and functioning with

particular emphasis on nutrients. Interciencia 3: 223—232.

Janzen, D. H., 1974. Tropical blackwater rivers, animals and

mast fruiting by the Dipterocarpaceae. Biotropica 6,2:

69-103.

Jordan, C. F. & Escalante, G., 1980. Root productivity in an

Amazonian rain forest. Ecology 61: 14—18.

Jordan, C. F. & Herrera, R., 1981. Tropical rain forests: are

nutrients really critical? Am. Nat. 117,2: 167—180.

Koen, J. H. 1981. A study of the distribution, population

composition, movements etc. of the Knysna elephants,

Loxodonta africana africana, Blumedach (1797). Unpubl.

Dept, of Environment Affairs.

Kruger, F. J. 1979. South African heathlands. In R. L. Specht,

Ecosystems of the world: heathlands and related shrublands,

19—80. Amsterdam: Elsevier.

Loveless, A. R., 1961. A nutritional interpretation of

sclerophylly based on differences in the chemical composi-

tion of sclerophyllous and mesophytic leaves. Ann. Bot.

N.S. 25,98: 168-184.

Loveless, A. R., 1962. Further evidence to support a nutritional

interpretation of sclerophylly. Ann. Bot. N.S. 26,104:

551-561.

McKey, D., Waterman, P. G., Mbi, C. N., Gartlan, J. S. &

Struhsacker. T. T., 1978. Phenolic content of vegetation in

two African rain forests: ecological implications. Science

202: 61-64.

Monk. C. D., 1966. An ecological significance of evergreenness.

Ecology 47: 504-505.

Monk, C. D., 1971. Leaf decomposition and loss of 45Ca from

deciduous and evergreen trees. Am. Midi. Nat. 86:

379-384.

Neethling, J. H., 1970. Classification of some forest soils of the

southern Cape. M.Sc. thesis. University of Stellenbosch.

Ojeniyi, S. O. & Agbede, O. O., 1980. Soil organic matter and

yield of forest and tree crops. PI. Soil 57: 61—67.

Orl6ci, L., 1975. Multivariate analysis in vegetation research. The

Hague: Junk.

Phillips, J. F. V., 1926. General biology of the flowers, fruits,

and young regeneration of the more important species of the

Knysna forest. S.Afr. J. Sci. 23: 366-417.

Pritchard, N. M. & Anderson, A. J. B., 1971. Observations on

the use of cluster analysis in botany with an ecological

example. J. Ecol. 59: 727-747.

Schimper, A. F. W., 1903. Plant- geography upon a physiological

basis. Oxford: Clarendon Press.

Seydack, A. H. W., 1981. Pilot investigation: possible inter-

actions between phenolic compounds and wildlife ecology in

the Southern Cape indigenous forests. Unpubl. Dept, of

Environment Affairs.

Small, E., 1972. Photosynthetic rates in relation to nitrogen

recycling as adaptation to nutrient deficiency in peat bog

plants. Can. J. Bot. 50: 2227—2233.

Sobrado, M. A. & Medina, E., 1980. General morphology,

anatomical structure and nutrient content of sclerophyllous

leaves of the ‘Bana’ vegetation of Amazonas. Oecologia 45:

341-345.

Stark, N. & Jordan, C. F. , 1978. Nutrient retention by the root

J. C. VAN DAALEN

239

mat of an Amazonian rain forest. Ecology 59: 434-437.

Tanner, E. V. J., 1977. Four montane rain forests of Jamaica:

quantitive characterization of the floristics, the soil and the

foliar mineral levels, and a discussion of the interrelation-

ships. J. Ecol. 65: 883—918.

Thomas, W. A. & Grigal, D. F., 1976. Phosphorus conservation

by evergreenness of mountain laurel. Oikos 27: 19—26.

Van Daalen, J. C., 1980. The colonization of fynbos and

disturbed sites by indigenous forest communities in the

southern Cape. M.Sc. thesis. University of Cape Town.

Von Breitenbach, F., 1965. The indigenous trees of southern

Africa. 5 vols, unpubl. Dept, of Environment Affairs.

Webster. R., 1979. Quantitative and numerical methods in soil

classification and survey. Oxford: Clarendon Press.

APPENDIX

ANALYTICAL METHODS USED FOR THE PHYSICAL

AND CHEMICAL SOIL ANALYSES

(a) pH in H20 and CaCl2 (O.OIM) was determined by means

of a 1: 2 soil: liquid ratio, measured in the supernatant after one

hour, using a combination glass electrode.

(b) Particle size analyses were made using the hydrometer

method (Day, 1965). All soils were pretreated by H202 to remove

the organic matter.

(c) Organic carbon of the A, horizon was determined by the

Walkley-Black method (Allison, 1965), using a correction factor

of 1,33 for recovery.

(d) Exchangeable cations (K, Ca, Mg, Na): The determination

was made on a IN NH4C1 extract of the less than 2 mm fine earth

by atomic absorption spectrophotometry. The S-value per 100 g

clay was calculated.

(e) Exchangeable acidity was determined on a IN KC1 extract

(extraction time = 4 minutes) by titration with NaOH. After

addition of NaF exchangeable A1 was determined by back

titration using HC1.

(f) Available phosphorus was determined colorimetrically by

the molybdenum blue procedure on a Bray No. 2 extract.

(g) Total phosphorus: After digestion with perchloric acid, the

colour development was determined spectrophotometrically using

ammonium vanadate.

(h) Total nitrogen was determined by a modified micro-

Kjehldahl method. After digestion with H2S04, the ammonia was

determined by titration with HCI.

Bothalia 15, 1 & 2: 241-244 (1984)

A cover meter for canopy and basal cover estimations

R. H. WESTFALL* and M. D. PANAGOS*

Keywords: basal cover, canopy cover, grassland, method, woodland

ABSTRACT

A simple, inexpensive, pocket-sized cover meter for estimating both canopy and basal cover is described. The

cover meter is based on the visual superimposition of canopy-to-gap ratio scales, on to vegetation. Tests indicate

that, in certain vegetation types, accuracy is comparable to the wheel-point apparatus in estimating basal cover.

Canopy cover is estimated in classes according to the Domin-Krajina cover-abundance scale.

INTRODUCTION

The accuracy of visually estimating canopy cover

in Braun-Blanquet vegetation analyses is dependent

on the observer’s experience in estimating cover.

The Braun-Blanquet cover-abundance scale

(Mueller-Dombois & Ellenberg, 1974) has large

class intervals to facilitate cover estimation. The

large class intervals, however, do not permit

detection of significant differences in vegetation

cover. Consequently, interpretation of Braun-

Blanquet tables places greater emphasis on presence

or absence of species than on species cover.

In a recent vegetation study (Westfall, 1981) the

Domin-Krajina cover-abundance scale (Mueller-

Dombois & Ellenberg, 1974) was used to determine

veld condition and trend, because of the smaller

class intervals than in the Braun-Blanquet cover-

abundance scale. However, the smaller class

intervals increase the difficulty of cover estimation.

This difficulty led to the development of a simple,

inexpensive, pocket-sized cover meter for canopy-

cover determinations. The success of the cover

meter was such that its application was extended to

basal-cover determinations.

DESCRIPTION AND USE OF THE COVER METER

The cover meter is based on cover estimation by

the canopy-to-gap ratio method (Edwards, 1983)

where cover is given in terms of the ratio of the mean

canopy diameter to the mean distance, as number of

canopy diameters, between the canopies of the

plants. Two black-and-white 35 mm positive slides

(transparencies) with various canopy-to-gap ratios

(Figs 1 & 2) depicted by horizontal bars and

corresponding spaces, to the right of each bar, are

used. Basal cover is estimated with the slide of Fig. 1

and canopy cover is estimated with the slide of Fig.

2. Slides may be constructed by photographing Figs

1 & 2 directly. The appropriate slide is placed in a

Cenei F3, or similar, pocket slide viewer.

A canopy or basal diameter together with the gap

to a random neighbour, to the right, is observed with

the free eye while the ratios observed in the slide

viewer are visually superimposed on the image

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

observed by the free eye. A filter, made of paper or

exposed film, and placed in front of the slide,

reduces the light input which can facilitate superim-

posing the two images. The observer selects a bar

and corresponding space to the right of the bar

which, when superimposed on the canopy or basal

diameter and gap, produce the best fit, and reads the

appropriate cover percentage, canopy: gap ratio (0)

or Domin-Krajina cover-abundance value (Figs 1 &

2). The proportions between bars and spaces for a

given percentage cover remain the same, so that

backward or forward movements by the observer to

match the images, are limited. The observer should,

however, remain at right angles to the canopies or

basal diameters observed.

The cover meter was tested to determine limits of

accuracy and minimum samples required.

STUDY AREAS

The estimates of basal cover were determined in

grassland plots at the Botanical Research Institute in

Pretoria, during March 1983. The canopy cover

estimates were determined in Acocks’s (1975) Mixed

Bushveld at the Nylsvley Nature Reserve, near

Naboomspruit, Transvaal, also during March 1983.

The stands were all selected to provide variation in

total cover.

METHODS

Basal cover

Basal cover was determined for each of the three

20 m x 20 m grassland plots by:

(i) A wheel-point apparatus (Tidmarsh & Haven-

ga, 1955), with points 0,38 mm in diameter and 1 m

apart. A sample of 1 000 points was taken in each

plot.

(ii) The cover meter, with 30 random readings in

each plot based on a grid with lines 1 m apart. Grid

co-ordinates were selected by random numbers

(Fisher & Yates, 1949). The nearest grass beneath

the grid intersection points was selected together

with the nearest neighbour intercepted by the grid

lines. The observer faced in a constant direction for

all recordings and used the right-hand grid line to

determine the nearest neighbour, for convenience.

Although grass, forbs and shrubs were recorded

242

A COVER METER FOR CANOPY AND BASAL COVER ESTIMATIONS

Fig. 1. — The basal cover-scale showing percentage basal cover and ratios (0) of basal diameter to gap. The horizontal bars

represent basal diameters and the corresponding spaces to the right represent gaps. The width of the vertical lines at the

extreme left represents bars.

^ 1

0 DOMIN - KRAJINA SCALE

0 I \ 91

9

0,10| L l_ I I I 76

8

0,33 L— L I I 51

7

0.661 L 1 I 34

6

0,90| I I I 26

5

2.001 I _ I 11

4

3,25l I I 6

3

8,501 1 I 1

2

29.001 I O’1

1

Fig 2. — The canopy-cover scale showing canopy-to-gap ratios (0) and percentage canopy cover. The Domin-Krajina

cover-abundance values are for the classes between the canopy-to-gap ratios or percentage cover. The horizontal bars

represent canopies and the corresponding spaces to the right represent gaps.

R. H. WESTFALL AND M. D. PANAGOS

243

separately for each method, only the grass

component is considered here.

The basal cover estimates with the cover meter

were tested for normal distribution according to

Southwood’s (1966) method as used by Van Ark

(1981). The number of samples required for

different limits of accuracy were then determined for

the three basal cover plots individually and jointly,

according to the formula (Van Ark, 1981):

where

N = number of samples

t = table value for the t- distribution with

(n-1) degrees of freedom

S = standard deviation with (n-1) degrees of

freedom

d = level of accuracy expressed as a decimal

x = mean.

Canopy cover

Eight stands, each with a different degree of

canopy cover, were selected from 1: 4 000 scale

aerial photographs. A plot of 2 cm x 2 cm was

demarcated on the aerial photographs in each stand,

representing approximately 80 m x 80 m in the field.

A single cover meter reading with the Domin-

Krajina cover-abundance scale was recorded of a

visually representative sample of the whole plot

within each plot. Only trees above 2 m in height

were taken into account. The 2 cm x 2 cm

demarcated areas on the aerial photographs were

traced on to film together with the tree (above 2 m

tall) canopies. The traced canopies were cut out and

weighed, the remaining film was weighed as also the

total 2 cm x 2 cm film with canopies as a control.

The results were expressed as percentage canopy

cover.

TABLE 1. — Comparison of estimates of total grass basal cover

with the wheel-point method and the cover meter in three

grassland plots

TABLE 2. — The minimum number of cover-meter samples

required in three grassland plots for different limits of

accuracy with 95% confidence limits

„ Minimum number of samples required

Limits of

Fig. 3. — A simple test for normal distribution on probability

graph paper (Southwood, 1966). A straight line through the

plotted points indicates a normal distribution. Plots 1 ( )

and 2 (ooooo) are approximately normally distributed but

plot 3 (xxxx) is not a normal distribution.

RESULTS

Basal cover

The comparison of total grass basal cover with the

wheel-point method and cover meter is given in

Table 1. The values for the cover meter are the

means of 30 samples. The cover meter samples are

approximately normally distributed for Plots 1 and 2

according to Southwood’s (1966) method of deter-

mining normal distribution- (Fig. 3). Spatial heter-

ogeneity of grass tufts accounts for the lack of a

normal distribution of samples in Plot 3 (Fig. 3). The

minimum samples required with 95% confidence

limits for different limits of accuracy with the cover

meter are given in Table 2 for each of the three

plots.

Canopy cover

The comparison of tree canopy cover, for trees

above 2 m in height, with the canopy cover

determined from aerial photographs and the cover

meter is given in Table 3.

TABLE 3. — Comparison of tree canopy cover, for trees above

2 m in height, with canopy cover determined from aerial

photographs and the cover meter

DISCUSSION AND CONCLUSIONS

The accuracy of the cover meter, as in the

wheel-point apparatus, (Tidmarsh & Havenga,

1955) decreases for a given number of samples, with

decreasing basal cover (Tables 1 & 2). For 10%

244

A COVER METER FOR CANOPY AND BASAL COVER ESTIMATIONS

basal cover, 26 samples are required with the cover

meter for a 20% limit of accuracy (Table 2). The

wheel-point apparatus requires 1 000 points for a

similar limit of accuracy (Tidmarsh & Havenga,

1955). However, 26 cover-meter samples takes

approximately a third of the time required for 1 000

points. A 20% limit of accuracy means that for 10%

basal cover there is 95% certainty that the results

will be in the range of 8-12%. The grass canopies in

the three grassland plots were knee-high. This did

not detract from estimating basal cover with the

cover meter, because the gap between basal tufts,

being greater than the bases, was of greater

importance in estimating basal cover than tuft basal

diameter. Where the basal tuft size is greater than

the spaces between tufts, tall grass could prevent use

of the cover meter. In practice, it was seldom

possible to obtain perfect fits of basal diameter to

gap and bar to space. The best fit was recorded with

emphasis placed on the gap and space fit, where the

ratio of basal diameter to gap ratio was greater than

one.

The discrepancies of the cover meter in estimating

canopy cover in classes for the 38% and 56% cover

values obtained from aerial photographs (Table 3)

may be attributed to incorrect selection of a

representative sample. In phytosociological surveys,

the problem of a single cover value representing an

entire community does not arise, because cover is

expressed as a raftge of cover values represented by

each releve in the community. The range may be

summarized as mean cover degree (Mueller-

Dombois & Ellenberg, 1974). As in the case of basal

cover, the best fit between canopy to gap and bar to

space is obtained when emphasis is placed on the gap

and space fit, where the canopy to gap ratio is

greater than one.

Although visually superimposing two different

images requires a little practice, the cover meter is

an extremely simple and portable device for rapidly

estimating both basal and canopy cover. The

accuracy of the cover meter depends on the user’s

requirements which may be a quick estimate of

cover or a more time-consuming assessment. Unlike

the single wheel-point apparatus, the cover meter

can be used by a single observer and is also very

inexpensive.

ACKNOWLEDGEMENTS

The authors thank Dr D. Edwards for his support,

Dr J. C. Scheepers for comments and suggestions

and Dr H. van Ark for assistance with the statistical

analyses.

UITTREKSEL

’n Eenvoudige, goedkoop sakbedekkingsmeter vir

die bepaling van beide kruin- en basalebedekking

word beskryf. Die bedekkingsmeter is gebaseer op die

visuele ooreenle van kruin- tot gapingskale op die

plantegroei. Toetse dui aan dat by sekere plantegroei

tipes, akkuraatheid is met die wielpunt-apparaat vir

basalebedekking vergelykbaar. Kruinbedekking

word in klasse volgens die Domin-Krajina bedek-

kingsgetalsterkte skaal bepaal.

REFERENCES

Acocks, J. P. H., 1975. Veld types of South Africa. 2nd edn.

Mem. bot. Surv. S. Afr. 40: 1 — 128.

Edwards, D., 1983. A broad-scale structural classification of

vegetation for practical purposes. Bothalia 14: 705—712.

Fisher, R. A. & Yates, F., 1949. Statistical tables for biological,

agricultural and medical research. London: Oliver & Boyd.

Mueller-Dombois, D. & Ellenberg, H., 1974. Aims and

methods of vegetation ecology. New York: John Wiley.

Southwood, T. R. E., 1966. Ecological methods with particular

reference to the study of insect populations. London:

Chapman & Hill.

Tidmarsh, C. E. M. & H/?venga, C. M., 1955. The wheel-point

method of survey and measurement of semi-open grasslands

and Karoo vegetation in South Africa. Mem. bot. Surv. S.

Afr. 29: 1-49.

Van Ark, H., 1981. Eenvoudige biometriese tegnieke en

proefontwerpe met spesiale verwysing na entomologiese

navorsing. Wet. Pamf. Dep. Landb. Vis. Rep. S. Afr. No

396.

Westfall, R. H., 1981. The plant ecology of the farm Groothoek,

Thabazimbi District. M.Sc. thesis. University of Pretoria.

Bothalia 15, 1 & 2: 245-258 (1984)

A vegetation survey of the Cape of Good Hope Nature Reserve. I. The

use of association-analysis and Braun-Blanquet methods*

H. C. TAYLOR**

Keywords: association-analysis, Braun-Blanquet, fynbos, phytosociology, plant community

ABSTRACT

The survey aimed to establish broad vegetation units that could be mapped on an extensive scale in the Cape of

Good Hope Nature Reserve at the southern tip of the Cape Peninsula. This paper compares the performance of

two methods, association-analysis and the Braun-Blanquet method as developed by the Zurich-Montpellier School

of Phytosociology. One hundred 50 m2 sample plots, covering the whole Reserve, were placed systematically at

grid intersections on the 1:18 000 topographical map, at 1 000-yard (914 m) intervals. Species lists, recording

merely presence of all species with permanently recognizable aerial parts, were made for each plot.

The association-analysis resulted in a classification of 23 final groups of sample plots, of which only five groups

showed high floristic and ecological homogeneity. Of the remainder, eight groups contained some anomalous,

misplaced plots, and ten represented small, isolated fragments of natural units. The original data were then

analysed using Braun-Blanquet methods to provide an independent classification for comparison with the former.

The Braun-Blanquet communities were found to be more homogeneous in terms of previously defined habitat

groupings and showed floristic relationships consistent with these groupings.

It is concluded that, with the type of sampling used, the synthetic phytosociological Braun-Blanquet method

provides a more natural classification of plant communities of the Reserve than does the monothetic divisive

association-analysis method.

INTRODUCTION

When Acocks (1975) first wrote his Veld types of

South Africa in 1953, little was known of the ecology

of the complex fynbos vegetation of the Capensis

region (Taylor, 1978). Acocks was therefore unable

to subdivide fynbos to the same extent as he did the

vegetation of other parts of South Africa. Though

there had been general descriptions of fynbos by

Marloth (1908) and Adamson (1938) and a few

quantitative studies like those of Wicht (1948) and

Rycroft (1951), there was still scant information on

the response of different types of fynbos to

treatments like veldburning and grazing. To

determine the effects of such treatments, experi-

mental research was initiated locally but these

isolated projects could not be satisfactorily com-

pared with one another, or extrapolated to other

areas, in the absence of a synoptic account of fynbos.

To fulfil this need, it was decided in the early 1960’s

to conduct a primary survey of Cape Mountain

Fynbos (Acocks’s Veld Types 69 and 70) and, for

this purpose, a suitable method had to be found. The

survey of such a large area — some 37 000 km2 in

rugged mountain terrain — would have to be divided

into components, and the method would have to be

sufficiently formalized and uniform to allow valid

comparison of data from each component. In

seeking a method suitable for a major survey of this

kind, a trial on a smaller area of fynbos was

required.

The Cape of Good Hope Nature Reserve (Fig. 1)

was chosen as the site for the trial, because its flora

* Based on an M.Sc. thesis. University of Cape Town.

** Botanical Research Unit, Department of Agriculture, P. O.

Box 471, Stellenbosch 7600.

was comparatively well known (Adamson & Salter,

1950), and its area large enough (77 km2) and its

vegetation sufficiently diverse to provide a represen-

tative sample of fynbos. The work was begun by the

author in 1966 and presented as a thesis to the

University of Cape Town (Taylor, 1969).

Study area

0 5 10 15 km

19°E

_l

Fig 1. — Geographical location of the study area.

246 A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE. I. THE USE OF

ASSOCIATION-ANALYSIS AND

The results of this trial, the first of its kind in

fynbos, led to the adoption by many subsequent

workers of the Braun-Blanquet phytosociological

method which has proved eminently suitable for

standard primary surveys of South African vegeta-

tion. This account compares the performance of two

methods used in the trial.

METHODS

The aim of the survey of the Cape of Good Hope

Nature Reserve was to establish broad vegetation

units that could be mapped on an extensive scale.

For this, a classification was needed. The first

requirement was to test a method that could permit

comparison of the vegetation of different areas.

Before the development of the electronic computer,

classifications based on statistical methods were

unsuccessful (Blackman, 1935; Ashby, 1936). By

1966, however, statistical and computing techniques

had been improved to provide practical methods

that could be interpreted comparatively easily.

Association-analysis (Williams & Lambert, 1959;

1960) appeared to be one suitable method. It had

given promising results on small areas of relatively

simple vegetation in Britain, by recording presence

of all species on small quadrats closely and

systematically spaced in a grid pattern. It would be

physically impossible to follow an identical proce-

dure in the large area of complex fynbos.

Nevertheless, it was thought that the method might

be adaptable by using larger plots spaced a kilometre

or more apart, and it was decided to use the method,

thus adapted, in the trial on the Reserve.

After the association-analysis results were inter-

preted, the original data were used in another

analysis applying the technique of the Zurich-

Montpellier School of Phytosociology, here referred

to as the Braun-Blanquet method (Braun-Blanquet,

1932; Becking, 1957; Whittaker, 1962; Kuchler,

1967; Werger, 1974), to provide an independent

classification for comparison with that obtained by

association-analysis.

Sampling and recording procedures

To provide a reference framework before samp-

ling commenced, habitat categories were classified.

BRAUN-BLANQUET METHODS

on physiographic/physiognomic criteria, in the

system shown in Table 1.

For the statistical analysis, sample plots were

placed systematically in a grid pattern over the

whole area, following the objective procedure of

Williams & Lambert (1959).

In a pilot enumeration conducted to determine the

best size and spacing of plots, twenty plots, each 10

m x 10 m, were placed at random intersections of

the grids on the 1: 18 000 map of the Cape

Peninsula. Each plot was subdivided into ten

sub-plots, each 2 m x 5 m. Separate lists for each

sub-plot were made of the permanently recognizable

species (discussed below), and the cumulative

number of species occurring in each successive

sub-plot was calculated. Regression equations for

plot sizes of 20 m2, 50 m2 and 100 m2 were derived

and, by extrapolation, the number of species which

one could expect to record in 350 plots of these sizes,

was calculated. (About 350 intersections of the grid

on the above-mentioned map occur in natural

vegetation on the Cape Peninsula.)

The results showed that, using plots of 100 m2, 734

species could be expected to occur in the samples. If

the plot size were reduced to 50 m2 one could expect

to record 624 species, whereas if the plot size were

further reduced to 20 m2 the expected number of

species would be 469. Therefore, a 50% reduction in

plot size would result in only a 15% expected

reduction in species present. This did not mean that

the survey could be completed in half the time,

however, since location and laying out of plots, and

the recording of observational data, took equally

long in both cases. Therefore, with further reduction

in plot size the time saved becomes less and less. The

use of a 20 m2 plot would effect very little further

saving in time while the expected number of species

would be reduced by 36%.

To check these calculations, the mean number of

additional species actually found in the ten sub-plots

(1, 2, 3 10), was determined. The results agreed

closely with the expected figures calculated theoreti-

cally.

Since rectangular plots are likely to cover greater

variations in pattern, and to include a greater

number of species than square plots (Brown, 1954;

TABLE 1 . — Habitat categories and equivalent vegetation units

Habitat categories

Vegetation units*

1. Sites exposed to recurrent fires

1 . 1 Marine sands and coastal Table Mountain Sandstone soils

1.1.1 well-drained marine sands of the coastal shelf

1.1.2 well-drained aeolian dunes of marine origin (including inland dunes)

1.1.3 poorly-drained coastal flats of marine origin

1.1.4 well-drained cliffs and ledges of the rocky sandstone coast

1.2 Acid sands derived from inland Table Mountain Sandstone soils

1.2.1 well-drained rocky hills, mountains and lower sandy slopes

1.2.2 well-drained ferricrete slopes

1.2.3 well-drained steep, rocky, northerly slope

1.2.4 drained level plateaux

1.2.5 seasonally inundated flats on the plateau

1 .2.6 permanently moist seepage steps demarcating the terraced flats of the plateau

FYNBOS

Coast Fynbos

(Coast-shelf Fynbos)

Dune Mixed Fynbos Degrading to Dry Sandflats and Metalasia-Passerina Communities

(Coast Marsh Sedgeland)

Rocky Coast Fynbos (including Coleonema Fynbos)

Inland Fynbos

Upland Mixed Fynbos including Western Hillveld, Dry Hillveld, Mountain Fynbos and

Metalasia-Passerina Fynbos; degrading to Dry Sandflats Community

Tall Fynbos (including Protea lepidocarpodendron Community)

(Protea nitida Woodland)

Plateau Fynbos

Tussock Marsh Community

Seepage Fynbos

2. Sites protected from recurrent fires, occurring on

2.1 dunes of marine origin

2.2 slopes and screes of sandstone origin

BROAD-LEAVED THICKET

Sideroxylon Thicket

Maurocenia- Ch ion an thus Thicket

* Vegetation units in parentheses are not mentioned in this text

247 -248

ASSOCIATION-ANALYSIS HIERARCHY : CAPE OF GOOD HOPE NATURE RESERVE

452 species

100 plots

80

394

47

394

53

40

c*e

- 30

-20

349

16

in -

74

41

O

LEGEND

Species code number (with bar = species absent)

Number of sample plots in subdivision (=N)

Groups, final (e.g. 21) or recombined (e.g. a, A or I)

- 60

349

31

cn 2

3 ■£

■5 -

166

4

166

27

-10

132

10

--©-

132

6

c- V)

oj o

3

rtl X

339

6

339

21

74

12

74

41

Subdivision terminated at

X2= 3,84 (P = 0,05)

Group of seven plots or less

automatically declared final

Criterion of subdivision

"nT without Yates' correction

fJ neri

Where subdivision occurs on

more than one species the

ambiguities are indicated by

333

20

333

21

— / b

232

3

— — — — -4 F

361

7

232

18

361

11

■- => —

I — U u

(B ^

•55 -2 ™

™ 2 w ■£

— —(9

E G

™ £

409

_2_

•1= e

o .2

409

9

130

4

12

4

12

16

130

5

253

4

— i J

277

3

277

1—131

147

4

305

5

15

305

16

■SI 144

5

144

11

a >-l_r_J 1 ■

Gh ~ <°> ~ 7 5

n» o

U) X

O UJ

o 4.

108

3

108

8

115

3

115

5

ft) (i) (•) d) (2) (*) (b ~(b ~(b & & & &&&&&(*) ® ® 6 ® ©

pIG 2. — Association-analysis hierarchy of the Cape of Good Hope Nature Reserve.

H. C. TAYLOR

249

Cain & Castro, 1959; Greig-Smith, 1964; Kershaw,

1964) it was decided to use a rectangular plot of 5 m

x 10 m as being the most efficient size. The

examples given by Werger (1972) confirm that a plot

of 50 m2 is usually adequate for sampling fynbos.

Therefore, the sample plots on the Cape of Good

Hope Nature Reserve were placed systematically at

intersections of the grids on the 1 : 18 000 topogra-

phical map, at 1 000 yd (914 m) intervals. One plot,

in the middle of an extensive thicket of the alien

Acacia cyclops, was discarded. Three specially

chosen summit samples were added, giving 100 plots

for the Reserve.

After locating the plot positions in the field by

pacing distances from fixed points scaled on the

map, each plot was laid out with the long side facing

true north-south.

Species lists were made for each sample plot.

Since annuals and geophytes appearing temporarily

at different seasons would obscure differences

between species lists made at different times of year,

lists for the computer analysis included only those

species with permanently recognizable aerial parts.

Plants appearing seasonally were noted for descrip-

tive purposes. A supplementary list was made of

additional species characteristic of the community

but occurring only outside the plot. These ‘surround

lists’ facilitated subsequent interpretation of the plot

data, but were not used in the association-analysis.

Whereas plants were positively identified to

species wherever possible, some species-complexes

had to be amalgamated, either because they could

not be specifically identified in the field or because

their taxonomy had been inadequately worked out.

Examples are certain minor genera of the Ericaceae

and some members of the Mesembryanthemaceae.

Alien species, the subject of a separate study, were

excluded from the association-analysis. An assess-

ment of abundance using Acocks’s (1975) symbols

was made for each species on each plot. This was

another valuable aid in interpreting the data, though

not needed for the association-analysis.

Post-burn ‘age’, based as far as possible on the

branching of Proteaceous shrubs described by Hall

(1959), was recorded, and descriptive notes on

dominance (if any), prevalent life-form, average

height and crown cover, and evident successional

trends, were made on each plot to help to identify

the groups revealed by association-analysis.

The following habitat data were recorded at each

sampling site: altitude, aspect and degree of slope;

percentage cover of stones of various size classes on

a five-point scale; and soil moisture also estimated

on a five-point scale (Taylor, 1969). Notes were

made on soil texture, soil colour, amount of humus,

and geological formation. A black-and-white photo-

graph was taken of the vegetation of each plot.

The field enumeration of the 100 samples was

done in 45 days during the period January to

November 1966. I was accompanied in the field by

one assistant who laid out the plots and helped with

the collection of environmental data.

The association-analysis method

The use of interspecific associations for classifying

vegetation was pioneered by Goodall (1953) who

examined presence-or-absence data from random

quadrats. Based on his work, Williams & Lambert

(1959 & 1960) developed the methods of ‘normal

association-analysis’, ‘inverse analysis’ (Williams &

Lambert, 1961) and ‘nodal analysis’ (Lambert &

Williams, 1962). Differences between association-

analysis and Goodall’s method are discussed by

Taylor (1969). By 1966, association-analysis had

been used to classify vegetation in South Africa by

Van der Walt (1962), Grunow (1965), Roberts

(1966), Downing (1966) and Miller (1966). All these

were done in the floristically poorer karroid

grassveld and bushveld vegetation.

Fynbos, by contrast, has a much richer flora

occurring in rugged terrain with many intergrading

habitats — conditions that were thought to differ

sufficiently to justify testing the method for its

applicability to this particular vegetation type.

Because time and suitable modifications to

computer programmes were lacking, the present

study was confined to normal analysis. The criterion

used for subdivision was the highest aggregate

which Williams & Lambert (1959; 1960) considered

to result in the most efficient subdivision. Subdivi-

sion continued until no individual x 2 value

equalled or exceeded 3,84 (corresponding to

ps£0.05). The relevant class was then designated a

‘final group’. Since a group must exceed seven

quadrats before this value can be attained, any

group containing less then eight quadrats was

automatically declared a final group. Where

subdivision occurred on more than one species, the

species occurring first in alphabetical sequence was

used (though the ambiguities, prefixed by +, are

also shown in Fig. 2).

Out of the total of 452 species that occurred on the

plots, the highest number per plot was 83, the lowest

5, and the average 38, 26. Computation was done in

1968 on an ICT 1900 computer using the Fortran

programme for association-analysis (A ANAL).

The Braun- Blanquet method

Association-analysis is a monothetic, divisive

method in which successive subdivisions of the

sample set take place on the presence or absence of a

single species each time. The Braun-Blanquet

method, in contrast, is a polythetic agglomerative

method, using the joint occurrence of more than one

species to characterize each community. Because of

this basic difference in the treatment of data, the

Braun-Blanquet method commended itself as an

independent test of the relative merits and demerits

of association-analysis as compared with the

Braun-Blanquet method. The original data were

therefore re-examined to determine whether a

community classification that showed better correla-

tion with environmental factors could be obtained.

The Braun-Blanquet method had been successful-

ly used by Continental phytosociologists for over

half a century but had in the past been poorly

250 A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE. I. THE USE OF

ASSOCIATION-ANALYSIS AND

received by workers not trained in Europe. This was

largely due to the fact that earlier published work

contained no precise English description of the

actual method of synthesis which was consequently

misunderstood and severely critized. Braun-

Blanquet’s techniques have since been revised and

comprehensively described in English, amongst

others by Becking (1957), Kiichler (1967), Mueller-

Dombois & Ellenberg (1974), and Werger (1974).

After the results of the present study showed the

success of the method, it has been increasingly

applied in South Africa in diverse types of

vegetation including Transvaal woodlands and

grasslands (Coetzee, 1974a, 1975; Coetzee et al.,

1976; Van der Meulen, 1979; Bredenkamp &

Theron, 1978, 1980), forests and bogs of the

Drakensberg (Van Zinderen Bakker Jr, 1973; Van

Zinderen Bakker Sr & Werger, 1974), karroid

vegetation and grassland in the Orange Free State

and northern Cape (Werger, 1973; Leistner &

Werger, 1973; Werger & Coetzee, 1977; Werger,

1980), aquatic vegetation in Zululand (Musil,

Grunow & Bornman, 1973) and in Cape fynbos

(Werger, Kruger & Taylor, 1972; Boucher &

Jarman, 1977; McKenzie, Moll & Campbell, 1977;

Van der Merwe, 1977; Boucher, 1978; Glyphis, Moll

& Campbell, 1978; Laidler et al., 1978).

The application of the Braun-Blainquet technique

in the present study deviated from standard practice.

In the first place, from the phytosociologist’s

viewpoint the original data were faulty because

sampling was systematic so that transitions between

communities were also sampled. The data were also

incomplete, because cover-abundance estimates for

the species in each plot were not available; the coded

data, assembled for association-analysis, were given

only in the form of presence-or-absence. In addition,

all species were not included in the tables. To do so,

when working in such a rich flora, was beyond the

scope of the trial: it would have greatly increased the

complexity of the tables, and at that time no

mechanical sorting apparatus (cf. Muller et al. , 1972)

and no computer programmes to assist in sorting

tabulated data were available. Therefore, the

construction of the tables also differed from normal

practice.

Instead of a single ‘raw table’ that included all

plots, each plot was first assigned to the most

appropriate of three broad habitat/vegetation types

that had been erected from knowledge and

experience gained in the original survey. Doubtful

plots were placed in more than one type to establish

where they fitted best. Species to differentiate each

type were chosen partly with the help of a Species

Constancy Table (Taylor, 1969) which showed the

percentage occurrence of each of the 452 species in

each of the 23 association-analysis groups. All

positive dividing species of the plots selected, and

other species that experience suggested might be of

character value, were also included. Surround

occurrences of the species selected, were included in

the tables. The three raw tables were then

manipulated in the usual way until ‘blocks’ of plots,

or noda, represented by differential species,

emerged. The plots within these noda were

BRAUN-BLANQUET METHODS

subjected to a Peculiarity Index test (Hall, 1965;

Taylor, 1969) to determine whether they were

appropriately placed. A Sneath’s Modulus test

(Sneath, 1962; Hall, 1968; Taylor, 1969) and a

Fidelity Factor developed by Taylor (1969) were

used to determine which species were most

characteristic of the units already distinguished, and

thereby to assess the efficiency of the species already

chosen. The improved arrangements resulting from

these tests were depicted in the three Final Tables

(see Table 3).

RESULTS AND DISCUSSION

Association-analysis

The analysis resulted in a classification containing

23 final groups of sample plots (Fig. 2). Because of

the low sampling density, all final groups contained

seven or less plots; subdivision was therefore

automatically terminated before a common level of

heterogeneity was reached. The mapped final

groups (Taylor, 1969, Map 2) did not yield entities

that fitted the pattern of environmental variation

satisfactorily. The final groups were therefore

recombined and mapped at successively higher

levels of association.

In some association-analysis studies in South

African vegetation (e.g. Grunow, 1965; Downing,

1966) mapping of the first division made an

ecologically interpretable pattern by separating the

wetter from the drier sites. In the present study this

was not the case. The first ten groups characterized

by presence of Struthiola ciliata, for instance,

contained not only the Inland Fynbos of rocky hills

and mountains (category 1.2.1 of Table 1) but also

the communities of the moist flats of the central

plateau (Group 5) and communities transitional to

Coastal Fynbos found on wind-blown sand (Group

4). In the same way, the major right-hand leg of the

hierarchy shown in Fig. 2 characterized by absence

of Struthiola ciliata, contained the remainder of hill

and mountain vegetation, and the major portion of

the moist flats and coastal communities. Therefore,

no fusion of groups based solely on level of

association could give a satisfactory interpretation of

the hierarchy. The same phenomenon, that several

closely related communities were widely separated

and consequently obscured in the association-

analysis hierarchy, was reported by Coetzee (1974b)

for the Central Bankenveld of the Transvaal. In a

recent comparable study of kwongan vegetation, an

Australian counterpart of fynbos, Hnatiuk &

Hopkins (1981) tried but soon discarded a monothe-

tic divisive classification of site data, because it

produced very heterogeneous results.

To interpret the association-analysis hierarchy,

the composition of each final group was described

and evaluated in detail and the habitat characteris-

tics of each plot in each group tabulated in a

summary of plot characteristics (Taylor, 1969). The

variation in ecological and floristic content within

each group was then summarized (Table 2) and the

homogeneity of groups assessed both subjectively

and by means of a Homogeneity Index (Taylor,

1969). The final groups did not, in general, combine

TABLE 2. — Summary of group characteristics

251-252

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

1

4

3

20

11

10

15

2

1

12+

3

4

15

3

2

3

Over

3

10

10+

24

20+

21 +

14

21

8

20

25

18

20

25

18

20+

30+

30

19

20+

2

0

3

1

0

3

0

3

0

5

0

3

5

4

0

3

4

5

4

5

3

5

0

5

0

3

0

3

0

3

0

40

7

0

5

0

30

0

80

0

15

49

60

0

15

40

43

60

77

37

6

0

5

0

6

0

55

0

20

1

1-3

1- 5

1

1,2

1

1-2

1.3

2- 4

1,2

3.4

1

1,2

1

3,4

1-5

T

T+M

T

T+M

T

M

T

T+M

6

4

6

3

7

2

4

5

4

3

5

4

3

4

2

7

5

3

159

107

88

108

60

70

79

135

99

92

76

65

46

61

99

70

37

40

47

23

39

32

39

31

35

33

26

46

44

15

34

27

24

28

18

24

14

7

14

75

48

57

49

51

36

42

41

65

60

36

42

45

34

35

24

50

36

24

22

13

57

45

42

45

46

35

37

36

53

51

26

38

34

30

32

21

32

24

16

19

11

52

9

25

10

20

1

9

15

19

16

21

8

18

10

7

6

26

22

17

8

6

9

16

15

19

18

19

21

5

13

3

7

8

11

37

3

8

11

8

5

Low

Med.

High

Low

High

Low

Med.

Low

LoW'

Med.

High

Med.

High

Med.

Low

Tussock Marsh elements.

Low Dry hill slopes and ridges with outcropping bedrock, sometimes iron-

impregnated; depauperate or atypical Western Hillveld or Tall Fynbos.

Med. Patches of deep sand on coastal shelf or inland; Dry Sandflats Community

with added coastal elements.

High Poorly drained flats with Plateau Fynbos; plot 79, on a cool, moist slope,

bears Restionaceous Upland Mixed Fynbos.

High High, cool, eastern slopes, mostly with iron-impregnated bedrock; long-

unburnt Mountain Fynbos with Dry Hillveld elements.

High Warm, dry, rocky hills bearing homogeneous Dry Hillveld.

Med. Heterogeneous dry-habitat ecotones; transitional or depauperate Tall Fynbos

or Upland Mixed Fynbos.

Dry, rocky ridges or sheetrock; depauperate Upland Mixed Fynbos.

Dry, mainly rocky ridges with coarse, white sand; moribund (over-mature)

Plateau Fynbos or Upland Mixed Fynbos.

Either rocky hills with depauperate Western Hillveld; or sandy flats or dunes

with Dune Mixed Fynbos.

Cool boulder-slopes or summits; Mountain Fynbos with Broad-leaved

Thicket elements where protected from fire.

Heterogeneous: rocky summits and slopes; deep sandy plateau; calcrete

dune; vegetation correspondingly varied.

Gentle sandy slopes with Restionaceous Upland Mixed Fynbos; except for

plot 5 1 on rock outcrop with depauperate Dry Hillveld.

Sandy coastal plain with Dry Sandflats Community, except for plot 14 on

bedrock with atypical Tussock Marsh/Upland Mixed Fynbos transition.

Moist flats with Restionaceous Tussock Marsh, bushy in plot 22.

Boundary between rocky hills and moist flats; vegetation containing

elements of both Upland Mixed Fynbos and Tussock Marsh.

Moist flats with simple or complex Tussock Marsh; or seepage steps with

Seepage Fynbos.

Inland dunes, sandy slopes or coastal terraces; Dune Mixed Fynbos or

Metalasia-Passerina Fynbos.

Stabilized littoral dunes with woody Sideroxylon Thicket.

Rocky littoral with semi-succulent variety of the Rocky Coast Fynbos.

Moist flats with simple or complex Tussock Marsh; or seepage step with

Seepage Fynbos.

Low Completely heterogeneous as to both habitat and vegetation.

High

Med.

Low-

High

Low

Med.

High

Med.

High

Med.

Symbols and abbreviations

* exceptional values

Columns 4, 5 and 8: For explanation of estimate-symbols, see Taylor (1969)

Column 9: T = Table Mountain Sandstone soils; M = transported sands of marine origin

Column 16: For explanation of Homogeneity Index values, see Taylor (1969)

Column 17: Ecological assessment

Column 18: Floristic assessment

250 A VEGE

received by w<

largely due tc

contained no

actual methoc

misunderstooi

Blanquet’s tec

comprehensive

others by Bee

Dombois & E

After the rest

success of th

applied in f

vegetation in

grasslands (C

1976; Van d

Theron, 1976

Drakensberg i

Zinderen Bal

vegetation an<

and northern

Werger, 1973

1980), aquati

Grunow & B

(Werger, Kri

Jarman, 1977;

Van der Merw

& Campbell,

The applicai

in the present :

In the first

viewpoint the

sampling was s

communities v

incomplete, be

the species in e

data, assemble

only in the forr

all species wer

when working

scope of the tri

complexity of

mechanical soi

and no compi

tabulated dat

construction o

practice.

Instead of £

plots, each p

appropriate of

that had be*

experience gai

plots were plat

where they fitt

type were cho

Constancy Tal

percentage occurrence of each of the 452 species in

each of the 23 association-analysis groups. All

positive dividing species of the plots selected, and

other species that experience suggested might be of

character value, were also included. Surround

occurrences of the species selected, were included in

the tables. The three raw tables were then

manipulated in the usual way until ‘blocks’ of plots,

or noda, represented by differential species,

emerged. The plots within these noda were

To interpret the association-analysis hierarchy,

the composition of each final group was described

and evaluated in detail and the habitat characteris-

tics of each plot in each group tabulated in a

summary of plot characteristics (Taylor, 1969). The

variation in ecological and floristic content within

each group was then summarized (Table 2) and the

homogeneity of groups assessed both subjectively

and by means of a Homogeneity Index (Taylor,

1969). The final groups did not, in general, combine

H. C. TAYLOR

253

into a classification that accorded closely with the

ecological patterns in the Reserve. Only groups 1,7,

16, 20 and 21 showed high floristic and ecological

homogeneity. Groups 2, 4, 6, 10, 12, 18, 19 and 22

contained some anomalous misplaced plots, whereas

the remaining ten groups represented small, isolated

fragments of natural units. All groups except the five

first mentioned could not be readily interpreted in

terms of habitat factors because each group included

samples of vegetation with different ecological

characteristics. In order to obtain a classification

that accorded with habitat, these final groups could

not be combined by simply raising the level of

heterogeneity; instead, their constituent plots had to

be regrouped selectively on extrinsic characters of

the habitat. This introduced a subjective element.

Fynbos has long been considered a vegetation

having few discernible clear-cut communities.

Although some fynbos species do show great

ecological tolerance, most species have an optimum

habitat in which they are more common than in

others. If a finer measure of abundance had been

used, a closer correspondence with habitat might

have been obtained. In association-analysis, how-

ever, the only record of abundance is the crude

‘present’ or ‘absent’ rating, resulting in great loss of

information. When fairly large sample plots are

used, as in the present study, a species is likely to be

recorded as present irrespective of whether it is

abundant or occasional, so the optimum habitat of

tolerant species will not be revealed by this

technique. Many of the dividing species responsible

for forming the hierarchy in this study appeared to

be tolerant or wide-ranging species. Since this could

explain the scattering of ecological units among the

final groups, the autecological characteristics of the

dividing species were closely examined (Taylor,

1969). It was found that, of the 32 dividing species of

the hierarchy, only 11 were restricted to specialized

habitats. The remaining 21 possessed wide ecologi-

cal amplitudes, therefore having little or no

character value for natural vegetation units. Divi-

ding species, by definition, effect the most efficient

subdivision of sets of plots. In the present study,

those subsets that do not represent recognizable

plant communities, may reflect the response of the

vegetation to factors such as fire or grazing. (All

plots in Group 2, for instance, had been burnt less

than six years before the survey.) Analysis of

variance and other tests failed, however, to reveal a

simple correlation.

In a monothetic, divisive method like association-

analysis, the final groups are determined solely by

presence or absence of a small set of dividing

species. No account is taken of the finer degrees of

abundance of different species, which might

constitute important differences between stands; nor

of other species that might help to characterize the

groups.

If natural vegetation units were discrete, with no

transitions between them, the final groups of an

association-analysis should theoretically reflect these

natural units exactly. If transitions occur, they will

be composed of tolerant species common to the

intergrading units. Where tolerant species act as

dividing species in an association-analysis, transi-

tional plots will fall into different groups in which a

dividing species is present in the one group and

absent from the other. If presence-or-absence of

these tolerant dividing species differentiates be-

tween heterogeneous subsets at a high level of

the transitional groups will be widely separated in

the hierarchy. The greater the degree of intergrada-

tion between vegetation units and the greater the

number of tolerant dividing species (especially at

high levels of y* ) the greater will be the

fragmentation of legitimate communities into unin-

terpretable groups situated far apart in the

hierarchy.

This would seem to explain some anomalies of the

present association-analysis hierarchy. Fynbos com-

munities in general intergrade freely. Many of the

dividing species on the Cape of Good Hope Nature

Reserve were shown to have wide ecological

amplitude; many of the final groups include

transitions and anomalous heterogeneous variations

of natural units widely separated in the hierarchy.

However, this does not mean that natural units do

not exist. There are indeed natural units characte-

rized by assemblages of ‘faithful’ species (as shown

later) but if the ‘faithful’ species are not dividing

species, the natural units will not be revealed in an

association-analysis.

Anomalous groups may arise not only from

transitions between natural units but also from

variations within them, provided the dividing species

for these groups are sensitive to the factors causing

the variations. It was shown, for instance, that

Leucadendron xanthoconus, a wide-ranging dividing

species, seemed to be a ‘stray’ in Plot 67 rather than

a true indicator of its habitat, thus placing Plot 67 in

the ‘wrong’ group. Again, Struthiola ciliata, Roella

ciliata and others are sensitive to fire, being

abundant in recently burnt stands but only

occasional in long-unburnt stands. Such species will

occur in all plots in recently burnt stands, but will be

absent from some plots in unburnt stands; the latter

plots will be relegated to a second group, often

widely separated from the first, although both

groups represent basically the same community. In

addition, the second group may contain plots

representing transitions between this community

and another in which the dividing species are missing

for reasons other than fire, so that this second group

will comprise a mixture of natural units.

On examining the manner of regeneration of

dividing species it appeared that many of those

which regenerate from seed, e.g. the three

mentioned in the preceding paragraph, were poor

character species for the groups which they

differentiate, whereas most of those which regene-

rate from coppice or sprouts were good character

species for the groups which they differentiate.

Seed-regenerating species are more sensitive to fire

history of the vegetation than sprouting plants, being

more abundant or rarer in often-burnt veld

depending on their regeneration behaviour and the

frequency of fires. Abundance of sprouting plants is

much less affected by fire. Other factors that could

254 A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE. I THE USE OF

ASSOCIATION-ANALYSIS AND BRAUN-BLANQUET METHODS

affect the abundance of dividing species, and

therefore the grouping of plots, without radically

altering the community, might be the intensity and

rotation of grazing, the depredations of baboons, the

influence of man, and local changes in habitat

occasioned by biotic factors or small climatic

variations, e.g. blown sand or desiccation of

seepages.

Another cause of misclassified plots is the human

factor, i.e. the accuracy of the field worker in

recording data. This factor is independent of the

method of data capture, but where mere presence or

absence of a single species is of prime importance, its

effect on the results will be much more serious than

where several degrees of abundance, or alternatively

a group of species, are used to characterize a

vegetation unit. The effect is enhanced where, as at

Cape Point, some of the key species — the dividing

species in association-analysis — are inconspicuous

or difficult to identify at certain stages or seasons.

Normally, the human factor is not regarded as a

serious source of error if the field worker is trained

and observant. However, the rich fynbos flora

contains many closely related plants often difficult to

distinguish in the field, and the systematic positions

of some taxa are still uncertain. The danger of

inaccuracies resulting from misidentifications will

therefore be greater than in simpler, better known

floras.

Braun- Blanquet

The three Final Tables represent three broad

vegetation types that can be identified in the original

habitat classification (Table 1) as follows: (1) the

Coast Fynbos and Broad-leaved Scrub (categories

1.1 and 2); (2) the Inland Fynbos of the moist flats

(categories 1.2.5 and 1.2.6); and (3) the Inland

Fynbos of well-drained slopes (categories 1.2.1 and

1.2.2). These tables (one of which, Table 3, is

reproduced here as an example) show the noda or

communities that were synthesized objectively and

emerged naturally from the data matrix by the

Braun-Blanquet method of manipulation. The

objectivity of the method is borne out by an

example: some refinements to Table 3 were effected

by a colleague (M.J.A. Werger) who at that time

had never seen the vegetation or the habitat, and

this resulted in the subdivision of Community 2 into

two floristically defined sub-communities that

showed clear differences in structural and habitat

features.

Each community, floristically defined by a group

of differential species, could be satisfactorily

interpreted in terms of habitat, as shown in Table

3 by the reference to the original habitat classifica-

tion. There are, of course, transitional situations in

any classification of vegetation where habitats

intergrade, and this effect is enhanced where

sampling has been rigidly systematic. Despite this,

the tables revealed the existence of reasonably

discrete communities and the floristic relationships

between them.

On the other hand, the tables showed that most of

the communities comprised a miscellany of

association-analysis final groups, and conversely,

that most final groups were scattered among a

number of communities (see Table 3). Thus, either

the association-analysis groups or the Braun-

Blanquet communities do not accurately reflect the

ecological patterns on the Reserve. Detailed

examination and statistical tests have shown many of

the association-analysis groups to be heterogeneous

in terms of habitat — and indeed in floristics also,

except for the dividing species. The Braun-Blanquet

communities were found to be more homogeneous

in terms of previously defined habitat groupings, and

showed floristic relationships consistent with these

groupings.

CONCLUSIONS

The advantage of association-analysis is that it

produces a rapid, objective classification of units

that can be mapped. However, in a monothetic,

divisive system some plots are bound to be

misplaced and this effect is enhanced if the

occurrence of the dividing species depends on

factors other than the habitat, as often appears to be

the case in this study. Therefore, if the classification

cannot be related to habitat without lengthy and

complicated re-interpretation by someone with a

detailed knowledge of the ecology of the area, the

prime advantage of association-analysis is nullified

and the map based upon it will be equally subjective.

Indeed, if the original, crude ‘block map’ is

transformed merely by rounding the outlines, the

position of boundaries — especially where distances

between points are large as in the present case —

will be much less accurate than where the map is

redrawn from an orthodox descriptive survey. On

the other hand, if boundaries are interpolated by

field re-examination, as much ‘subjectivity’ is

employed as in orthodox mapping.

In association-analysis, where the path of subdivi-

sion depends solely upon the presence or absence of

a single dividing species at each stage, the size,

location and number of samples are crucial factors in

determining the classification that results. In some

instances, the plot size used in this survey appears

too large to reflect a single vegetation unit; in others,

it may be too small to represent the stand adequately

where the flora is rich, the vegetation pattern coarse

and the plot density low. The systematic location of

plots makes for the inclusion of many transitional

stands and, with the sampling technique used, the

number of plots (representing 0,006% of the area of

the Reserve) is too small to provide a picture of

communities that can be accurately mapped. If

stratified by major landscape features, the use of 100

plots would have been more efficient while

subjectivity would have been minimized.

The present association-analysis is unusual in that

the number of attributes (species) exceeds the

number of individuals (samples). This has two

disadvantages. Firstly, the effect of misplacement of

a few plots owing to chance or to ‘degree of absence’

(absence in the plot through reduced abundance in

the stand) will be relatively great. Secondly, because

of the low number of constituent plots, final groups

are terminated not by the limiting level of

TABLE 3. — Final synthesis table for moist flats

255-256

17

46

O

X

o

X

o

X

X o

X

o

X X

X

X X

O X

X X

o

X

X X

X

o o

X

o

X

o

X

X X

X

X X

X

X X

X

o

X X

o

O X

X X

o

X X

o

► (1)

Seepage

Fynbos

(2)

Tussock Marsh

Community

(2a)

(2b)

* Companion

Species

X present within plot

O present only in plot surround

H. C. TAYLOR

257

heterogeneity (usually highest single x 2 ^3,84) but

by the limiting number of seven plots. Their

individual x 2 values therefore equal or exceed

3,84 but have not been determined. Consequently,

while the final groups are not equally heterogene-

ous, their variation is not known; some might be

capable of many further subdivisions, others of

none, before reaching the limiting level. These

drawbacks could to some extent be relieved by

adopting the techniques of Crawford & Wishart

(1968), who used a rapid agglomerative method

after the initial divisive process, to check for any

misclassifications.

The authors of association-analysis claim it to be a

very useful tool in primary survey (Williams &

Lambert, 1959). The aim of primary survey is not to

elucidate minor factors causing community variation

but to expose major habitat discontinuities reflected

in major vegetational changes. It appears that

association-analysis, as applied in this survey, makes

a premature attempt at the former without

accomplishing the latter. It has, perforce, produced

a detailed and clear-cut classification which does not

represent nature.

Association-analysis has two major drawbacks:

the nature of the method, involving a rigid series of

successive subdivisions on a single species each time,

and the location of the plots on an equally rigid grid

pattern. The adverse effect of misclassifications,

poor sampling strategy and other drawbacks

discussed above could be minimized by using an

agglomerative method and a stratified location of

plots.

The Braun-Blanquet method is agglomerative and

polythetic. It starts from individual plots and

combines units that are most similar, fusing groups

at successively higher levels of the parameter used.

In this way, a picture is obtained of the degree of

similarity rather than the amount of difference

between units. Moreover, as pointed out by Coetzee

& Werger (1975), agglomerative classifications are

regarded as more stable and as having higher

extrapolative value when applied to regions outside

the study area, than divisive classifications.

In the present study, when the Braun-Blanquet

method was used in an independent assessment of

the association-analysis, despite the drawbacks

inherent in the data, the communities were found to

be better correlated with habitats than either the

final or recombined association-analysis groups.

One must conclude that, with the type of sampling

used, the synthetic phytosociological Braun-

Blanquet method provides a more natural classifica-

tion of communities of the Cape of Good Hope

Nature Reserve than does the monothetic, divisive

association-analysis method.

ACKNOWLEDGEMENTS

My grateful thanks are due to the following:

various student assistants, especially Mr S. Birch, for

help and companionship during field work in 1966 -

1967; the Divisional Council of the Cape for

providing accommodation on the Reserve; the

Department of Agriculture for permission to use this

study for an M.Sc. thesis (1969); my supervisors and

many colleagues, especially Dr D. Edwards, for

discussion and advice during preparation of the

thesis; Dr J. C. Scheepers for helping with the

preparation of this text for the press, and Miss E. N.

Pare for typing.

UITTREKSEL

Met die opname is gepoog om bree plantegroei-

eenhede wat op ’n ekstensiewe skaal karteerbaar is, in

die Kaap die Goeie Hoop Natuurreservaat op die

suidelike punt van die Kaapse Skiereiland, tot stand te

bring. Hierdie referaat vergelyk die doeltreffendheid

van twee metodes, naamlik assosiasie-analise en die

Braun-Blanquet metode soos deur die Ziirich-

Montpellier denkwyse van fitososiologie ontwerp.

Eenhonderd 50 m2 proefpersele wat die hele

Reservaat bedek, was sistematies op ruitnet kruispun-

te op die 1: 18 000 topografiese kaart met intervalle

van 1 000 tree (914 m) geplaas. Spesieslyste, waar die

aanwesigheid van alle spesies met permanent herken-

bare bogrondse dele in ag geneem is, was vir elke

proefperseel aangeteken.

Drie-en-twintig finale groepe proefpersele waarvan

net vyftekens van floristiese en ekologiese homogeni-

teit getoon het, was vanuit die assosiasie-analise

klassifikasie gekry. Van die oorblywende proefperse-

le is daar agt groepe met enkele onreelmatige persele

en tien groepe wat klein geisoleerde brokstukke van

natuurlike eenhede verteenwoordig. Die oorspronk-

like data was geanaliseer volgens die Braun-Blanquet

metode om ’n onafhanklike klassifikasie daar te stel

wat met die assosiasie-analise metode, vergelyk kan

word. Die Braun-Blanquet gemeenskappe was meer

homogeen wat vooraf gedefineerde omgewingsgroepe

betref en het floristiese verwantskappe wat verenig-

baar is met die habitatsgroeperings getoon.

Die gevolgtrekking is dat die sintetiese fitososiolo-

giese Braun-Blanquet metode gee ’n meer natuurlike

klassifikasie van plantgemeenskappe van die Reser-

vaat as die monotetiese verdelings assosiasie-analise

metode met die tipe monsterneming in ag geneem.

REFERENCES

Acocks, J. P. H., 1975. Veld types of South Africa. Mem. bot.

Surv. S. Afr. 40.

Adamson, R. S., 1938. The vegetation of South Africa. London:

British Empire Vegetation Committee.

Adamson, R. S. & Salter, T. M., 1950. Flora of the Cape

Peninsula. Cape Town: Juta.

Ashby, E., 1936. Statistical ecology. Bot. Rev. 2: 221-235.

Becking, R. W., 1957. The Zurich-Montpellier School of

Phytosociology. Bot. Rev. 23:411—488.

Blackman, G. E., 1935. A study by statistical methods of the

distribution of species in grassland associations. Ann. Bot.

49: 749-777.

Boucher, C., 1977. Cape Hangklip area. I. The application of

association-analysis, homogeneity functions and Braun-

Blanquet techniques in the description of south-western

Cape vegetation. Bothalia 12: 293-300.

Boucher, C., 1978. Cape Hangklip area. II. The vegetation.

Bothalia 12: 455-497.

Boucher, C. & Jarman, M. L., 1977. The vegetation of the

Langebaan area, South Africa. Trans. R. Soc. S. Afr. 42,3:

241-272.

258 A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE. I. THE USE OF

ASSOCIATION-ANALYSIS AND BRAUN-BLANQUET METHODS

Braun-Blanquet, J., 1932. (Transl. G. D. Fuller and H. S.

Conard). Plant sociology. New York - London: McGraw

Hill.

Bredenkamp, G. J. & Theron, G. K., 1978. A synecological

account of the Suikerbosrand Nature Reserve. I. The

phytosociology of the Witwatersrand geological system.

Bothalia 12: 513-529.

Bredenkamp, G. J. & Theron, G. K., 1980. A synecological

account of the Suikerbosrand Nature Reserve. II. The

phytosociology of the Ventersdorp Geological System.

Bothalia 13: 199—216.

Brown, D., 1954. Methods of surveying and measuring

vegetation. Bull. 42, Commonw. Agric. Bur.

Cain, S. A. & Castro. G. M. de O., 1959. Manual of vegetation

analysis. New York: Harper.

Coetzee, B. J., 1974a. A phytosociological classification of the

vegetation of the Jack Scott Nature Reserve. Bothalia 11:

329-347.

Coetzee, B. J., 1974b. Improvement of association-analysis

classification by Braun-Blanquet technique. Bothalia 11:

365-367.

Coetzee. B. J., 1975. A phytosociological classification of the

Rustenburg Nature Reserve. Bothalia 11: 561-580.

Coetzee, B. J. & Werger, M. J. A., 1975. On association-

analysis and the classification of plant communities.

Vegetatio 30,3: 201-206.

Coetzee, B. J., Van der Meulen, F., Zwanziger, S.,

Gonsalves, P. & Weisser. P., 1976. A phytosociological

classification of the Nylsvlei Nature Reserve. Bothalia 12:

137-160.

Crawford, R. M. M. & Wishart, D., 1968. A rapid

classification and ordination method and its application to

vegetation mapping. J. Ecol. 56: 385-404.

Downing, B. H., 1966. The plant ecology of Tabamhlope Vlei,

Natal. M.Sc. thesis. University of Natal.

Glyphis, J., Moll, E. J. & Campbell, B. M., 1978.

Phytosociological studies on Table Mountain, South Africa:

1. The Back Table. Jl S. Afr. Bot. 44,3: 281-289.

Goodall, D. W., 1953. Objective methods for the classification

of vegetation. I. The use of positive interspecific correlation.

Aust. J. Bot. 1: 39-63.

Greig-Smith, P., 1964. Quantitative plant ecology. London:

Butterworths.

Grunow, J. O., 1965. Objective classification of plant communi-

ties: a synecological study in the sour-mixed bushveld of

Transvaal. D.Sc. (Agric.) thesis, University of Pretoria.

Hall, A. V., 1959. Observations on the distribution and ecology

of Orchidaceae in the Muizenberg Mountains, Cape

Peninsula. Jl S. Afr. Bot. 25: 265-278.

Hall, A. V., 1965. The Peculiarity Index, a new function for use

in numerical taxonomy. Nature 206, 4987: 952.

Hall, A. V., 1968. Methods for showing distinctness and aiding

identification of critical groups in taxonomy and ecology.

Nature 218, 5137: 203-204.

Hnatiuk, R. J. & Hopkins. A. J. M., 1981. An ecological analysis

of kwongan vegetation south of Enneabba, Western

Australia. Aust. 7. Ecol. 6, 423—438.

Kershaw, K. A., 1964. Quantitative and dynamic ecology.

London: Arnold.

Kuchler, A. W., 1967. Vegetation mapping. New York: Ronald

Press.

Laidler, D., Moll, E. J., Campbell. B. M. & Glyphis, J., 1978.

Phytosociological studies on Table Mountain, South Africa:

2. The Front Table. Jl S. Afr. Bot. 44,3: 291-295.

Lambert, J. M. & Williams, W. T., 1962. Multivariate methods

in plant ecology. IV. Nodal analysis. 7. Ecol. 50: 775-802.

Leistner, O. A. & Werger, M. J. A., 1973. Southern Kalahari

phytosociology. Vegetatio 28, 5-6: 353-399.

Marloth, R., 1908. Das Kapland. Jena: Fischer.

McKenzie. B., Moll. E. J. & Campbell, B. M., 1977. A

phytosociological study of Orange Kloof, Table Mountain,

South Africa. Vegetatio 34: 41—53.

Miller, P. . 1966. An objective survey of plant communities in four

catchments in the Highland Sourveld. M.Sc. (Agric.) thesis,

University of Natal.

Mueller-Dombois, D. & Ellenberg, H., 1974. Aims and

methods of vegetation ecology. New York: Wiley.

Muller, P. J., Werger, M. J. A., Coetzee, B. J., Edwards, D.

& Jarman, N. G., 1972. An apparatus for facilitating the

manual tabulation of phytosociological data. Bothalia 10:

579-581.

Musil, C. F., Grunow, J. O. & Borman, C. H., 1973.

Classification and ordination of aquatic macrophytes in the

Pongola River pans, Natal. Bothalia 11: 181-190.

Roberts. B. R., 1966. The ecology of Thaba 'Nchu — a statistical

study of vegetation! habitat relationships. Ph.D. (Agric.)

thesis. University of Natal.

Rycroft. H. B., 1951. A quantitative ecological study of the

vegetation of Biesievlei catchment, Jonkershoek. Ph.D.

thesis, University of Cape Town.

Sneath, P. H. A., 1962. The construction of taxonomic groups.

In G. C. Ainsworth & P. H. A. Sneath, Microbial

classification. Twelfth Symp. of the Soc. for Gen. Microbiol.

Cambridge University Press.

Taylor, H. C., 1969. A vegetation survey of the Cape of Good

Hope Nature Reserve. M.Sc. thesis, University of Cape

Town.

Taylor, H. C., 1978. Capensis. In M. J. A. Werger,

Biogeography and ecology of southern Africa, 171—229. The

Hague: Junk.

Van der Merwe. C. V., 1977. 'n Plantegroei opname van die De

Hoop-Natuurreservaat. Cape Dept. Nature & Environmen-

tal Conservation internal report: 1—29.

Van der Meulen, F., 1979. Plant sociology of the Western

Transvaal Bushveld, South Africa. A syntaxonomic and

synecological study. Dissertationes Botanicae 49: 1-192.

Van der Walt, J. L., 1962. Ondersoek na 'n objektiewe metode

van plantopname in die Sneeubergreeks. M.Sc. (Agric.)

thesis, University of Pretoria.

Van Zinderen Barker. E. M. Jr, 1973. Ecological investigations

of the forest communities in the Eastern Orange Free State

and the adjacent Natal Drakensberg. Vegetatio 28, 3-6:

299-334.

Van Zinderen Barker. E. M. Sr & Werger, M. J. A., 1974.

Vegetation and phytogeography of the high-altitude bogs of

Lesotho. Vegetatio 29: 37-49.

Werger, M. J. A., 1972. Species— area relationship and plot size:

with some examples from South African vegetation.

Bothalia 10: 583-594.

Werger. M. J. A. , 1973. An account of the plant communities of

Tussen die Riviere Game Farm, Orange Free State. Bothalia

11: 165-176.

Werger, M. J. A., 1974. On concepts and techniques applied in

the Zurich-Montpellier method of vegetation survey.

Bothalia 11: 309-323.

Werger. M. J. A., 1980. A phytosociological study of the Upper

Orange River valley. Mem. bot. Surv. S. Afr. 46.

Werger. M. J. A. & Coetzee, B. J., 1977. A phytosociological

and phytogeographical study of Augrabies Falls National

Park, Republic of South Africa. Koedoe 20: 11-51.

Werger, M. J. A., Kruger, F. J. & Taylor. H. C., 1972. A

phytosociological study of the Cape fynbos and other

vegetation at Jonkershoek, Stellenbosch. Bothalia 10:

599-614.

Whittaker, R. H., 1962. Classification of natural communities.

Bot. Rev. 28: 1-239.

Wicht, C. L., 1948. A statistically designed experiment to test the

effects of veldburning on a sclerophyll scrub community. I.

Preliminary account. Trans. R. Soc. S. Afr. 31,5: 479-501.

Williams, W. T. & Lambert, J. M., 1959. Multivariate methods

in plant ecology. I. Association-analysis in plant communi-

ties. 7. Ecol. 47: 83-101.

Williams, W. T. & Lambert J. M., 1960. Multivariate methods

in plant ecology. II. The use of an electronic digital

computer for association-analysis. 7. Ecol. 48: 689-710.

Williams. W. T. & Lambert, J. M., 1961. Multivariate methods

in plant ecology. III. Inverse association-analysis. 7. Ecol.

49: 717-729.

Bothalia 15, 1 & 2: 259-291 (1984)

A vegetation survey of the Cape of Good Hope Nature Reserve. II.

Descriptive account*

H. C. TAYLOR**

Keywords: broad-leaved thicket, fynbos, plant community

ABSTRACT

This paper describes the reconnaissance survey of the Cape of Good Hope Nature Reserve that formed the

background to the trial of methods reported in the first paper. A vegetation map is provided and an account given

of the physical setting, climate, history and, more fully, the vegetation of the Reserve.

Two structural formations are recognized. Broad-leaved Thicket allied to the coastal forests of the Knysna

region, occupies’ scarcely 3% of the area of the Reserve and is confined to dunes, screes and scarps that are usually

well protected from fire. The remainder of the plant cover is fynbos. Its two floristic categories. Coast and Inland

Fynbos, reflect the two major soil types present: transported coastal sands which are usually neutral or alkaline in

reaction, and inland acid soils derived from the Table Mountain Sandstone substratum. Four communities of Coast

Fynbos and six of Inland Fynbos are described. Fire is the most important biotic factor influencing the vegetation.

Many coastal communities, if protected from fire, tend to develop into Broad-leaved Thicket, whereas some

Inland Fynbos communities exhibit a cyclical, self-perpetuating succession.

The vegetation of the Reserve appears to have greater floristic affinity with Langebaan and Cape Hangklip than

it does with Table Mountain despite the fact that the latter is geographically closer.

CONTENTS

1 Introduction 259

2 The study area 259

3 Methods 263

4 Vegetation 263

4.1 Fynbos 264

4.1.1 Coast Fynbos 264

4. 1.1.1 Eriocephalus Coast-Shelf Fynbos 267

4. 1.1.2 Dune Mixed Fynbos 268

4. 1.1. 3 Plecostachys—Scirpus Sedgeland 271

4. 1.1. 4 Coleonema Fynbos 271

4.1.2 Inland Fynbos 271

4. 1.2.1 Upland Mixed Fynbos 272

4. 1.2. 2 Protea lepidocarpodendron Tall

Fynbos 278

4. 1.2.3 Protea nitida Woodland 279

4. 1.2.4 Restionaceous Plateau Fynbos 279

4. 1.2. 5 Restionaceous Tussock Marsh 280

4. 1.2. 6 Berzelia—Osmitopsis Seepage Fynbos 282

4.2 Broad-leaved Thicket 283

4.2.1 Sideroxylon Thicket 284

4.2.2 Maurocenia— Chionanthus Thicket 285

5 Discussion and conclusions 286

Acknowledgements 287

Uittreksel 290

References 290

1 INTRODUCTION

Although the fynbos of the Capensis Region

(Taylor, 1978) is becoming increasingly well known

(cf. Boucher & McDonald, 1982), little has been

published on the vegetation of the southern Cape

Peninsula. Taylor (1969a) carried out a study of the

vegetation of the Cape of Good Hope Nature

Reserve in order to find a method of semi-detailed

* Partly based on an M.Sc. thesis. University of Cape Town.

** Botanical Research Unit, P. O. Box 471, Stellenbosch 7600.

survey suitable for use in fynbos generally. One part

of this study, the quantitative analysis of systemati-

cally collected data, is described by Taylor (1984a);

the other, a descriptive reconnaissance of the

vegetation, is the subject of the present paper.

2 THE STUDY AREA

2.1 The physical setting

2.1.1 Location

The Cape of Good Hope Nature Reserve (34° 15'

south latitude; 18° 25' east longitude) occupies the

southern end of the Cape Peninsula which is the

most south-westerly portion of the African conti-

nent, jutting southward some 64 km into the sea (see

Fig. 30; also Fig. 1 in Taylor, 1984a). The area of the

Reserve is approximately 7 750 ha. The land

boundary from Schusters Bay on the Atlantic coast

to Smitswinkel Bay in False Bay is 13,5 km long. The

coastline measures 40 km, the maximum length of

the Reserve from Schusters Bay to Cape Point is

about 21 km and the greatest width is some 9 km.

2.1.2 Geology and topography

The whole Reserve is composed of level or gently

inclined sandstone beds of the Table Mountain

Group of the Cape Supergroup, resting on Cape

granite which is visible only at the base of the Cape

Point cliffs (Fig. 16). The sedimentary Table

Mountain Group consists mainly of a hard, resistant

quartzitic sandstone. When freshly fractured, it is

white or pale grey in colour, weathering usually to a

darker grey. It may, however, be brown or red due

to the leaching of iron that is deposited in the surface

layers as brown hydroxide or red oxide. The

weathered sandstone surface is often highly irregu-

lar, giving rise to typical joint and bedding-plane

weathering. The block-shaped boulders seen on the

summit of Bonteberg (Fig. 11) and the escarpment

260

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

edge above Olifantsbos (Fig. 12) have been formed

in this way. The lower strata of the Table Mountain

Group are softer, the grains smaller and bound

together by a bright red or purple muddy sand.

These basal shales are seen in places along the foot

of the cliffs, as at Cape Maclear (Fig. 8).

The diagonal land boundary of the Reserve

follows the line of a fault which caused a vertical

displacement of ‘several hundred feet’ (Walker,

1952). The chain of hills south of the fault valley

merges eastward into the mountain range of the

False Bay coast (Fig. 19), with summits up to 360 m

at Paulsberg, the highest point on the Reserve. The

eastern slopes of these mountains, oversteepened by

wave attack, drop almost sheer to the sea (Fig. 7).

The coast here is steeply shelving, with deep water

close inshore, and this part of the coastline follows

the joints in the Table Mountain Group with a

north-east to south-west trend. The western slopes

of these mountains fall less steeply to a central

plateau (Figs 20 & 21) shelving gently westward for

5—6 km to its western edge at about 60 m above sea

level and terminating in a steep sandstone scarp

(Fig. 12).

The plateau forms the most nearly level surface of

the Reserve. It is devoid of rocks except for

scattered low sandstone outcrops which show that

bedrock is generally not far below the surface. It is

interrupted by a series of shallow seepage steps,

usually less than 3 m high, formed where a band of

tilted sandstone approaches the surface, damming

the soil water behind it. Drainage is poor: marshy

flats, flooded in winter, are connected by a few

meandering streams. Sirkelsvlei, a permanent body

of water, lies between sandstone ridges on high

ground west of the central plateau. It appears to be

fed from beneath by springs, the overflow through

marshy ground to the north-west draining into Die

Kloof, a stream which reaches the sea at Olifants-

bos. The water of Sirkelsvlei, and of all streams, is

dark brown in colour. The factors causing the dark

colour of the standing waters which are a feature of

many fynbos areas in the southwestern Cape, are at

present being investigated by Day (1981).

Most streams and ridges on the Reserve run

roughly north-west or south-east, or at right angles

to these directions, following lines of weakness

offered by joints or faults. The main drainage

system, the Krom River, rises in the Smitswinkel

Flats on the northern part of the plateau, and flows

north-west in a broad valley to the coast plain where

it is joined by the Klaasjagers River which orginates

outside the Reserve. At the coast these confluent

rivers form a lagoon whose outlet is deflected

southward by ‘beach drift’ (Mabbutt, 1952).

The present coast shelf on the Atlantic side may

extend half a kilometre westward of the plateau rim

but is usually narrower. The coastline of the Cape

Peninsula would have been altered with eustatic

changes in sea level in the Pleistocene and prior

times. The most recent recovery of sea level would

have been the transgression from a level some 140 m

lower than the present (Dingle & Rogers, 1972)

roughly 20 000 years ago, contemporary with the last

glacial maximum, to a level up to 2 m above the

present in the Holocene (Mabbutt, 1955) at about

6 000 years B. P. (Flemming, 1977). Van Donk

(1976) suggests that sea levels have been as low as

those of the last glacial maximum at least four times

during the last million years, so that the coast shelf

surrounding the peninsula has been periodically

exposed and the whole of False Bay has been a land

area, thus making available sources of sands for

transport as dunes across the peninsula.

The older preserved dunes are now stabilized and

cemented. They originated as plumes extending

inland in a north-westerly direction from sandy bays

along the east coast as a consequence of the

prevailing south-east summer winds (see Fig. 30;

also Fig. 1 in Taylor, 1984a). Variation in dune

structure and in vegetation suggests that the dunes

differ quite widely in age. The calcrete-cemented

dunes may be the oldest; certainly their flora is the

most distinctive. Mabbutt (1955) suggests that the

main phase of dune building was contemporary with

the Middle Stone Age and therefore broadly Late

Pleistocene in age (12 500-10 000 years B.P.).

2.1.3 Soils

A reconnaissance survey determined the distribu-

tion of soil types in relation to land-form. In order to

verify the field classification, 16 pre-selected type

profiles were sampled by auger for chemical

analysis. The dominant soil on the rocky hills and

mountains is the Mispah Series of the Mispah Form,

a shallow skeletal soil of in situ weathered and

colluvial materials (MacVicar et al., 1977). On the

lower slopes, deep, medium-coarse colluvial layers

of the Fernwood Series are formed. The Cartref

Series, a leached, iron-free soil with a weakly

developed lithocutanic B horizon, is found in small,

shallow, poorly drained depressions on the plateaux.

On sites where iron can accumulate, both on the

plateaux and in the larger stream valleys, humo-

ferric podzols of the Houwhoek Form may develop.

More strongly ferruginized soils of the Sandvlei

Series (Wasbank Form) are found on lower slopes

near the basal shales of the Table Mountain Group.

Soils developed on marine surfaces belong to the

Mispah and Fernwood Forms. They are thus

morphologically similar to the skeletal soils of rocky

hills and to the deeper sand of lower slopes. The

skeletal coast soils, however, although acid, are

more or less base saturated, whereas the skeletal hill

soils are base unsaturated. The remaining coast soils

are neutral or alkaline.

2.2 Climate

The Cape Peninsula possesses the equable,

Mediterranean type of climate characteristic of the

western margin of continents in comparable latitu-

des. The summers are warm and dry, the winters

cool with rain, but these so-called summer and

winter phenomena are not entirely confined to these

seasons, a fact which accentuates the equable nature

of the climate; and the gradual transition between

summer and winter gives rise to prolonged spring

and autumn seasons.

H C. TAYLOR

261

2.2.1 Temperature

The equable temperatures are most marked at

Cape Point itself which has the lowest average

annual range in mean daily maxima and minima

(4,2°C) of all weather stations in the Republic

(Weather Bureau, 1954). By comparison, the

average annual range for Simon’s Town is 8,5°C and

for Cape Town 10,5°C (see Fig. 30; also Fig. 1 in

Taylor, 1984a). At Cape Point the lowest monthly

range occurs during summer (December, 3,6°C)

while at Simon’s Town the highest monthly range

occurs during the same season (February, 9,0°C).

This is because Cape Point is continually exposed to

the cooling south-east winds, making summer

temperatures consistently low, while Simon’s Town,

on the north side of a range of mountains, not only

receives greater insolation but is also sheltered from

some of the summer winds and consequently

experiences some hot days. The exposed weather

station at Cape Point is not representative of general

conditions in the Reserve but no other temperature

data are available. The xerophytic vegetation of

north slopes in the Reserve suggests that their

temperature regimes are more akin to those of

Simon’s Town than to Cape Point. South-east

slopes, in contrast, are cool and moist, not only on

account of their shady exposure but also because

they face directly into the south-east winds.

Seasonal temperature variation at Cape Point is

also small. Summer and winter temperatures differ

by only about 5°C:

absolute daily max. + absolute daily min.

~ 2

for February is 18,3°C, for July 13,4°C. At

Simon’s Town the corresponding figure is nearly

17°C (March 31,3°C, July 14,6°C). Here again the

Cape Point figures would be representative only of

exposed bluffs and headlands, whereas many inland

parts of the Reserve would be comparable to

Simon’s Town. Highest summer temperature at

Cape Point is only 20,3°C (mean of daily maximum

for February) while at Simon’s Town it is 27,1°C

(highest monthly mean of daily maximum). The

lowest mean daily maximum temperatures are in

July: 15,5°C for Cape Point, 18,5°C for Simon’s

Town. The winter maxima show smaller differences

than the summer because the cool south-east winds

are rare during winter.

Frost was never recorded at Cape Point during the

period 1921-1950. It is unlikely that any part of the

Reserve has ever experienced frost.

The differing ocean temperatures on the east and

west coasts may also affect air temperatures near the

coast. Ocean temperatures differ most widely during

summer when the warm Agulhas current flows more

strongly, penetrating westward into False Bay and

around Cape Point. The summer south-east winds

accentuate the difference by piling up warm surface

waters against the False Bay coast and blowing them

away from the Atlantic coast where their place is

taken by the cold Benguella Current welling up from

beneath. Thus, summer sea temperatures on the

west coast are about 6°C colder than in False Bay

and on calm days air temperatures along the western

hills are noticeably cooler than on the eastern side.

In winter the warm Agulhas current is weaker and

the north-west wind blows the surface waters away

from the east coast while piling them up on the west

coast, thus reducing the temperature difference.

2.2.2 Wind

Wind plays a prominent role in moulding the

climate of the Reserve. The summer southeasters

blow for longer periods than the winter northwesters

— up to a week or more. At Cape Point southeasters

of 15—40 km per hour occur regularly during

summer, gusting up to 100 km per hour. The peaks

of False Bay, exposed to the full force of these

winds, may be enveloped in cloud above an altitude

of about 250 m during ‘black’ southeasters. Because

these strong persistent winds blow chiefly in the fire

season they are instrumental in the formation of

dunes, especially in recently-burnt coastal areas.

Also, evidence from the vegetation and the soil

profile suggests that soil below the rocky plateau rim

was blown by the southeaster over the rim where it

formed deep sand deposits bearing a characteristic

vegetation.

The winter winds seldom form dunes but have a

greater scorching effect on the vegetation of the

coastal shelf.

2.2.3 Precipitation

In the Mediterranean-type climate of the

south-western Cape, less than 10% of the annual

total rain falls between November and January

(Fuggle, 1981) and over 70% of the rain falls during

the ‘winter’ half-year April to September (Gold-

blatt, 1978: Fig. 5) when the cyclones moving

eastward over the southern Ocean bring in cool,

moist air off the Atlantic. This air condenses as rain

which may last for several days, the rainfall being

heavier near the mountains. As the depression

passes, the north-west wind backs to south-westerly

and later to southerly, with clearing showers and a

drop in temperature. Fine weather following a

depression may last for over a week (Schaffer,

1952).

The rainfall at Smitswinkel Bay, just outside the

north-eastern corner of the Reserve, averages 698

mm per annum while Cape Point (on the ridge at an

altitude of 227 m) receives only about 355 mm per

year (Weather Bureau, 1965). This large discrepan-

cy over a few kilometres is remarkable, but no

intervening records are available to show whether a

gradient exists. The peaks south of Smitswinkel

probably receive a higher rainfall than the Bay itself,

whereas the lowlands south of Smith’s Farm, with

their modified vegetation, may be drier. The

exceptionally low figure for Cape Point may be due

partly to inaccurate recording as a result of marked

air turbulence on the windy knife-edge. Annual

rainfall figures recently obtained from the Weather

Bureau for 1960 to 1982 (D. Clark, pers. comm.)

indicate that the mean annual rainfall for the

north-western part of the Reserve is about 550 mm

to 575 mm. This shows that the rainfall of the whole

Reserve is in general lower than that of the northern

part of the Peninsula.

262

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

Air moving eastward over the Atlantic is relatively

humid from its contact with fairly warm waters far

out to sea. When it meets the cold Benguella shore

current, the air cools and its moisture condenses as

fog. The west coast of the Reserve is subject to fog at

all seasons but particularly during autumn. Figures

for the Reserve are lacking, but at Cape Town

docks, on the Atlantic coast, fogjis experienced on

about 25 days a year, whereas Simon’s Town, on the

False Bay coast, has only about five days of fog

annually (Schaffer, 1952).

2.3 Historical background

2.3.1 Prehistory

Evidence from scatters of stone artefacts at open

sites, shell middens and rock shelter deposits shows

that from the Middle Pleistocene onwards, some

200 000 years or more ago, Stone Age peoples

periodically made their home in the southern Cape

Peninsula (Mabbutt, 1955). The shell fish remains,

fish vertebrae, rodent, ungulate and carnivore bones

and ostrich eggshell associated with Holocene

middens show that a variety of protein foods was

available and that marine resources were intensively

exploited seasonally (Maggs & Speed, 1967;

Grindley, 1967). These better-quality foods were no

doubt supplemented by the carbohydrates from

fleshy bulbs, roots and corms of the rich geophytic

flora and, to a lesser extent, by the edible fruits and

seeds that were available in the seasonal round. The

Stone Age peoples would have been transient

visitors and their main impact on the vegetation

would have been through their use of fire in hunting

and by conscious burning of the veld to push back

the succession to promote the abundance of

geophytes. In this, the incidence of natural fires

from lightning strikes and other natural causes

would have been increased.

There is good ethnographic evidence that both

Bushman hunter-gatherers and the later Hottentot

herders who were established in the Cape some

2 000 years ago, used fire in hunting before

European settlement. Vasco da Gama, for instance,

named the East Coast as the Land of Fire ‘in

reference to the multitude of veld fires observed

from the sea’ (Acocks, 1975). Thus fire was probably

used regularly and with increasing intensity as an

aboriginal veld management technique long before

European occupation. Formal hearths at Middle

Stone Age sites dating to the last interglacial some

100 000 years ago indicate the time depth of man’s

control of fire and his ability to generate it at will in

the region.

2.3.2 European settlement

In 1742, nearly a century after the Dutch East

India Company established a permanent settlement

at the present site of Cape Town, the Company

developed Simon’s Bay as an anchorage protected

from winter storms.

Until this time the indigenous fauna in the

southern Peninsula was seldom hunted by Euro-

peans. Duiker, grysbok, ribbok, klipspringer, red

hartebeest, bush buck, eland, Cape buffalo, zebra,

hippopotamus, chacma baboon, ostrich, silver fox,

spotted hyaena, leopard and lion were all found

(Opie, 1967). From the second half of the eighteenth

century this game was ruthlessly shot out and by the

end of the nineteenth century only a few grysbok,

ribbok and baboons remained.

After the founding of Simon’s Bay, farmers

started settling south of the new town. Until early in

the nineteenth century most of the area now

comprising the Cape of Good Hope Nature Reserve

was unoccupied and was consequently classified as

Crown Land. From about 1810, this land was

granted to farmers on a quitrent basis right down to

Cape Point. Each grant was subject to conditions

allowing passage of stock to adjoining farms and

access of fishermen to the coast. Tracks also led to

the lime kilns at Bordjiesrif and Buffels Bay, and to

the new lighthouse at Cape Point. With increasing

use by heavy wagons these tracks became deeply

rutted and were replaced by new ones running

parallel to the old. Even today, where these parallel

routes traverse the sandy plateaux, they can be

traced by the distinctive Restionaceous vegetation

that developed in the moister soil of the hollows,

closer to the underlying water table.

The early farmers found no wood supply other

than kreupelhout ( Leucospermum conocarpoden-

dron) which, though unsuitable for building,

provided excellent firewood as well as tannin for

curing hides (Adamson & Salter, 1950). Firewood

and rough timber for shelters were also obtained

from the littoral scrub by fishermen living semi-

permanently along the coast.

During these pioneering days of the early

nineteenth century, when many more ships were

rounding the Cape than ever before, shipwrecks

were not uncommon along the poorly charted west

coast. The hidden reefs off Olifants Point were

especially treacherous, and picturesque names such

as Matroosdam, Duitserstent and Mast Bay testify to

the frequency of wrecks in these waters. The

stranded ships often carried much valuable mer-

chandise which had to be salvaged at all costs.

Salvaging operations on such distant coasts were

major undertakings, sometimes lasting many

months. Temporary hutments were erected and

draught animals concentrated in great numbers. The

natural vegetation must have been locally destroyed

or at least severely trampled and overgrazed.

Patches of anomalous secondary vegetation along

the coast may date from these events.

Stock farming developed on an extensive scale.

The natural veld was frequently burnt to provide

young pasturage for horses, cattle and sheep.

Conditions proved very suitable for horses and,

during the South African War at the turn of this

century, 6 000 remounts were grazed on the area

(Opie, 1967). Stock concentrated on the better veld

in the Schusters and Krom River valleys and the

plateau south of Smith’s Farm, places which today

bear evidence of considerable disturbance to the

natural vegetation. Arable land was limited to richer

pockets of alluvial soil with a permanent water

supply, as at Klaasjagers River and Olifantsbos.

Here, crops such as potatoes, tomatoes and green

H. C. TAYLOR

263

vegetables were grown for the local market along

with barley as a supplementary feed for stock. Fresh

meat, dairy products and small orchards of fruit

helped local farmers to eke out a subsistence

livelihood. As communications improved and large-

scale commercial agriculture developed in the

hinterland, prices became too competitive for the

smallholdings of the southern Peninsula to pay their

way. In the first decades of this century many of the

southern farmers abandoned their land or used it

sporadically for seasonal grazing. In the 1930’s only

a few hundred cattle ranged over the abandoned

farms.

During this time, introduced woody species that

had been planted as windbreaks, shade trees and

sand stabilizers developed into impenetrable

thickets on the abandoned lands and then, aided by

fire, spread unchecked into the adjoining veld.

2.3.3 Creation of a Nature Reserve

With the decline of agriculture in the southern

Peninsula, farming was no longer an economic

proposition. In 1938-39 the Divisional Council of

the Cape acquired the two southernmost properties.

Smith’s Farm and Brightwater, to form the nucleus

of the Cape of Good Hope Nature Reserve. During

the ensuing 27 years the Council bought all the

remaining land south of the Plateau Road and a

small area to the north of it, to bring the present

extent of the Reserve to 7 750 ha.

The Divisional Council aims eventually to

re-establish the natural conditions prevailing at the

time of the early settlers, while catering for the

recreational needs of an increasing number of

visitors. It is hoped to resolve these conflicting aims

by developing recreation areas and picnic sites in

selected localities mainly along the coast, keeping

the rest of the Reserve in a wild state.

The intention is to re-introduce only those animals

known to have been present before settlement,

encouraging them to congregate near roads within

easy view of tourists. To this end, areas of level,

stone-free ground near the main road were

experimentally bush-cut (Fig. 28) or ploughed,

fertilized and planted with pasture species, but this

has been discontinued following damage by wind

erosion. Considerable restocking of animals has

already been done (Fig. 29).

A serious management problem facing the

authorities is the eradication of the introduced

invasive species. Because of the competitive

advantage that these plants have over the native

flora, especially by their rapid regeneration and

efficient establishment after fire, each successive

burn increases their range and density to the

detriment of the natural vegetation. Chief aliens on

the Reserve are the Australian wattles Acacia

cyclops (Figs 2, 12 and 26), A. saligna (Fig. 4) and A.

longifolia (Fig. 15), followed by Pinus pinaster,

Leptospermum laevigatum, Eucalyptus species and

Hakea suaveolens. During the 1966/67 survey, only

10% of the sampling sites on the Reserve, each with

a 183 m radius, were completely free of alien plants

(Taylor, 1969b). A ^gramme to combat their

spread has been initiated and in recent years the

Council has greatly increased the tempo of this

essential operation.

3 METHODS

In the quantitative survey of the Cape of Good

Hope Nature Reserve described by Taylor (1984a),

100 sample plots each 5 x 10 m in size were spaced

systematically in a grid pattern over the entire area.

Complete floristic lists were made for each plot,

using Acocks’s (1975) symbols as an estimate of

cover-abundance of each species. The data were

subjected to an association-analysis and later to a

Braun-Blanquet type of synthesis. Before and

during the quantitative data-gathering, additional

observations on the vegetation and habitat were

made, in order to help in interpreting the data. The

present descriptive account of plant communities has

been compiled from these observations, augmented

by information gained during the synthesis of the

sample plot data. The accompanying map (Fig. 30)

depicts only those vegetation units that are large

enough to be shown at this scale.

For reasons given elsewhere (Taylor, 1984a), the

original phytosociological tables (Taylor, 1969a)

were constructed from data which, from the

phytosociologist’s viewpoint, were incomplete. The

abridged tables presented here are therefore mere

abstracts of the originals, intended to illustrate and

strengthen the descriptive account with some

quantitative data. They show only some of the

species diagnostic for the main plant communities

(not character or differential species in the

Braun-Blanquet sense), omitting transitional sam-

ples and many companion species. Hence, certain

dominant and conspicuous species mentioned in the

text, for example Elegia parvi flora in Restionaceous

Tussock Marsh, do not appear in the tables because

of their low fidelity to the communities described.

4 VEGETATION

In his map of the Veld Types of South Africa,

Acocks (1975) depicts the greater part of the

Reserve as Macchia (Veld Type 69), with a narrow

strip of Strandveld (Veld Type 34) running down the

west coast. Floristically, the distinction between the

two types is clear. Physiognomically, both types

comprise a mixture of two structural formation

types, fynbos and ‘thicket’ (i.e. ‘thicket and

bushland’ of Edwards, 1983). In the present

account, the vegetation has been divided firstly into

these two formations. The fynbos element comprises

a homogeneous structural unit, the Cape sclerophyll

shrubland of the Mediterranean-type climate of the

Capensis region (Taylor, 1978). The thicket element

is represented by a simplified, floristically impover-

ished form of the coast forests and thickets of the

Knysna region (Phillips, 1931) here termed Broad-

leaved Thicket. The first subdivision of each

formation reflects the two major soil parent

materials on the Reserve. Further subdivisions can

be correlated firstly with edaphic factors, secondly

with those physiographic factors that determine local

264

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

climate, and thirdly with the effect of man’s activities

resulting in secondary vegetation units.

A provisional check list for the Reserve (Taylor,

1984b) records a total of 1 093 species of flowering

plants and ferns including 27 introduced species that

have become naturalized. This represents 41% of

the flora of the Cape Peninsula as recorded by

Adamson & Salter (1950). The eleven largest

families, with percentages of the total flora that they

comprise are: Asteraceae (11,0%), Iridaceae

(7,1%), Cyperaceae (6,8%), Fabaceae (6,5%),

Poaceae (5,9%), Restionaceae (5,5%), Orchidaceae

(5,4%), Ericaceae (4,5%), Liliaceae (3,9%), Me-

sembryanthemaceae (2,6%) and Proteaceae (2,2%).

When compared with other check lists of fynbos and

related vegetation (Taylor, 1979) these figures show

that the flora of the Reserve consists of both

Mountain and Coast Fynbos (Taylor, 1978) with

some elements of West Coast Strandveld (Acocks,

1975). Species richness in the vegetation sampled on

the Reserve varies from 83 species per 50 m2 plot in

Western Hillveld, to 5 species in Seepage Fynbos,

with a mean of 38,2 species, out of the total of 452

species recorded during enumeration of the 100

sample plots.

At least nine species are known to be endemic to

the Reserve and its immediate environs. No less

than five of these are Erica spp. (E. blancheana, E.

capensis, E. clavisepala, E. eburnea and E. font ana),

two are orchids (Disa salteri and Pterygodium

connivens) while the families Restionaceae (Elegia

fenestrata) and Bruniaceae (Staavia dodii) account

for one each. In addition, two members of the

Proteaceae, Leucadendron macowanii and L. flori-

dum, are today confined to the Reserve and its

immediate vicinity through destruction of their

habitats elsewhere. Aspects of the ecology of some

of these rare species are discussed by Taylor (1977)

and Moll & Gubb (1981).

4.1 Fynbos

After long protection from fire, fynbos may

develop into thicket or forest. In the south-western

Cape, however, one may regard fire as a natural

habitat factor recurring frequently enough to

maintain fynbos as a self-perpetuating community

(cf. Kruger, 1977), that is, a climax in this particular

habitat.

Fynbos is by far the more widespread of the two

formations on the Reserve. It agrees closely with

Adamson’s (1938) description of ‘Sclerophyll Bush’

summarized by Taylor (1972a). When fully develo-

ped it has three layers. On the central plateau and in

parts of the adjoining hills, Leucadendron laureolum

forms a discontinuous upper layer 1— 2 m tall. On

specialized sites it is replaced by other Proteaceae,

for example Protea lepidocarpodendron on gravels.

In some communities on rocky ridges, the upper

layer of scattered tall Proteaceous shrubs 2—3 m tall,

e.g. Leucospermum conocarpodendron and Mimetes

fimbriifolius, gives a picturesque woodland-like

appearance characteristic of the southern Cape

Peninsula (Figs 11 & 17). The middle layer consists

of a rich variety of small wiry shrubs 0,5 — 1 m tall,

usually with heath-like leaves, among which the

families Asteraceae, Ericaceae, Rutaceae, Thyme-

laeaceae and Rosaceae (Cliff ortia) are conspicuous.

Generally these shrubs occur evenly mixed but

locally one species may dominate a small area and be

replaced elsewhere by another, often of the same

genus or family. Precisely which species assumes

local dominance appears to be determined largely by

fire history — frequency, intensity and season of

past burns. The lower layer of tufted hemicrypto-

phytes and spreading undershrubs is usually not over

0,5 m tall. Thamnochortus dichotomus, the most

frequent plant of the lower layer, is found in a

variety of habitats. Geophytes are common but

many appear above ground only after a burn.

Annuals are usually infrequent but may become

abundant after a burn in Coast Fynbos.

Adamson’s observations that the layers are

variable in height, density and floristic composition

and that much fynbos consists of simpler communi-

ties which are single- or two-layered (Adamson,

1938, p. 89), are well exemplified in the Reserve.

Sometimes the shrubs of the middle layer merge

with the hemicryptophytes of the lower, forming a

single layer of variable height. The chief factor

causing this diversity is fire, which is responsible for

a sere ranging from the early regenerative stages

of open, low, predominantly Restionaceous vegeta-

tion, through the later stages when ericoid shrubs

predominate, to the climax of tall Proteaceous shrubs.

Often these stages are structurally very different from

each other but all the important species are present in

all stages except the earliest, though their size and

abundance may differ. Soil and drainage factors can

also give rise to communities of simpler structure. In

much of the Coast Fynbos, for example, tufted

Restionaceae are scarce or absent, resulting in a

shrubby, clumped community with a relatively open

ground layer. Alternatively, impeded drainage may

cause the absence of subshrubs, so that a closed

Restionaceous community, often dominated by a

single species, is formed (e.g. the Elegia parvi flora

Variant of Tussock Marsh).

Those tufted hemicryptophytes of the Restiona-

ceae and Cyperaceae that sprout rapidly from the

base after a fire, usually persist throughout the life of

the community. They are therefore of greater value in

characterizing a community than seed-regenerating

shrubs such as most Proteaceae and Ericaceae which,

through repeated burning and grazing, may have

disappeared from certain areas , or through prolonged

protection have become dominant in others. Certain

members of the Restionaceae are also more restricted

in habitat than many of the shrubs. Over large tracts in

the central and northern parts of the Reserve, for

example, Staberoha banksii is confined to dry rocky

ridges, Elegia stipularis to well-drained sandy slopes

and Elegia parvi flora to poorly-drained plateaux while

the Proteaceous shrub Leucadendron laureolum

ranges freely over the whole area. Restionaceae have

thus been widely used to characterize plant

communities in this study.

4.1.1 Coast Fynbos (Figs 1-8)

Coast Fynbos occurs on all transported sands of

H. C. TAYLOR

265

Fig. 1. — Stable Dune Mixed Fyn-

bos near Krom River lagoon.

Metalasia muricata and Indi-

gofera brachystachya in de-

pressions; wind-moulded

humps of pioneer Broad-

leaved Thicket (mainly Pte-

rocelastrus tricuspidatus) to 2

m tall.

Fig. 2. — Dwarf Dune Mixed Fyn-

bos, Platboom Road. Tham-

nochortus fraternus c. 30 cm

on calcrete ridge in centre,

flanked by Leucadendron

coniferum to 2,5 m in dune

slacks to left and right, with

darker-foliaged alien Acacia

cyclops invading, especially

to right.

Fig. 3. — Plot 69. Almost pure

stand of Thamnochortus er-

ectus on inland dune above

Brightwater. -Crown cover

barely 50%.

266

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

Fig. 4. — Plot 24. Sand Fynbos

near Klaasjagers with Staavia

radiata, left foreground.

Alien Acacia saligna inva-

ding right; Aspalathus-Phy-

lica Community on conical

hill left of centre.

Fig 5.— Plot 91. Plecosiach-

ys-Scirpus Sedgeland at

Potbank. Grey Plecostachys

serpyllifolia and erect Scirpus

nodosus in foreground. Be-

yond and right, Metalasia

muricata and Myrica cordifo-

lia in Coast-shelf Fynbos.

Left, old calcrete dunes with

Broad-leaved Eriocephalus

Thicket developing on the

southerly aspect. Centre, the

open dune at Platboom is

just visible.

Fig. 6.— Plot 60. Dry variant of

Coleonema Fynbos at Batsa-

ta Cove on the precipitous

False Bay coast.

H. C. TAYLOR

267

Fig. 7. — Oversteepened peaks of

the False Bay coast.

Paulsberg (centre) is the

highest point on the Reserve

(366 m). Below this, the

rounded calcrete hummocks,

pitted with caves, bear Dwarf

Dune Mixed Fynbos, with a

sward of Stenotaphrum se-

cundatum between these and

the rocky coast. Note the

wind-sheared thicket at left.

Fig. 8. — Cliff-ledges formed by

weathering of the soft basal

shales at Cape Maclear (76

m). To right, exposed calcre-

te; in foreground. Broad-

leaved Thicket.

marine origin which have formed the old raised

beaches and aeolian dunes; some of the latter extend

far inland. It occurs also, as Coleonema Fynbos,

where Table Mountain Sandstone crops out close to

the sea. The sandstone soils are acid while the

marine soils are more alkaline unless they have been

strongly leached.

4.1. 1.1 Eriocephalus Coast-shelf Fynbos on well-

drained marine sands of the coast-shelf

The coast-shelf comprises the old, more or less

level, raised beaches at 6m and 18m above mean sea

level, chiefly on the west coast. Coast-shelf Fynbos is

found mainly on the younger raised beach at the 6m

level. It consists principally of elements of the

Strandveld (Veld Type 34) that Acocks (1975)

depicts as a narrow band along the west coast of the

Reserve, but here at the southern end of its range it

is a fragmentary community that cannot be mapped

separately from the dune and thicket communities

with which it intergrades. Typical sites of Coast-shelf

Fynbos are therefore few and no typical samples

were represented in the plots. Perhaps the best site

on the Reserve is at Springholsgat just south of

Olifantsbos; fragments occur southwards, as at

Gifkommetjie, Ribboksdam and Neptunes Dairy.

The soil is a coarse white sand of the Mispah Form

with pebbles scattered on the surface. Bare patches

up to lm in diameter result from the burrowing of

Cap^dune moles ( Bathyergus suillus).

The vegetation varies in height, giving an irregular

canopy. Clumps of the tufted Thamnochortus

spicigerus about 1,5 m tall occur as pure stands of

variable size. These are probably southern outliers

268

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

of the T. spicigerus Community described for the

Langebaan area by Boucher & Jarman (1977, p.

257). Fine-leaved short shrubs like Eriocephalus

racemosus, Metalasia muricata and Passerina vulga-

ris, shaped by salt-laden winds into domed

hummocks about 60 cm high, form a discontinuous

layer. Below this is a low scrub 15-30 cm tall with

creeping Mesembryanthemaceae and other succu-

lent shrubs (e.g. species of Cotyledon, Euphorbia,

Othonna and Pelargonium ) and woody shrubs such

as Nylandtia spinosa, Euclea racemosa, Cassine

maritima, Maytenus lucidus and even Pterocelastrus

tricuspidatus, all stunted and spreading, with the

grass Stenotaphrum secundatum spreading between.

The woody shrubs tend to form short thickets of the

Sideroxylon Thicket (Community 4.2.1) scarcely

exceeding 90 cm in height.

With increasing distance from the sea the

vegetation becomes richer in species, but the paucity

of Restionaceae and the predominance of woody

shrubs separates it from Inland Fynbos. Where the

shelf is wide, as at Brightwater, the Coast-shelf

Fynbos is heterogeneous, being mixed with Inland

Fynbos elements on Table Mountain Sandstone soil

washed down from the plateau.

4. 1.1.2 Dune Mixed Fynbos on well-drained aeolian

dunes mainly of marine origin

The dunes vary in age from those still in the

process of formation to those formed since the

recession of the sea from the 6m-elevation shoreline

some 20 000 years ago. The soil consists of deep fine

sand, often mixed with shell fragments, and enriched

further inland, especially in the slacks, by humus,

which gives the soil a dark grey colour. The

vegetation varies greatly between the youngest and

oldest dunes but intergrades between. Four varia-

tions or ‘noda’ may, however, be distinguished.

(a) A pioneer community occurs along the edges

of a few bare dunes, for example, near Platboom

(Fig. 5) and at Buffels Bay. Such dunes, open to the

sea at one end, are being constantly blown back and

forth by the two prevailing winds. The dunes are not

enlarging, nor are they being actively colonized by

plants. Pioneers of the psammosere along the edges

are Agropyron distichum, Ammophila arenaria,

Ehrharta villosa and Myrica cordifolia.

(b) Stable Dune Mixed Fynbos (Fig. 1) comprises

a typical coast form occurring on stable dunes within

about a kilometre of the sea, and an inland variation

which is richer in species and taller. The coast form

is best developed around the Krom River lagoon,

local patches diminishing in extent southwards. On

the slopes and crests of the dunes, with their

Fernwood Form soils, fine-leaved elements predo-

minate. Metalasia muricata is dominant on the

littoral dune, but further from the sea typical

constituents are Indigofera brachystachya, Passerina

paleacea, Pelargonium betulinum, Phylica ericoides

and Psoralea fruticans, forming a rather open

canopy about 60 cm tall, with Myrica cordifolia

emerging in sheltered places to 1,5 m, and a lower

layer of hemicryptophytes like Restio eleocharis,

Chondropetalum microcarpum, Ficinia lateralis and

F. ramosissima about 20 cm tall. In sheltered slacks,

the woody elements of the Sideroxylon thicket

(Community 4.2.1) appear and may develop into

dense thickets where protected from fire. The coast

form of Stable Dune Mixed Fynbos was not sampled

by the grid-sampling technique.

The inland form is found principally on the large

dune traversing the Reserve between Smith’s Farm

(now known as ‘The Homestead’) and Rietveld.

There are smaller examples near Ribboksdam and

Bloubergstrand. The soil is a deep sand of the

Fernwood or Langebaan Series (Fernwood Form)

usually over calcrete. The vegetation becomes taller

and more diverse than in the coast form although

some of the coast species persist. Dense colonies of

the tufted thatching reed, Thamnochortus erectus,

reaching a height of 1,3 m, dominate the flanks of

the dune (Fig. 3) as they do on the Cape Flats

(Taylor, 1972). Scirpus membranaceus, of similar

growth form but shorter, is found locally in the same

habitat. Leucadendron coniferum, 1—2,5 m tall,

becomes dominant on deep sand with Myrica

quercifolia in the lower layer. On drier sand of the

ridges, Diosma hirsuta, Phylica stipularis, Ficinia

secunda and the creeping scented Leucospermum

hypophyllocarpodendron are more common. This

community is represented in Table 1 by Plots 69, 70

and 77, and can be compared with the Cape

Hangklip Leucadendron coniferum Community

(Boucher, 1978; Fig. 9, p. 476).

(c) Dwarf Dune Mixed Fynbos is confined to

calcrete ridges thinly overlain with sandy, shelly soil.

It occurs from Buffels Bay northward along the base

of the hills to Paulsberg (Fig. 7), from Meadows

tapering north-west to near Diaz Beacon (Fig. 2),

and inland of the bare dune at Platboom (Fig. 5). At

places where the south-east wind is very strong, e.g.

above Diaz Beach (Fig. 8), the sand has been blown

away and the limestone base of the dune exposed.

Where the main road crosses the calcrete dune

near Meadows, the fine-grained sandy soil up to 30

cm in depth supports a distinctive dwarf duneveld of

divaricate shrubs about 30 cm high, quite rich in

species. Further inland at the highest point on the

Platboom road where the soil is shallower, only the

hardiest species survive. The vegetation of this ridge

(Fig. 2) is reminiscent of depauperated garrigue in

Spain and southern France (Prof. D. A. Webb, pers.

comm.) and is similar in structure but not in floristics

to the dwarf fynbos of the Elim flats ( Acocks, 1975) .

Many plants absent or rare elsewhere in the

Reserve occur in Dwarf Dune Mixed Fynbos, e.g.

the conspicuous Thamnochortus fraternus which is

confined to this community, and Ursinia tenuifolia

subsp. ciliaris, a rare plant, found only on limestone

outcrops. Plants more common in Dwarf Dune

Mixed Fynbos than in other variations of Dune

Fynbos include Asparagus stipulaceus, Chascanum

cernuum, Erica coar data, E. coccinea (yellow form),

Hermannia trifoliata, Senecio arnicaeflorus and

Struthiola salteri. Because of insufficient quantitative

data, the relationship of this community to other

limestone communities along the south coast from

the Cape Flats to Algoa Bay (e.g. Taylor, 1972b;

Van der Merwe, 1976, 1977; Boucher, 1978; Taylor

H. C. TAYLOR

269

TABLE 1. — Abstract table for Coast Fynbos and Sideroxylon Thicket

Species diagnostic for the dry variant of Coleonema Fynbos

(1.1.4)

Drosanthemum candens

Exomis microphylla var. axyrioides

Species diagnostic for Sideroxylon Thicket (2.1)

Euclea racemosa

Ficinia ramosissima

Pterocelastrus tricuspidatus

Helichrysum dasyanthemum

Passerina paleacea

Colpoon compression

Cassine maritima

Kedrostis nana

Galium tomentosum

Sideroxylon inerme

Olea exasperata

Phylica ericoides

Species common to 1.1.4 and 2.1

Rhus glauca

Cineraria geifolia

Limonium scabrum

Species diagnostic for inland form of Stable Dune Mixed

Fynbos (1.1.2b)

Leucadendron coniferum

Indigo f era brachystachya

Pelargonium betulinum

Species diagnostic for Ifloga-Petalacte Sand Fynbos

(1.1. 2d)

Ifloga ambigua

Pentaschistis thunbergii

Phamaceum lanatum

Petalacte coronata

Manulea tomentosa

Crassula cymosa

Species common to 1.1.2b and 1.1. 2d

Phylica stipularis

Ficinia secunda

Aspalathus hispida subsp. hispida

Thamnochortus erectus

Species common to 2.1, 1.1.2b and 1.1. 2d

Ehrharta villosa

Carpobrotus acinaciformis

Helichrysum patulum

Salvia aurea

Thesium aggregatum

XXX

O

o

X X X X

XXX

O X O X

X O O X

0X0

XXX

O O X

O X

OX X

X o

O X X o

X

o

X X

o

XXX

0X0

X

X X X X

X X X X X

X X X X

X X

XXX

X X X X

X X

O X X X

O X O X

0X0

X o

O X

X

o

X

X present within plot

O present only in plot-surround

& Morris, 1981) is uncertain. As pointed out by

Acocks (1975) limestone fynbos is probably a

separate veld type that needs more detailed study.

The limestone communities on the west coast

described by Boucher & Jarman (1977) are

Strandveld: they bear no floristic relationship to the

south coast limestone fynbos.

(d) Ifloga - Petalacte Sand Fynbos (Fig. 4),

unlike other dune communities, occurs on local

inland sites where the acid sandstone soil has been

transported by wind to form small ‘islands’ of

dune-like deposits (Table 1, Plot 20) and also near

the contact between Coast Fynbos and Inland

Fynbos where the marine sand is admixed with

sandstone soil washed down from the adjacent

plateau (Table 1, Plots 86 and 92). Plot 24 is an

anomalous situation where Coast Fynbos occurs

locally in an inland valley. The community is closely

related to the Dry Sandflats Nodum of Upland

Mixed Fynbos (4. 1.2.1 and Table 2).

Sand Fynbos is characterized by Ifloga ambigua, a

straggly rhizomatous low shrub with minute grey

leaves, and Petalacte coronata, a compact, grey-

leaved low shrub about 30 cm tall, which together

give the community its distinctive appearance.

Depending on the location, extent and soil depth of

the sand patch, it contains an admixture of species

from other communities. Many of these sandy

patches, often less than 0,1 ha in extent, are

scattered throughout the Reserve. Their presence

within Inland Fynbos seems to indicate a period of

drier climate, probably during the last interpluvial

TABLE 2. — Abstract table for Upland Mixed Fynbos

NB. For notes see Table 1

270

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

1 £

XXXX

x x O x

X O X

o

O

X X X X

O x x O

O O O x

O

xxxxx xxxx

xxxxxx xxxx

xOxxx xxxx

XXXX X xxxx

XXXXX O X

xxxx Ox

xxxxxx xxxx

X xxxx

X XXX

x x x O

O x x O

XXX

xxxx

X XXX

x x x x O

O xxxxxx

X xxxx

xxxxxx

xxxxx

XXX XX

xxxxxx

X X X X O

xxxxx

X X XX

x x O x x

xxxxxx

x x O

.a £

o

c °

Z

5 «

§ a

X 13

c -c>

S o -I -§

13 k c a

^■3.3

■a -ail

5 .a*

C G G Cq

(N :

= lit

o ”2 .s

BZ S-s

2 ’c

Co G

.a £

% 3

o •a

e ©

5pZ

a ■§

£ §3

afS

a.

£ 5

o o

?.s 2 £

o.

C 8

£ 5

,Hh 0 o-SjS.ij.S

o Co;

£ 3 H

E <j »

£ -2 53

8T§

o H;

.a £

% 3

o ■©

c ©

eoZ

a og-a.gs,,

^ ^ ^ r* q -2

•SS C S 13

§ 2 x !§. 5 & y 'c

§ 0.0 g

•c 8 •§ -a q .0 « 'S

u X C g ^ ^

Co tfl Q JO n 3 O

8 « •» ^ ^ S *c

ft, o cok, ^ ^ cotq

.S -a

Xi >.

•5 *5

£ o

c c

<L> <L>

<D <D

t-i M

Ou D-

x O

H. C TAYLOR

271

when the soil was more easily eroded and

transported by wind because of an impoverished

cover of vegetation. The same process, on miniature

scale, can be seen in operation today, after a

summer fire.

4. 1.1. 3 Plecostachys-Scirpus Sedgeland on poorly-

drained coastal flats

Small depressions sometimes occur in the thin

layer of marine sand overlying the sandstone

bedrock of the coastal shelf. Here, drainage is so

poor that the soil is almost permanently moist;

during winter it is waterlogged. The coarse marine

sand, containing a high percentage of shell

fragments, is white on the surface but dark brown

beneath.

Plecostachys-Scirpus Sedgeland, the only Com-

munity of this habitat, is often fragmentary but

wherever Plecostachys serpyllifolia and Scirpus

nodosus are both present this marsh community

exists in rudimentary form. Similar vegetation is to

be found in marshy areas of the Cape Flats near

Strandfontein (Taylor, 1972b) and at Langebaan,

described as the Nidorella— Senecio form of J uncus

kraussii Dense Sedgelands (Boucher & Jarman,

1977).

At Potbank, the well-developed marsh vegetation

(Fig. 5) consists of rounded Plecostachys serpyllifolia

low shrubs about 45 cm tall, with the erect Scirpus

nodosus and J uncus kraussii up to 60 cm tall,

between which semi-scandent herbs and moisture-

loving suffrutices form a dense matted cover. Apium

graveolens, Chironia decumbens, Conyza pinnatifida

and Samolus valerandi are confined to the wettest

parts of this community while Chondropetalum

nudum, C. tectorum, Merxmuellera cincta , Orpheum

frutescens and Samolus porosus are shared with

other marshy communities inland.

4. 1.1. 4 Coleonema Fynbos on well-drained rocky

sandstone coasts

The soil of the rocky coast cliffs, ledges and

screes, with its substratum of sandstone bedrock

close to the surface, is genetically related to the soil

of Inland Fynbos habitats. However, the vegetation,

directly exposed to salt-laden sea winds, shows close

affinity with Coast Fynbos. On the False Bay coast,

the narrow ledges and precipitous slopes are hot and

dry, but the Atlantic slopes are often gentler and

many southwest-facing slopes are cooler, so that the

vegetation, if protected, develops into Broad-leaved

Thicket.

Coleonema Fynbos, occurring on both the east

and west coasts, covers a variety of habitats with a

corresponding variation in species composition. Like

Boucher’s (1978) Coleonema Fynbos, which it

greatly resembles, the Community is rich in

narrow-leaved shrubs but poor in the restioid

element. The uneven canopy of wind-pruned low to

short shrubs varies in height from about 15—75 cm.

Coleonema album, a rounded, bright green short

shrub about 75 cm high, is the most characteristic

plant. It is sometimes dominant on cooler aspects.

Elsewhere it is associated with other shrubs such as

Eriocephalus africanus, Metalasia muricata, Passeri-

na vulgaris, Phylica ericoides and Thesium viridifoli-

um. Restio cincinnatus, the only restioid, sprouts

occasionally from rock crevices.

The community of sprawling succulents common

on hot, dry rock ledges (e.g. Drosanthemum

candens, Exomis microphylla var. axyrioides, Tetra-

gonia fruticosa and sometimes Carpobrotus acinaci-

formis ), with its ground layer of Limonium scabrum

and Cineraria geifolia, may warrant separate status

(Fig. 6). It is represented in Table 1 by plots 60, 80

and 97. The typical Coleonema Fynbos was not

sampled.

4.1.2 Inland Fynbos (Figs 9—24)

In contrast to Coast Fynbos, Inland Fynbos is

confined to acid soils formed from the underlying

sandstone bedrock. On the hills the sand grains are

coarse, white and angular, but on the plateau the

grains are finer, smoother and mixed with humus

which gives the soil a grey or black colour near the

surface. Soil depth varies from nil on bedrock to

more than one metre on the plateau. Ferricrete

crops out locally along some of the mountain slopes.

On the Reserve, Inland Fynbos occupies a far

greater area than Coast Fynbos and covers a wider

range of habitats. Altitude varies from the summit of

Paulsberg (366 m) to nearly sea level in the Krom

River valley. All aspects, and all slopes from

perpendicular cliffs to level flats, are represented.

The southerly and easterly aspects are cooler than

usual because of the effect of the cool, cloud-bearing

south-east winds of summer. The considerable

variation in rainfall, its markedly seasonal nature

and the great differences in slope and soil depth

combine to produce soil conditions which vary from

permanently waterlogged to almost permanently

dry. Despite this great range in habitats, it is seldom

possible to draw sharp boundaries between them,

especially in the rocky hills and mountains where

drainage is relatively uniform throughout. The plant

communities are correspondingly ill-defined.

In an attempt to subdivide the complex plateau,

hill and mountain vegetation, 33 sample plots, each

5 x 10 m, were examined in eight tentative

‘habitat-community’ units. Detailed examination of

these units in a rough synthesis table, described

elsewhere (Taylor, 1969a), revealed that only four of

them, numbered 4. 1.2.1 to 4. 1.2.4 in the following

account, could be upheld.

Nevertheless, Inland Fynbos as a whole is so rich

in species and so varied in habitat that no attempt is

made at this level of description to subdivide it

further into formal units. The partial Braun-

Blanquet analysis of systematically collected floristic

data (Tables 2 & 3) suggests the existence of ‘noda’,

but in the absence of more data, gathered, analysed

and synthesized in the recognized Braun-Blanquet

manner, it cannot be contended that these noda

represent plant associations as defined by the

Zurich-Montpellier phytosociologists (e.g. Braun-

Blanquet, 1932). Thus, the general term ‘communi-

ty’ is used here.

During field work, it became apparent that

Thamnochortus dichotomus was widespread, often

common and even locally dominant in Inland

272

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

Fynbos, but rare or absent in Coast Fynbos. This

was confirmed by the synthesis table which showed

that Thamnochortus dichotomus had the highest

constancy of all species in the table. It is therefore

the best indicator species for Inland Fynbos on the

Reserve.

4. 1.2.1 Upland Mixed Fynbos on well-drained rocky

hills and mountains (Figs 9—18)

Upland Mixed Fynbos is the most extensive and

variable vegetation on the Reserve. Habitat varies

greatly with slope, aspect and altitude. The skeletal

Mispah Form soil, composed of coarse white sand

over bedrock, is usually shallow and this, where

combined with steep slopes, results in rapid

drainage.

Three main habitat variations can be distinguished

but, in addition to the variability in physical habitat,

other factors such as fire, wind and previous grazing

by domestic stock have caused great variation in

vegetation structure. Some aspects of this variation

are described below.

Fire succession

Floristically, the community is richest two to five

years after fire when annuals, geophytes, hemicryp-

tophytes and sprouting and seed-regenerating shrubs

all occur together (Fig. 14). Stratification, on the

other hand, is simplest in these early stages and

becomes more complex with advancing age. Starting

from a single low layer of sprouting and seed-

regenerating plants, it increases to a three-layered

community of (1) emergent shrubs, often Proteaceae

with broad leathery leaves, (2) canopy shrubs,

usually fine-leaved, mixed with taller Restionaceae,

and (3) a ground layer of the lower tufted

Restionaceae, Cyperaceae, Poaceae and herbs.

After some twenty years of protection from fire all

three layers become moribund; the emerge nts and

canopy shrubs slowly die out, littering the ground

with dead twigs and leaves which decompose slowly,

suppressing even the sprouting element of the lower

layer. If fire is excluded for a longer period, fynbos

in steep, rocky places may give way to Aspala-

thus — Phylica shrubland. On more level and less

rocky ground, fynbos may be self-perpetuating by

regeneration of the early stages in openings caused

by the death of shrubs in the mature community.

Examples of the early stages in the fire sere were,

in 1968, to be found near Kommetjieberg and in the

Bonteberg-Teeberg area. The highest number of

species recorded on all the one hundred 5 x 10 m

plots of the methodological study (Taylor, 1984a)

occurred on Bonteberg where 83 species were listed

in vegetation burnt three years before (Fig. 11). The

average count of species in vegetation one to three

years after burning was 72 per 50 m2. One year after

burning, the vegetation was one-layered and not

over 30 cm tall; the Restionaceae -Cyperaceae

element was regenerating vigorously and annuals

and biennials were common; seedlings of the larger

bushes were present but not yet contributing much

to the ground cover, and over 50% of the surface

was still bare. In three-year veld, the Restionace-

ae - Cyperaceae element, especially Elegiastipularis,

was becoming dominant, forming a canopy 30—45

cm high; annuals and biennials, including Selago

spuria and creeping Aspalathus species, were

disappearing; fine-leaved shrubs were appearing and

seed-regenerators ( Leucadendron laureolum, Serru-

ria villosa) were beginning to emerge sparingly to a

height of 60—90 cm. Cover had increased to about

75% on dry slopes and to more than 85% on

sheltered sites.

Intermediate stages in the fire sere were rare

because fire had been effectively excluded from the

Reserve for ten years or more. However, examples

of vegetation protected from fire for ten to twenty

years were found in all three of the major habitats.

Mature vegetation

(1) The western hills comprise the lower slopes of

Bonteberg in the north-western corner of the

Reserve and the western edge of the central plateau

from Krom River southwards to Pegrams Point. This

is a broad upland belt with a sandy soil over broken

rock, about 75 m elevation rising to 115 m at

Kommetjieberg (Fig. 27) and falling steeply to the

coastal plain on the west. Outliers penetrate the

plateau around Sirkelsvlei, at Russouwskop and

Lean Rock (122 m), with many smaller scattered

outcrops. These western hills close to the Atlantic

Ocean receive moisture not only from rain but also

from the autumn mists. This may make them

relatively cool and moist.

Mature Mixed Fynbos of the western hills is

typically three-layered. Leucospermum conocarpo-

dendron, a rounded, densely-branched shrub, and

Mimetes fimbriifolius, a small tree with umbrageous

crown, form an erratic woodland-like emergent

layer about 3 m high, the individuals about 4— 6 m

apart (Figs 11 & 17). Below this is a layer of

Leucadendron laureolum, a globose shrub 90—120

cm high, its yellow-green leaves giving a characteris-

tic tinge to the plateau vegetation, especially during

the winter flowering period. In drier areas of greater

relief it is an infrequent component, but along the

western escarpment it is common in Upland Mixed

Fynbos. Its distribution and mode of regeneration

suggest that it is a site-tolerant species which has

spread beyond its natural range (Plateau Fynbos) as

a result of repeated burning in the past. The L.

laureolum layer, sometimes dense but more often

scattered, usually affords less than 50% of the total

cover. The lowest layer, densely filling these

interspaces and extending more sparsely beneath the

Leucadendron, is dominated by Restionaceae 30—75

cm tall, among which Thamnochortus dichotomus,

Hypodiscus albo-aristatus and Restio cuspidatus are

common (Figs 16 & 17). Elegia stipularis is less

vigorous than in younger veld. Minor Ericaceous

genera and other fine-leaved shrubs are present. The

Upland Mixed Fynbos of the western escarpment is

represented in Table 2 by the Western Hillveld

nodum with Roella ciliata, Lobelia setacea, Thesium

virgatum, T. acuminatum and Osteospermum poly-

galoides as characteristic species.

Further down the escarpment, nearer the coast,

the emergent pseudo-woodland shrubs are absent

and in the lowest layer Restionaceae take second

place to a dense mixture of fine-leaved Ericaceous

H. C. TAYLOR

273

Fig. 9. — Plot 98. Summit of Judas

Peak. Mountain Fynbos with

admixture of Dry Hillveld.

Tetraria thermalis, Elegia

juncea; Leucadendron xan-

thoconus left and right fore-

ground.

Fig. 10. — Plot 66. Typical Dry

Hillveld, Kanonkop. On

plot, Tetraria thermalis (lax,

coarse-leaved clumps), Sta-

beroha banksii (dark, erect

tufts). Beyond, at base of

boulders, Aspalathus capen-

sis. Post-burn period at least

16 years.

Fig. 11. — Plot 8. Typical Western

Hillveld of the Upland Mixed

Fynbos, south slope of Bon-

teberg. The richest plot in

the survey (83 species), burnt

c. 3 years previously. The

rounded bushes of Mimetes

fimbriifolius (up to 2 m tall)

form stem-sprouts after a

fire. Note block weathering

of the Table Mountain sand-

stone.

274

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

Fig. 12. — Plot 19. Dry Sandflats

Community on deep wind-

blown sand of western

escarpment near Klein

Rondevlei, over 17 years

since last burn. Metalasia

muricata (white flowers) and

Salaxis flexuosa co-

dominant. Isolated Acacia

cyclops invading, left middle

distance. Aspalathus—Phyli-

ca Community between the

scattered Table Mountain

sandstone blocks of the

escarpment.

Fig. 13. — Dry Sandflats Commu-

nity near Groot Rondevlei,

Metalasia- Salaxis again co-

dominant in long-unburnt

vegetation. Circle indicates

position of Plot 15 (Fig. 14).

Fig. 14. — Plot 15. The same

community as in Figs 12 and

13, but burnt about three

years previously. This plot,

with 66 species compared to

27 in Plot 19, contains some

diagnostic species of Western

Hillveld and the companion

species which are absent in

the long-unburnt vegetation

of Plot 19.

H. C. TAYLOR

275

Fig. 15. — Plot 12. Somewhat de-

pauperate Western Hillveld

on rocky shelf with bedrock

close beneath; southern foot

of Bonteberg. The silvery

tufts of Merxmuellera cincta

(e.g. near figure) are conspi-

cuous on rock outcrops three

years after fire. Note en-

croachment of Pinus pinaster

and Acacia longifolia along

the Krom River, left middle

distance.

Fig. 16. — Plot 79. Restionaceous

Hillveld with some Tussock

Marsh species admixed. He-

lichrysum vestitum flowering.

Cape Point in the distance,

with granite boulders at sea

level.

Fig. 17. — Restionaceous Hillveld

on the southern plateau near

Diaz Beacon. Scattered

bushes of Leucospermum co-

nocarpodendron and Mime-

tes fimbriifolius to 2,5 m tall.

276

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

Fig. 18. — The biotic Metalasia—

Passerina Community sup-

planting the Restionaceous

Community shown in Fig. 17;

Leucospermum conocarpo-

dendron moribund. Near

Platboom road turnoff.

shrubs such as Salaxis flexuosa and other represen-

tatives of the minor genera, with Erica pulchella, E.

imbricata, E. corifolia and, among rocks, E.

mammosa.

In ‘islands’, where Western Hillveld occurs on

rock outcrops within the plateau, the structure is

reduced by shallow soil and strong winds to a low

shrubland or ‘heath’ of divaricate shrubs some 30 cm

tall, in which the minor Ericaceae are dominant.

Elegia stipularis is the most frequent of the

Restionaceae and Leucadendron laureolum is rare

and stunted, not exceeding 60 cm in height.

(2) The dry hills, with very shallow sand over

bedrock, flank the central plateau on its northern

and eastern sides, with scattered patches on bedrock

on the western side, at altitudes ranging from about

75-200 m.

In the north, the line of high hills forming the land

boundary of the Reserve along the Plateau Road

follows the northwest-trending joint plane of the

Table Mountain sandstone, intersected by the

Klaasjagers River valley. Detached in the north but

uniting progressively south-eastward, the hills form

a hummocky ridge ending in its highest point,

Rooihoogte (283 m), near the main entrance gate.

The central portion of the ridge appears to be hot

and dry while the south-eastern end is exposed to the

full force of the cool south-easterly winds. These

hills have a simpler vegetation than the western hills.

Leucadendron laureolum is conspicuous on the

lower slopes but less abundant on the intermediate

slopes which bear a Restionaceous mixture. Restio

egregius occurs frequently in this area. Where

sheltered by boulders, the crests and higher slopes

support the pseudo-woodland of the tall shrubs

Leucospermum conocarpodendron and Mimetes

fimbriifolius, while Tetraria thermalis is occasional to

fairly frequent. On screes and warm boulder-slopes,

Aspalathus capensis and Phylica buxifolia sometimes

form the tall closed shrubland described later.

On the east side of the plateau, the dry hills

comprise the hot lower western slopes of the

mountains along the False Bay coast. Here, the

pseudo-woodland shrubs af*e almost absent while

Tetraria thermalis is common, regularly spaced and

conspicuous. The tufted Staberoha banksii is

characteristic, interspersed with small-leaved low

shrubs such as Lobelia pinifolia and Saltera

sarcocolla (Fig. 10). These four species are

characteristic of the Dry Hillveld Nodum (Table 2)

which is the most distinctive community of the dry

hills.

(3) The mountains of the False Bay coast, from

Judas Peak to Cape Point, rise from about 180 m

along the upper edge of the Smitswinkel Flats, to

summits up to 366 m high at Paulsberg. These

mountains fall sheer into False Bay on the east (Fig.

7). Along the northern boundary hills, outliers of the

mountain habitat are found on Bonteberg, Teeberg,

Wolfkop and Rooihoogte. The upper slopes and

summits, especially along the False Bay coast, are

often covered in south-east cloud in summer. The

mountains are thus cooler and moister than the dry

hills below them.

The vegetation is lower and denser than on the dry

hills. Elegia juncea is a typical restiad of the upper

slopes (Fig. 9) while Alciope tabularis is conspicuous

in rock outcrops. Characteristic species of the

Mountainveld Nodum in Table 2, besides Elegia

juncea, are Clutia alaternoides, Berkheya barbata

and Leucadendron xanthoconus, but the mixed

composition of some of the transitional plots not

shown in the Abstract Table suggests that the

mountain vegetation is closely related to Dry

Hillveld.

Special communities

Three communities that may represent separate

successional stages within the Upland Mixed Fynbos

complex, differing in structure and composition both

from each other and from the general matrix of

Upland Mixed Fynbos, are described below. They

each have only one or two local dominants.

(a) Aspalathus-Phylica Community

The rockiest places not only afford prolonged

H. C. TAYLOR

277

protection from fire but also appear to provide a

suitable microclimate for fynbos to develop eventu-

ally into a type of tall closed shrubland. Suitable sites

for this development are found along the upper

western edge of the plateau escarpment (Fig. 12)

and on inland rocky screes (Fig. 4).

Along the escarpment edge at Olifantsbos

Aspalathus capensis, an almost tree-like woody

shrub 2, 0-2, 5 m tall with single stem and flat

spreading crown, forms dense closed stands. Like

most Aspalathus species, this plant regenerates with

vigour from seed when burnt but, unlike most other

members of the genus, it persists for long periods,

probably twenty or thirty years, and has a place in

the mature community. Phylica buxifolia, a plant

with similar form, is strongly associated, and the two

together sometimes form grotesque low forests with

a canopy not exceeding 2,5 m and a scant cover of

Myrsine africana as undershrub in the subordinate

layer. On screes and bedrock where the canopy is

open, fine-leaved shrubs such as Coleonema album,

Cliffortia spp., Erica mammosa and Aspalathus

carnosa partly fill the gaps, with succulents like

Carpobrotus acinaciformis and other Mesembryan-

themaceae in cracks between the rocks, and a few

moribund Restionaceae between the shrubs. Dipo-

gon lignosus and sometimes Bolusafra bituminosa

occur as climbers.

After long protection, shrubby trees ( Cassine

barbara, Colpoon compressum, Olea capensis subsp.

capensis, Pterocelastrus tricuspidatus , Rhus lucida,

Tarchonanthus camphoratus) begin to invade until,

if fire is excluded for perhaps a few decades, the

Community will develop into the Broad-leaved

Thicket.

(b) Metalasia— Passerina Community

This community occurs chiefly south of the

Smith’s Farm homestead (Fig. 18), an area that is

mostly lower in altitude than the typical localities of

Mixed Fynbos (30—75 m) with a more level

substrate of deeper sand with unimpeded drainage,

resulting in a drier soil.

The Community is typically three-layered, the

upper layer consisting of the two dominants,

Metalasia muricata and Passerina vulgaris, 90—120

cm tall, with some local admixture of Salaxis

flexuosa. Where this layer forms a closed canopy,

the lower layers are sparse, but generally there is

room for development of an impoverished middle

layer, 45-60 cm tall, of Mixed Fynbos plants

particularly Elegia stipularis and Serruria villosa. On

sand-patches, which are common in this locality,

some of the species of the Sand Fynbos of the dunes,

e.g. Leucospermum hypophyllocarpodendron, Pro-

tea scolymocephala and the annuals Gymnodiscus

capillaris and Manulea tomentosa, appear in the

middle layer of the Metalasia— Passerina Communi-

ty. In addition, a few sand-loving constituents of

Mixed Fynbos such as Stylapterus fruticulosus and

Thamnochortus fruticosus are more common than

they are in typical Mixed Fynbos. Where the upper

layers are very open, the ground layer contains

creeping succulents like Carpobrotus acinaciformis

and Ruschia sarmentosa, and smaller tufted Resti-

onaceae and Cyperaceae such as Restio cuspidatus

and Ficinia ramosissima. Some of these species

suggest an affinity with Coast Fynbos, Indeed, the

boundary of this community towards the sea is by no

means clear-cut and both dominants are frequently

found in the sandy littoral.

Both Metalasia muricata and Passerina vulgaris

are frequent on rocky ridges in the south. Here,

where one would have expected to find the richer,

more varied, typical Mixed Fynbos of the hills, only

a few of the most characteristic saxicolous plants like

Erica coccinea, Tetraria compar, T. cuspidata and

Saltera sarcocolla occur, giving the appearance of a

hardy remnant community resisting the advance of

invaders. Various stages in impoverishment can be

traced until, in the final stage, a dense stand of M.

muricata and P. vulgaris to 2 m tall, has excluded

almost all the original species. Both M. muricata and

P. vulgaris are vigorous colonizers of bare ground

along roadsides and in other disturbed areas. The

strong dominance of these two species in vegetation

where dominance is seldom a feature in the pristine

state suggests an invasion of these semi-pioneers into

a community that has been disturbed in some way.

The relic of Mixed Fynbos on rocky ridges lends

support to this view. Martin (1965), working at

Grahamstown near the eastern limit of fynbos,

describes a ‘heath’ community, probably owing its

origin to disturbance, dominated by Metalasia

muricata. He considers that burning may give a

competitive advantage to certain species, of which

Metalasia muricata is one.

It is suggested that, before the advent of

European man, Metalasia muricata and Passerina

vulgaris were concentrated in the sandy lower parts

of the Reserve and that their spread to higher rocky

ground has been encouraged by repeated burning of

the veld combined with grazing by domestic

livestock. This degeneration could well have been

accelerated if, as appears likely, rainfall, soil

moisture and other habitat factors are less favour-

able than in typical Mixed Fynbos areas, so that the

vegetation is inherently less stable. Smith’s Farm,

one of the oldest settlements on the Reserve, carried

considerable herds of horses and cattle in the early

days of this century (Opie, 1967). At that time, the

veld close to the homestead would probably have

been frequently burned to provide fresh grazing for

these animals.

(c) Salaxis flexuosa Community

Adjoining the Metalasia- Passerina Community

on the Platboom road and extending for about three

kilometres in a north-westerly direction from Diaz

Beacon, a community similar to the foregoing in

over-all species content but differing in physiogno-

my, is found. It is dominated by the fine-leaved

Salaxis flexuosa which, with Elegia stipularis as

subdominant, forms an even single-layered, rather

open canopy about 60 cm tall. Other Mixed Fynbos

species occur sparingly, some of them up to 90 cm

tall, while occasional Leucospermum conocarpoden-

dron and Mimetes fimbriifolius emerge.

The Community occurs on the broad, windswept,

coarse-sandy ridge forming the watershed of the

southern end of the Reserve, here at about 120 m

278

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

4. 1.2. 2 Protea lepidocarpodendron Tall Fynbos on

well-drained, ferricrete slopes

altitude. To the west, the Salaxis flexuosa Commu-

nity merges into Restionaceous Tussock Marsh of

the central plateau. On rock outcrops, Salaxis

flexuosa is replaced as dominant by other minor

Ericaceae (species of Blaeria, Syndesmanthus and

Scyphogne muscosa ) and the Mixed Fynbos is

represented by species such as Anaxeton laeve,

Alciope tabularis, Staberoha banksii and Tetraria

thermalis. As in the Metalasia— Passerina Communi-

ty, these Mixed Fynbos species take a subsidiary

place and appear to have been invaded by the

dominant Ericaceae. The same Ericaceous species,

including Salaxis flexuosa, appear to be also

invading the Tussock Marsh. What is thought to be

the original community is still found in the

north-western part near the Gifkommetjie road,

where a greater mixture of Ericaceous species

including Erica corifolia, E. imbricata and E.

pulchella, together with Leucadendron laureolum,

form a typical rich Plateau Fynbos.

In this southern part of the Reserve where both

the Metalasia- Passerina Community and the Salaxis

flexuosa Community exhibit the same symptoms of

biotic disturbance, the Salaxis Community appears

to have been derived from Plateau Fynbos by the

same means and for the same reasons as the

Metalasia— Passerina Community was derived from

Mixed Fynbos.

Elsewhere on the Reserve, Salaxis flexuosa is

found together with Passerina vulgaris in what

appears to be a more natural but ill-defined unit

transitional between Coast and Inland Fynbos,

depicted in Table 2 as the Dry Sandflats Nodum.

This occurs where the sandstone bedrock is overlain

by deep white sand, principally on the lower slopes

and flats below the rim of the western escarpment

(Figs 12-14). Floristically, it is closely allied to Sand

Fynbos (4.1.1.2(d)) and the habitats are similar. The

sandy substrate seems unstable and subject, after

fire, to erosion or to deposition by wind. Both

Salaxis flexuosa and Passerina vulgaris appear to be

pioneers of unstable sandy habitats, hence their

dominance in areas of natural as well as man-made

disturbance.

The chief concentration of Tall Fynbos is found in

the north-eastern corner of the Reserve from

Rooihoogte along the slopes of Judas Peak to Die

Boer ridge at 120-300 m elevation (Fig. 19).

Outliers occur on Teeberg and the lower slopes of

Vasco da Gama Peak at 60—120 m. Though Protea

lepidocarpodendron individuals occur rarely on

white Table Mountain Sandstone sands, the com-

munity as a whole is confined to strata containing

iron oxides in solution. These oxides are deposited

in the subsoil as impervious layers of ferricrete. In

the Reserve, these gravels of the Sandvlei Series

(Wasbank Form) occur chiefly on slopes and the

shallow soil becomes very dry in summer. The

surface soil is a fine yellow-brown sand with an

admixture of small brown stones, gravelly pebbles or

sometimes larger stones 10-20 cm in diameter,

stained light brown to dark purple from lateritic

concretion. The litter layer, consisting largely of

undecomposed Protea leaves and Restionaceous

stems, is thick. Sometimes the ferricrete crops out

on the surface.

The vegetation is two-layered. The upper layer,

1,75-2,5 m tall, consists of the single species Protea

lepidocarpodendron giving a cover of over 85%

when mature. This erect narrow-crowned Protea

with its dark-green oblong leaves and black-bearded

bracts gives a characteristically sombre appearance

to the community. The lower layer of Restionaceae

and shrubs 45-60 cm tall varies in vigour and

density depending on the cover afforded by the

proteas. The chief species of this layer, and the most

faithful species for the Community, is the tufted

Hypolaena digitata which is abundant throughout

the ferricrete areas. Two other plants which occur

sparingly, Peucedanum sieberianum and Tetraria

fimbriolata, seldom found elsewhere on the Re-

serve, are reminiscent of the granite or shale slopes

around Stellenbosch. Leucadendron xanthoconus

which emerges from the lower layer to 90 cm or

more, is characteristic of the Tall Fynbos but ranges

beyond this to surrounding slopes. Other species

Fig. 19. — Protea lepidocarpoden-

dron Tall Fynbos on western

slopes of Judas Peak and Die

Boer Ridge. Dark-foliaged

P. lepidocarpodendron 2 m

tall in right foreground, with

the lighter-coloured, lower

Leucadendron xanthoconus

in front.

H. C. TAYLOR

279

with a similar distribution are Cliffortia stricta,

Tetraria bromoides and Leucadendron salignum —

species which are again reminiscent of the vegetation

of the mainland mountains. Most species of the

lower layer, which is floristically rich and varied,

show similarities to Mixed Fynbos communities.

Elegia juncea and Erica cerinthoides, for example,

are found in high, cool, habitats; Hypo discus

albo-aristatus, Elegia stipularis and Serruria vallaris

on the lower, warmer hill slopes; Petalacte coronata

in deep sand pockets; and Rhus lucida on boulder

scree.

There are indications of two phytogeographical

elements on the Cape Peninsula, a northern one

related to the mainland mountains from Stellen-

bosch northwards and a southern element related to

the coast mountains on the eastern side of False Bay

(Taylor, 1981). The Tall Fynbos and the Protea

nitida Woodland (4. 1.2. 3 below) contain species

from both elements and are thought to be southern

outliers of the northern element. Certainly the two

diagnostic species of Tall Fynbos, viz Protea

lepidocarpodendron and Hypolaena digitata, both so

common and so distinctive in their respective layers,

give this community a characteristic physiognomy

which serves to set it clearly apart from other

communities on the Reserve.

4. 1.2.3 Protea nitida Woodland (‘Waboomveld’) on

a steep, well-drained, rocky, northerly slope

This community is limited to about 25 ha of the

escarpment in the extreme north-eastern corner of

the Reserve, from the slopes of Judas Peak to the

False Bay coast. Altitude ranges from 120 m to

almost sea-level within a distance of some 275 m on a

very steep slope with a uniform north-easterly

aspect. Level bands of Table Mountain Sandstone

tinged red-brown with iron oxide, traverse the slope;

the shallow soil of coarse sand and the surface stones

and pebbles are of the same colour.

Protea nitida (‘waboom’), the diagnostic species of

the Community, occurs at an even espacement of

about 6 m. It is a stoutly-branched, broadly-

spreading, woody shrub 1,75-2,5 m tall, with thick

white stem and broad blue-grey leaves. Its light

colour, large size and regular spacing make it a

conspicuous feature of the landscape to which it

gives the appearance of an orchard-like woodland.

On most inland mountain ranges, waboom is

characteristic of well-drained soil-covered screes or

colluvial deposits near the base of steep slopes, not

hard plinthic gravel. Nevertheless, Boucher (1978)

reports the occurrence of communities resembling

both this and the previous one on yellow plinthite

soils in the Kogelberg, a coastal range on the eastern

side of False Bay. Drainage may be the primary

factor determining the distribution of these commu-

nities on coastal ranges.

In structure, the Protea nitida Community

comprises a variable field layer 45-90 cm tall, with

Protea nitida and the similar Leucospermum

conocarpodendron emerging. The field layer con-

tains many species common to Mixed Fynbos. The

Community thus resembles a low woodland with a

field layer of fynbos plants instead of grass. As in the

case of Tall Fynbos, representatives of the higher

mountain slopes and the lower hills are present, but

species of deep sand, which is not found in this

habitat, are absent. At lower altitudes, an increa-

singly large proportion of coast-thicket species

intrude, e.g. Salvia aurea, Ficinia ramosissima,

Cassine barbara, Pterocelastrus tricuspidatus, Rhus

lucida and Tarchonanthus camphoratus, owing to

the proximity of a small but well-developed patch of

Maurocenia—Chionanthus Thicket at The Chair.

Species nearly confined to the ‘pseudo-woodland’

here and found regularly in the inland community,

are Montinia caryophyllacea, Rhus rosmarini folia,

Themeda triandra and Hermannia cuneifolia. The

fact that these are also found occasionally in Tall

Fynbos lends strength to the view that the two types

of Protea Veld are related. Cymbopogon marginatus

and Bobartia indica, regularly occurring in Tall

Fynbos, are also found in Waboomveld; both are

common constituents of Protea communities on

inland mountains.

4. 1.2.4 Restionaceous Plateau Fynbos on the drained

level plateau

Plateau Fynbos is found on the level central

plateau excluding the marshy areas and seeps which

are usually sharply delineated. From Suurdam in the

south, it fans out northwestward to include the

Smitswinkel Flats and the shallow drainage system

of the Krom River. Altitude ranges from 90-135 m

in the southern and central parts, decreasing to 30 m

in the lower parts of the Krom River drainage

system. Drainage is neither very poor nor very good:

where standing water is encountered in winter,

Plateau Fynbos is replaced by Tussock Marsh. The

Houwhoek Form humo-ferric podzol has a fine-sand

A horizon, usually grey from partially decomposed

organic matter and is much deeper than the soil of

the hills. Stones and rocks are rare, either on the

surface or close beneath it. The soil is acid (pH

3, 9-5, 6) and litter does not decompose readily.

The main cover is afforded by a closed uniform

layer of Restionaceae mixed with low or sprawling

shrubs 30-45 cm in height. Above this, throughout

most of the plateau, Leucadendron laureolum forms

a conspicuous but discontinuous layer about 90-120

cm tall (Fig. 20).

Plateau Fynbos contains no species that distin-

guish it from Upland Mixed Fynbos on the one hand

and Tussock Marsh on the other. It might be

described as a mosaic of these two communities with

the hill flora predominating on the slightly raised

ground, the marsh flora predominating on level or

lower areas. The following considerations led to the

decision to separate it from Mixed Fynbos: (1) the

presence of species characteristic of Tussock Marsh;

(2) the absence of certain groups of Mixed Fynbos

species; (3) the rarity of other Mixed Fynbos shrubs

and sub-shrubs; (4) the greater proportion of

Restionaceae in the lower layer; and (5) the

dominance in the upper stratum throughout the

Plateau (except in the southernmost parts) of

Leucadendron laureolum. From this community, L.

laureolum has spread into, indeed dominated, large

adjacent areas of Mixed Fynbos where the habitat is

280

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

Fig. 20. — Plot 36. Restionaceous

Plateau Fynbos with Leuca-

dendron laureolum about 1 m

tall, at least 14 years since

last burn. Near Sirkelsvlei.

not dissimilar from Plateau Fynbos, e.g. along the

western escarpment to Brightwater. Its spread might

have been encouraged by veldburning since regene-

ration of many of the re-seeding Leucadendron

species is greatly stimulated by fire. Becking (1957)

draws a distinction between the terms ‘character

species’ which should be used only to indicate

sociological fidelity, and ‘indicator species’ which

should be reserved for species showing fidelity to an

ecological factor. If L. laureolum is more indicative

of a burning regime than faithful to a particular

community, it could be termed a ‘fire-indicator.’

As a mosaic of two communities, Plateau Fynbos

is floristically rich. No species are strongly faithful to

it, the highest fidelity rating being ‘Preferents’

(fidelity III; cf. Becking, 1957), e.g. Leucadendron

laureolum, Macrostylis villosa subsp. villosa, Penta-

schistis colorata and Staberoha cernua. ‘Indifferents’

or wide-ranging species found often in Plateau

Fynbos as well as in other communities, are Diastella

divaricata subsp. divaricata, Diosma hirsuta, D.

op positi folia, Erica imbricata, Haplocarpha lanata,

Hypodiscus aristatus, Phylica imberbis, Penaea

mucronata, Pentaschistis curvifolia, Restio cincinna-

tus (ecotype?), Tetraria microstachys and Thamno-

chortus dichotomus. Species occurring more often in

other communities fall into two groups: (a) species

with their optimum occurrence in Tussock Marsh

e.g. Chondropetalum nudum, Elegia neesii, E.

parviflora, Restio bifurcus, R. quinquefarius, Tetra-

ria fasciata, T. flexuosa and Serruria glomerata; and

(b) species with their optimum occurrence in Mixed

Fynbos e.g. Adenandra villosa subsp. umbellata,

Castalis nudicaulis, Corymbium africanum, Elegia

stipularis, Erica pulchella, Tetraria eximia and

Watsonia tabularis var. concolor. Absent from

Plateau Fynbos are all the non-Proteaceous woody

components ( Rhus etc.), the preferents of the

Metalasia — Passerina Community and the Ifloga

— Petalacte Community, and all species confined to

rocky slopes such as Leucospermum conocarpoden-

dron, Mimetes fimbriifolius, Staberoha banksii and

Tetraria thermalis.

4. 1.2.5 Restionaceous Tussock Marsh on seasonally

inundated flats

Tussock Marsh is found where water is stagnant

during winter but dries up in summer. This occurs in

shallow depressions within the central plateau, and

on the broad flats of the Krom and Schusters River

valleys in the north.

Tussock Marsh, strongly dominated by one or two

species, is usually sharply delineated but occasional-

ly, where the flats slope very gently, there is a broad

transition from Tussock Marsh via Plateau Fynbos to

the drier communities. This transition can be seen

between the Krom River and Teeberg, and in the

Garslandskloof area.

The simplest Tussock Marsh Community [Table 3,

variant (a)] is found in shallow depressions or ‘pans’

varying from about 40 m2 to a few ha. in extent, to

which there are no outlets. With bedrock close to the

surface, drainage is so poor that standing water is

often present throughout the rainy winter months,

though the soil can become very dry in summer. The

Cartref Form soil has a medium-sand A horizon,

dark grey on the surface from decomposed organic

matter. The vegetation consists typically of a dense

stand of Elegia parviflora about 30 cm tall, with

occasional sprouting shrubs, particularly Rafnia

crassifolia, and with an open field layer of

decumbent dwarf shrubs, e.g. Ursinia tenuifolia

subsp. tenuifolia, and prostrate annuals, e.g.

Prismatocarpus sessilis. A few Restionaceae and low

shrubs of the moister Tussock Marsh occur sparingly

towards the edge of the Community (Table 3).

The moister Tussock Marsh Community [Table 3,

variant (b)] is found in the larger plateau depressions

and river flats where the soil (Witsand Series,

Westleigh Form) is deeper and very wet. It has three

layers, in all of which Restionaceae are conspicuous.

Elegia parviflora is still dominant in the lowest layer,

Restio bifurcus, R. quinquefarius and R. dodii are

conspicuous in an erratic middle layer about 60 cm

tall, and the black-spiked Elegia cuspidata emerges

H. C. TAYLOR

281

TABLE 3. — Abstract table for Tussock Marsh and Seepage Fynbos

NB. For notes see Table 1)

X present within plot

O present only in plot-surround

to 90 cm (Fig. 21). The most varied layer is the

middle one which often contains the cyperoid

hemicryptophytes Tetraria fasciata and T. flexuosa,

soft- and small-leaved shrubs like Cliffortia subseta-

cea, the sprawling Diastella divaricata subsp.

divaricata, Erica gnaphaloides, E. bruniades, E.

capensis and E. clavisepala (the last two endemic),

Scyphogyne muscosa, and Serruria glomerata (Fig.

22). Bobartia indica is conspicuous after a burn and

Leucadendron laureolum may be present in the

mature phase as an intruder from Restionaceous

Plateau Fynbos. The latter species is absent from the

early post-fire phase. This phase structurally

resembles the pan community because Elegia

parviflora grows faster than E. cuspidata which

regenerates only from seed after a fire.

Fig. 21. — Plot 65. Restionaceous

Tussock Marsh near Bright-

water road turnoff. Tall,

black-spiked Elegia cuspidata

in foreground and beyond

plot; many other Restiona-

ceae, with Elegia parviflora

dominant, comprise the low-

er layers.

282

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

Fig. 22. — Plot 83. Variety of

Restionaceous Tussock

Marsh with many ericoid

shrubs. Upland Mixed Fyn-

bos on the low rocky ridge in

the background, with glo-

bose bushes of Leucosper-

mum conocarpodendron.

Suurdam.

Occurring usually at the low-altftude margins of

Inland Fynbos where it merges with Coast Fynbos is

a single-layered Tussock Marsh Community 30—40

cm high. In its simplest form near Ribboksdam it

consists of two species evenly intermixed, Chondro-

petalum nudum and Erica subdivaricata. Beneath

these are a few inconspicuous Restionaceae hardly

contributing to the cover. At Hestersdam this

community is more complex and contains, among

others, the rhizomatous Elegia vaginulata.

4. 1.2. 6 Berzelia-Osmitopsis Seepage Fynbos on

permanent seeps

The plateau drops gradually into the broad

stream-valleys in a series of terraces separated by

abrupt shallow steps about 1-3 m high. The seepage

zones formed on these steps are permanently moist

or wet with percolating water. The soil beneath the

30 cm deep peaty litter is fine black silty sand. The

vegetation differs markedly from that of Tussock

Marsh. It is tall, shrubby and dense with no

conspicuous Restionaceae.

In the south, Berzelia abrotanoies do-

minates the steps, sparsely mixed with a few

Osmitopsis asteriscoides plants, a lean erect Compo-

site, both shrubs being about 1,75 m tall. Beneath

this closed canopy only a few Cyperaceae, e.g.

Tetraria sp. cf. T. cuspidata and Ficinia filiformis,

occur in an open understorey, and sparse rosettes of

Drosera spp. dot the ground. Northwards, the

seepage steps are richer in species (Table 3). The

canopy contains, besides the above, Berzelia

lanuginosa, Mimetes hirtus, Psoralea aphylla and

sometimes Psoralea pinnata. In the understorey,

additional species include the hemicryptophytes

Restio compressus, R. dodii, R. ambiguus, Merx-

muellera cincta and Tetraria flexuosa, and the shrubs

Chironia decumbens, Cliffortia subsetacea, Erica

bruniades and E. capensis. Patches of Utricularia

capensis occur on wet open ground. Two endemic

Erica spp., viz E. eburnea and E. fontana, are quite

common in certain seeps. One locality near Plot 29

has a complex and very dense seep vegetation

containing, in addition, banks of pure Watsonia

tabularis var. concolor and Pteridium aquilinum.

Fig. 23. — Seepage Fynbos about

1-2 m tall, at edge of Groot

Rondevlei: mainly Berzelia

lanuginosa (round white

flower-heads) and Psoralea

aphylla. On pan floor,

creeping Sporobolus virgini-

cus and Laurembergia repens

with (left) three black-spiked

tufts of Chondropetalum tec-

torum.

H. C. TAYLOR

283

Fig. 24. — Plot 29. Distinctive

woodland-like variety of

Seepage Scrub with scattered

Psoralea pinnata to 2,5 m

above a dense field layer of

grasses and Restionaceae.

Smitswinkel Flats.

At Plot 29 there is an open scrub of Psoralea

pinnata to 2,5 m high, with a dense field layer of

grasses and Restionaceae in which Pentaschistis

curvifolia and Staberoha distachya are common, the

former locally dominant (Fig. 24). This distinctive

woodland-like community, only a few hectares in

extent, has not been seen elsewhere on the Reserve.

Its relationship is uncertain. It occurs on almost level

ground just below the complex seepage step

mentioned above, which apparently feeds it with a

permanent supply of slowly percolating soil moist-

ure.

A seepage zone on the steep north-east slope

below The Camp contains the very local endemic

Leucadendron macowanii, emergent to 3 m, and

Neesenbeckia punctoria in the 2 m canopy layer. In

the subordinate layer is Carpacoce spermacocea (1,2

m), a soft-leaved sprawling herb smelling strongly of

hydrogen sulphide when crushed. The last two

species, like members of the surrounding Protea

nitida Woodland, are more characteristic of the

northern Cape Peninsula and the seepages on the

coastal plain across False Bay (cf. Boucher, 1978;

pp. 477-478).

On stream banks, Berzelia abrotanoides is usually

common, Psoralea pinnata occasional, and Indigofe-

ra filifolia locally dominant in the northern areas.

Along the margins of vleis, e.g. Skilpadvlei in the

south and Groot Rondevlei in the north (Fig. 23), is

a richer flora with more complex structure. Besides

the Tussock March Restionaceae there are tufts of

Chondropetalum tectorum almost 2 m high, both

Berzelia species ( B . abrotanoides and B. lanugino-

sa), Myrica quercifolia and Cliff ortia ferruginea

(sprawling undershrubs), and Merxmuellera cincta, a

tufted grass 90 cm or more in height. Where J uncus

kraussii is present, this community resembles

Boucher’s (1978) Chondropetalum- Juncus Vlei-

fringing Tussock but at Cape Point where the vleis

have less permanent surface water than across False

Bay, Juncus is not common.

In general, many of the variations of seep

vegetation described above have counterparts in the

Erica- Osmitopsis Seepage Fynbos described by

Boucher (1978) and may be placed under Adam-

son’s (1938) broad category, Wet Sclerophyll Bush.

4.2 Broad-leaved Thicket (Figs 25—27)

On the Reserve, the only representative species of

the mountain forests of the Knysna region is

Rapanea melanophleos which is confined to the

steep scrubby slopes overlooking False Bay. The

patches of woody thicket that do occur on the

Reserve are allied to the coastal forests of the

Knysna region, but they are impoverished in species

and simpler in structure. This Broad-leaved Thicket

occupies perhaps 2 or 3% of the area of the Reserve

and is confined to dunes, screes and scarps between

the plateau rim and the sea.

In the following account a distinction is made

between the Sideroxylon Thicket which develops on

calcareous marine sand, and the Maurocenia— Chi-

onanthus Thicket on acid sandstone substrates. This

distinction finds support in the work of Campbell &

Moll (1977) who showed that Maurocenia is a

constituent of certain forest communities on the acid

rocky substrates of Table Mountain. These authors

did not survey the Sideroxylon Thicket abutting the

western slopes of Table Mountain, but Boucher

(1978) did record the presence of Maurocenia

frangularia in his Sideroxylon inerme Dune Scrub in

the Cape Hangklip area. In this instance, however,

Maurocenia was found only in two localities, both on

older stable dunes where the substrate of marine

sand might be more acidic owing to the leaching

action of rain and to the influence of nearby

outcrops of Table Mountain sandstones (Boucher,

pers. comm.).

More quantitative work on coast-thicket commu-

nities is needed to determine their finer floristic and

successional relationships with the woody element of

284

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

Fig. 25. — Plot 25. Coast Thicket

with Euclea racemosa to 1,75

m tall in foreground, Cassine

barbara 2,5 m tall to rear.

Escarpment edge on right

skyline, with seepage step

below it. Olifantsbos.

western Strandveld on the one hand and with

eastern forest on the other.

4.2.1 Sideroxylon Thicket on marine sand of the

coastal dunes

In sheltered slacks of the coastal dunes the woody

element invades and gradually ousts Dune Mixed

Fynbos, forming a dense thicket about 1,5 -2,5 m

tall (Table 1). Typically found are Colpoon

compressum, Euclea racemosa, Cassine maritima,

Olea exasperata, Pterocelastrus tricuspidatus and

Rhus glauca. Chionanthus foveolatus occurs sparsely

in the best-developed thicket. Openings are filled

with 90 cm shrubs of Chrysanthemoides monilifera,

Helichrysum dasyanthemum and Metalasia muricata,

while an open field layer is formed by Carpobrotus

acinaciformis, Cineraria geifolia, Ficinia ramosissi-

ma and some grasses. Climbers are Antizoma

capensis, Asparagus aethiopicus, Cynanchum africa-

num var. africanum, Dipogon lignosus, Galium

tomentosum and Kedrostis nana; Solanum guineense

and 5. quadrangulare are also sometimes scandent.

The adequate litter formed from leaf-fail of Euclea

racemosa breaks down into a rich, well-developed

mull humus 5-8 cm deep (Fig. 25). This dense

mixed thicket is floristically similar to the Colpoon-

Rhus Dune Scrub of the Hangklip area (Boucher,

1978).

At a later stage in the succession, Sideroxylon

inerme appears in the ‘Mixed Scrub’. With

prolonged protection this small tree becomes

dominant, forming the dense ‘Melkhout’ low

thickets still found at Buffel’s Bay (Fig. 26),

Bordjiesrif, Platboom and Bloubergstrand. Here

Sideroxylon inerme occurs as almost pure stands

3-4,5 m tall, with Cussonia thyrsiflora, a thick

woody scrambler, as its chief associate besides the

climbers listed above. When directly exposed to the

southeaster, these thickets scarcely attain a height of

Fig 26 .—Sideroxylon Thicket

about 4 m tall, wind-moulded

into a humped, impenetrable

thicket. Alien Acacia cyclops

encroaching below the white

cliffs on ridge. Buffels Bay.

H. C. TAYLOR

285

2,5 m and are wind-moulded into a humped,

impenetrable low thicket (Fig. 26). Parts of the

thicket have abrupt margins indicating their retreat

from a former larger area, probably as a result of

fire. Human trampling and picnicking, if continued,

will further damage this community.

The Sideroxylon Thicket at Cape Point closely

resemble Boucher’s (1978) Sideroxylon inerme Dune

Scrub but the latter appear to be structurally more

developed.

4.2.2 Maurocenia— Chionanthus Thicket on sand-

stone colluvium near the sea

With protection from fire, the Coleonema Fynbos

of the rocky littoral (4. 1.1. 4) develops into a dense,

tall, closed shrubland, scarcely 1,5 m tall and quite

impenetrable. The following species are conspicu-

ous: Cassine maritima, Euclea racemosa, Maytenus

lucida, Pelargonium gibbosum, Pterocelastrus tricus-

pidatus, Rhus laevigata, Rhus lucida and Tar-

chonanthus camphoratus. At a later stage Olea

exasperata, Cassine barbara and Cussonia thyrsiflora

appear. Sideroxylon inerme sometimes occurs,

though it does not become dominant.

With longer protection, Euclea racemosa and

Tarchonanthus camphoratus grow to over 4,5 m tall.

Their spreading crowns and slender stems produce

sheltered conditions, and their decaying leaves form

humus that creates a seedbed in which the two

climax species, Maurocenia frangularia and Chio-

nanthus foveolatus, can germinate.

The highest development of thicket on the

Reserve, indicated by the presence of these two

species, occurs on the steep, bouldery, southerly

slopes above Gifkommetjie where an encircling arc

of sandstone cliffs has afforded protection from fire

possibly for centuries (Fig. 27). Other examples of

well-developed thicket occur on the Atlantic coast

among large boulders on the slopes above Blou-

bergstrand, and on the south-western aspects of the

rocky escarpment from Brightwater to Hestersdam.

On the False Bay coast there is a small patch of scrub

forest at the foot of the steep northeast-facing slopes

at The Chair, about 800 m south of Smitswinkel Bay.

A modified stunted thicket occurs on the southeast-

facing precipices in the vicinity of Batsata Cove.

The following account of the climax low thicket is

based on observations made at Gifkommetjie,

supplemented by data gathered at The Chair.

Tarchonanthus camphoratus, a grey-leaved many-

stemmed spreading tree to 4,5 m tall, is common,

often dominant, frequently occurring in small pure

stands. Maurocenia frangularia, a heavy-stemmed

tree to 4,5 m tall, with spreading crown of broad,

dark green, glossy, leathery leaves, is fairly common

to co-dominant, almost confined to the climax.

Chionanthus foveolatus, a small erect tree, is only

found in the climax community. Woody associates

are Cassine barbara, Colpoon compressum, Cusso-

nia thrysiflora, Euclea racemosa, Olea capensis

subsp. capensis, O. exasperata, Phylica buxifolia,

Pterocelastrus tricuspidatus, Rhus glauca, R. lucida,

R. laevigata, R. tomentosa and Sideroxylon inerme.

Occasionally, Maytenus heterophylla is found on the

east coast.

The shrub layer contains Myrsine africana where

the canopy is open, e.g. along the margins. Shrubs

commonly found in openings within the forest, such

as Pelargonium gibbosum, Polygala myrtifolia and

Salvia aurea, indicate the coastal affinities of this

woody community. The field layer is only developed

in the rare places where there is an open glade under

the trees. Here a dense cover of soft annual herbs

(including Australina lanceolata ) and grasses, about

15 cm tall, develops on the richly humified soil,

interspersed with scattered herbaceous perennials

and semi-scandent shrubs such as Knowltonia

capensis and species of Sutera, Solanum and

Tetragonia. Among the climbers, Asparagus aeth-

iopicus is common; Cynanchum obtusifolium and

Kedrostis nana are less so. Astephanus triflorus,

Cynanchum africanum var. africanum, Dipogon

lignosus (sprawling over crowns of trees in disturbed

forest) and Bolusafra bituminosa are more common

in early stages of the thicket.

Fig. 27. — Broad-leaved Thicket on steep talus slopes at Gifkommetjie. Top right, Maurocenia frangularia (dark foliage); centre,

Tarchonanthus camphoratus (pale foliage). Lower slopes, tangled, wind-moulded Coast Thicket on marine sand of the coastal

shelf. The boundary between the two communities, following the soil distinction, is well-marked. Circle indicates the position

of Plot 81, an ill-defined bushy Hillveld assemblage, long unburnt.

286

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

Fig. 28. — Bush-cutting Leucaden-

dron coniferum on fringe of

inland dune to encourage

grass regeneration for game.

Near Plot 78, Gifkommetjie

road turn-off.

Fig. 29. — Bontebok on artificial

pasture, Gifkommetjie road

turn-off.

Two special thicket communities, each dominated

by a single species, are rare but noteworthy since

they have affinity with the Broad-leaved Thicket.

(1) Tarchonanthus camphoratus forms pure

stands in warm sheltered places near the False Bay

coast, for example at Platbank where it grows to 3,5

m tall.

(2) Philippia chamissonis, an ericoid-leaved

rounded shrub about 2,5 m tall with dense crown

reaching to the ground, occurs in stands of about one

hectare near Pegrams Point, Gifkommetjie and

Matroosdam. Usually the bushes are spaced 2-3 m

apart and form pure stands but sometimes, where

long protected from fire, moribund P. chamissonis

has thinned to about double this espacement, the

spaces between being filled with woody shrubs

typical of thicket, and undershrubs of Coast Fynbos.

From this it appears that the Philippia chamissonis

Community may be transitional to Broad-leaved

Thicket. Nevertheless, the presence of a typical

stand of P. chamissonis at an altitude of 213 m on the

Red Hill plateau (outside the Reserve) 5 km from

the sea and surrounded by short mountain fynbos, is

evidence that it is not always associated with coast

vegetation.

5 DISCUSSION AND CONCLUSIONS

The vegetation of the study area is diverse,

comprising both Inland and Coast Fynbos as well as

elements of forest and Strandveld. The flora is

correspondingly rich, with nearly 1 100 species in an

area of 7 750 ha. The Inland Fynbos of the Reserve,

however, is not very typical of Mountain Fynbos —

the name now used (Taylor, 178) for Acocks’s (1975)

Macchia and False Macchia. Much of the area

consists of a central plateau where soil depth,

drainage and other habitat features intergrade and

differ from those commonly found in the rest of the

VEGETATION MAP OF THE

CAPE OF GOOD HOPE

NATURE RESERVE

1969

H. C. TAYLOR

289

mountain terrain of the Capensis region. These

factors, and nearly two centuries of disturbance by

modern man, blur the boundaries of plant commu-

nities and render some of them difficult to interpret.

The harsh climate of the Reserve affects mainly

the physiognomy of the vegetation. The strong,

often gale-force, south-east winds of summer cause a

blanket of moisture-laden cloud to form on the high

peaks at a season when the lower hills and plateau

receive little or no precipitation, hence the peaks

bear a montane vegetation lower in stature and more

mesophytic than that of the hills and plateau.

Similarly, the southeast-facing slopes of the

mountains overlooking False Bay support low

thicket despite their shallow rocky substratum and

precipitous gradient.

Soil conditions, on the other hand, which affect

floristic composition, largely determine the plant

communities. On the whole, the correlation between

soil, topography and vegetation is remarkably good.

This is all the more striking because there are few

major differences in parent material of Cape Point

soils. The minor differences that exist are determi-

ned by varying soil climate which in turn is affected

by topography. That even these differences are

closely reflected in vegetational changes shows that

Cape Point vegetation is a sensitive indicator of

edaphic factors.

Of the biotic factors, fire is the most important;

and fire can here be regarded as a biotic factor

because most of the fires in this climate are, and for

millenia have been, caused by man. From the early

days of European colonization up to the middle of

this century, repeated burning combined with

overgrazing had degraded parts of the Reserve to

form secondary communities determined more by

fire history than by the physical habitat. With the

creation of a nature reserve, an attempt was made to

exclude fire and much of the fynbos, adapted

through millenia to regenerate by burning, became

moribund and highly inflammable. The present

management plan for the Reserve makes provision

for periodic burning of the veld at sufficiently long

intervals to ensure that both species diversity and

community diversity are maintained.

Alien woody plants that were introduced during

the farming period have, over the last half-century,

invaded the veld to such an extent that they have

become a serious management problem. The

authorities are now vigorously tackling this problem

by systematic weeding of the areas burnt, coupled

with mass eradication of the dense infestations that

occur chiefly in the extreme northern and southern

parts of the Reserve.

The vegetation units described here have not been

studied over their whole range and are not based on

quantitative floristic data. Hence, no attempt has

been made to give them either formal name or rank

in the standard hierarchy of syntaxonomic nomen-

clature recognized by the Braun-Blanquet school.

Most of these communities would seem to be

roughly equivalent to the rank of association, but

this cannot be established with certainty until a

full-scale phytosociological survey has been done.

Moreover, until such a survey has been done, the

Cape Point communities cannot be fully equated

with those distinguished in other fynbos studies (cf.

Werger, Kruger & Taylor, 1972; Boucher & Jarman,

1977; Van der Merwe, 1977; McKenzie et al., 1977;

Campbell & Moll, 1977; Glyphis et al., 1978; Laidler

et al., 1978 and Boucher, 1978) who have floristically

analysed vegetation and shown the composition of

each syntaxon in phytosociological tables. Never-

theless, certain preliminary observations and com-

parisons may be made.

Firstly, Cape Point vegetation on the whole

appears to have little in common with the vegetation

of Table Mountain, especially the Front Table.

Species characteristic of Cape Point Hillveld and

Sandflats are poorly represented on the Back Table

and virtually absent from the Front Table. Two

species characteristic of Cape Point Mountainveld

(Elegia juncea and Leucadendron xanthoconus ) are

constituents of Back Table communities but only the

former occurs on the Front Table. Cape Point

Tussock Marsh species are poorly represented on

Table Mountain, but the presence of Restio

compressus, Osmitopsis asteriscoides, Berzelia lanu-

ginosa and Psoralea pinnata in some Table Mountain

communities, especially those of the Back Table,

suggests the existence there of a high-altitude

equivalent of Cape Point Seepage Fynbos. Thus,

what little affinity does exist between Cape Point

and Table Mountain seems to be stronger with the

Back Table than with the Front; and it might be

significant to observe that altitude, precipitation and

distance all increase as one travels northward from

Cape Point via the Back Table to the Front.

Both the Langebaan area (Boucher & Jarman,

1977) and the Cape Hangklip area (Boucher, 1978),

though further away from Cape Point than is Table

Mountain, appear to have stronger vegetational

affinities with Cape Point. Since local workers had

access to the Cape Point study in thesis form since

1969, these affinities were discussed in their texts

and will not be repeated here. The Langebaan

strand plant communities are linked to Cape Point

by the direct pathway that exists along the Cape west

coast, a pathway already recognized and mapped by

Acocks in 1953. Affinities with Cape Hangklip are

not as easily explained, at least in so far as the

mountain communities are concerned. The southern

Cape Peninsula and the Cape Hangklip area are

separated by some 40 km of dunefields in the

northern part (the Cape Flats) and by about 30 km

of sea in the southern part (False Bay); both of these

could act as considerable barriers to the migration of

mountain plants. But at various times, both during

and before the Pleistocene, the sea level was as

much as 140 m lower than at present (Dingle &

Rogers, 1972), False Bay was dry land and the Cape

Point and Cape Hangklip sandstone massifs were

extended towards one another, thus creating a more

direct route for migration than is possible today.

ACKNOWLEDGEMENTS

Besides those whose help was acknowledged in

the first paper, I am indebted to my colleagues at

Stellenbosch, especially Mr C. Boucher for stimula-

290

A VEGETATION SURVEY OF THE CAPE OF GOOD HOPE NATURE RESERVE.

II. DESCRIPTIVE ACCOUNT

ting discussion over many years, and the late Mrs

Mary Rand for her patient and meticulous

determination of my specimens, many of which,

from Cape Point, were scrappy ‘problems’.

UITTREKSEL

Hierdie referaat beskryf die verkenningsopname

van die Kaap die Goeie Hoop Natuurreservaat wat as

agtergrond gedien het vir die proef van die metodes

waaroor daar in die eerste referaat verslag gedoen is.

'n Plantegroei kaart is ingesluit en ’n beskrywing is

gegee van die fisiese eienskappe, klimaat, geskiede-

nis, en meer volledig die plantegroei van die

Reservaat.

Twee strukturele formasies word herken. ’n

Breeblaar struikgewas wat verwant is aan die

kuswoude van die Knysnastreek, neem ca 3% van

die Reservaat in beslag. Hierdie formasie word

beperk tot duine, glyhellings en skarpe wat gewoon-

weg goed beskerm is teen vuur. Die oorblywende

plantegroei is alles fynbos. Die twee floristiese klasse

van fynbos naamlik Kus- en Binnelandsefynbos, dui

die twee hoof grondtipes aan: verplaasde kusgronde

wat gewoonlik neutraal of alkalies is, en binnelandse

suurgronde afkomstig van die Tafelbergsandsteen

substratum. Vier Kusfynbos en ses Binnelandsefyn-

bos gemeenskappe word beskryf. Vuur is die

belangrikste biotiese faktor wat die plantegroei

beinvloed. Baie van die kusgemeenskappe neig om in

Breeblaar struikgewas te ontwikkel as hulle teen vuur

beskerm word, terwyl sommige Binnelandsefynbos

gemeenskappe ’n sikliese selfinstandhoudende suk-

sessie toon.

Dit blyk asof die plantegroei van die Reservaat ’n

groter floristiese verwantskap met Langebaan en

Kaap Hangklip toon as met Tafelberg nieteenstaande

die feit dat laasgenoemde geografies nader aan die

Reservaat gelee is.

REFERENCES

Acocks, J. P. H., 1975. Veld Types of South Africa. Mem. bot.

Surv. S. Afr. 40.

Adamson, R. S., 1938. The vegetation of South Africa. London:

British Empire Vegetation Committee.

Adamson, R. S. & Salter, T. M., 1950. Flora of the Cape

Peninsula. Cape Town & Johannesburg: Juta.

Ashby. E., 1936. Statistical ecology. Bot. Rev. 2: 221—235.

Becking, R. W., 1957. The Zurich-Montpellier School of

Phytosociology. Bot. Rev. 23: 411-488.

Boucher, C., 1978. Cape Hangklip area. II. The vegetation.

Bothalia 12: 455-497.

Boucher, C. & Jarman. M. L., 1977. The vegetation of the

Langebaan area. South Africa. Trans. R. Soc. S. Afr. 42:

241-272.

Boucher, C. & McDonald, A. E., 1982. An inventory of plant

communities recorded in the western, southern and eastern

Cape Province, South Africa up to the end of 1980. S. Afr.

Nat. Sc. Programmes Report 57, CSIR pp. 62.

Braun-Blanquet, J., 1932. Plant sociology. New York &

London: McGraw Hill.

Campbell, B. M. & Moll, E. J., 1977. The forest communities of

Table Mountain, South Africa. Vegetatio 34: 105-115.

Day,J. A., 1981, Limnology of standing waters in the fynbos areas

of the south-western Cape. Project Report, Fynbos Biome

Project, Co-operative Scientific Programmes, CSIR.

Deacon, H. J., 1972. A review of the Post-Pleistocene in South

Africa. S. Afr. Archaeol. Soc. Goodwin Series 1: 26— 45. »

Dingle, R. V. & Rogers, J., 1972. Effects of sea-level changes

on the Pleistocene palaeoecology of the Agulhas Bank. In E.

M. van Zinderen Bakker, Palaeoecology of Africa and of the

surrounding islands and Antarctica 6: 55-58. Cape Town:

Balkema.

Edwards, D., 1983. A broad-scale structural classification of

vegetation for practical purposes. Bothalia 14: 705-712.

Flemming, B. W., 1977. Distribution of recent sediments in

Saldanha Bay and Langebaan Lagoon. Trans. R. Soc. S.

Afr. 42: 317-340.

Fuggle, R. E., 1981. Macroclimatic patterns within the Fynbos

Biome. Nat. Progr. for Envir. Sciences Fynbos Biome

Project, Final Report pp. 116. Pretoria: CSIR.

Glyphis, J., Moll, E. J. & Campbell, B. M., 1978.

Phytosociological studies on Table Mountain, South Africa:

1. The Back Table. Jl S. Afr. Bot. 44: 281-289.

Goldblatt, P., 1978. An analysis of the flora of southern Africa:

its characteristics, relationships and origins. Ann. Mo. bot.

Gdn 65: 369-436.

Kruger, F. J., 1977. Ecology of Cape fynbos in relation to fire.

Proceedings of the Symposium on the Environmental

Consequences of Fire and Fuel Management in Mediter-

ranean Ecosystems. General Technical Report WO-3:

230—244. Washington, D.C.: USDA Forest Services.

Laidler, D., Moll, E. J., Campbell, B. M. & Glyphis, J., 1978.

Phytosociological studies on Table Mountain, South Africa:

2. The Front Table. Jl S. Afr. Bot. 44: 291-295.

Mabbutt, J. A., 1952. The physical landscape. In J. A. Mabbutt,

The Cape Peninsula, 13—32. Cape Town: Maskew Miller.

Mabbutt, J. A., 1955. The physiography of an area adjoining

Cape Maclear. Trans. R. Soc. S. Afr. 34: 61-67.

Macvicar, C. N., De Villiers, J. M., Loxton, R. F., Verster,

E., Lambrechts, J. J. N., Merryweather, F. R., Le

Roux, J., Van Rooyen, T. H. & Harmse, H. J. von M.,

1977. Soil classification: a binomial system for South Africa.

Pretoria: Soil and Irrigation Research Institute, Department

Agricultural Technical Services.

Maggs, T. & Speed, E., 1967. Bonteberg shelter. S. Afr.

Archaeol. Bull. 22: 80—93.

Martin, A. R. H., 1965. Plant ecology of the Grahamstown

Nature Reserve. 1. Primary communities and plant succes-

sion. Jl S. Afr. Bot. 31: 1-54.

McKenzie, B., Moll, E. J. & Campbell, B. M., 1977. A

phytosociological study of Orange Kloof, Table Mountain,

South Africa. Vegetatio 34: 41-53.

Moll, E. J. & Gubb, A. A., 1981. Aspects of the ecology of

Staavia dodii in the south-western Cape of South Africa. In

H. Synge, The biological aspects of rare plant conservation.

New York: Wiley.

Opie, F. W. G., 1967. The ecology and geographical development

of the Cape Point area. Geography Department project.

University of Cape Town.

Phillips, J. F. V., 1931. Forest-succession and ecology in the

Knysna region. Mem. bot. Surv. S. Afr. 14.

Schaffer, W., 1952. The climate. In J. A. Mabbutt, The Cape

Peninsula, 33-45. Cape Town: Maskew Miller.

Taylor, H. C., 1969a. A vegetation survey of the Cape of Good

Hope Nature Reserve. M.Sc. thesis, University of Cape

Town.

Taylor, H. C., 1969b. Pesplante en natuurbewaring. For. S. Afr.

10: 41-46.

Taylor, H. C., 1972a, Fynbos, Veld Flora 2: 68-75.

Taylor, H. C., 1972b. Notes on the vegetation of the Cape Flats.

Bothalia 10: 637-646.

Taylor, H. C. , 1977. Aspects of the ecology of the Cape of Good

Hope Nature Reserve in relation to fire and conservation.

Proceedings of the Symposium on the Environmental

Consequences of Fire and Fuel Management in Mediter-

ranean Ecosystems. USDA Forest Services General Techni-

cal Report WO-3, 483-487.

Taylor, H. C., 1978. Capensis. In M. J. A. Werger,

Biogeography and ecology of southern Africa, 171—229. The

Hague: Junk.

Taylor, H. C., 1979. Observations on the flora and phytogeo-

graphy of Rooiberg, a dry fynbos mountain in the southern

Cape Province, South Africa. Phytocoenologia 6: 524—531.

Taylor, H. C., 1981. Plant communities along the False Bay

coast. In B. Gasson, Proceedings of a seminar on the future

management of False Bay, 59—67. Cape Town: False Bay

Conservation Society.

Taylor, H. C., 1984a. A vegetation survey of the Cape of Good

H. C. TAYLOR

291

Hope Nature Reserve. I. The use of association-analysis and

Braun-Blanquet methods. Bothalia 15: 245-258.

Taylor, H. C., 1984b. An analysis of the Flora of the Cape of

Good Hope Nature Reserve. 5. Afr. J. Bot. In press.

Taylor, H. C. & Morris, J. W., 1981. A brief account of the

coast vegetation near Port Elizabeth. Bothalia 13: 519-525.

Van der Merwe, C. V., 1976. Die plant ekologiese aspekte en

bestuursprobleme van die Goukamma-Natuurreservaat.

M.Sc. thesis. University of Stellenbosch.

Van der Merwe, C. V., 1977. ’n Plantegroei opname van die De

Hoop-Natuurreservaat, Cape Dept. Nature & Environmental

Conservation internal report, 1-29.

VanDonk, J., 1976. 018 record of the Atlantic Ocean for the

entire Pleistocene Epoch. In R. M. Cline & J. D. Hays,

Investigation of the late Quaternary palaeo-oceonography

andpalaeoclimatology. Mem. geol. Soc. Am. 145: 147-163.

Walker, F., 1952. The geology. In J. A. Mabbutt, The Cape

Peninsula, 1 — 12. Cape Town: Maskew Miller.

Weather Bureau, 1954. Climate of South Africa. Part 1: climate

statistics. WB 19. Pretoria: Weather Bureau.

Weather Bureau, 1965. Climate of South Africa. Part 9: average

monthly rainfall up to the end of I960. WB 29. Pretoria:

Weather Bureau.

Werger, M. J. A., Kruger, F. J. & Taylor, H. C., 1972. A

phytosociological study of the Cape fynbos and other

vegetation at Jonkershoek, Stellenbosch. Bothalia 10:

599-614.

Bothalia 15, 1 & 2: 293-295 (1984)

Miscellaneous ecological notes

VARIOUS AUTHORS

GRASS ROOT PATTERN IN AN ORANGE FREE STATE FLOODPLAIN

The floodplain downstream of the confluence of

the Vet and Sand Rivers, Orange Free State, was

visited from the 27th to the 29th June 1979 to

determine possible correlations between root activi-

ty and soil moisture conditions. Although no

correlation between root activity and the soil

moisture regime could be found at the time of the

investigation, the results proved interesting as few

data concerning grass root activity are available.

According to Acocks (1975), the area consists of Pan

Turf Veld (Veld Type 51) where the principal

species are Themeda triandra, Panicum coloratum,

Eragrostis spp., Setaria woodii, Sporobolus fimbria-

tus and Digitaria argyrograpta with Echinochloa

holubii, Sporobolus tenellus, Platycarpha parvifolia,

Diplachne fusca, Panicum laevifolium and Eragrostis

bicolor in the wetter low-lying parts.

Twenty-four newly-drilled boreholes, each ap-

proximately two metres deep and one metre in

diameter, distributed over an area extending from

the confluence of the Vet and Sand Rivers to 50 km

downstream on the floodplain, were investigated. A

random vertical transect, in which a 50 mm wide

strip of soil was removed, from soil surface to

borehole floor, to expose the grass roots, was

analysed in each borehole. The total number of

observable grass roots, in each 100 mm segment of

the transect, was counted to determine root

distribution patterns. The dominant plant species in

the vicinity of each borehole were noted where the

vegetation condition made identification possible.

Photographs illustrating vegetation cover were also

taken at the borehole sites.

ROOT - COUNT

Fig. 1. — The average number of

grass roots counted in each

100 mm of the borehole

transects.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 IS

MANY ROOTS / MATTED

Fig- 2. — Vegetation with a high

canopy cover, taken next to a

borehole.

294

MISCELLANEOUS ECOLOGICAL NOTES

Fig. 3. — Vegetation with a low

canopy cover taken next to a

borehole.

The soil forms present in the area, which were

revealed by the boreholes, were Arcadia, Bonheim,

Oakleaf, Rensburg and Valsriver.

The maximum depth of observable grass roots

varied between 400 mm and 2 000 mm. In most cases

the majority of grass roots were observed in the

uppermost 100 mm of each transect with a

decreasing root-count with increasing depth. The

only exception to this trend was the rooting pattern

of a single sample in the Oakleaf Form where the

majority of roots occurred above the neocutanic B

horizon, at 200 to 300 mm depth. The average

root-count for all the transects is illustrated

graphically in Fig. 1. The vegetation surrounding the

boreholes was mainly grass with species of Cynodon,

Diplachne, Eragrostis, Panicum and Themeda

identifiable and only isolated forbs present. It was

possible in most cases to follow the grass roots from

the above-ground plant to the bottom of the

borehole. The grass cover in the vicinity of the

boreholes varied from high to low (Figs 2 & 3).

Correlation co-efficient tests on the correlation

between grass root-count and maximum observed

grass root-depth, soil forms present in the area,

vegetation cover and species present in the vicinity

of each borehole, indicated little or no correlation

between these factors. It would therefore seem that

the root-count and root-depth of the abovemention-

ed grass species are independent of the described

soil forms as well as the vegetation cover.

Furthermore, grass roots of the species observed are

confined mainly to the first 100 mm of soil but may

penetrate to depths of two metres or more. The

authors thank Dr J.C. Scheepers for comments and

suggestions.

REFERENCE

Acocks, J. P. H., 1975. Veld types of South Africa. Mem. bot.

Surv. S. Afr. No. 40: 1-128.

R. H. Westfall* and R. Drewes**

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

** Agricultural College, Private Bag X804, Potchefstroom 2520.

SECTORS OF THE TRANSVAAL PROVINCE OF SOUTH AFRICA

Natural science field researchers usually describe

the precise location or boundaries of their areas

under study (for example, Edwards, 1981). Concise

reference to the broad location of regions studied is

often also necessary, particularly for use in the

abstract of the published accounts of results. Past

references to broad regions of the Transvaal

province of South Africa have revealed considerable

inconsistency and confusion, for example, when the

term ‘northern Transvaal’ is used with reference to

work done in the vicinity of Pretoria and to that on

the border with Zimbabwe.

Although common language usage may influence

the choice of terms to describe various sectors of the

Transvaal, the information conveyed by regional

terms, especially to other researchers unfamiliar

with Transvaal, should preferably be consistently

used across the different natural scientific discipli-

nes.

In an attempt to place the choice of terms to

describe broadly but concisely different sectors of

the irregularly outlined Transvaal on a neutrally

acceptable, logical footing, the following simple

approach was used. The main point of reference

taken, was the centre of gravity of the plane surface

representation of the Transvaal (1976 boundary).

This point was found to be 25°02'S; 28°53'E (± T).

The proposed divisions are made relative to this

point. The scheme is flexible to cater for specific

needs by allowing overlapping entities.

The proposed terms for the sectors are given

below in menu form, with their boundaries indicated

in Fig. 4.

north half Transvaal

south half Transvaal

Choice 1 (halves):

VARIOUS AUTHORS

295

Fig. 4. — Sectors of the Transvaal. The circle that runs through Pretoria represents the limits of the central Transvaal. Base map by

courtesy of Dr J. M. Anderson.

Choice 2 (halves):

Choice 3 (quadrants):

Choice 4 (quadrants):

Choice 5:

east half Transvaal

west half Transvaal

north-eastern Transvaal

north-western Transvaal

south-eastern Transvaal

south-western Transvaal

northern Transvaal

southern Transvaal

eastern Transvaal

western Transvaal

central Transvaal

there is a choice of at least four defined sector

descriptions, whereas in the central part there is a

choice of five. The choice of an appropriate term will

normally be determined by that sector in which the

locality(ies) of interest is (are) most centrally placed.

ACKNOWLEDGEMENTS

The authors thank Miss A. Stadler and Mrs J.

Schaap for assistance with the drawing and Dr J. C.

Scheepers for comments and suggestions.

The last-mentioned central Transvaal is designed

to overcome the problem of the high degree of

sector divisibility in the region surrounding the

central point and equals an area one ninth that of the

Transvaal. This assumes a linear centre third

fraction that converts to area in the form of (V3)2 =

V9 with a resulting radius of 100 km.

It follows that for any given point in the Transvaal

REFERENCE

Edwards, D., 1981. A note on the extension of the degree

reference system for citing biological distribution records to

north of equator and west of Greenwich meridian. Bothalia

13: 574-575.

M. C. Rutherford* and R. H. Westfall*

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

Bothalia 15, 1 & 2: 297-313 (1984)

Exotic woody plant invaders of the Transvaal

L. HENDERSON* and K. J. MUSIL*

Keywords: alien, checklist, distribution, woody invader, survey, Transvaal

ABSTRACT

The frequency and abundance of exotic, woody plant invaders were recorded in 60% of the quarter degree

squares in the study area. Sixty-one invaders were encountered of which the most important and aggressive were

Acacia dealbata, Populus spp., Melia azedarach, Opuntia ficus-indica, Salix babylonica and Acacia mearnsii.

Invasion patterns are discussed and an attempt is made to correlate distribution with environmental factors.

Attention is drawn to the areas of greatest invasion and the areas that are liable to show the greatest expansion in

the future.

1 INTRODUCTION

1.1 Survey history and objectives

The objectives of this survey are: to produce a

checklist of the major exotic, woody invaders of

streambank, roadside and veld habitats in the

Transvaal (see Appendix); to determine the pattern

of exotic woody invasion as a whole and for

individual species; to attempt to correlate distribu-

tion with environmental factors; to determine which

are the most important and aggressive invaders.

The first part of the Transvaal that was surveyed

was the south central region, 10 000 km2 (Wells et

al., 1980). The survey was then extended to include

the section east of Pretoria to the Kruger National

Park, and stretching from Volksrust in the south to

Louis Trichardt in the north (Duggan & Henderson,

1982). The survey has now been completed with the

inclusion of 113 samples in the western sector and a

further 58 samples in the eastern sector (Fig. 1).

1.2 The study area

The Transvaal covers an area of approximately

262 449 km2. The greater part of the province

consists of table lands (highveld) ranging from 600

— 1 800 m in altitude, the highest point being in the

east where the Drakensberg overlooks the lowveld.

The lowveld (altitude range of 200 — 600 m) and the

low-lying portions of the northern and central

sections of the province have a tropical to

sub-tropical climate. The southern highveld and the

upper slopes of the escarpment are more temperate

and experience frost in winter.

The highest rainfall in the Transvaal occurs along

the upper slopes of the Drakensberg escarpment and

the average rainfall can exceed 2 000 mm per

annum.

The driest parts of the Transvaal occur along the

Limpopo River Valley in the north and surrounding

country, the eastern lowveld and the south-western

Transvaal. These parts experience a mean annual

rainfall of 250 — 500 mm.

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

The rest of the Transvaal receives a mean annual

rainfall of 500 — 900 mm, with the rainfall generally

increasing from west to east.

Indigenous woody vegetation is concentrated on

the wet escarpment face, where tropical forests

occur (Acocks Veld Types 8, 9) and in the hot, drier

lowveld and middleveld where bush and savanna

types occur (Acocks Veld Types 10, 11, 12, 13, 14,

15, 16, 18, 19, 20). The cooler highveld is mostly

covered by grassveld (Acocks Veld Types 48, 50, 52,

53, 54, 55, 57) and false grassveld (Acocks Veld

Types 61, 62, 63, 64, 67) with scrub and woodland

fringing the streams (Fig. 2) (Acocks, 1975).

2 METHOD

2.1 Sampling method and intensity

The sampling method was specifically designed to

make use of otherwise unproductive travelling time

whilst officers are engaged on other projects.

The method consisted of making continuous

recordings of roadside and veld invaders from a

moving vehicle and of streambank invaders at

water-course crossings (preferably while the vehicle

was stationary). Abundance ratings were based on

frequency of encounter within each sample unit (the

quarter degree square). Abundance ratings for

streambank habitats were based on estimates at

specific sites.

2.2 Abundance ratings

Since the last publication (Duggan & Henderson,

1982) changes have been made to the abundance

ratings for roadside and veld habitats. Two ratings

have been inserted between the original 5 and 6

ratings and 2 have been inserted below the old rating

1, i.e. there are now 11 abundance ratings. This was

done because the gap between the 5 and 6 ratings

was too large and the rating 1 was not low enough to

accommodate the most infrequently occurring

species.

The abundance ratings for roadside and veld

habitats and streambank habitats are given in Table

1.

298

EXOTIC WOODY PLANT INVADERS OF THE TRANSVAAL

TABLE 1. — Abundance ratings

2.3 Sampling level envisaged and achieved

The minimum sampling level that was envisaged

for the Transvaal was 40% of the total quarter

degree squares at an average of 64 km per square.

This was based on an analysis of the effectivity of

sampling at various levels (Henderson & Duggan,

unpublished report).

The sampling level that has been achieved is: 60%

(248 of the total 410 quarter degree squares) at an

average of 33 km per square. That is, a greater

number of sampling units was sampled but at a less

intensive level. Twenty-two out of the total of 24

veld types recognized by Acocks (1975) as being

present in the Transvaal were sampled. The smallest

veld types, numbers 53 and 55 were unsampled.

TABLE 2. — Veld type coverage, excluding pilot study area

Invader species were recorded separately in each

veld type passed through in each sample unit.

The sampling coverage for each Acocks veld type

in terms of quarter degree squares sampled and total

kilometres travelled is shown in Table 2. The

number of water-course recordings, quarter degree

squares sampled and kilometres travelled in each of

the four broad veld type categories, tropical veld

types, temperate veld types and the whole survey

area are shown in Table 3.

TABLE 3. — Veld type coverage, excluding pilot study area

3 RESULTS

Sixty-one invaders were encountered in the

Transvaal. These species are listed in the checklist

together with a further 17 which were obtained from

various literature and other sources (see Appendix) .

3.1 The streambank habitat

Four hundred and ten recordings were made in

which 40 different exotic, woody invader species

were recorded, with up to 9 species in one sample.

3.1.1 Analysis according to veld type

The number and abundance of invader species

recorded in streambank samples in each of the four

broad veld type categories are given in Table 4.

L. HENDERSON AND K. J. MUSIL

299

TABLE 4. — The number and abundance of invader species

recorded in streambank samples

3.1.2 Analysis according to species

3. 1.2.1 Frequency: The percentage frequency of

occurrence of exotic woody invader species along

streambanks in forest, bushveld, false grassveld,

pure grassveld, as well as in tropical and temperate

veld types and in the whole survey area is shown in

Figs 3 — 9. Fig. 9 (whole survey area) confirms the

earlier finding (Duggan & Henderson, 1982) that the

most widespread invaders are: Salix babylonica,

Populus albalcanescens (see checklist for explana-

tion), Acacia dealbata and Melia azedarach. The

latest sampling, however, has shown that:

1. Salix babylonica is not the most frequent

invader throughout the temperate veld types but in

the false grassveld 'it ' is superseded by Acacia

dealbata and joined by Populus albalcanescens and

Acacia mearnsii (Fig. 5),

2. Melia azedarach is not the most frequent

invader throughout the tropical veld types, there

being several more frequent invader species in the

forest veld types,

3. Acacia dealbata and Acacia mearnsii are more

frequent, particularly in the temperate veld types,

than was previously recorded. Although these

species (together with Psidium guajava and Caesal-

pinia decapetala) were not recorded in fhe western

Transvaal — a total of 113 quarter degree squares —

this is most likely due to unavailability of source

plants and not unsuitability to the environmental

conditions,

4. An additional 18 exotic, woody invader species

have been recorded since our last report. These

include: Agave spp., Eucalyptus spp., Pinus spp.

including P. patula, Opuntia ficus-indica, Toona

ciliata, Populus nigra, Pyracantha angustifolia and

Bambusa balcooa.

3. 1.2.2 Importance: The importance value for each

species was calculated as follows:

total abundance of a sp. 100

x

sum of the abundance values of all spp. 1

importance +

value = frequency of a sp. 100

x

sum frequency of all spp. 1

The species with the highest importance values

and which collectively added up to a total of

approximately 160 points (out of a maximum of 200

points) were plotted in Figs 10, 11 & 12.

A comparison of the percentage frequency

histogram Fig. 9 and importance histogram Fig. 12

for the whole survey area shows that Salix

babylonica retains its position at the top, but Acacia

dealbata supersedes Populus albalcanescens to take

second place. The gap between the latter species and

Melia azedarach is lengthened and Acacia mearnsii

supersedes Ricinus communis to take 5th position.

In the temperate veld types Salix babylonica

retains top position but the gap between this species

and Acacia dealbata is reduced. Populus albalcane-

scens retains third position. In the tropical veld types

Melia azedarach and Populus albalcanescens retain

first and second positions respectively. Acacia

dealbata supersedes Salix babylonica and Ricinus

communis to take third position.

3. 1.2.3 Aggressiveness: If species aggressiveness is

gauged by its ability to penetrate and suppress

existing vegetation (Duggan & Henderson, 1982)

then the most aggressive species overall, in order,

are Acacia dealbata, Melia azedarach, Populus

albalcanescens, Salix babylonica and Acacia mearn-

sii. Locally aggressive species include Caesalpinia

decapetala, Psidium guajava, Arundo donax, Jaca-

randa mimosifolia and Sesbania punicea.

3.2 Roadside and veld habitats

Exotic woody invaders were encountered in 212

(86%) of the 248 quarter degree squares sampled.

Sixty invader species were recorded, with an

additional 24 species being recorded since the last

publication. The new species include in the western

Transvaal: Opuntia cultivar with round, blue-green

cladodes, Gleditsia triacanthos, Prosopis velutina,

Schinus molle and Trichocereus sp. and in the

eastern Transvaal: Toona ciliata, Tecoma stans,

Bauhinia variegata, Bambusa balcooa, Passiflora

edulis and Rosa sp.

3.2.1 Analysis according to veld type

The number and abundance of invader species

recorded in roadside and veld samples in each of the

four broad veld type categories are shown in Table

5.

TABLE 5. — The number and abundance of invader species

recorded in roadside and veld samples

3.2.2 Analysis according to species

3.2.2. 1 Frequency: The recorded percentage fre-

quency of occurrence of invader species in the veld

type groupings and in the whole survey area are

shown in Figs 13 — 19. The most frequent and

widespread invaders are Opuntia ficus-indica, Melia

azedarach and Eucalyptus spp. In the temperate veld

types Eucalyptus spp., Opuntia ficus-indica, Acacia

dealbata, A. mearnsii, Melia azedarach and A.

decurrens are the most widespread species. Melia

azedarach and Opuntia ficus-indica are by far the

300

EXOTIC WOODY PLANT INVADERS OF THE TRANSVAAL

most widespread invaders in the tropical veld types,

particularly the bushveld which occupies the greatest

area.

A comparison of Fig. 13 (forest) and Fig. 14

(bushveld) reveals great differences:

1. In forest, approximately 24 species have a

percentage frequency greater than 10% compared

with only about 10 species in bushveld.

2. In bushveld only Opuntia ficus-indica and Melia

azedarach were recorded in more than 30% of the

sample units, whereas in forest at least 11 species

were recorded in more than 30% of the sample

units.

3. A number of species which are well represented

(27% frequency and higher) in forest namely

Solarium mauritianum, Rubus spp., Caesalpinia

decapetala, Pinus spp. and Acacia mearnsii are

poorly represented in bushveld.

These differences indicate that invasion has

progressed furthest in forests and that the bushveld

has a greater resistance to invasion. The greater

degree of invasion in forest can be attributed mainly

to the more favourable growing conditions

throughout and the greater disturbance that these

veld types have been subjected to. Large tracts of

bushveld are very dry and inhospitable to the more

mesic invader species such as those mentioned under

point 3 above.

3. 2. 2. 2 Importance: The importance values for

roadside and veld invader species were calculated in

the same way as for streambank invaders. Again, the

species with the highest values and which collectively

added up to a total of approximately 160 points were

plotted (Figs 20, 21 & 22).

A comparison of the percentage frequency

histogram Fig. 19 and importance histogram Fig. 22

shows that for the whole survey area Opuntia

ficus-indica, Melia azedarach and Eucalyptus spp.

retain first, second, and third positions respectively.

Solanum mauritianum climbs from about 11th

position to 4th position. Acacia dealbata and A.

mearnsii shift from 6th to 5th position. Both Agave

spp. and Ricinus communis drop considerably.

Solanum mauritianum climbed from obscurity in

the percentage frequency histogram to prominence

in the importance histogram because of the very

large numbers of individuals that were recorded in a

few quarter degree squares along the Drakensberg

escarpment. In contrast the Agave spp. and Ricinus

communis plummeted in position because they

generally occurred in small numbers.

In the temperate veld types Acacia dealbata and

A. mearnsii overtake Eucalyptus spp. to take first

and second positions respectively, thus indicating

that they are very much more abundant than the

Eucalyptus spp. In the tropical veld types Opuntia

ficus-indica and Melia azedarach retain first and

second positions respectively. Solanum mauritianum

and Eucalyptus spp. climb to third and fifth positions

respectively. Jacaranda mimosifolia drops one

position to fourth place and Agave spp. and Ricinus

communis drop to tenth and eighth positions

respectively.

3. 2. 2. 3 Aggressiveness: Overall, the most aggressive

species are Melia azedarach, Acacia dealbata and

Opuntia ficus-indica. Locally aggressive species

include: Lantana camara, A. mearnsii, A. mela-

noxylon, Pinus patula, Solanum mauritianum,

Jacaranda mimosifolia, Psidium guajava, Toona

ciliata and Eucalyptus spp.

3.3 Invasion patterns

The pattern of invasion for the Transvaal as a

whole is uneven with respect to both diversity and

abundance of species. The high diversity and

abundance areas occur in and around major towns

and cities, and in the eastern Transvaal along the

foothills and slopes of the Drakensberg escarpment.

The pattern of species diversity for the Transvaal

is illustrated in Fig. 23, which gives the distribution

of quarter degree squares in which 6 or more

roadside and veld invader species were recorded. A

similar pattern was obtained for streambank

invaders.

The distribution maps of the most important

species (Figs 23 - 32) also illustrate an uneven

invasion pattern. The high abundance areas of all

the species overlap to the greatest extent in the

degree squares 2330 (Tzaneen District), 2430

(Pilgrims Rest District) 2530 (Lydenburg District)

and 2630 (Carolina District).

Overall the degree square with the greatest

abundance and diversity of invader species was 2530

— which includes the towns of Lydenburg, Belfast,

Machadodorp, Sabie and Nelspruit. This degree

square is also one of the most variable in the

Transvaal with respect to topography, climate and

indigenous veld types. It has a high rainfall and has

been greatly disturbed.

The western parts of the Transvaal have been

subject to much less invasion than the eastern parts.

Contributing factors include less disturbance, the

absence of large and extensive plantations of invader

species, a drier climate and a landscape that offers a

smaller range of habitats and niches for invading

species.

4 DISCUSSION AND CONCLUSIONS

4.1 Sampling

As mentioned in the previous paper (Duggan &

Henderson, 1982), the sampling method has its

limitations such as the undersampling of certain

habitats and the less distinctive species. The method,

however, has proved successful and economical in

obtaining information that otherwise would have

been unobtainable. It has also provided an initial

broad survey which can serve as a basis for further

intensive work on more restricted areas.

One recommendation that we have with regard to

the sampling method is that a standard roadside

transect length of about 10 km be used within each

veld type or sample unit in recording abundance

values of roadside and veld species. This would

enable all the records obtained to be compared

directly and simplify the computation of average and

total abundance values. If this recommendation is

adopted, the abundance ratings <1 and <<1 will fall

L. HENDERSON AND K. J. MUSIL

301

away. In addition, the use of many short standard

transect lengths instead of one or a few long ones,

will also provide more recordings which is statistical-

ly desirable and would also reveal variation that

otherwise would not be apparent.

4.2 Important species

This survey has underlined the invasive importance

of Acacia dealbata, Melia azedarach, Opuntia

ficus-indica, Acacia mearnsii, Salix babylonica and

Populus alba/canescens.

All these species, with the exception of Opuntia

ficus-indica, are invaders of the streambank habitat

and two of them (Salix babylonica and Populus

alba/canescens ) are virtually restricted to the

streambank habitat. Acacia dealbata, A. mearnsii,

Melia azedarach and Opuntia ficus-indica are all

important invaders of roadside and veld habitats.

Further sampling since the last publication has

confirmed the observation that Acacia dealbata is an

aggressive and dangerous invader. Acacia mearnsii

has proved to be of greater importance than was

previously believed, particularly in the moist

forested and false grassveld areas. Both these

species are either absent or very rarely encountered

in the western Transvaal. If seed was made

available, however, both species could extend their

range into these parts mainly along water-courses

and into the moister bushveld areas, e.g. the

Waterberg.

Melia azedarach and Opuntia ficus-indica have

proved to be the most widespread invader species in

the Transvaal as a whole. Both species are highly

penetrative and are found far from any plantings.

Several potentially important invaders have been

recorded since the last publication. These include

Pinus patula, Toona ciliata, Prosopis velutina,

Schinus molle and Trichocereus sp.

Pinus patula is an extremely aggressive invader of

the moist and cool, forested regions of the

Drakensberg escarpment. This species seeds prolifi-

cally, colonizing roadsides, road cuttings and other

disturbed places, plantations of other tree species

and indigenous forest, scrub and grassland. At least

one other species of pine is also invasive in these

parts. (According to the Sabie regional forestry

office, P. taeda and P. elliotii are also naturalized in

places.)

Toona ciliata is a tree similar in appearance to

Melia azedarach and belonging to the same family,

Meliaceae. T. ciliata is naturalized in the Barberton,

Nelspruit and Duiwelskloof areas and possibly

elsewhere. Like Melia azedarach, it has become a

roadside weed and has invaded watercourses,

bushveld and forest. This species, however, is less

drought and frost resistant than Melia azedarach and

therefore has a potentially smaller distribution.

Passiflora edulis is another species that has

become naturalized in the moist bushveld and

forests of the eastern Transvaal. The climbing

growth habit of the species, however, made it

difficult to record and therefore we have little

information on its percentage frequency and

abundance. Mr M. J. Wells of the Botanical

Research Institute (pers. comm.) reports that it is

very frequent, particularly in remnant forests that

survive between plantations.

Prosopis velutina, Schinus molle and Trichocereus

sp. have been recorded in the dry south western

Transvaal. At present these species are very

localized but have the potential to spread.

Pyracantha angustifolia, which was recorded in

the pilot study area, has been recorded since the last

publication. At present this species occurs in small

numbers at scattered localities, mainly in the

southern grassland areas of the Transvaal. This

species, together with the Cotoneaster spp. (C.

pannosa recorded in pilot study area) are potentially

important invaders in the colder parts of the

Transvaal where they have been planted because of

their resistance to cold and for their attractive

berries.

The berries of these species are well-liked by birds

particularly since they are produced in winter when

there is very little other food available on the cold

highveld. The dispersal of the fruits by birds enables

these species, and other suitable fruit-bearing plants

such as Ligustrum spp., Lantana camara, Melia

azedarach, Morus alba, Psidium guajava and

Solanum mauritianum, to spread far afield and to

reach inaccessible places from which reinfestation

can take place.

Species which have assumed greater importance,

as a result of better recording, since the last

publication include Acacia melanoxylon, Solanum

mauritianum and Eucalyptus spp. All these species

are the most abundant in the cool, moist forested

regions of the eastern Transvaal.

Solanum mauritianum is a species which we

suspect has been undersampled. It was seldom seen

except in the eastern Transvaal where it occurred in

greater numbers than any other exotic woody

invader in the Transvaal. It is particularly bad as an

understorey weed in pine plantations, being able to

tolerate considerable shading.

At present Acacia melanoxylon is fairly localized

but has a potentially wide distribution coinciding

with the cool, moist parts of the Transvaal. Small

plants were often seen far from any planting,

particularly along roadsides and in clearings in

plantations and forests.

The Eucalyptus spp. are, on the whole, relatively

unaggressive species. In the forestry areas of the

eastern Transvaal, however, they assume their

greatest importance. Several species seed themsel-

ves, of which the most prolific is E. grandis.

4.3 Correlation of invasion with environmental

factors

Overall the warmer tropical forest, bush and

savanna veld types were generally more subject to

invasion by exotic, woody species than were the

cooler, temperate grassveld and false grassveld

types.

In the streambank habitat, 36 of the 40 species

recorded occurred in the tropical veld types

compared with 20 species in the temperate veld

types.

302

EXOTIC WOODY PLANT INVADERS OF THE TRANSVAAL

In roadside and veld habitats, 53 of the 60 species

occurred in the tropical veld types compared with 42

species in the temperate veld types.

Further sampling since the 1982 publication has

allowed for a more in-depth study of the correlation

between invasion and veld type. The results have

shown that considerable differences exist between

the inland tropical forest types, tropical bush and

savanna types, false grassveld and pure grassveld

types.

If the degree of invasion is gauged by the

abundance and species diversity of invaders per

sampling unit and water-course recording then the

inland tropical forest types show the greatest

invasion in streambanks, roadside and veld habitats.

The false grassveld types take second place in

roadside and veld habitats followed by the tropical

bush and savanna and pure grassveld types.

Although the tropical bush and savanna types

have the most recorded invader species, they occupy

the largest area and account for 63% of the samples.

Invasion of the bushveld is also very patchy — with

large tracts of dry veld relatively undisturbed while

other parts again have been severely disturbed, for

example the moist bushveld, Acocks 10, in the

eastern Transvaal.

In the streambank habitat there is little difference

in the abundance and species diversity values for the

bushveld and false grassveld types. The pure

grassveld types have the least number of exotic,

woody invader species, and the lowest species

abundance and diversity values.

4.4. Prospects for the future

Invasion is expected to increase in all parts of the

Transvaal. The greatest expansion is liable to occur

in the moist and warm areas which are favourable

for plant growth — namely the forests, moist

bushveld and warmer false grassveld areas (south-

eastern Transvaal).

The stability of the forests has been so weakened

by clearing for afforestation that they have been

made exceedingly vulnerable to invasion by a wide

spectrum of exotic invaders. The moist bushveld has

been disturbed to a lesser extent, but the climatic

conditions are favourable for the growth of many

exotic species. The false grassveld is favourable for

fewer exotic species, but offers much less competi-

tion to invading species because of its relatively few

indigenous woody species.

The very dry bushveld areas of the Transvaal are

probably in the least danger from invading exotics.

The combination of unfavourable climatic condi-

tions and a well-adapted indigenous woody compo-

nent act as barriers to exotic species. Rivers and

streams, however, are more vulnerable to invasion

and permit the entry of moisture-loving species into

an otherwise inhospitable environment.

Although severe frosts help to limit the spread of

many potential invader species in the cold grassveld

areas, these veld types are vulnerable because of

their lack of a natural woody component. Frost-

hardy, exotic species are therefore able to invade

unoccupied niches with relatively little competition.

The ultimate prospects for the future if action is

not taken against the invading species are very bleak

indeed:

1. The few remaining relics of natural forest are

liable to be almost totally replaced by exotics.

2. In the moist bushveld exotic species will replace

many of the indigenous species, altering the basic

appearance and character of the veld.

3. In the dry bushveld one can expect fewer exotic

invaders but one or a few dominant species which

will take over large tracts of land at the expense of

the indigenous species.

4. The grassveld and false grassveld areas will

probably undergo the greatest change in appearance

by acquiring a woody element that was either totally

lacking or poorly represented before.

5. The indigenous streambank vegetation

throughout the Transvaal is in grave danger of being

almost totally replaced by exotic, woody invaders.

UITTREKSEL

Die frekwensie en getalsterkte van uitheemse

houtagtige plantindringers was in 60% van die

kwartgradevierkante in die studiegebied opgeteken.

Een-en-sestig indringers is aangetref waarvan die

belangrikste en die aggressiefste Acacia dealbata,

Populus spp., Melia azedarach, Opuntia ficus-

indica, Salix babylonica en Acacia meamsii was.

Indringingspatrone word bespreek en ’n poging

word aangewend om verspreiding met omgewings-

faktore in verband te bring. Die gebiede van grootste

indringing en die gebiede waar die grootste uitbrei-

ding in die toekoms sal voorkom is onder die aandag

gebring.

REFERENCES

Acocks, J. P. H., 1975. Veld types of South Africa. Mem. bot.

Surv. S. Afr. 40: 1-128.

Bailey, L. H., 1963. The standard cyclopedia of horticulture.

Vols 1—3. New York: Macmillan.

Bailey, L. H. & Bailey, E. Z., 1976. Hortus third. New York,

London: Macmillan.

Brice Bruce, A. P., 1979. A key to the eucaiypts in southern

Africa. Pietermaritzburg: University of Natal.

Britton, N. L. & Rose, N. J., 1963. The Cactaceae; descriptions

and illustrations of plants of the cactus family. Vol 2. New

York: Dover Publications.

Burger, G., 1972. Rochdale. Trees S. Afr. 24: 58-66.

De Winter, B., Vahrmeuer, J. & Von Breitenbach, F., 1978.

The national list of trees. Pretoria: Van Schaik.

Duggan, K. J. & Henderson, L., 1982. Progress with a survey of

exotic, woody plant invaders of the Transvaal. Proceedings

of the fourth National Weeds Conference of South Africa,

1981. Cape Town: Balkema.

Henderson, L. & Duggan, K. J., 1981. The minimum effective

sampling level for the determination of the major exotic,

woody plant invaders, the frequency of occurrence and

relative importance and major invasion trends in the central

Transvaal. Unpublished.

Hubbard, C. S., 1926. A review of the species of Populus

introduced into South Africa. S. Afr. J. Sci. 23: 340—365.

Palgrave, K. C., 1977. Trees of southern Africa. Cape Town:

Struik.

Purseglove, J. W., 1968. Tropical crops, dicotyledons. Vols 1 &

2. London: Longmans.

Ross J. H., 1975. The naturalized and cultivated exotic Acacia

species in South Africa. Bothalia 11: 463-470.

L. HENDERSON AND K. J. MUSIL

303

Ross, J. H., 1975. Mimosoideae. In J. H. Ross, FI. S. Afr. 16, 1:

6-7.

Stirton, C. H., 1978. Plant invaders, beautiful but dangerous.

Cape Town: Department of Nature and Environmental

Conservation of the Cape Provincial Administration.

Wells, M. J., Duggan, K. J. & Henderson, L., 1980. Woody

plant invaders of the central Transvaal. Proceedings of the

third National Weeds Conference of South Africa, 1979.

Cape Town: Balkema.

APPENDIX

SPECIES CHECKLIST

* Species recorded in pilot study area only

** Species recorded in pilot study area subsequent to survey

+ Personal communication — not supported by herbarium specimens

+ + Cited on national herbarium specimen labels

x Cited in various literature sources

304

EXOTIC WOODY PLANT INVADERS OF THE TRANSVAAL

Fig. 1. — Study area and sampling

routes.

Fig. 2.— The four broad veld type

categories in the Transvaal.

Inland tropical forest types

| J False grassveld types

Tropical bush and savanna types

Pure grassveld types

L. HENDERSON AND K. J. MUSIL

305

3. Less frequent species: Agave spp., Arundo donax, Bambusa bal-

cooa. Cassia didymobotrya, C. floribunda, Cupressus sp., Grevillea

robusta, Jacaranda mimosifolia, Nicotiana glauca, Prunus persica,

Rosa sp., Toona ciliata.

4. Less frequent species: Acacia dealbata, Caesal-

pinia decapetala. Cassia didymobotrya, C. floribun-

da, Crotalaria agatiflora, Eucalyptus spp., Lantana

camara, Nicotiana glauca, Opuntia imbricata, Popu-

lus deltoides, P. nigra, Prunus persica, Robinia

pseudo-acacia, Rubus spp.

5. Less frequent species: Acacia baileyana, A. decur-

rens, Citrus sp., Eucalyptus spp., Lantana camara, Melia

azedarach, Opuntia ficus-indica, Pinus spp., Populus

deltoides, Psidium guajava, Pyracantha angustifolia,

Sesbania punicea.

6. Less frequent species: Acacia decurrens, A. mearn-

sii. Agave spp., Caesalpinia decapetala. Eucalyptus spp.,

Melia azedarach, Opuntia ficus-indica, Pinus spp., Pop-

ulus deltoides, Prunus persica, Pyracantha angustifolia.

Figs 3—6. — Percentage frequency of streambank species in: 3, forest veld types; 4, bushveld veld types; 5, false grassveld veld types;

6, pure grassveld veld types.

306

EXOTIC WOODY PLANT INVADERS OF THE TRANSVAAL

7. Less frequent species: Bambusa balcooa. Cassia didymobotrya, C. floribun-

da, Crotalaria agatiflora, Cupressus sp., Grevillea robusta, Nicotiana glauca,

Opuntia imbricata, Populus deltoides, P. nigra, Prunus persica, Robinia pseudo-

acacia, Rosa sp., Rubus spp., Toona ciliata.

8

5

_o

-O

03

X>

8. Less frequent species: Acacia baileyana, A. decur-

rens. Agave spp., Caesalpinia decape tala. Citrus sp.,

Eucalyptus spp., Lantana camara, Melia azedarach,

Opuntia ficus-indica, Pinus spp., Populus deltoides,

Prunus persica, Psidium guajava, Pyracantha angusti-

folia. Sesbania punicea.

9. Less frequent species: Acacia baileyana, A. decurrens.

Agave spp., Arundo donax, Bambusa balcooa, Cassia didy-

mobotrya, C. floribunda. Citrus sp., Crotalaria agatiflora,

Cupressus sp., Eucalyptus spp., Grevillea robusta, Jacaranda

mimosifolia, Lantana camara, Moms alba, Nicotiana glauca,

Opuntia ficus-indica, O. imbricata, Pinus spp., Populus del-

toides, P. nigra, Pmnus persica, Pyracantha angustifolia,

Robinia pseudo-acacia, Rosa sp., Rubus spp., Sesbania puni-

cea, Toona ciliata.

Figs 7— 9. — Percentage frequency of streambank species in: 7, tropical veld types; 8, temperate veld types; 9, the whole survey area.

L. HENDERSON AND K. J. MUSIL

307

30

w 25

u

^ 20

H

% 15

2

S 10

S E 3 B £

=2 6 F -2 «

« o c >, .3

■s S 6 -s 1

«j 3 3 X> 3

'§ .5 § * g

y ^ "3 53 «

< DC to 1/5 u

S < < w ™

10. Importance values add up to 200. Only those adding up

to 160 are plotted. The others include: Arundo donax,

Bambusa balcooa, Cassia didymobotrya, C. floribunda, Crot-

alaria agatiflora, Cupressus sp., Grevillea robusta, Lantana

camara, Nicotiana glauca, Opuntia ficus-indica, O. imbricata,

Pinus spp., Populus deltoides, P. nigra, Prunus persica, Rosa

sp., Rubus spp., Toona ciliata.

80

75

70

65

60

55

50

w

^ 45

H 40

C6

g 35

§30

25

20

15

10

5

« 5

11. Less important species: Acacia baileyana

A. decurrens A. mearnsii. Agave spp., Caesalpinki

decapetala Citrus sp., Eucalyptus spp., Lantana

camara, Melia azedarach, Opuntia ficus-indica

Pinus spp., Populus deltoides, Prunus persica Psi-

dium guajava, Pyracantha angustifolia, Ses’bania

punicea.

UJ

40

35

30

u

Z 25

06 20

O

Oh 1 c

s 15

10

•S § i

S < 06

12. Less important species: Acacia baileyana, A. decurrens.

Agave spp., Arundo donax, Bambusa balcooa, Cassia didymobot-

rya, C. floribunda, Citrus sp., Crotalaria agatiflora, Cupressus sp.,

Grevillea robusta, Jacaranda mimosi folia, Lantana camara. Moms

alba, Nicotiana glauca, Opuntia ficus-indica, O. imbricata, Pinus

spp., Populus deltoides, P. nigra, Pmnus persica, Pyracantha

angustifolia, Rosa sp., Rubus spp., Toona ciliata.

Figs 10-12. — Importance of streambank species in: 10, tropical veld types; 11, temperate veld types; 12, the whole survey area.

308

EXOTIC WOODY PLANT INVADERS OF THE TRANSVAAL

13. Less frequent species: Bauhinia variegata, Casuarina sp., Crotalaria agatiflora, Ligustrum japonicum, Pyracantha angustifolia

Vitis sp.

3

•3

.5

3

U

C

14. Less frequent species: Acacia baileyana, A. dealbata, A. decur-

rens, A. mearnsii, A. podalyriifolia, Caesalpinia decapetala, Carica

sp., Cassia didymobotrya, C. floribunda, Cereus peruvianus, Crotal-

aria agatiflora, Grevillea robusta, Opuntia imbricata, Pinus spp.,

Populus deltoides, Prosopis velutina, Prunus arrneniaca, Rosa sp.,

Rubus spp., Salix babylonica, Schinus mode, Tecoma stans, Tipu-

ana tipu, Toona ciliata, Trichocereus sp.

Figs 13—14. — Percentage frequency of roadside and veld species in: 13, forest veld types; 14, bushveld veld types.

Populus alba/canescens

Acacia decurrens

L. HENDERSON AND K. J. MUSIL

309

04

15. Less frequent species: Acacia longifolia, A. melanoxylon, A. pod-

alyriifolia. Citrus sp., Gleditsia triacanthos, Grevillea robusta, Lantana

camara, Moms alba, Nicotiana glauca, Psidium guajava, Robinia pseudo-

acacia, Rubus spp., Sesbania punicea. Solarium mauritianum, Yucca sp.

\6 Less frequent species: Acacia baileyana, A. podaly-

rujolia, Amndo donax, Cereus pemvianus, Cupressus sp

Gleditsia triacanthos, Nicotiana glauca. Populus deltoides,

Prosopis velutina, Ricinus communis, Robinia pseudo-

acacia, Salix babylonica, Sesbania punicea Solanum

mauritianum. Yucca sp.

17. Less frequent species: Acacia baileyana, A. decurrens, A. melanoxylon, A. podalyriifolia,

Bambusa balcooa, Bauhinia variegata, Carica sp., Cassia floribunda, Casuarina sp., Cereus peruvia-

nus, Crotalaria agatiflora, Ligustmm japonicum, Opuntia imbricata, Passiflora edulis, Populus

deltoides, Prosopis velutina, Prunus armeniaca, Pyracantha angustifolia, Rosa sp., Salix babylonica.

Schinus molle , Tecoma stans, Tipuana tipu, Toona ciliata, Trichocereus sp., Vitis sp.

Figs 15-17. — Percentage frequency of roadside and veld species in: 15, false grassveld veld types; 16, pure grassveld veld types; 17,

tropical veld types.

310

EXOTIC WOODY PLANT INVADERS OF THE TRANSVAAL

Q.

O.

a

18. Less frequent species: Acacia longifolia, A. melanoxy-

lon, A. podalyriifolia, Arundo donax, Cereus peruvianus.

Citrus sp., Cupressus sp., Gleditsia triacanthos, Grevillea ro-

busta, Jacaranda mimosifolia, Lantana camara, Morns alba,

Nicotiana glauca, Opuntia sp. (blue-green cultivar), Populus

deltoides, Prosopis velutina, Psidium guajava, Robinia pseudo-

acacia, Rubus spp., Schinus molle, Sesbania punicea. Solarium

mauritianum, Yucca sp.

8

-3

c

3

O

19. Less frequent species: Acacia baileyana, A. longifolia, A. melanoxylon, A. podalyrii-

folia, Bambusa balcooa, Bauhinia varicgata, Caesalpinia de cape tala, Carica sp.. Cassia didvmo-

botrya, C. floribunda, Casuarina sp., Cereus peruvianus. Citrus sp., Crotalaria agatiflora, Cu-

pressus sp., Gleditsia triacanthos, Grevillea robusta, Ligustrum japonicum, Opuntia imbricata,

Passiflora edulis, Populus deltoides, Prosopis velutina, Primus armeniaca, Pyracantha angusti-

folia, Robinia pseudo-acacia, Rosa sp., Salix babylonica, Schinus molle, Tecoma stans, Tipu-

ana tipu, Toona ciliata, Trichocereus sp., Vitis sp.. Yucca sp.

Figs 18—19. — Percentage frequency of roadside and veld species in: 18, temperate veld

types; 19, the whole survey area.

L. HENDERSON AND K. J. MUSIL

311

20. Importance values add up to 200. Only those adding up to 160 are plot-

ted. The rest include: Acacia baileyana, A. dealbata, A. decurrens, A. melan-

oxylon, A. podalyriifolia, Arundo donax, Bambusa balcooa, Bauhinia varie-

gata, Caesalpinia decapetala, Carica sp.. Cassia didvmobotrya, C. floribunda,

Casuarina sp., Cereus peruvianus, Crotalaria agatiflora, Grevillea robusta,

Ligustrum japonicum, Moms alba, Nicotiana glauca, Opuntia sp. (blue-green

cultivar), O. imbricata, Passiflora edulis, Populus alba/ cane scens, P. deltoides,

Prosopis velutina, Pmnus anneniaca, P. persica, Pyracantha angustifolia, Rob-

inia pseudo-acacia, Rosa sp., Salix babylonica, Schinus molle, Sesbania puni-

cea, Tecoma stans, Tipuana tipu, Toona ciliata, Trichocereus sp., Vitis sp.

21. Less important species: Acacia baileyana, A. longifolia,

A. melanoxylon, A. podalyriifolia, Agave spp., Amndo donax,

Cereus pemvianus, Citms sp., Cupressus sp., Gleditsia triacanth-

os, Grevillea robusta, Jacaranda mimosifolia, Lantana camara,

Moms alba, Nicotiana glauca, Opuntia sp. (blue-green cultivar),

Populus alba/canescens, P. deltoides, Prosopis velutina, Psidium

guajava, Pyracantha angustifolia, Ricinus communis, Robinia

pseudo-acacia, Rubus spp., Salix babylonica, Schinus molle,

Sesbania punicea, Solanum mauritianum. Yucca sp.

22. Less important species: Acacia baileyana, A. decurrens, A. longifolia,

A. melanoxylon, A. podalyriifolia, Arundo donax, Bambusa balcooa, Bauhinia

variegata, Caesalpinia decapetala, Carica sp.. Cassia didymobotrya, C. floribun-

da, Casuarina sp., Cereus pemvianus, Citms sp., Crotalaria agatiflora, Cupres-

Slil SP ■ , Gleditsia triacanthos, Grevillea robusta, Ligustmm japonicum, Moms

ajba, Nicotiana glauca, Opuntia sp. (blue-green cultivar), O. imbricata, Passi-

flora edulis, Populus alba/canescens, P. deltoides, Prosopis velutina, Pmnus

armeniaca, P. persica, Pyracantha angustifolia, Robinia pseudo-acacia, Rosa

sp., Rubus spp., Salix babylonica, Schinus molle, Sesbania punicea, Tecoma

stans, Tipuana tipu, Toona ciliata, Trichocereus sp., Vitis sp., Yucca sp.

23

4 Degree squares in which 6 or more species occurred.

Figs 20-23. — Importance of roadside and veld species in: 20, tropical veld types; 21, temperate veld types; 22, the whole survey

area; 23, areas of high species diversity in roadside and veld habitats.

312

EXOTIC WOODY PLANT INVADERS OF THE TRANSVAAL

Figs 24—29. — Distribution and high abundance areas of the most important species. 24, Acacia dealbata; 25, Acacia mearnsii; 26,

Eucalyptus spp; 27, Jacaranda mimosifolia ; 28, Melia azedarach; 29, Opuntia ficus-indica and spineless cultivar. □, roadside

and veld habitats, abundance values 3 or more per 20 km or 4 or more per 15 km; A, streambank habitat, abundance values 4

or more; ■, streambank, roadside and veld habitats with aforementioned values.

L. HENDERSON AND K. J. MUSIL

313

26* 27* 28* 29* 30* 31* 32*

25* 28* 27* 28* 29* 30* 31* 32*

28* 27* 28* 29* 30* 31*

26* 27* 28* 29* 30* 31*

Figs 30-32. — Distribution and high abundance areas of the most

important species. 30, Populus albalcanescens; 31, Salix

babylonica; 32, Solarium mauritianum. □, roadside and veld

habitats, abundance values 3 or more per 20 km or 4 or more

per 15 km; A, streambank habitat, abundance values 4 or

more; ■, streambank, roadside and veld habitats with

aforementioned values.

Bothalia 15, 1 & 2: 315-321 (1984)

New species and combinations in Parmelia (Lichenes) from southern

Africa

F. BRUSSE*

Keywords: new combinations, new species, Parmelia, southern Africa

ABSTRACT

Eight new species are described and six new combinations are made in the lichen genus Parmelia from southern

Africa. The new species are Parmelia astricta Brusse, P. clivorum Brusse, P. marroninipuncta Brusse, P. patula

Brusse, P. scitula Brusse, P. spargens Brusse, P. unctula Brusse and P. vernicosa Brusse. The new combinations

are P. aggregata (D. Knox) Brusse, P. cedrus-montana (Brusse), Brusse, P. dysprosa (Brusse & D. Knox) Brusse, P.

exornata (Zahlbr.) Brusse, P. karoo (D. Knox & Brusse) Brusse and P. leucostigma (Brusse) Brusse.

NEW SPECIES

Parmelia astricta Brusse, sp. nov.

Thallus minute foliosus, adnatus, saxicola, usque

ad 12 mm diametro, omnino lobatus. Lobi lineares,

0,1— 0,8 mm lati, subimbricati, sat adpressi, 75 — 120

pm crassi. Thallus superne cinereus, nitidus,

epicortice poroso, emaculatus, isidiis sorediisque

nullis. Cortex superior 10—20 pm crassus. Stratum

gonidiale 15—25 pm crassum. Medulla alba, 25—90

pm crassa. Cortex inferior 5 — 10 pm crassus,

brunneus. Thallus inf erne piceus. Rhizinae simpli-

ces, sparsae vel moderatae, piceae, 30—90 pm

crassae. Apothecia et pycnidia ignota. Thallus

atranorinum et acidum alectoronicum continens.

Type. — Cape, 3321 (Ladismith): 17 km N of

Riversdale, Garcia’s Pass, SW side of Kareekop, in

the Langeberg, on TMS on a steep S slope, Brusse

3626, 1981.05.10 (PRE!). (With olivetoric acid

accessory and medulla C+ red.) Fig. 1.

This species is probably most closely related to P.

mongaensis Elix with grayanic and colensionic acids

(Elix, 1981), orcinol depsidones like alectoronic

acid.

The following specimen is tentatively assigned

here too:

Cape.— 3322 (Oudtshoorn): 19 km S of Prince Albert, top of

Swartberg Pass, on TMS, on steep S slope (-AC), Brusse 3444,

1981.05.07 (PRE). (Contains atranorin, gyrophoric, alectoronic

+ acids.)

Parmelia clivorum Brusse, sp. nov.

Thallus foliosus, saxicola, sat adnatus. Lobi

elongati, 1—3 mm plerumque 1,5— 2,5 mm lati,

100-300 pm crassi. Thallus superne flavovirens,

subnitidus, isidiatus, isidiis grossis, globosis, usque

ad 0,3 mm crassis. Cortex superior 18-28 pm

crassus. Stratum gonidiale 30—70 pm crassum.

Medulla alba, C + rubra, 40—200 pm crassa. Cortex

inferior 12—17 pm crassus, brunneus. Thallus inferne

* Botanical Research Institute, Department of Agriculture,

Private Bag X101, Pretoria 0001.

piceps, sparse vel sat rhizinosus. Apothecia ignota.

Pycnidia globosa, 70—190 pm dipmetris. Pycnidios-

porae hyalinae, bacillares, rectae, 5—9 x 1 pm.

Thallus acidum usnicum et acidum lecanoricum

continens.

Type. — Cape, 3218 (Clanwilliam): 3 km W of

Olyvenboskraal, Witelskloof, on large TMS boulder

outcrop (-BD), Brusse 3069, 1981.05.02 (PRE).

Fig. 2.

Isidiate Xanthoparmeliae are unusual in the Cape

and it was a surprise to find this new unusual species,

particularly as it is closest to P. joranadia Nash

(1974a) from the state of New Mexico, USA. P.

clivorum, however, is clearly distinguished from the

latter by its pitch black lower surface. At present this

species is known only from the type specimen.

Parmelia marroninipuncta Brusse, sp. nov.

Thallus foliosus, adnatus, saxicola, usque ad 8 cm

diametro. Lobi elongati, 1,5—6 mm plerumque 2—4

mm lati, 80-200 pm crassi. Thallus superne

virescens, nitidus, isidiis sorediisque nullis, dissite

atropunctatus. Cortex superior 7—15 pm crassus.

Stratum gonidiale 25—65 pm crassum. Medulla alba,

sub atropunctis marroninis, 30-120 pm crassa.

Cortex inferior 5-10 pm crassus, pallidus. Thallus

inferne pallidus, sat rhizinosus. Apothecia usque ad

10 mm diametris. Hymenium 33—40 pm altum.

Subhymenium 5-10 pm crassum. Hypothecium

20-40 pm crassum. Ascosporae octonae, simplices,

hyalinae, ellipsoideae, 8—10 x 4—6 pm. Pycnidia

globosa, 100-150 pm diametris. Pycnidiosporae

hyalinae, bacillares, rectae, 4-6 x 1 pm. Thallus

acidum usnicum, acidum protocetraricum et duo

usque ad quatuor pigmenta anthraquinona ignota

continens.

Type. — Cape, 3221 (Merweville): 40 km NW of

Merweville, Uitspannings River Pass, on a low

mudstone kranz, on the S verge of a mountain

(-CA), Brusse 3417, 1981.05.07 (PRE, holo!; LD,

iso!). Fig. 3.

This species is very close to P. dichromatica Hale

(1971), with which it is identical in all respects,

except that it has the pigments concentrated in

316

NEW SPECIES AND COMBINATIONS IN PARMELlAt (LICHENES) FROM SOUTHERN AFRICA

Fig. 2. — Parmelia clivorum, Bnis*>yo

se 3069, holotype. Scale in

mm.

Fig. 3. — Parmelia marroninipunc-

ta, Brusse 3417, holotype.

Scale in mm.

Fig. 1. — Parmelia astricta, Brusse

3626, holotype. Scale in mm.

F. BRUSSE

317

pockets immediately below the black spots on the

upper surface. SEM examination of these black

spots, which are normally somewhat raised, reveal

that they are pored areas on a non-pored epicortex

(Figs 4 & 5).

Fig. 4. — Parmelia marroninipuncta, Brusse

3417, holotype. SEM micrograph of a

young spot. Bar = 100 pm.

Fig. 5 — Parmelia marroninipuncta, Brusse 3417,

holotype. SEM micrograph of an older

spot. Bar = 100 pm.

The upper cortex is also missing at these spots, so

that the maroon medulla is found immediately below

the epicortex. A light touch opens them, with the

result that older parts appear to possess maroon

pseudocyphellae. P. marroninipuncta appears to be

restricted to the southern Cape in a broad sense. P.

dichromatica, on the other hand, has a smooth

unspotted upper surface, with the pigments spots in

the medulla. It also has a different distribution,

occurring in the eastern interior of the Cape

Province to Lesotho and environs. The distribution

of these two species probably overlaps in the eastern

Cape but, as yet, the two have not been found at the

same locality. The pigments involved in both these

species are the same as those in P. endomiltodesNyl.

Cape — 3124 (Hanover): 58 km from Graaff-Reinet to

Cradock, Farm Bluegum House, on rock on a gentle NW slope

(-DD), Relief & Loock 364, 1978.07.23 (PRE). 3320 (Montagu):

8 km S of Laingsburg, on S slopes of an E-W ridge with strata

exposed near vertically (-BB), Brusse 3503, 1981.05.09 (PRE,

LD). 3321 (Ladismith): Seven Weeks Poort, about 9 km from the

main Calitzdorp-Ladismith road, on S faces of TMS exposures

and rocks, on a steep E slope (-AD), Brusse 3472, 1981.05.08

(PRE).

Parmelia patula Brusse, sp. nov., thallo ut in P.

weberi Hale, sed inferne piceo.

Thallus saxicola, arete adnatus, foliosus. Lobi

elongati 0,6-3 mm plerumque 1—2 mm lati,

120—300 pm crassi. Thallus superne viridus, nitidus,

isidiatus. Isidia grossa, globosa, 0,1 -0,2 mm

diametris, 0,1 -0,3 mm alta. Cortex superior 15—25

pm crassus. Stratum gonidiale 30—50 pm crassum.

Medulla alba, 75—200 pm crassa. Cortex inferior

6 — 10 pm crassus, brunneus. Thallus inferne piceus,

sat rhizinosus. Apothecia ignota. Pycnidia globosa,

100—200 pm diametris. Pycnidiosporae hyalinae,

bacillares, rectae, 5—7x1 pm. Thallus acidum

usnicum, acidum hypoprotocetraricum et acidum

4— 0-demethylnotaticum continens.

Type. — Transvaal, 2529 (Witbank): between

Middleburg and Loskop Dam, ‘Cycad trail’, on N

side of Olifants River gorge (-CB), Brusse 1337,

1981.03.27 (PRE, holo.; LD, COLO, MEL, iso.).

Fig. 6.

This species is similar to P. weberi Hale, except

that it has a black lower surface. The non-isidiate

counterpart is P. domokosii Gyeln., a Cape species.

The complete complement of species in this group is

now known: P. perspersa Stiz , — P. neocongensis

Hale; P. domokosii Gyeln.— P. patula Brusse; P.

encrustans Hale— Xantho parmelia hedbergii D.

Knox; and P. subdomokosii Hale-P. weberi Hale.

However, the members of these pairs are not always

absolutely identical and this is particularly true for

the last pair. This is a fairly common species in the

eastern half of the country and appears to extend

into the south-western Cape along the mountains.

Parmelia scitula Brusse, sp. nov.

Thallus suberustosus, saxicola, usque ad 6 cm

diametro. Lobi elongati, 0,5 — 1,5 mm plerumque

circa 1 mm lati, 70-150 pm crassi. Thallus superne

viridus, nitidus, isidiis sorediisque nullis. Cortex

superior 12 — 14 pm crassus. Stratum gonidiale 12-30

pm crassum. Medulla alba, C -1- rubra, 35-100 pm

crassa. Cortex inferior pallidus vel hyalinus, 5-10

pm crassus. Thallus inferne pallidus, sparse rhizino-

sus. Apothecia usque ad 0,8 mm diametris.

Hymenium 35-45 pm altum. Subhymenium circa 10

pm crassum. Hypothecium 20-40 pm crassum.

Ascosporae octonae, simplices, hyalinae, ellip-

soideae, 8—9,5 x 5—6 pm. Pycnidia globosa,

90—150 pm diametris. Pycnidiosporae bacillares,

rectae, hyalinae, 5-7 x 1 pm. Thallus acidum

usnicum et acidum lecanoricum continens.

318

NEW SPECIES AND COMBINATIONS IN PARMELIA (LICHENES) FROM SOUTHERN AFRICA

Fig. 6. — Parmelia patula, Brusse

1337, holotype. Scale in mm.

Type. — Cape, 3320 (Montagu): 8 km S of

Laingsburg, on S slopes of an E— W ridge, with

strata exposed near vertically (— BB), Brusse 3522,

1981.05.09 (PRE, holo.; LD, iso.). Fig. 7.

This species is the subcrustose counterpart of P.

worcesteri J. Steiner & Zahlbr. (Zahlbruckner,

1926) and is less common than the latter. It seems to

be restricted to the Karoo. P. scitula Brusse

superficially resembles any of the other subcrustose

Xanthoparmeliae such as P. adhaerens Nyl. (stictic

acid), P. perplexa Stiz. (salazinic acid), P. perspersa

Stiz. (hypoprotocetraric acid, lower side black), P.

unctula Brusse (fatty acids) and even sometimes

Lecanora leptoplaca Zahlbr. (salazinic acid).

Cape. — 2923 (Douglas): 19 km NE of Strydenburg, Elands

Mountain, on dolerite (-DD), Brusse 768 5-1-5 (J).

Parmelia spargens Brusse, sp. nov.

Thallus foliosus, saxicola, sat adnatus, saepe

dispersus. Lobi elongati, 1—4 mm plerumque 1,5—3

mm lati, 120—170 pm crassi. Thallus superne viridus,

nitidus, isidiatus. Isidia grossa, globosa vel elongata,

0,1 -0,3 mm crassa et usque ad 0,8 mm longa.

Cortex superior 10—15 pm crassus. Stratum gonidiale

30-55 pm crassum. Medulla alba, 50-120 pm

crassa. Cortex inferior 8-20 pm crassus. Thallus

inferne pallidus, sat rhizinosus. Apothecia sparsa,

usque ad 3 mm diametris, marginibus isidiatis.

Hymenium 50—60 pm crassum. Hypothecium 35—40

pm crassum. Ascosporae octonae, simplices, hyali-

nae, ellipsoideae, 8-11 x 5—6 pm. Pycnidia

globosa, 100—150 pm diametris. Pycnidiosporae

bacillares, rectae, hyalinae, 5-7 x 1 pm. Thallus

acidum usnicum et acidum lichesterinicum con-

tinens.

Type. — Cape, 3321 (Ladismith): Seven Weeks

Poort, about 9 km from the main Calitzdorp-

Ladismith road, on S faces of TMS exposures and

rocks, on a steep E slope (—AD), Brusse 3481,

1981.05.08. (PRE, holo.; LD, iso.). Fig. 8.

This species occurs on the southern Cape

mountains in higher rainfall areas, together with

other isidiate species which are otherwise uncom-

mon in the Cape Province [cf. Europe (Hale,

1964) ], Arizona (Nash, 1974b), Texas (Egan, 1977),

Fig. 7. — Parmelia scitula, Brusse

3522, holotype. Scale in cm

and mm.

F. BRUSSE

319

Fig. 8. — Parmelia spargens, Brus-

se 3481, holotype. Scale in

mm.

New Zealand (Galloway, 1980), Australia (Kuroka-

wa, 1969; Filson, 1982)). The identity of the fatty

acid in Parmelia spargens Brusse was confirmed by

co-chromatography in three solvent systems (Cul-

berson, 1972) with the acetone extract of Cetraria

ericetorum Opiz (Arnold Lichen Exsiccati no. 1609,

in PRE).

Cape. — 3320 (Montagu): 4 km SW of Montagu, Kogmans

Kloof near the old British Fort of 1899, on S sides of an E— W

ridge (-CC), Brusse 3736 (PRE), 3743 (PRE).

Parmelia unctula Brusse, sp. nov.

Thallus subcrustosus, saxicola, usque ad 5 cm

diametro. Lobi elongati, 1—6 mm longi, 0,3 — 1,5

mm lati, 80—110 pm crassi. Thallus superne viridus,

nitidus, isidiis sorediisque nullis. Cortex superior

10-13 pm crassus. Stratum gonidiale 20—35 pm

crassum. Medulla alba 35—70 pm crassa. Cortex

inferior 6-10 pm crassus. Thallus inferne pallidus,

sparse vel sat rhizinosus. Apothecia usque ad 1,2 mm

diametris. Hymenium 45—55 pm altum. Subhyme-

nium 5 — 10 pm crassum. Hypothecium 50—75 pm

crassum. Ascosporae octonae, simplices, hyalinae,

ellipsoideae, 8—10 x 5—6 pm. Pycnidia globosa,

circa 100 pm diametris. Pycnidiosporae bacillares,

rectae, hyalinae, 5—7x1 pm. Thallus acidum

usnicum, atranorinum et acidum aliphaticum acido

protolichesterinico affine continens.

Type. — Cape, 3321 (Ladismith): Seven Weeks

Poort, about 164 km from the main Calitzdorp-

Ladismith road, on large TMS boulders in the cove

of a high water trickle (waterfall), Brusse 3582

(PRE, holo.). Fig. 9.

This species may be the perfect morph of P.

globulifera Kurok. & Filson, another small species

containing fatty acids, but the latter is globose-

isidiate and has so far only been found in Australia.

(Kurokawa & Filson, 1975; Filson, 1982). P. unctula

Brusse may also be regarded as the subcrustose

counterpart of P. subdecipiens Vain, ex Lynge,

which contains the same fatty acids in the medulla.

The latter species is widespread in the Cape

Province and Lesotho and environs, and is clearly

foliose, resembling P. subdomokosii Hale most

closely in morphology. At present this species is

known only from the type locality.

Fig. 9. — Parmelia unctula, Brusse

3582, holotype. Scale in cm

and mm.

320

NEW SPECIES AND COMBINATIONS IN PARMELIA (LICHENES) FROM SOUTHERN AFRICA

Parmelia vernicosa Brusse, sp. nov.

Thallus foliosus, saxicola, usque ad 5 cm

diametro. Lobi elongati, 0,5 -2,0 mm lati, 120—250

pm crassi, aequaliter adpressi. Thallus superne

brunneus, nitidus, epicortice poroso, isidiis sorediis-

que nullis. Cortex superior 10-12 pm crassus.

Stratum gonidiale 20—35 pm crassum. Medulla

marronina vel violacea, 70-200 pm crassa. Cortex

inferior pallidus, 6 — 10 pm crassus. Thallus inferne

pallidus, sat rhizinosus. Apothecia usque ad 3 mm

diametris, numerosa. Hymenium 40-55 pm altum.

Subhymenium 5 — 10 pm crassum. Hypothecium

20—40 pm crassum. Ascosporae octonae, simplices,

hyalinae, ellipsoideae, 71—10 x 5-6 pm. Pycnidia

globosa, circa 120-130 pm diametris. Pycnidiospor-

ae bacillares, rectae, hyalinae, 5-7x1 pm. Thallus

quatuor pigmenta anthroquinona ignota (ut in

Parmelia endomiltodes Nyl.) solum continens.

Type. — Cape, 3219 (Worcester): 17 km W of

Citrusdale, top of Middleberg Pass, on W facing

TMS (-CA), Brusse 3039, 1981.05.02 (PRE). Fig.

10.

Rhyn’s Dorp Division (Cape Province), on ground (-BC),

Stokoe 7721, Sept. 1941 (BOL, holo.; US, iso.).

Cape. — 3118 (Vanrhynsdorp): Knersvlakte, towards Douse the

Glim, on ground (-DD), Stirton 9368, 1981.09.21 (PRE).

Parmelia cedrus-montana (Brusse) Brusse,

comb. nov.

Xanthoparmelia cedrus-montana Brusse in J1 S. Afr. Bot. 49:

145 (1983). Type: Cape, 3319 (Worcester): Ceres (—AD), T.B.

Leslie (TUR-VAIN 33499, Hb. Vain. 34575!).

As noted by Knox & Brusse (1983), the species is

a member of the P. hypoleia Nyl. group, even

though the type specimen is only faintly maculate.

The distinct to faint variation in maculation has been

observed in most of the Cape species of this group.

The chemistry is unusual for a Parmelia, containing

thamnolic and squamatic acids with the unknowns

Th— 1 and Th— 2 sometimes present as well. Some

specimens contain Th-1 and Th-2 as the only

medullary substances. P. cedrus-montana is named

after the Cedarberg where it was first rediscovered

Fig. 10 — Parmelia vernicosa,

Brusse 3039, holotype. Scale

in mm.

The upper surface of this distinctive brown

Parmelia does not react very clearly with concen-

trated nitric acid, but certainly belongs to the

Neofuscae (Esslinger, 1977, 1978). Despite this, P.

vernicosa is probably most closely related to the

grey, P. violacea Kurok., a species containing

succinoprotocetraric and fumarprotocetraric acids in

the medulla, with or without (as in the type)

atranorin in the upper cortex. Both these species are

reminiscent of P. endomiltodes Nyl., a Xanthopar-

melia with the same four medullary pigments, but

with salazinic acid present instead.

NEW COMBINATIONS

Parmelia aggregata (D. Knox ) Brusse, comb.

nov.

Xanthoparmelia aggregata D. Knox in J1 S. Afr. Bot., 49: 144

(1983). Type: Cape, 3118 (Vanrhynsdorp): Salt River, Van

(Brusse, 1980). The citation of Brusse 772 8-2-19,

8/2/77 (J) by Knox & Brusse (1983) is unfortunate,

as this species does not conform to the type in

chemistry: the J specimen from Seven Weeks Poort

cited below would have been a better choice.

Cape. — 3218 (Clanwilliam): 3 km W of Olyvenboskraal,

Witelskloof, on large TMS boulder-outcrop ( — BD), Brusse 3091

(PRE, LD), 3098 (PRE, COLO), 3099 (PRE), 3100 (PRE),

1981.05.02; top of Versveld Pass near Piketberg, on TMS

boulders on W aspect (-DC), Brusse 2993, 1981.05.01. (PRE).

3318 (Cape Town): 4 km N of Mamre, on granite outcrops on a

gentle W slope in hilly terrain ( — AD), Brusse 2814, 1981.04.29

tPRE, LD). 3319 (Worcester): 7 km W of Tulbagh, Nuwekloof

Pass, Obiqua mountains, on TMS on a steep SW slope ( — AC),

Brusse 2708 (PRE, LD), 2715 (PRE), 1981.04.28. 12 km E of

Wellington, Bains Kloof near the Convict Graves, on W TMS

cliffs (-CA), Brusse 2586, 1981.04.26. (PRE). 3321 (Ladismith):

Seven Weeks Poort near Ladismith, on TMS (-AD), Brusse 772

11-3-9, 1977.02.11 (J); Seven Weeks Poort, about 9 km from

the main Calitzdorp-Ladismith road, on S faces of TMS exposures

and rocks on a steep E slope, ( — AD), Brusse 3485, 1981.05.08

(PRE, LD); Seven Weeks Poort, about 16| km from the main

Calitzdorp-Ladismith road, on large TMS boulders in the cove of

F. BRUSSE

321

a high water trickle (waterfall), ( — AD), Brusse 3584, 1981.05.09

(PRE, COLO, LD).

Parmelia dysprosa (Brusse & D. Knox ) Brusse,

comb. nov.

Xanthoparmelia dysprosa Brusse & D. Knox in J1 S. Afr. Bot.

49: 148 (1983). Type: Cape, 3318 (Cape Town): Platteklip,

Vlottenberg ( — AD), on granite rock, Garside 5035 (a) (BOL,

holo.; US, iso.).

Cape. — 3218 (Clanwilliam): 3 km W of Olyvenboskraal,

Witelskloof, on large TMS boulder outcrop (— BD), Brusse 3102,

1981.05.02. (PRE).

Parmelia exornata (Zahlbr.) Brusse, comb.

nov.

Parmelia conturbata Miill. Arg. var. exornata Zahlbr.

Zahlbruckner, A., Ann. Cryptog. exot. 5: 251 (1932).

Xanthoparmelia exornata (Zahlbr.) Brusse & D. Knox in J1 S.

Afr. Bot. 49: 150 (1983). Type: Cape, 2917 (Springbok):

Namaqualand, Steinkopf ( — BD), Leg. Pastor G. Meyer (Van der

Bijl 950, W 315, holo!).

This is a very common species and is easily

recognized at drier sites by its discrete, convex and

maculate lobes. The lower surface may be pitch

black as in the type specimen, or tan which is very

common, but the two are not distinct enough to

warrant an additional name. In this case, the extent

of blackening varies from specimen to specimen.

The specimens seen are too numerous to cite

individually. This species is allied to P. fistulata Tayl.

of South America.

Parmelia karoo (D. Knox & Brusse) Brusse,

comb. nov.

Xanthoparmelia karoo D. Knox & Brusse in J1 S. Afr. Bot. 49:

154 (1983). Type: Cape, 3219 (Wuppertal): 32 km NE of

Clanwilliam, Klipfonteinrand ( — AA), Brusse 768 10—3—7 (J.

holo!).

The similarity between this and P. hypoprotocet-

rarica Kurok. & Elix was mentioned by Knox &

Brusse (1983). Further evidence for the maintenance

of this taxon as a species is that hypoprotocetraric

acid is the only substance found in this form,

whereas in the P. hypoleia Nyl. group, protocetraric

acid predominates (P. hypoleia Nyl.), followed by

hypoprotocetraric acid (P. hypoprotocetrarica

Kurok. & Elix), thamnolic acid and related

substances ( P . cedrus-montana (Brusse) Brusse),

barbatic acid (P. burmeisteri Elix) and finally evernic

acid [P. dysprosa (Brusse & D. Knox) Brusse]. The

species is common in Namaqualand and extends as

far south as Clanwilliam District. The specimens

seen are cited by Brusse (1980).

Parmelia leucostigma (Brusse) Brusse, comb.

nov.

Xanthoparmelia leucostigma Brusse in J1 S. Afr. Bot. 49: 155

(1983). Type: Cape, 3322 (Oudtshoorn) 18 km N of De Rust,

Meiringspoort (-BC), Brusse 772 14-1-17 (PRE, holo!); 772

14-1-13 (LD, iso!).

This species is somewhat reminiscent of Parmelia

exornata (Zahlbr.) Brusse, but is not as coarsely

lobed and contains gyprophoric instead of salazinic

acid and related compounds in the medulla. The

species seems to be intermediate between P.

exornata and the more usual Xanthoparmeliae. The

two species are not related to the P. hypoleia group

despite the maculation. Their closest relatives are P.

fistulata Tayl. and P. pachyderma Hue with the same

lobe anatomy (Culberson & Culberson, 1981).

Cape. — 3320; (Montagu): 4 km SW of Montagu, Kogmans

Kloof near the Old British Fort of 1899, on S side of an E-W

ridge (-CC), Brusse 3710, 1981.05.12 (PRE).

UITTREKSEL

Agt nuwe soorte word beskryf en ses nuwe

kombinasies word gemaak in die ligeen genus

Parmelia in suidelike Afrika. Die nuwe soorte is

Parmelia astricta Brusse, P. clivorum Brusse, P.

marroninipuncta Brusse, P. patula Brusse, P. scitula

Brusse, P. spargens Brusse, P. unctula Brusse en P.

vernicosa Brusse. Die nuwe kombinasies is P.

aggregata (D. Knox) Brusse, P. cedrus-montana

(Brusse) Brusse, P. dysprosa (Brusse & D. Knox)

Brusse, P. exornata (Zahlbr.) Brusse, P. karoo (D.

Knox & Brusse) Brusse en P. leucostigma (Brusse)

Brusse.

REFERENCES

Brusse, F. A., 1980. A taxonomic and geographic study of the

genus Xanthoparmelia in the Karoo. M.Sc. thesis, Universi-

ty of the Witwatersrand, Johannesburg.

Culberson, C. F., 1972. Improved conditions and new data for

the identification of lichen products by a standardized

thin-layer chromatographic method. 1. Chromatogr. 72:

113-125.

Culberson, W. L. & Culberson, C. F., 1981. The genera

Cetrariastrum and Concamerella (Parmeliaceae): a chemo-

systematic synopsis. Bryologist 84: 273—314.

Egan, R. S., 1977. New and additional lichen records from Texas

II. Bryologist 80: 136—142.

Elix, J. A., 1981. Further new species of Parmelia (lichens) from

Australia. Aust. J. Bot. 29: 17—23.

Esslinger, T. L., 1977. A chemosystematic revision of the brown

Parmeliae. J. Hattori bot. Lab. 42: 1—211.

Esslinger, T. L., 1978. A new status for the brown Parmeliae.

My cotaxon 7: 45—54.

Filson, R. B., 1982. A contribution on the genus Parmelia

(lichens) in southern Australia. Aust. J. Bot. 30: 511—582.

Galloway, D. J., 1980. Xanthoparmelia and Chondropsis

(Lichenes) in New Zealand. N. Zeal. J. Bot. 18: 525—552.

Hale, M. E. Jr, 1964. The Parmelia conspersa group in North

America and Europe. Bryologist 67: 462—473.

Hale. M. E. Jr, 1971. Studies on Parmelia subgenus

Xanthoparmelia (Lichenes) in South Africa. Bot. Notiser

124: 343-354.

Knox, M. D. E. & Brusse, F. A., 1983. New Xanthoparmeliae

(Lichenes) from southern and central Africa. Jl S. Afr. Bot.

49: 143-159.

Kurokawa, S., 1969. On the occurrence of norlobaridone in

Parmeliae. J. Hattori bot. Lab. 32: 205—218.

Kurokawa, S. & Filson, R. B., 1975. New species of Parmelia

from South Australia. Bull. natn. Sci. Mus. Tokyo, ser. B,

Bot. 1: 35-48.

Nash, T. H., 1974a. New and additional Xanthoparmelia from

New Mexico. Bryologist 77: 72—73.

Nash, T. H., 1974b. Chemotaxonomy of Arizonian lichens of the

genus Parmelia subgen. Xanthoparmelia. Bull. Torrey bot.

Club 101: 317-325.

Zahlbruckner, A., 1926. Afrikanische Flechten (Lichenes).

Bot. Jb. 60: 468-552.

REVIEW OF THE WORK OF THE BOTANICAL RESEARCH INSTITUTE, 1982/83

1st April 1982 — 31st March 1983

CONTENTS

Introduction 323

Reports of sections 324

Staff list 332

Publications by the staff 336

INTRODUCTION

The aftermath of the very successful 1982

AETFAT Congress was still felt during the past

year. It can confidently be said that the Congress

was a stimulus to co-operation between botanists in

South Africa and their overseas counterparts. For

the Botanical Research Institute an improvement of

relations overall was noted. The mammoth-sized

Proceedings (699 pages), which occupied the

energies of many of the staff during the year, will

serve as a confirmation of the success of this

meeting.

The proposed creation of three national working

groups namely, the National Working Groups for

‘Vegetation Ecology’, for ‘Phytosociological No-

menclature’ and for the Flora of Southern Africa, is a

development which could stimulate botanical re-

search in the Department. With the aid of these

working groups it is intended to broaden research in

the appropriate field by full use of all available

manpower in South Africa and elsewhere.

The great success achieved by mobilizing a large

proportion of manpower available in the Institute, as

well as nationally and internationally, for the

preparation of the Flora of Southern Africa, suggests

that a similar approach to ecological research could

have great advantages.

The Flora of Southern Africa is making steady

progress in spite of continuing acute manpower

shortages. Much additional work stimulated by the

BRI with the support of the Department is being

undertaken in this country and overseas. Taxonomic

research is thus advancing on a broad front and

publications on the flowering plants, the ferns, the

mosses and the green- and marine algae are being

finalized for the printer. Work on the lichens is

ongoing and expanding. The magnitude of the task

of writing the ‘Flora’ makes it imperative that the

manpower for flora research within the BRI be

expanded. Limited funds were granted for this

purpose for 1983 but more is urgently needed.

Except for check-lists which have appeared or are in

preparation, it is regrettable that very little

taxonomic research is being undertaken on the fungi

and algae (with the exception of the diatoms),

groups which are no longer the responsibility of the

BRI. Their inclusion in the FSA in the near future

can therefore not be anticipated.

Ecological research is progressing steadily but at

insufficient pace to keep up with demands. The

effect of the expansion of the research team

responsible for collecting vegetation information for

the National Resources Data Bank is not yet being

fully experienced, due to the lack of suitable

ecologists to fill the posts. Prospects for filling the

Fig 1. — View of the main building of the Botanical Research Institute, Pretoria, from the north-east.

324

posts are nevertheless improving now that it is

known that vacancies exist. The staff position on the

whole remained fairly good. Only Ecology and Plant

Exploration showed any vacancies in the profession-

al ranks and the other ranks were also practically all

filled towards the end of the review period.

In spite of the severe loading of particularly senior

professional staff with a diversity of tasks which has

caused severe stresses, the publication record of the

Institute has remained constant at about 90 papers

per annum. The institute journals have maintained

an extremely high standard.

The R. Allen Dyer Book Prize for the most

meritorious scientific publication during 1982/83 was

awarded to Dr M. C. Rutherford for his paper

‘Growth rates, biomass and distribution of selected

woody plant roots in Burkea africana - Ochna

pulchra savanna’ published in Vegetatio 52: 45—63

(1983). This is the second time that Dr Rutherford

has won this prize.

REPORTS OF THE SECTIONS

HERBARIUM SERVICES SECTION

The four herbaria of the Institute continued to

identify plants and provide information for a wide

range of people including officers of the Institute,

various State and Provincial Departments, universi-

ties and the public both in the Republic of South

Africa and its neighbouring states.

National Herbarium, Pretoria (PRE)

The herbarium section continued to be adminis-

tered in an acting capacity by Mrs E. van Hoepen as

curator and Miss W. G. Welman (finances).

A total of 17 153 specimens was named and about

750 visitors were dealt with. During the year 55 loans

(11 056 specimens) were sent to other Institutes, and

29 loans (2 930 specimens) were received. PRE

received 2 184 specimens in exchange, but did not

send out duplicates during the year.

Several collecting expeditions were undertaken

during the year, namely to northern Natal, including

the Ingwavuma area. South West Africa/Namibia

(mainly around Windhoek), the north-western Cape

as far as Aughrabies, the north-western and

north-eastern Transvaal and the Reitz-Frankfort

area of the Orange Free State, which is poorly

represented in the National Herbarium. Numerous

one-day excursions closer to Pretoria for the purpose

of collecting specific taxa were also undertaken.

Because of the drought, conditions for collecting

were generally poor and collecting later in the

summer was confined to areas which had received

some rain.

The replacement of old herbarium cabinets by

new modular steel cabinets continued. None of the

old cabinets has been sent to Durban or Stellenbosch

this year, but three cabinets have been given to the

Plant Structure and Function and four to the Plant

Exploration sections, to house specimens presently

being investigated by these sections. Very few

vacant spaces remain in the four main herbarium

wings, but 205 units still need replacing. The lichen

collection, which is at present housed in the

basement light well, will be moved to B16 and

modular steel cabinets installed. This is, however,

being held up until an alteration to B16 has been

completed. A further major alteration to take place

in the near future will be the installation of air

conditioning in the four herbarium wings for easier

control of insects and protection of the collection, as

well as reducing the fire hazard. This involves

subdividing each herbarium wing by a north-south

wall across the centre.

The number of visitors was about 750, some

coming from overseas and neighbouring states. Dr

Juliet Prior from the Imperial College, University of

London, paid one of periodic visits, as did Dr Peter

Goldblatt from St Louis, Missouri, Mr & Mrs K.

Coates Palgrave, Mr R. B. Drummond and Mr D.

C. H. Plowes were among visitors from Zimbabwe.

Dr C. G. Vosa from Oxford and Dr J. H. Ross from

Australia were other old acquaintances. Mr P.

Halliwell from Kew, together with Mrs Sally Walker

from the USA spent some time collecting in Lesotho

and Natal and brought their specimens to PRE for

identification.

Wing A: Miss C. Reid who is responsible for

identifications of Pteridophytes and all Monocotyle-

dons with the exception of Poaceae, is continuing

her studies on Cyperaceae and worked on a review

of Cyrtanthus for Herbertia as well as a guide to the

plants of the Magaliesberg.

Miss L. Smook, who identifies Poaceae, worked

on her ‘collecting’ project, filling in gaps in our

collections from under-collected areas, and was able

to go on several collecting expeditions and one-day

trips for this purpose. Work on her booklet on

Transvaal grasses continues, as does her recording of

common names of grasses.

Wing B: Mr. G. Germishuizen has completed his

work on the Polygonaceae, which has been accepted

for an M.Sc. degree at the University of Pretoria. He

supplied the text for Mrs A. Fabian’s Transvaal Wild

Flowers and the plant descriptions for Dr F. W. Fox

and Mrs M. E. Norwood Young’s Food from the

Veld, both books having been published during the

year. Mr Germishuizen is responsible for the

identification of mainly Leguminosae in Wing B and

for curating the spirit collection.

Mrs P. M. Olivier helped with identification of

early families of Dicotyledons and was responsible

for research on a number of problems which cropped

up during her work, as well as an article on

Tylecodon grandiflorus.

Wing C: Miss E. Retief is in control of this wing. She

is a member of the Seminar Committee, which has

promoted a number of very interesting talks and

325

slide shows. During the past year she described two

new species namely Raphionacme dyeri and Euclea

dewinteri, and gave a talk on Cyphostemma at the

1983 SAAB Congress. She assists with identifica-

tions in this wing.

Mr P. P. J. Herman rounded off his work on

Pavetta for his M.Sc. thesis, which has been accepted

by the University of Pretoria. He organized a course

on plant collecting, helped with numerous transla-

tions and the moving and installation of new

herbarium cabinets — in addition to his normal work

of identification of specimens.

During the year Mr L. C. Leach worked at PRE

and gave a great deal of valuable assistance in the

identification of Euphorbia and Stapelieae. Mr E. G.

H. Oliver visited PRE from Stellenbosch for a

fortnight to give much-needed assistance in the

identification of Erica specimens.

Wing D: Miss W. G. Welman who is in charge of this

wing, continues as regional abstractor for Excerpta

Botanica (Taxonomica). In addition to identifica-

tions of mainly Asteraceae, Miss Welman has given

assistance to a large number of visitors and

colleagues on various problems, ranging from

identifications, descriptions, training of new staff,

and French translations.

Mrs M. J. A. W. Crosby assists with identifica-

tions in this wing and in addition, administers the

Staff Gift Fund and helps organize social functions.

Cryptogams: Mr J. van Rooy, who is acting as

curator of the moss herbarium, has been granted

study leave and is working for his B.Sc. Hons

degree, spending his spare time and university

holidays at the herbarium, where he is ably assisted

by Mrs S. M. Perold. Mrs Perold is working on

Ricciaceae, in addition to being in charge of the

SEM. She has given much valuable aid 'to various

taxonomists and during the past year has produced

4 327 SEM micrographs for Institute staff.

Mr F. A. Brusse continues to build up the lichen

collection. During the year he identified more than

900 specimens and spent much of his time keeping

up with the literature. Loans of all lichen material

from other South African herbarium have been

requested for future study. The move of the lichen

specimens to room B16 is anticipated in the near

future, but is dependent on certain building

alterations.

Service Room: The Herbarium Service Room, which

may be regarded as the nerve centre of identification

services, continues to be ably administered by Mrs

M. Dednam, through whose hands thousands of

specimens pass monthly.

Natal Herbarium, Durban (NH)

A total of 3 438 specimens was identified, 375

visitors were dealt with plus five student groups.

Accessions to the herbarium numbered 1 108, and

412 specimens were sent out on loan.

Mr B. D. Schrire, the officer in charge of the unit,

has continued his work on Desmodieae, in addition

to his administrative duties. Mrs M. Jordaan has

been responsible for most of the identifications of

specimens. Mr A. Ngwenya started work as

additional herbarium assistant on the retirement in

December of Mr J. Nzuza after 22 years as gardener

at the unit.

Renovations to all the buildings have been carried

out, including complete repainting, sanding of

floors, replacement of rotten timber and broken roof

panels. This involved the complete evacuation of

each building in turn, but was accomplished with the

minimum of disorganization.

Albany Museum Grahamstown (GRA)

2 593 specimens were identified, 991 visitors and 5

groups of students, were dealt with. There were 844

accessions and 15 loans totalling 518 specimens were

sent out. 680 donations of specimens were received.

Mrs E. Brink is the officer in charge of the

herbarium and is assisted by Dr A. F. M. G. Jacot

Guillarmod.

Nine displays were arranged in the Museum foyer

and a number of lectures were given at schools and

to students. Special attention was given to threaten-

ed plants and weeds.

Miss G. V. Britten was away on sick leave for

three months, but has fortunately recovered

sufficiently to return to duty.

A great deal of time has been spent preparing for

the move to temporary quarters while extensions are

being made to the Museum building. The herbarium

has been cleared of many odds and ends and the

garden has been completely cleared with the help of

workers from the 1820 Settlers Garden. The historic

glass-house has been taken down and stored for

re-erection once building operations are completed.

Valuable living plant specimens will in the meantime

be housed in the 1820 Settlers Garden. A volunteer

worker, Mr Neil Abrahams, has given much help in

checking old, valuable collections in the herbarium

prior to packing. This is much appreciated.

Government Herbarium, Stellenbosch (STE)

A total of 4 654 specimens was identified, 354

visitors were dealt with, accessions to the herbarium

numbered 4 286 and 1 193 specimens were sent out

on loan.

Miss L. Hugo, curatrix of the herbarium, married

during the year and is now Mrs van Zyl. She is

assisted in the herbarium by Mrs C. M. van Wyk and

Mrs A. C. Fellingham.

Mrs R. Wikner, who resigned at the end of

December, was replaced by Miss J. Fourie and Jan

Ambraal was replaced by Claude Paulse in

November.

A number of interesting collecting trips were

undertaken and some valuable material collected,

among others a new genus in Proteaceae and a new

species of Cliffortia.

326

Fig. 2. — The Reynolds Gate, designed and made by the late Hans Briigger, one of the two southern entrances to the Pretoria

National Botanical Garden. The main building appears in the background.

FLORA RESEARCH SECTION

Flora of Southern Africa (FSA)

On the recommendation of the Advisory Com-

mittee for Botanical Research to the Minister of

Agriculture and with the support of the South

African Association of Botanists (SAAB) an

Advisory Committee for the FSA is being created.

This Committee will contribute towards the more

efficient production of this important work. During

a special session on the FSA held during the SAAB

Congress in January 1983 it was decided to make

such sessions a regular feature of future annual

congresses. It was also decided to issue a FSA

newsletter.

Two fascicles are in press and will be published

during 1983: (1) Vol. 7, 2, 2 dealing with the genera

Syringodea and Romulea of the Iridaceae which

were contributed by Prof. M. P. de Vos of the

University of Stellenbosch, and (2) Vol. 33, 7, 2

dealing with part of the Gnaphaliinae (Asteraceae)

which comprises Helichrysum, one of the largest and

most difficult genera of the South African flora. This

fascicle was written by Prof. O. M. Hilliard of the

University of Natal.

Manuscripts of four volumes or fascicles covering

a total of about 860 species are at an advanced stage

of preparation and editing and should go to the press

during 1983 or early 1984. They include the volume

on Pteridophyta, written under contract to the

Department by Prof. E. A. Schelpe of the

University of Cape Town, and Vol. 14 on

Crassulaceae, contributed by Dr H. R. Tolken, a

former staff member, now attached to the State

Herbarium of South Australia in Adelaide.

A catalogue of South African green, brown and

red marine algae, compiled by Prof. S. C. Seagrief of

Rhodes University in Grahamstown has been

composed at the Institute and will be sent to the

printer during 1983.

Members of the Institute reported as follows on

progress with research fascicles on volumes of the

Flora:

Vol. 2: Register of names and types for Poaceae.

Since the register was completed in 1982 some

1 200 original descriptions of species have been

photocopied and added to the index. Comparing

this register to a similar one compiled for

Mesembryanthemaceae has led to the conclu-

sion that a register of all southern African plants

should be compiled in two steps: (1) All relevant

names should be drawn from certain works of

great general importance, such as the Flora

Capensis and Kew Bulletin; (2) Complete

registers should then be compiled for each

family when it is revised.

Meliceae: As this group is too small for publication

as an independent fascicle it was re-written and

published in Bothalia.

327

Oryzoideae, Centostecoideae and Bambusoideae:

Work was devoted largely to the genus

Ehrharta. Twenty species, including a new one,

and 12 infraspecific taxa are recognised. A key

and descriptions to most species were compiled.

Key to Southern African grasses: Keys to a further 7

genera and 15 species were drawn up. Certain

vegetative characteristics were correlated with

evolutionary history, flowering phenology, veld

type and successional class.

Vol. 4: Restionaceae. Continued work on the

checklist showed that there are about 320

species in Africa of which 40 are undescribed.

Numerous species complexes were noted. The

survey indicated possible problems with apo-

mixis, where species are only known from

female plants or where male plants may be

absent from certain areas of the distribution

range. Collection of data towards a generic

revision was almost completed. Pollen morpho-

logy, culm anatomy and seed-coat morphology

all provided taxonomically valuable characteris-

tics.

Vol. 5: Liliaceae — Asparagoideae. Twelve of the

species hitherto placed, under Asparagus have

been put in Myrsiphyllum, a genus largely

confined to the southern Cape. The other

species were placed into Protasparagus with two

subgenera: Protasparagus in which the species

represent modified branches and Asparagopsis

in which the spines are modified leaves. The

subgenus Asparagopsis comprises about 50

species.

Vol. 8: Orchidaceae.The family is written up in

Flora format by Prof. E. A. Schelpe of the Bolus

Herbarium under contract, but Dr H. P. Linder,

the present Liaison Officer at Kew, has made a

major contribution to the publication Wild

Orchids of Southern Africa, which deals with all

species known from the region.

Vol. 11: Mesembryanthemaceae - Ruschiinae.

Some 25 000 items of data were gathered by Dr

H. Glen from herbarium specimens and SEM

photographs of seed surfaces. This brings the

total amount of numerical data obtained to over

52 000 items of a desired 65 098, or slightly over

80%. A dendrogram calculated from such a

matrix was shown to be reliable at the 95%

confidence level. Provided that this work is

controlled vigorously by fieldwork the matrix is

considered adequate for the purpose. Distribu-

tion records were mapped using a computer

programme developed by Dr H. Glen.

Vol. 21: Tiliaceae. Dr L. E. Codd updated the

account by the late Prof. H. Wild, and it was

sent to press.

Vol. 25: Ericaceae. Studies by Mr E. G. H. Oliver in

the 22 South African minor genera continued in

two main fields: (1) investigation of inflorescen-

ces and bract/bracteole relationships, and (2)

revisionary work. The inflorescences of all

southern African minor genera so far examined

occur on non-innovating branches, probably an

advanced condition. The philippioid genera in

which the bract is completely recaulescent, with

the abaxial sepal reduced or absent, have been

placed together as the Salaxideae. A formula

expressing inflorescence type and structure and

bract/bracteole relationship has been developed.

Revisions of Philippia, Ericinella, Coilostigma

and Nagelocarpus were largely completed.

Pollen and fruit/seed types have proved to be

taxonomically useful.

Vol. 28: Lamiaceae. The 67 species of the genera

Plectranthus, Solenostemon, Syncolostemon,

Orthosiphon and Thorncroftia, which had been

previously revised, were written up in Flora

format. The following were revised and written

up: Tetradenia (= Iboza ) with 3 species,

Holostylon (1 sp.), Hoslundia (1 sp.), Acrocep-

halus (1 sp.), Geniosporum (1 sp.), Basilicum (1

sp.), Ocimum (2 spp.) and Becium (4 spp.). The

revisionary work in the family is now complete

except for the small genus Leonotis, which is

being revised overseas.

Vol. 30: Acanthaceae — Justicia. All taxonomic

decisions have been taken, descriptions of most

species have been completed and a key to

species has been drawn up. Of the 25 species

recognized from the region only 9 are restricted

to southern Africa. The micromorphology of

pollen and seed surfaces supports the sections

into which Justicia has traditionally been

divided. It also indicates that the division of this

large genus into segregate genera should be

considered.

Ceropegia and related genera

This semi-popular account of Ceropegia, Brachys-

telma and Riocreuxia written by Dr R. A. Dyer is at

an advanced stage of publication and is expected to

appear before the end of 1983.

History of plant collecting

Since the publication in 1981 of the standard

reference work on the subject, Plant Exploration of

Southern Africa by Mary Gunn and L. E. Codd,

biographical information on a further 50 collectors

has been gathered and partly published. A lengthy

publication was devoted to Anton Rehmann

(1840-1971).

Pretoria Flora

Most of the work was translated into Afrikaans

and camera-ready copy of 327 text pages was

prepared at the Institute.

Guide to Drakensberg Flora.

Mrs R. C. Holcroft has almost completed the

several hundred illustrations and Dr D. J. B. Killick

has started on the text.

Palaeoflora of Southern Africa

Drs J. M. and H. M. Anderson who are producing

this major work have made excellent progress:

camera-ready copy of Volume 1 on the Molteno

formation, which deals with the gymnosperm genus

Dicroidium, was completed. Publication by A. A.

Balkema is being negotiated. Volume 2, which will

describe all other gymnosperms of the Molteno

328

formation, is at an advanced stage of preparation. A

Prodromus volume, which presents an overview of

the megaplant fossils from the late Silurian to

Middle Cretaceous, is near to completion.

Freshwater algae

A Catalogue of Freshwater Chlorophyceae of

Southern Africa, which will comprise an estimated

2 000 names, is being compiled by Mrs R. P. Glen,

an ex-staff member.

Liaison officer, Kew

Dr H. P. Linder, the present incumbent, provided

information on taxonomy and nomenclature, as well

as related subjects of southern African plants, for

the Institute and other research centres in South

Africa and overseas. His research was directed

mainly towards the Restionaceae, one of the

dominant families of the winter rainfall region.

PLANT STRUCTURE AND FUNCTION SECTION

During May 1982 the laboratories and staff were

moved to Velcich House in the grounds of the

Pretoria National Botanical Garden. This move has

proved beneficial as additional laboratory and office

space was acquired and a potential fire threat was

removed from the National Herbarium building.

Comparative grass leaf anatomy

In June 1982 Dr R. P. Ellis, who is in charge of

this Section, undertook a three week study tour to

the Smithsonian Institution, Washington and the

Museo Argentino de Ciencias Naturales, .Buenos

Aires. A short stopover was made at the Jodrell

Laboratory, Kew, United Kingdom en route.

Possible co-operation with the revision of the

Gramineae volume of Metcalfe’s Anatomy of the

Monocotyledons was discussed with Dr C. R.

Metcalfe and Dr D. F. Cutler at Kew. Preliminary

findings of the bamboo anatomy studies were

evaluated with Dr T. R. Soderstrom of the

Smithsonian Institution with expansion of the

project in mind. In Argentina Dr E. Sanchez was

visited to discuss our joint interests in grass leaf

anatomy. In addition, over 200 freshly fixed

specimens of grass leaves wert acquired for

anatomical study.

Locally 250 grasses were collected and prepared

for anatomical study. Collections were made in

Maputaland, the Kalahari and South West Afri-

ca/Namibia. The leaf anatomy of Asthenatherum was

written up and the final paper on the genus

Merxmuellera was published. A paper on the leaf

anatomy and taxonomy of Lintonia nutans was also

completed.

Grass identification by epidermal structure for

herbivore food preference studies

Mrs R. Botha has continued with her examination

of variation in epidermal characters of three widely

distributed South African grasses. The results

confirm that geographical variation is mainly

quantitive in nature and, consequently, does not

affect the identification of the grass species.

However, distinct differences were found to exist in

the epidermal structure of different parts of the same

plant. The leaf lamina, leaf sheath, inflorescence

bracts and culms were compared for this purpose, as

were the ab- and adaxial surfaces. Surface differen-

ces were most apparent on the leaf sheaths and

bracts. This observation implies that abaxial leaf

blade characteristics, which are generally assumed to

be representative of the whole plant, are not typical

of that portion of the plant which is grazed by

herbivores and thus, basic assumptions in the

technique of microscopic faecal analysis are ques-

tioned.

The unevenness of cuticle distribution on a single

plant is another aspect of epidermal structure which

affects the quantification of this method. Young

tissue was seen to possess little or no cuticle and in

vitro digestibility studies showed that the digestibili-

ty of grass tissue decreases with age and that the leaf

sheaths are less digestible than the leaf blades.

Cytogenetic studies

Very good progress has been made with the

cytogenetic studies of Lantana camara and Mr J. J.

Spies has been able to show that L. camara only

reproduces sexually. Hybridization within, and

between, different polyploid levels is of general

occurrence resulting in new gene combinations in

the hybrid offspring. This implies that this already

aggressive weed has the potential to spread at an

increased tempo and additional eradication meas-

ures must, therefore, be urgently implemented. It

has also been demonstrated that L. camara has a

basic chromosome number of x =5+6, and not x

= 11, as has been generally accepted in the past.

In addition, Mr Spies and Miss J. C. P. Loots have

initiated an embryo sac study of the Eragrostis

curvula complex in an attempt to gain a better

understanding of speciation and hybridization in this

taxonomically very complex group of grasses. This

work is being carried out in conjunction with the

National Transport Commission which is evaluating

a wide range of E. curvula ecotypes for use in the

stabilization of road cuttings and embankments.

DATA SUB-SECTION

During the course of the year two new major

systems were developed on the burroughs 7800

computer belonging to the Department of Agricul-

ture. The revised form of the Herbarium specimen

data base, PRECIS, was handed over to the Institute

in April 1982. Under the guidance of Mr P.

Gonsalves, it has functioned with great success. All

the outstanding backlog of new specimens which had

accumulated over the past two years, plus the

current year’s accessions, were loaded, as well as

seven years’ backlog of corrections to species names,

plant collectors and specimen data. For the Ecology

Section, Mr R. H. Westfall has transferred to the

Burroughs his package of programs to classify

vegetation types, known as PHYTOTAB. Mrs B. C.

de Wet adapted the system to run on the Burroughs

machine. Both systems are more flexible in their

output than previous systems, and their operating

costs are less.

329

Systems that have continued from previous years

on the Burroughs are Garden Records, under the

control of Mrs de Wet; Register of Names and Types

in Poaceae, developed by Dr G. E. Gibbs Russell;

and sorting programs for vegetation data used by

ecologists at the Stellenbosch Unit.

The Hewlett-Packard 9845B desktop microcom-

puter was updated with a floppy disk drive, and with

this increased memory, new data banks have been

established. Mr J. J. Spies has developed records of

chromosome numbers in Poaceae from a range of

literature references and Mr T. H. Arnold has

developed a data base of locality records for

primitive crop cultivars. Previous uses of the

Hewlett Packard for address labels, word proces-

sing, plotting of graphs and diagrams, and taxono-

mic applications such as Determinavit slips and

distribution maps has continued.

ECOLOGY SECTION

Transvaal bushveld studies

Mr R. H. Westfall has completed his pilot study of

the Sour Bushveld (Veld Type 20) and a number of

papers on various aspects of the work have been

published or are being prepared for publication.

Suggestions for the application of veld condition

assessment in respect of grazing potential in

bushveld vegetation include the use of frequency in 1

m2 quadrats to determine percentage composition.

Preparations for the next phase of these studies have

reached an advanced stage.

South African Savanna Ecosystem Project

The first two phases of this project having been

completed, Dr M. C. Rutherford has published

numerous papers on a variety of aspects of the

research results. He is currently engaged in rounding

off his contributions on the first two phases, as well

as working on the third phase of setting up predictive

models of relevant components of the system and

ascertaining the limits over which the modelling is

applicable. Much of this work has been done during

his study leave at the University of Osnabriick in

West Germany.

Transvaal forest survey

Mr G. B. Deall has distinguished approximately

40 different floristic associations (representing three

major physiognomic classes) distributed throughout

six broad physiographic zones in the Sabie area.

Forests (11 associations), dominate the Eastern

Escarpment Slopes. Closed and Sparse Woodlands

(11 associations) dominate the Escarpment Foothills

and Valley Lowlands. Open Woodlands (four

associations) occur on the exposed Escarpment

Ridge and Upper Mountain Slopes, whereas

grasslands (11 associations) dominate the Escarp-

ment Plateau and Upper Mountain Slopes.

Coastal studies

The report by Dr P. J. Weisser on vegetation and

conservation priorities in Reserve 10 (KwaZulu) has

been finalized, submitted and accepted. New map

and air-photo coverage of the Mtunzini Area has

enabled him to do further research on dune

advancement and vegetation rating to be completed.

He has started fieldwork and air-photo interpreta-

tion of the vegetation of Reserve 7 (between Tugela

River and Matigulu Lagoon, KwaZulu).

Mr M. G. O’Callaghan has studied and mapped

the aquatic, semi-aquatic and adjoining terrestrial

vegetation of 17 Cape estuaries. Of these, five

accounts have been published in CSIR reports.

During this period, work was concentrated on the

south-western and southern Cape Coast. It was

found that very few of these rivers have much in the

way of undisturbed natural vegetation. Most

disturbance is due to the encroachment of alien

plants and development.

Aquatic ecology

Aquatic vegetation data collected in coastal and

inland water bodies in Natal were resynthesized by

Dr C. F. Musil using PHYTOTAB, a computer-

based classificatory approach. He has prepared a

report on this work which will provide the basic text

for a series of papers due for publication in the near

future. He is also preparing a series of papers on the

development and refinement of a model for

predicting yields, growth rates, and amounts and

frequencies of harvest of Eichhornia crassipes to

control both nutrient inputs and excessive growth of

this plant in eutrophied water systems.

Cape fynbos studies

va) Vegetation survey of the Cape of Good Hope

Nature Reserve

Mr H. C. Taylor has prepared papers on this

survey for publication. Analysis of the check list

reveals that, whereas the Nature Reserve occupies

only 16% of the area of the Cape Peninsula, the

flora comprises some 41% of the flora of the whole

Peninsula.

(b) The vegetation of Swartboschkloof, Jonkershoek

The field sampling and data processing for this

facet have been completed by Mr D. J. McDonald.

The forest and fynbos vegetation have been treated

separately, because they are structurally and

floristically different. Analysis of the data shows that

this fynbos vegetation can be classified into 15

groupings (noda), whereas the forest vegetation can

be grouped into six. These plant-community types

are being ranked and described.

(c) A study of the vegetation along transects through

the Western Cape forelands

Mr C. Boucher has completed the analysis of

floristic data collected along three transects through

the western Cape foreland between the Berg River

and False Bay and distinguished 43 plant communi-

ties. Analysis of structural data, collected together

with these floristic data, showed that veld types

could be readily identified on a non-floristic basis,

but that finer community differences could not be

differentiated clearly. Mr Boucher has also incor-

porated plant-community data from other published

and unpublished sources into the transect-data

matrix. Ninety-four plant communities have been

identified and uniformly named in the area. Draft

proposals, in collaboration with other authors, have

330

Fig. 3. — Part of the Encephalartos

(cycad) collection in the Pre-

toria National Botanical

Garden.

been produced for a new sub-division of the Fynbos

Biome as a whole. It is now possible to place the

western lowland communities in a wider and

reasonably natural framework.

(d) Ecophysiological research

The effects of competitive stress between Protea

repens (suikerbos) and an invasive woody alien,

Acacia saligna (Port Jackson), are being investigated

by Miss F. M. Pressinger. Results to date indicate

that A. saligna has greater tolerance of moisture

stress and ability to compete for limited space and

moisture than P. repens in high-density stands of

seedlings. The presence of A. saligna increases the

mortality rate but also increases the growth rate

resulting in taller and more massive P. repens

seedlings. The initial height of the seedling plays a

major role in determining the fate of the plant.

Ecological literature indexing

The plant ecological bibliography for southern

Africa up to 1975 (Vol. 1) is being finally proofread

under the supervision of Miss A. P. Backer and Mr

R. H. Westfall. Work has commenced on the next

volume covering 1976 to 1980.

PLANT EXPLORATION SECTION

This section, under Mr M. J. Wells, is concentra-

ting its attention on food plants research, and related

aspects such as the conservation of germ plasm.

Other utilization facets, as well as weed research

(taken over by the Plant Protection Research

Institute) are being phased out.

Crop plants of African origin

A total of 732 seed and herbarium collections of

primitive crops of African origin were made by Mr

T. H. Arnold, Mrs K. J. Musil and Mr A. A.

Balsinhas. 245 seed samples of Sorghum were tested

for tannin content and the percentage frequency of

occurrence of tannins in each of the five Sorghum

races was calculated. A form, which contains tannin

(i.e. is ‘bird resistant’), but in which the tannins are

‘non-reactive’ during beer-brewing has been found.

Another potentially valuable collection is of a

number of apparently virus-resistant Citrullus (wa-

termelons)— one of 69 Citrullus collections made. 95

collections of Lagenaria (gourds) were examined for

character associations, but only a few loosely-linked

characters were found. In a taxonomic study of

Pennisetum (pearl millet) Mrs B. Pienaar has thus

far, examined 90 specimens for 21 character states.

Indigenous food plants

A national food-plant data-bank is being built up

by Mr A. A. Balsinhas and Mr T. H. Arnold. About

140 elements of information are gathered about each

species. To date 322 species in 40 plant families have

been processed.

331

Conservation of germ plasm

Over 200 collections of seed of primitive crop

plants were incorporated in the germ plasm bank by

Mrs K. J. Musil. A similar number were grown in

order to increase the seed available. Seed was

distributed to other users/germ plasm banks,

including 109 samples to ICRISAT.

130 families of Africans were interviewed regar-

ding their crop-preferences, in order to determine

which crop variants were coming under pressure.

These data have been computerized to aid in

analysis and mapping.

Wood use

A survey of wood use by the Tsonga in a portion

of Gazankulu was completed by Miss C. A.

Liengme. It revealed that the average annual

timber-use per family is 5 ,4 tonnes of firewood and

0,23 tonnes of building timber. Almost all the timber

used is Colophospermum mopane.

Border Cave

Re-dating of the strata of the deposit by the

co-operating archaeologist may open the way to

publication of the botanical part of this survey,

which was completed by Dr J. M. Anderson some

years ago.

Tree distribution in the Transvaal

The field work for this study by Dr J. M.

Anderson has been virtually completed and only a

few extra field recordings were added in the last

year. A further 148 of the species have been

illustrated by Mrs J. van Gogh, bringing the total of

378 of the ± 900 woody species included in the

study.

Barrier plants

Miss L. Henderson has completed data forms for

200 out of a total of 210 priority barrier species, and

109 species have been photographed for the

projected publication on barrier plants.

Information service

Our scientific information service co-ordinated by

Mrs D. M. C. Fourie has handled 556 requests for

information about economic plants and their

utilization or control. These included 101 identifica-

tions, 157 telephonic enquiries and 298 postal

requests. A topical investigation carried out by Mrs

Fourie in conjunction with the CSIR was on drought

resistant plants for gardens in South Africa. Our

Public Relations Officers, first Mrs B. Pienaar and

later Mrs S. D. Hewitt took 51 groups, comprising

2 104 visitors on tours of the Institute and garden.

Mr Balsinhas continued to curate the colour slide

collection and to assist in collecting requested

material.

National Weed List

Classification of the ± 1 600 species in the list was

completed by Mrs V. Engelbrecht of the PPRI and

needs only to be checked before it can be published.

Woody invaders

A survey of exotic woody invaders in the

Transvaal was completed by Miss L. Henderson and

Mrs K. J. Musil, with samples along 2 000 km of

roads and at 217 river crossings, in 64 x \ degree

squares. A total of 60 invader species was recorded.

Their distribution, frequency and aggressiveness

were analysed and the main dangers of invasion

summarized.

Taxonomy of Ficinia

A publication on the morphology of the section

Bracteosae was completed by Mr T. H. Arnold.

PRETORIA NATIONAL BOTANICAL GARDEN

A total of 1 135 accessions, including 343 research

accessions (mainly Eragrostis, Lantana and Citrul-

lus) was received and accessioned by the records

team: Mrs B. C. de Wet and Mrs K. P. Clarke.

Many plants, including Madagascan species, flou-

rished in the new ‘rare and endangered’ house, to

which they were transferred by Mr D. S. Hardy.

Garden staff under Mr H. J. de Villiers concentrated

on maintenance under drought conditions, but also

expanded biome plants. We were all very sad to lose

one of our farm foremen, Mr G. J. Stolz, who died

as a result of an accident whilst on the way to work.

332

BOTANICAL RESEARCH INSTITUTE

Scientific, Technical and Administrative Staff

(31st March 1983)

Director

National Herbarium, Pretoria (PRE)

B. de Winter, M.Sc., D.Sc. (Taxonomy of Poaceae,

especially Eragrostis and of Hermannia; plant

geography)

Deputy Director

D. J. B. Killick, M.Sc., Ph.D., F.L.S. (General

taxonomy, nomenclature, mountain ecology and

editing)

Assistant Director

(Agricultural Research Leader)

D. Edwards, M.Sc., Ph.D. (Ecological methodolo-

gy, aquatic plants, remote sensing, vegetation

structure and physiognomy)

ADMINISTRATION

Chief Provisioning Admi-

nistration Clerk

State Accountant

Senior Administrative

Assistants

Provisioning Administra-

tion Clerks

Administrative Assistants

Personal Secretary to

Director

Senior Clerical Assistant

Clerical Assistant

Accounting Clerk

Receptionist

Typists

Technician: Editorial As-

sistant to Deputy

Director

D. F. M. Venter

Mrs J. Rautenbach

Mrs D. J. Gerber

Mrs J. M. Mulvenna

Mrs J. J. Van Niekerk,

B.A.

Mrs C. van Niekerk*

Miss H. M. Schreiider

Mrs I. A. Ebersohn

Mrs M. M. Loots

Mrs T. M. Creffield

Mrs I. J. Joubert*

Mrs C. A. Bester*

Mrs C. E. M. Engel-

brecht

Mrs S. S. Brink

Mrs J. Gerke*

Mrs S. M. Thiart*

Mrs M. P. M. C. van

der Merwe*

Mrs B. A. Momberg,

B.Sc.*

HERBARIUM SERVICES SECTION

Officer in Charge

(Acting) Mrs E. van Hoepen,

M.Sc.

* Half-day

Agricultural Researcher Mrs E. van Hoepen,

M.Sc. (Acting Cu-

rator; supervision of

identifications and

enquiries)

Administrative Assistant Mrs J. M. Mulvenna

(until Sept. 1982)

Wing A (Pteridophytes— Monocotyledons)

Agricultural Researcher Miss C. Reid, B.Sc.

Hons (Cyperaceae)

Research Technician Miss L. Smook, B.Sc.

(Poaceae)

Technical Assistants Mrs A. M. Fourie*

Mrs S. Burger (from

Jan. 1983)

Wing B (Piperaceae-Oxalidaceae)

Agricultural Researchers G. Germishuizen,

B.Sc. Hons (Polygo-

naceae)

Mrs P. M. Olivier,

B.Sc. Hons

Technical Assistant Mrs I. Smith* (from

July 1982 to Dec.

1982)

Wing C (Linaceae-Asclepiadaceae)

Agricultural Researchers Miss E. Retief, M.Sc.

P. P. J. Herman,

B.Sc. Hons

Technical Assistants Mrs J. I. M. Grobler*

Mrs M. Heymann*

(from Feb. 1983)

Wing D (Convolvulaceae-Asteraceae)

Agricultural Researcher Miss W. G. Welman,

M.Sc.

Research Technician Mrs M. J. A. W.

Crosby, B.Sc.*

Technical Assistants Mrs S. S. Human* (to

Aug. 1982)

Mrs R. Loggenberg*

(from Sept. 1982)

Cryptogamic Herbarium

Agricultural Researchers F. A. Brusse, M.Sc.

(Lichens)

J. van Rooy, B.Sc.

(Musci)

Technical Assistants Mrs L. R. Filter*

Mrs P. W. van der

Helde

Services

Research Technician Mrs S. M. Perold,

B.Sc.* S.E.M. and

laboratory work,

Ricciaceae)

Technical Assistant Mrs M. Dednam*

(herbarium records,

333

loans and exchanges

etc.)

Typist Mrs A. M. Verhoef

Laboratory Assistants J. Phahla (mounting)

G. Lephaka (prepa-

ration and packing)

Natal Herbarium, Durban (NH)

Agricultural Researcher

Research Technicians

Clerical Assistant

Laboratory Assistant

B. D. Schrire, B.Sc.

Hons (Curator, Fa-

baceae and general

identifications)

Mrs M. Jordaan,

B.Sc.*

Miss N. Buirski* (until

Dec. 1982)

Mrs M. K. Lynch*

C. Buthelezi

Albany Museum Herbarium, Grahamstqwn

(GRA)

Agricultural Researchers Mrs E. Brink, B.Sc.

(Curator, general

identifications)

Mrs A. F. M. G.

Jacot Guillarmod,

D.Sc.*

Technical Assistant Miss G.V. Britten*

Laboratory Assistant A. Booi

Government Herbarium, Stellenbosch (STE)

Agricultural Researchers

Research Technician

Technical Assistants

Clerical Assistant

Laboratory Assistant

Mrs L. van Zyl, M.Sc.

(Curator, general

identifications)

Mrs C. M. van Wyk,

M.Sc.

Mrs A. C. Fellingham,

B.Sc.

Mrs R. Wikner (until

Dec. 1982)

Miss W. J. Geldenhuys

Miss J. Fourie (from

Feb. 1983)

J. Ambraal (until Oct.

1982)

C. Paulse (from Nov.

1982)

Data Sub-Section

Data Officer

Datametrician

Research Technician

G. E. Gibbs Russell,

B.S., PH.D

P. Gonsalves, B.Sc.

Hons (Database man-

ager for PRECIS)**

M. D. Panagos, Dip.

Agric. (System ma-

nager for Hewlett-

Packard until Sep-

tember 1982)

Miss J. C. P. Loots,

B.Sc. (System ma-

nager for Hewlett-

Packard after Sep-

tember 1982)

Mrs J. C. Mogford,

B.Sc.* (Quality con-

trol for PRECIS)

Technical Assistants Mrs J. H. Jooste (Chief

encoder for PRE-

CIS)

Mrs E. Evenwel*

(Quality control and

update encoding for

PRECIS)

Mrs H. P. van der

Westhuizen* (Data-

Capture and enco-

ding for PRECIS)

Mrs U. Schoeman*

(Quality control for

PRECIS after Feb-

ruary 1983)

PL4 S. Makgakga (Quality

control for PRECIS)

FLORA RESEARCH SECTION

Officer in Charge O. A. Leistner, M.Sc.,

D.Sc., F.L.S.

Flora of Southern Africa Team

Researcher Leader O. A. Leistner, M.Sc.,

D.Sc., F.L.S. (Ge-

neral taxonomy)

Agricultural Researchers E. G. H. Oliver, M.Sc.

(Taxonomy of Eri-

caceae)

J. M. Anderson,

M.Sc., Ph.D. (Pala-

eobotany, plant

geography)

H. F. Glen, M.Sc.,

Ph.D., F.L.S. (Nu-

merical taxonomy,

Mesembryanthema-

ceae)

G. E. Gibbs Russell,

B.S., Ph.D.

(Taxonomy, especi-

ally grasses; compu-

terized data proces-

sing)

Miss K. L. Immelman,

M.Sc. (Taxonomy,

especially Justicia )

H. P. Linder, B.Sc.,

Ph.D. (Liaison Offi-

cer, Kew; taxonomy

especially Orchida-

ceae and Restiona-

ceae)

L. E. W. Codd, M.Sc.,

D.Sc. (Taxonomy,

especially Lamiace-

ae; history of plant

collecting)

Mrs A. A. Mauve,

M.Sc. (Taxonomy,

** Datametrics

334

Regional Studies

Research Technician

Graphic Artists

Administrative Assistant

Technical Assistant

Officer in Charge

Research Leader

Agricultural Researcher

Technical Assistants

Agricultural Researcher

Research Technician

Learner Technician

Photographer

Senior Librarian

Library Assistant

especially Monoco-

tyledons)

Mrs R. P. Glen,

M.Sc.* (Taxonomy

of algae)

Mrs H. M. Anderson,

M.Sc., Ph.D.* (Pa-

laeobotany)

Mrs R. C. Holcroft*

Miss G. C. Condy,

M.A.

Mrs I. M. M. Kock,

B.A.*

Mrs W. J. G. Roux*

R. P. Ellis, M.Sc.,

D.Sc.

R. P. Ellis, M.Sc.,

D.Sc. (Anatomy of

South African gras-

ses)

Mrs R. Botha, B.Sc.

Hons (Applied grass

anatomy studies)

Miss H. Botha (Micro-

technique)

Miss L. de Wet (until

December 1982)

J. J. Spies, M.Sc. (Cy-

togenetics of Lanta-

na, Rubus and Erag-

rostis curvula )

Mrs H. du Plessis,

B.Sc. (Microtechni-

que of Lantana and

Rubus )

Miss J. C. P. Loots,

B.Sc. (Embryo sac

studies of Eragrostis

curvula)+

Miss A. Alberts

Mrs A. J. Romanowski

Mrs E. Potgieter, B.

Libr.f

Mrs B. F. Lateganf*

Research Leader

Agricultural Researchers

Research Technicians

J. C. Scheepers, M.Sc.,

D.Sc. (Vegetation

ecology, especially

of forest/wood-

land/grassland rela-

tionships; conserva-

tion and land-use

planning; phytogeo-

graphy)

C. Boucher, M.Sc.

(O/C Botanical Re-

search Unit, Stellen-

bosch; lowland fyn-

bos ecology and

phytosociology; con-

servation and land-

use planning; Braun-

Blanquet approach

and techniques)

G. B. Deall, B.Sc.

Hons (Vegetation

ecology of forest/

woodland/grass-

land interrelation-

ships)

D. J. McDonald,

B.Sc. Hons (Moun-

tain fynbos ecology

and phytosociology;

Braun-Blanquet

approach and tech-

niques)

C. F. Musil, M.Sc.,

Ph.D. (Aquatic eco-

logy and survey of

aquatic plants, espe-

cially in Natal; eco-

physiological studies

on Eichhornia cras-

sipes and Salvinia

molesta )

M. G. O’Callaghan,

B.Sc. Hons (Estua-

rine ecology and

phytosociology)

H. C. Taylor, M.Sc.

(Mountain fynbos

and forest ecology;

Braun-Blanquet

approach and tech-

niques; conserva-

tion)

P. J. Weisser, Ph.D.

(Reedswamp ecolo-

gy; air-photo inter-

pretation and map-

ping; Zululand coast

dune vegetation;

conservation)

R. H. Westfall, M.Sc.

(Ecology and phyto-

sociology of Trans-

vaal Bushveld)

Miss A. P. Backer,

B.Sc. (Ecological

ECOLOGY SECTION

Officer in Charge J. C. Scheepers, M.Sc.,

D.Sc.

+ National Transport Commission, Institute of Ecology, Potchef-

stroom University for C.H.E.

t Library Services, Department of National Education

PLANT STRUCTURE AND FUNCTION

SECTION

Comparative Plant Anatomy

Cytogenetics

Photographic Services

Mary Gunn Library

335

data processing and

presentation; ecolo-

gical literature; na-

ture conservation;

air-photo interpreta-

tion and cartogra-

phy)

Miss B. J. Bosnian,

B.Sc. (For. Nat.

Cons.) (Estuarine

ecology; ecological

planning and man-

agement)

M. D. Panagos, N.

Dipl. Agric. (Bot.

Res.) (Computer

science; data proces-

sing: sampling and

monitoring vegeta-

tion and environ-

ment)

Mrs J. Schaap,

H.P.E.D.

(Draughtsmanship

and cartography:

artwork, layout and

design)

Miss A. M. Stadler,

B.Sc. (Ecological

data processing and

presentation; ecolo-

gical literature; na-

ture conservation;

air-photo interpreta-

tion and cartogra-

phy)

J. F. van Blerk, B.Sc.

(Grassland ecology ;

ecological literature;

pasture science)

Technical Assistant Mrs G. S. Nel* (tech-

nical, editorial and

administrative assis-

tance)

Experimental Ecology

Agricultural Researchers M. C. Rutherford,

M.Sc., Ph.D., Dipl.

Datamet. (Biomass

and production stu-

dies in woodland and

fynbos ecosystems;

competition and

stress in fynbos eco-

systems)

Miss F. M. Pressinger,

B.Sc. Hons (Eco-

physiological studies

in fynbos; competi-

tion and stress in

fynbos ecosystems)

PLANT EXPLORATION SECTION

Officer in Charge M. J. Wells, M.Sc.

Agricultural Research

Leader

Agricultural Researchers

Research Technician

Technical Assistants

M. J. Wells, M.Sc.

(Weeds research,

botanical horticul-

ture, fynbos utiliza-

tion and conserva-

tion)

T. H. Arnold, M.Sc.

(Plant utilization,

taxonomy of crop

plants — Sorghum,

conservation of germ

plasm, and threa-

tened indigenous

plants)

Miss L. Henderson,

B.Sc. Hons (Cover

and barrier plants,

woody exotic inva-

ders)

Miss C. A. Liengme,

B.Sc. Hons (Ethno-

botany)

Mrs K. J. Musil, B.Sc.

Hons (Conservation

of germ plasm,

woody exotic inva-

ders).

A. A. Balsinhas (Plant

collecting)

Miss K. Behr,

B.Sc. (Garden utili-

zation and research

collections)

Mrs D. M. C. Fourie,

B.Sc.* (Scientific in-

formation service,

identification of

exotics)

Mrs S. D. Hewitt,

B.A. (Public rela-

tions)

Mrs B. Pienaar,

B.Sc.* (Public rela-

tions)

Mrs K. P. Clarke

(Garden records)

Mrs B. C. de Wet*

B.A. (Garden re-

cords)

PRETORIA NATIONAL BOTANICAL

GARDEN

Curator

Acting Curator (garden)

Acting Curator (nursery)

Agricultural Research

Technicians

Vacant

H. J. de Villiers

D. S. Hardy

H. J. de Villiers, NTC

(Hort.) Dipl. Rec.

P.A. (Development

of savanna biome)

D. S. Hardy (Nur-

sery supervision,

succulents and orc-

hids)

336

Agricultural Research

Technicians T. A. Ankiewicz, Dip.

For. (Administra-

tion, stores and

purchases, and de-

velopment of coastal

forest biome)

Pupil Research

Technicians Mrs A. Celliers

Miss K. Kruger

Mr P. van Eeden

Farm Foremen H. N. J. de Beer

G. J. Stolz

TN 4 Supervisor L. C. Steenkamp

PUBLICATIONS BY THE STAFF

Allen, R., Bowmaker, A. P., Brain, C. K., Edwards, D.,

Newman, G., Owen-Smith, R. N., Ribbink, A. J. & Specht,

R. L., 1982. Goals of ecosystem conservation. In W. R.

Siegfried & B. R. Davies (eds), Conservation of ecosystems:

theory and practice. South African Scientific Programmes

Report No. 61: 3-9. Pretoria: CSIR.

Arnold, T. H., & Gordon-Gray, K. D., 1982. Notes on the

genus Ficinia (Cyperaceae): morphological variation within

the section Bracteosae. Bothalia 14: 15—26.

Arnold, T. H., 1983. Review: Food from the veld, edible wild

plants of Southern Africa by F. W. Fox and M. E. Norwood

Young. 5. Afr. J. Sci. 79: 112-113.

Botkin, D. B., Davies, B. R., Edwards, D., Frost, P. G. H.,

Lucas, G. L., Newman, G., Rabinovitch-Vin, A., Ray, G.

C., Specht, R. L., Van Der Elst, R. & Walker, B. H.,

1982. Ecological characteristics of ecosystems. In W. R.

Siegfried & B. R. Davies (eds), Conservation of ecosystems:

theory and practice. South African Scientific Programmes

Report No. 61: 17—27. Pretoria: CSIR.

Boucher, C., 1982a. Some thoughts about the West Coast

Strandveld vegetation. Veld & Flora 68: 108-109.

Boucher, C., 1982b. Floristic and structural features of the

coastal foreland vegetation, south of the Berg River,

western Cape Province, South Africa. In E. Moll,

Proceedings of a symposium on coastal lowlands of the

western Cape, 21-26. Bellville: University of the Western

Cape.

Boucher, C., 1982c. Western Cape Province lowland alien

vegetation. In E. Moll, Proceedings of a symposium on

coastal lowlands of the western Cape. Bellville: University of

the Western Cape.

Boucher, C., 1982d. Lowcon field excursion guide. In E. Moll,

Proceedings of a symposium on coastal lowlands of the

western Cape, 73-89. Bellville: University of the Western

Cape.

Boucher, C. & McDonald, A., 1982. An inventory of plant

communities recorded in the western, southern and eastern

Cape, up to the end of 1980. South African National

Scientific Programmes Report 57: 1-62.

Codd, L. E., 1982a. Notes on the genus Aptosimum. Bothalia 14:

80-81.

Codd, L. E., 1982b. Plectranthus ecklonii Benth. Flower. PI. Afr.

47: 1. 1854.

Codd, L. E., 1982c. Plectranthus ernstii Codd. Flower. PI. Afr.

47: t. 1855.

Codd, L. E. & Gunn, M. D., 1982a. The collecting activities of

Anton Rehmann (1840-1917). Bothalia 14: 1-14.

Codd, L. E. & Gunn, M. D., 1982b. More early records of Cape

plants. Veld & Flora 68: 93—94.

Cresswell, C. F., Ferrar, P., Grunow, J. O., Grossman, D.,

Rutherford, M. C. & Van Wyk, J. J. P. , 1982. Phytomass,

seasonal phenology and photosynthetic studies. In B. J.

Huntley & B. H. Walker, Ecology of tropical savannas,

476-497. Heidelberg: Springer.

Diamond, J. M., Breytenbach, J., Child, G., Cooper, K. H.,

Frost, P. G. H., Given, D. R., Heydorn, A. E. F.,

MacDoNALD, I. A. W., Ribbank, A. J., Robinson, G. A. &

Scheepers, J. C., 1982. Implications of island biogeography

for ecosystems conservation. In W. R. Siegfried & B. R.

Davies, Conservation of ecosystems: theory and practice.

South African National Scientific Programmes Report No.

61: 46-60.

Dyer, R. A., 1982a. Brachystelma caffrum (Schltr.) N.E. Br.

Flower. PI. Afr. 47: t. 1843.

Dyer, R. A., 1982b. Brachystelma comptum N.E. Br. Flower. PI.

Afr. 47: t. 1844.

Dyer, R. A. & Hardy, D., 1982a. Trichocaulon delaetianum

Dinter. Flower. PI. Afr. 47: t. 1845.

Dyer, R. A. & Hardy, D., 1982b. Trichocaulon alstonii N.E.

Br. Flower. PI. Afr. 47: t. 1846.

Dyer, R. A. & Lavranos, J., 1982. Ceropegia rupicola Defiers.

Flower. PI. Afr. 47: t. 1847.

Edwards, D. & Werger, M. J. A., 1982. Threatened vegetation

and its conservation in South Africa. In A. Schwabe-Braun,

Gefahrdete Vegetation und ihre Erhaltung (Ber. Internat.

Symp. Internat. Verein. Vegetationskunde, Rinteln),

573-583. Vaduz: Cramer.

Ellis, R. P., 1982a. Leaf anatomy of the South African

Danthonieae (Poaceae). VII. Merxmuellera dura and M.

rangei. Bothalia 14: 95-99.

Ellis, R. P., 1982b. Leaf anatomy of the South African

Danthonieae (Poaceae). VI. Merxmuellera arundinacea and

M. cincta. Bothalia 14: 89-93.

Germishuizen, G., 1982. Botanical descriptions. In F. W. Fox &

M. E. Norwood Young, Food from the veld. 440 pp.

Johannesburg: Delta Books.

Gibbs Russell, G. E., 1982a. Urelytrum agropyroides (Hack.)

Hack. Flower. PI. Afr. 47: t. 1841.

Gibbs Russell, G. E., 1982b. Hyparrhenia tamba (Hochst. ex

Steud.) Anderss. ex Stapf. Flower. PI. Afr. 47: t. 1842.

Gibbs Russell, G. E. & Ellis, R. P., 1982a. The genus Melica

(Poaceae) in southern Africa. Bothalia 14: 37-44.

Gibbs Russell, G. E. & Robinson, E. R., 1982b. Speciation

environments and centres of diversity in southern Africa. I.

Conceptual framework. Bothalia 14: 83-88.

Huntley, B. J. & Morris, J. W., 1982. Structure of the Nylsvley

savanna. In B. J. Huntley & B. H. Walker, Ecology of

tropical savannas, 433-455. Heidelberg: Springer.

Immelman, K. L., 1982. Name changes and changes of rank in

Triaspis Burch. Bothalia 14: 78.

Immelman, K. L., Van Wyk, A. E. & De Villiers, P. D., 1982.

Southern African Malpighiaceae — unexplored horticultural

potential. Veld & Flora 68,3: 75.

Jacot Guillarmod, A.F.M.G., 1982. Limnological bibliogra-

phy for Africa south of the Sahara. Nos. 27-29.

Jacot Guillarmod, A.F.M.G., 1982. Checklist of the aquatic

and floodplain vegetation of the Wilderness Lakes, southern

Cape. Bontebok 2: 41-51.

Jacot Guillarmod, A.F.M.G., 1982. A weed, yet beautiful and

useful. Naturalist 26,3: 17—18.

Jacot Guillarmod, A.F.M.G., 1982. A list of Sesotho plant

names with corresponding scientific names. 72 pp. Ministry

of Agriculture, Lesotho.

Jacot Guillarmod, A.F.M.G., 1983. Limnological bibliogra-

phy for Africa south of the Sahara. No. 30 & supplement.

Killick, D. J. B., 1982a. Corrections to Acocks’s Veld Types of

South Africa (1975). Bothalia 14: 143.

Killick, D. J. B., 1982b. Review: Pelargoniums of Southern

Africa by J. J. A. van der Walt & P. J. Vorster. Bothalia 14:

163.

Killick, D. J. B., 1982c. Review: Trees of Natal by Eugene Moll.

Bothalia 14: 163-164.

Leistner, O. A., 1982a. Review: Die natiirlichen Pflanzenfami-

lien nebst ihren Gattungen und wichtigsten Arten,

insbesondere den Nutzpflanzen. Volume 28bl, Angiosper-

mae: Order Gentianales, fam. Loganiaceae. Edited by A. J.

M. Leeuwenberg. Bothalia 14: 161.

Leistner, O. A., 1982b. Review: Thonner’s analytical key to the

families of flowering plants by R. Geesink, A. J. M.

Leeuwenberg, C. E. Ridsdale & J. F. Veldkamp. Bothalia

14: 161-162.

337

Leistner, O. A., 1982c. Review: The identification of flowering

plant families including a key to those native and cultivated

in north temperate regions, 2nd edn by P. H. Davis & J.

Cullen. Bothalia 14, 161 — 162.

McDonald, C. J., 1982. A spring visit to Zoo Ridge. Veld &

Flora 68: 79—80.

Morris, J. W., 1982. Review: Classification and ordination.

Advances in vegetation science edited by Z. E. van der

Maarel. Bothalia 14: 165-166.

Morris, J. W., Bezuidenhout, J. J. & Furniss, P. R., 1982.

Litter decomposition. In B. J. Huntley & B. H. Walker,

Ecology of tropical savannas, 535—553. Heidelberg:

Springer.

Obermeyer, A. A., 1982a. A new species of Gladiolus. Bothalia

14: 78.

Obermeyer, A. A., 1982b. Gladiolus cataractarum Oberm.

Bothalia 14: 78.

Obermeyer, A. A., 1982c. Rhadamanthopsis changed to

Drimioides. Bothalia 14: 78.

O’Callaghan, M. G., 1982a. Biotic characteristics. In A. E. F.

Heydorn & J. R. Grindley, Estuaries of the Cape. Report

No. 10, Kowie (CSEIO), 26—31. Stellenbosch: CSIR

Research Report 409.

O’Callaghan, M. G., 1982b. Biotic characteristics. In A. E. F.

Heydorn & J. R. Grindley, Estuaries of the Cape. Report

No. 11, Hartenbos (CMSI), 23—27. Stellenbosch: CSIR

Research Report 410.

O’Callaghan, M. G., 1982c. Biotic characteristics. In A. E. F.

Heydorn & J. R. Grindley, Estuaries of the Cape. Report

No. 13, Silvermine (CSW3), 16-20. Stellenbosch: CSIR

Research Report 412.

O’Callaghan, M. G., 1982d. Biotic characteristics. In A. E. F.

Heydorn & J. R. Grindley, Estuaries of the Cape. Report

No. 17, Lourens (CSW7), 14-15 & 17-18. Stellenbosch:

CSIR Research Report 416.

O’Callaghan, M. G., 1982e. Biotic characteristics. In A. E. F.

Heydorn & J. R. Grindley, Estuaries of the Cape. Report

No. 18, Botl Kleinmond System (CSW12), 24—29. Stellen-

bosch Research Report 417.

Ray, G. C., Cooper, K. H., Cottrell, C. B., Given, D. R.,

Grindley, J. R., Scheepers, J. C., Shaughnessy, G. &

Tinley, K. L., 1982. Obstacles to the conservation of

ecosystems. In W. R. Siegfried & B. R. Davies,

Conservation of ecosystems: theory and practice. South

African National Scientific Programmes Report No. 61:

10-16.

Ross, J. H., 1982a. Oxalis stenorhyncha Salter. Flower. PL Afr.

47: t. 1850.

Ross, J. H., 1982b. Sesbania macrantha Melw. ex Phill. & Hutch,

var. levis Gillett. Flower. PI. Afr. 47: t. 1856.

Ross, J. H., 1982c. Baphia racemosa (Hochst.) Bak. Flower. PI.

Afr. 47: t. 1857.

Ross, J. H., 1982d. Pterolobium stellatum (Forssk.) Brenan.

Flower. PI. Afr. 47: t. 1858.

Ross, J. H., 1982e. Elephantorrhiza burkei Benth. Flower. PI.

Afr. 47: t. 1859.

Ross, J. H., 1982f. Albizia tanganyicensis Bak. f. subsp.

tanganyicensis. Flower. PI. Afr. 47: t. 1860.

Rutherford, M. C., 1982a. Aboveground biomass categories of

woody plants in Burkea africana — Ochna pulchra Savanna.

Bothalia 14: 131-138.

Rutherford, M. C., 1982b. Annual production fraction of

aboveground biomass in relation to plant shrubbiness in

savanna. Bothalia 14: 139-142.

Rutherford, M. C., 1982c. Review: Heathlands and related

shrublands: analytical studies (Ecosystems of the world,

Vol. 9B edited by R. L. Specht) Bothalia 14: 166-167.

Rutherford, M. C., 1982d. Review: The radiation regime and

architecture of plant stands by J. Ross. Bothalia 14:

167-168.

Rutherford, M. C., 1982e. Review: Succession edited by E. van

der Maarel. Bothalia 14: 168.

Rutherford, M. C., 1982f. Woody plant biomass distribution in

Burkea africana savannas. In B. J. Huntley & B. H. Walker,

Ecology of tropical savannas, 120—141. Heidelberg: Sprin-

ger.

Rutherford, M. C., 1983. Growth rates, biomass and

distribution of selected woody plant roots in Burkea africana

- Ochna pulchra Savanna. Vegetatio 52: 45-63.

Rutherford, M. C. & Panagos, M. D., 1982. Seasonal woody

plant shoot growth in Burkea africana — Ochna pulchra

Savanna. 5. Afr. J. Bot. 1: 104-116.

Scheepers, J. C., 1982. Review: Seedlings of dicotyledons by E.

F. de Vogel. Bothalia 14: 164—165.

Soderstrom, T. R. & Ellis R. P. , 1982. Taxonomic status of the

endemic South African bamboo, Thamnocalamus tessella-

tus. Bothalia 14: 53—67.

Spies, J. J., 1982a. Stomatal area as an anatomical criterion for

the determination of chromosome number in the Eragrostis

curvula complex. Bothalia 14: 119-122.

Spies, J. J., 1982b. The ploidy status of Anthephora pubescens

(Panicoideae). S. Afr. J. Bot. 1: 77.

Spies, J. J. & Stirton, C. H., 1982a. Embryo sac development in

some South African cultivars of Lantana camara. Bothalia

14: 113-117.

Spies, J. J. & Stirton, C. H., 1982b. Meiotic studies of some

South African cultivars of Lantana camara (Verbenaceae).

Bothalia 14: 101-111.

Stirton, C. H., 1982a. The genus Medicago (Leguminosae) in

southern Africa. Bothalia 14: 27-35.

Stirton, C. H., 1982b. Catharanthus and Vinca in southern

Africa. Bothalia 14: 69.

Stirton, C. H., 1982c. A new species of Indigofera from the

south-west Cape. Bothalia 14: 69.

Stirton, C. H., 1982d. A new species of Otholobium in South

Africa. Bothalia 14: 72.

Stirton, C. H., 1982e. A new species of Rafnia from the Cape.

Bothalia 14: 74.

Stirton, C. H., 1982f. A new species of Rhynchosia from Venda.

Bothalia 14: 76.

Stirton, C. H., 1982g. The identity of Boerhavia pterocarpa in

South Africa. Bothalia 14: 79.

Stirton, C. H., 1982h. The correct name for plants called

Agrimonia odorata in southern Africa. Bothalia 14: 80.

Stirton, C. H., 1982i. Uncarina grandidieri (Baill.) Stapf.

Flower. PI. Afr. 47: t. 1852.

Taylor, H. C., 1982. Is fynbos the natural climax vegetation of

the south-western Cape Province of South Africa? In A.

Schwabe-Braun, Ber. Int. Symp. d. Int. Verein. Veget.,

385—391. Vaduz: J. Cramer.

Tolken, H. R., 1982a. Kalanchoe lanceolata (Forssk.) Pers.

Flower. PI. Afr. 47: t. 1848.

TOlken, H. R., 1982b. Adromischus cooperi (Bak.) Berger.

Flower. PI. Afr. 47: t. 1849.

Verdoorn, I. C., 1982. Harpagophytum zeyheri Decne. Flower.

PI. Afr. 47: t. 1853.

Weisser, P. J., 1982. Review: Bekenntnisse eines Okologen by

H. Walter. Bothalia 14: 166.

Weisser, P. J., Garland, I. F. & Drews, B. K., 1982. Dune

advancement 1937-1977 at the Mlalazi Nature Reserve,

Mtunzini, Natal, South Africa, and a preliminary

vegetation-succession chronology. Bothalia 14: 127-130.

Weisser, P. J. & Howard-Williams, C. , 1982. The vegetation of

the Wilderness Lakes system and the macrophyte encroach-

ment problem. Bontebok 2: 19—40.

Weisser, P. J. & Ward, C. J., 1982. Destruction of the

Phoenix-Hibiscus and Barringtonia racemosa communities

at Richards Bay, Natal, South Africa. BothalialA : 123-125.

Wells, M. J. & Stirton, C. H., 1982a. The agrestal weed flora

and vegetation of the world: South Africa. In W. Holzner &

M. Numata, Biology and ecology of weeds, 339—343. The

Hague: Junk.

Wells, M. J. & Stirton, C. H., 1982b. Weed problems of South

African pastures. In W. Holzner & M. Numata, Biology and

ecology of weeds, 429—434. The Hague: Junk.

Westfall, R. H., Dednam, G., Van Rooyen, N. & Theron, G.

K., 1982. PHYTOTAB — a program package for

Braun-Blanquet tables. Vegetatio 49: 35-37.

Westfall, R. H., Everson, C. S. & Everson, T. M., 1983. The

vegetation of the protected plots at Thabamhlope Research

Station. 5. Sfr. J. Bot. 2: 15-25.

339

Book Reviews

Ecology and Utilization of Desert Shrub Rangelands in

Iraq by D. C. P. Thalen. The Hague: Junk. 1979. Pp. viii + 448,

160 x 245 mm, 129 figures, 85 tables. Price Hardbound Dfl.

150/US $ 77,50.

Some 240 000 km2, constituting about three quarters of the

surface of Iraq, are covered by desert or semidesert. The

agricultural potential of this region lies solely in its (semi-) natural

vegetation and the animal life dependent on it. The total desert

flora of the region is estimated by Evan Guest in his classical

introductory volume to the Flora of Iraq at 250 to 300 species of

which some 60% are annuals. During a reconnaissance survey of

the region, Thalen encountered 42 perennial species in the 54

sample locations on which he based his preliminary vegetation

classification which distinguishes 15 community types.

In his introduction, the author provides an overview of the

existing literature. The part on ecology includes a very detailed

account of the environmental setting, a description of the range

cover types and a discussion of the ecology of three key species,

two of which are Asteraceae: Rhanterium epapposum and

Artemisia herba-alba, the other a member of the Chenopodiace-

ae, Haloxylon salicornicum. The chapter on utilization is divided

into sections on grazing, cutting of fuel, ploughing, the effect of

protection and, as in all other chapters, a summary. The fourth

part of the work, which deals with management, is followed by

summaries in English and Dutch, 17 pages of references, an

author index and a very useful general index with numerous

cross-references.

Thalen presents a depressing picture of the ever-accelerating

rate of destruction of the vegetation by grazing, fuel collecting

and ploughing. His observations have shown that in the desert

rangelands of Iraq virtually all plant species are being grazed or

browsed. Under the prevailing conditions heavy utilization by

stock does not appear to result in a change in floristic composition

and no invaders or even increasers are recognized. In the

protected areas recovery is slow and it will require long-term

studies to determine what species composition will result from

continued protection.

The old system under which man and his livestock migrated

according to the dictates of tribal custom and water resources has

been replaced by the present conditions under which, thanks to

modern technology, almost uninterrupted use can be made of

plant resources throughout the region.

Thalen’s main aim is to provide guidelines for improving range

management and for achieving sustained optimal land use. He

recommends a three-phase programme involving soil and

vegetation surveys of increasing intensity and quantitative land

evaluation taking into account socio-economic factors and

cost-benefit predictions. As he points out, implementation of the

required measures would require strong governmental control

but, as he concludes, the same technology which is contributing

towards the rapid destruction of the renewable natural resources

could and should be used to establish optimal land use practices.

The necessary procedures and techniques are largely known, and

with its oil revenues Iraq is in a better position to deal with the

problem of desertification than most other countries in arid

regions.

O. A. Leistner

Macroclimate and Plant Forms: an Introduction to Predic-

tive Modeling in Phytogeography by E. O. Box. The Hague:

Junk. 1981. Pp. xiii + 258, one figure, 25 maps, 19 tables and six

appendices. Price Dfl. 165/US $ 69,50.

This book boldly attempts to use climatic data to predict the

plant growth forms present at any given point on the terrestrial

surface of our planet. This task is so enormous that many may

question the feasibility of the attempt. It should be realized,

however, that even partial success in this type of modelling

venture is undoubtedly valuable in enhancing perspectives of

world vegetation.

The book is the first volume of a new series entitled Tasks for

vegetation science. The series editor, Professor Helmut Lieth of

the University of Osnabriick, has selected a former student of his

at the University of North Carolina to start the series using results

from his Ph.D. work.

Box’s model uses eight climatic variables to predict vegetation

units, namely, 77 plant forms excluding sub-types, at 1 225

terrestrial test sites throughout the world. A major assumption of

the model is stated in the first paragraph of the introduction

namely, that the effects of non-climatic, for example, edaphic,

factors ‘are superimposed on a general climatic situation and are

of secondary importance in determining the general structural,

functional and geographical patterns of land vegetation’. The

eight macroclimatic model inputs are three temperature

parameters, four precipitation parameters and an annual moisture

index derived from precipitation and temperature data. One of

the main parts of the author’s work was to divide the world’s

plants into 77 plant growth forms or 90 sub-types each of which

are described in a lengthy appendix. The author’s use of the terms

‘growth form’ and ‘life form’ does not conform to established

definitions and it is important that the reader does not miss the

explanatory footnote in the introduction.

In the design of the model, Box found that the ‘positive’

modelling of physiological processes with climatic driving forces

was less applicable than ‘negative’ modelling using limiting

factors. He consequently set limits for each of the climatic factors

for each plant growth form. Application of these limits constitutes

the first component of the working model. The other component

is a growth form dominance hierarchy which determines the

relative importance of growth forms at a given site. Results of the

model are plotted on small-scale world maps using the

well-known SYMAP programme with modifications. It is

interesting to note that southern Africa features prominently in

the large world data set, in having some of the highest

physiognomic diversities predicted at certain sites. Box tests his

model by withholding 74 sites from consideration in developing

the model to use subsequently as validation sites. In 50% of these

sites he succeeds in predicting all dominant (and sometimes

subdominant) forms. This degree of success he considers

remarkable.

The author acknowledges that one of the main problems of the

model is the difficulty in correctly predicting the succulent growth

forms. It should also be noted that the application of the model to

the South African validation sites is unsuccessful. The grassy

heathland on the Little Berg at Cathedral Peak, Drakensberg,

becomes ‘Diverse subtropical montane rain forest’ and the

grassland at Kroonstad becomes ‘Diverse savanna woodland with

shrubs’. The author concludes that his model is limited to areas

with little interference by man and to areas where non-climatic,

including edaphic, effects are insignificant.

I found some of the information contained in the book painfully

tantalizing, since I was unable to trace it through normal

information system channels. We learn from an appendix that

CLIMFIT and FITCHECK programmes were used but these are

not published in the open literature. We read: ‘A model

PHOSYN for simulating plant metabolic rates and production

balances from environmental conditions for periods ranging from

hours to years (unpublished)’ (p.30). There is also ‘see SOLWAT

documentation’ and extensive reference to ECOSIEVE, both of

which are unpublished. Some items of 'Box, E. O. (in

preparation)’ occur in the literature list and the text sometimes

contains ‘(Box, unpublished)’ (for example, p.29).

In parts, the style of writing and choice of words used, jar. We

have: ‘a fairly rigorous quantitative approach’ (p. viii) ; uncertain

meanings of parentheses, for example, ‘phytomass (plant size)’

(p.l) and expressions of ‘morphs’ (p.20), ‘target phenomena’

(p.33) and ‘raw distance’ (p.45). In contrast, there is ‘Small trees

(treelets) are most commonly simply smaller versions of larger

trees’ (p. 145). I am at a loss to understand why, in Appendix B of

an English language text, all names of countries cited are given in

local language form. Consequently we have, for example,

Myanma for Burma, Zhongguo for China and Maguarorszag for

Hungary and, to compound the effect, all are truncated to a

maximum of eight characters.

Definitions of plant growth forms are not helped by citing

vague examples such as, inter alia, ‘some Leguminosae’ for

tropical rainforest trees (p.137). Generally, the taxonomic

inclusions in this book seem to show unfamiliarity with plant

taxonomy. Genera are sometimes referred to as species, no

consistent taxonomic system is followed, families are abbreviated,

for example, on p.230, thus: Simaroub., Caesalp., Leg. and

340

Proteac., species of Cymbopogon and Festuca are not classed as

graminoid, whereas a species of Vernonia is, etc.

Plant structural and physiological attributes, although often

mentioned in the same breath, are seldom closely linked, let alone

systematically integrated in the book. Much of the discussion of

physiological responses is not substantiated from any cited

previous work. Later in the text, it appears that much of the

allusion to various physiological attributes and plant processes

could have been easily omitted. The author admits that the first

part of the model on the climatic limits ‘were essentially

empirically determined’ (p.36), whereas the other part of the

model ‘is not intended as an ecological theory but rather as a

convenient, empirically based means of predicting . . .’ (p.43).

Some of the above comments must be seen against the

background of the extremely broad scope of the problem that has

been tackled, virtually single-handedly by Dr Box. Despite

certain imperfections, often also pointed out by the author, this

type of work requires a starting point and the present book goes a

long way towards obtaining a balanced perspective of world

vegetation. It also provides a new, quantitative basis of prediction

upon which others, and undoubtedly the author himself, will

continue to build.

M. C. Rutherford

Contributions to the Ecology of Halophytes edited by D. N.

Sen and K. S. Rajpurohit. The Hague: Junk. 1982. Pp. viii +

272, 75 figures, 69 tables, 17 plates. Price Dfl 165,00/US $ 69,50.

This volume is the second in the series Tasks for Vegetation

Science (series editor: Prof. H. Lieth). According to the editors,

this book tries to give the reader an overall view of saline

environments and the ecology of their plant growth, and attempts

to elucidate new aspects of the problems faced by halophytes. The

paper and binding are of the high quality that we have come to

expect from the publishers. Photographs, figures and tables are

good and, for the most part, easily understood. Typographical

errors are present throughout the book, but although sometimes

irritating, they do not detract from the comprehension of the text.

The book is divided into three parts: Part 1. Biology and

biogeography of halophyte species and salinity-controlled

ecosystems (five chapters); Part 2. Ecological and physiological

problems (eight chapters); Part 3. Potentialities and uses of

halophyte species and ecosystems (one chapter).

Part 1 largely deals with regional studies of halophyte

communities, concentrating on India and Egypt. Although the

species composition of these communities may differ from region

to region, environmental parameters controlling the distribution

of halophytes are often similar. A good basic understanding of

these environmental parameters can be obtained from these

papers. Some rather novel and interesting ideas for the

classification of halophyte communities are also put forward.

The titles of some of the chapters in this section are somewhat

misleading. For example. Chapter 3 is entitled ‘The biogeography

of mangroves’, but only contains extensive lists of mangrove

species together with their broader world-wide distribution and

references. Very few other biogeographic aspects, such as origins,

dispersal methods or natural boundaries are discussed. Chapter 2,

entitled ‘Estuarine ecosystems of Egypt’ discusses some of the

mangroves of that country but not estuaries per se.

Part 2 starts with an excellent account of mangrove zonation by

S. C. Snedaker. Several hypotheses are reviewed in a critical and

interesting way with the conclusion being drawn that no single

factor can account for mangrove zonation, apart from individual

habitat preferences. The rest of this section deals with ecological

and ecophysiological adaptions by plants to saline environments.

The role of specialized salt-secreting structures is discussed

repeatedly here, whereas other factors such as the effect of salt on

mineral nutrition, biochemical interactions and senescence are

only mentioned briefly. Although salinity might be one of the

most important factors controlling halophyte distribution, other

parameters, such as wave action, temperature and inundation

period are also important. For a better understanding of the

ecology and ecophysiology of these plants, the roles of these

factors should have been included.

As is rightly stated in Part 3, about 7% of the Earth’s land

surface is covered by saline-affected areas or has become salinized

by agricultural practices (about 50% of such areas). These areas

are therefore important for a world with rapidly increasing food

requirements. In the last chapter of this volume on economic uses

of halophytes, Zahran and Abdel Wahid relate how Juncus

species can be used in the paper industry, Salsola species are

shown to be a potential source of useful drugs, and numerous

halophytes can be used for fodder. Because halophyte

communities are usually highly susceptible to disturbance, many

ecological impact studies should be carried out to understand the

long-term effects of this type of utilization.

Except possibly for the Section on ecophysiology, most studies

presented in this volume are too regional and specific to be of

much use as a reference text. The only mention of South Africa is

in connection with the distribution of some mangrove species, and

some of this information is suspect. An insight into some general

aspects of halophyte ecology can be obtained by reading this

book. However, better texts are available to review the ecology of

saline environments.

Keeping in mind the title of this volume, this book is successful

and the editor’s aims are achieved. However, if data were

presented to give a more holistic view, and the ecology of

halophytes in different parts of the world were compared, this

would have been a more useful reference book for those

interested in this fascinating group of plants.

M. G. O'Callaghan

The Vegetation of the Subantarctic Islands Marion and

Prince Edward by N. J. M. Gremmen. The Hague: Junk. 1982.

Pp. x + 149, 63 figures, 33 tables. Price Dfl 110,00/US $ 48,00.

This work is the third in the Geobotany series, published by W.

Junk. The subject covered in this volume is a phytosociological

description of the vegetation of Marion and Prince Edward

Islands. The author also deals with the environment and describes

the geology, topography, climate, soils, phytogeography, history

and human influence on the islands’ ecosystems. Although

construction of phytosociological tables presents problems in a

species-poor area, the author has nevertheless, succeeded in

differentiating phytocoena clearly, in well constructed tables.

The author describes his mapping strategy in detail but only

presents a single map of the vegetation on a portion of Marion

Island, in the Nellie Humps area. Direction and landmarks have

been omitted from the vegetation map and the locality of Nellie

Humps is not indicated on the topographic map of Marion Island.

The vegetation map is, therefore, of little use in locating plant

communities. The same applies to the maps illustrating elevation,

groundwater level and soil water content in the Nellie Humps

area. The absence of a vertical scale (Figs 31 & 55) and a

horizontal scale in all but Fig. 58 in the structure-diagrams,

similarly detract from the value of these diagrams.

Although Huntley (1971)* is quoted nine times in the text, it is

remarkable that no reference is made to the vegetation described

by Huntley on Marion and Prince Edward Islands. The dynamics

of the vegetation could have been investigated, in spite of an

absence of releve data, because Huntley published cover-

abundance and constancy values for species listed with his

community descriptions. It is regrettable that dynamics are often

overlooked in current phytosociological work in southern Africa

because of an absence of comparable data so that when data are

available it is surprising that they are overlooked.

The book is readable and the author is to be congratulated on

his presentation of phytosociological tables, the structure of which

can serve as an excellent example for those working in

species-poor areas.

R. H. Westfall

Application of Vegetation Science to Forestry (Handbook

of Vegetation Science. Part XII) edited by G. Jahn. The Hague:

Junk. 1982. Pp. xi + 405, 55 figures and 23 tables. Price Dfl

185,00/US $ 79,50.

This work represents the first part of the Handbook of

Vegetation Science Series to appear since the passing of the

* Huntley, B. J., 1971. Vegetation. In E. M. van Zinderen

Bakker Sr, J. M. Winterbottom & R. A. Dyer, Marion and Prince

Edward Islands , 98-160. Cape Town: Balkema.

341

founding editor-in-chief, the late Prof. Dr R. Tiixen, and the

assumption by Prof. Dr H. Lieth of the role of editor-in-chief of

the series. This book is the fifth part to be published so far. The

editor. Prof. Dr Gisela Jahn, deserves to be congratulated for

tackling the difficult task of putting this volume together and

overcoming the problems associated with its production.

The aim of this volume is to present a survey of the present

stand of vegetation science as applied to forestry in different

countries. The coverage of applications and approaches is

generally restricted to those continents and floristic regions where

forestry has developed along lines similar to the applications and

approaches practised in the northern temperate zone. Neverthe-

less, this book does deal with the history and development of

applications of vegetation science to forestry in 15 countries

distributed over nearly all continents and endowed with a wide

range of environmental conditions. For each country treated, an

introduction to forestry circumstances is given, followed by the

history of the application of vegetation science in forestry, a

description of the main fields of application (especially

classification of growth regions and forest sites) and a critical

evaluation. Unfortunately, there are no contributions from other

countries where the applications of vegetation science to forestry

are also relatively far advanced, such as the United States, Soviet

Russia, Poland and East Germany. However, there are certainly

indications that in these countries vegetation science is being

applied to forestry in ways similar to those covered in this book.

Much has been written about the possibilities for applying

vegetation science to forestry and these possibilities are described

in the introductory chapter by the editor. The contributions from

the different countries show how far these possibilities have been

used to date. Most of the articles have been translated into

English from other languages so that the contributors are being

quoted with the year of publication not reflecting dates of

preparation of their texts. Especially interesting are the

contributions, together with their copious references, from

behind the Iron Curtain.

Of particular interest to South Africans is the contribution on

Australian forests by R. G. Florence. South Africans can learn a

great deal from this account which in many ways suggests the

directions along which the applications of vegetation science to

forestry may develop in this country. This treatment is a very

useful review of the state of the art in Australia. Some of the more

significant points include a few that deserve mention here as

pointing to ways in which South African forest ecology could

profitably develop and the needs for some of this work have

already been indicated in the South African literature and

elsewhere. For instance, research into forest ecosystem processes

is beginning to gain considerable momentum in Australia. Of

particular significance is the growing understanding of the

eucalypt forest system — one of the most important features of

the eucalypts being their efficiency in utilizing and conserving

phosphorus. The retention of phosphorus within the biomass and

its internal or biochemical cycling are, therefore, important

mechanisms for maintaining ecosystem productivity on such a

site. Removal of up to 50% of this phosphorus in the form of

timber could, especially in a short-rotation coppice system of

logging, result in a steady reduction of site productivity. Although

some work on treating the symptoms of nutrient depletion has

taken place here, much basic work needs to be done in South

Africa as, indeed, in Australia. The related questions of the

quality of eucalypt litter and the problems arising from the

immobilization of nutrients in refractory components of soil

organic matter and shifts in microbial equilibrium are similarly

calling for basic research together with research towards better

understanding of the use of fire to deal with these problems. In

South Africa, as in Australia, the main areas of ecological

research required demand attention for better understanding,

planning and management of vegetation resources. These fields of

study include phytosociology, ecosystem processes including fire

ecology, and specific research on the nature and role of eucalypts

(and, possibly, other groups such as Australian Acacia spp. and

Proteaceae).

The European, Canadian and Japanese contributions reveal

how far South Africa lags behind in our knowledge of vegetation

science especially as applied to forestry. Although much of this

understanding is at a level too detailed for our present

requirements, it is clear that we have a long way to go simply to

obtain a better basis of knowledge and understanding, even at a

relatively elementary level, of vegetation science as applied

primarily to conservation forestry and sylviculture in indigenous

forest, but also to plantation forestry.

The editor rounds off the book with a final contribution

drawing together the main themes and presenting her conclu-

sions. Her discussion covers the requirements for the application

of vegetation science to forestry, fields of application including

site mapping, the uses of site maps and forest-type maps, and the

outlook for the future. It is clear that, though vegetation science is

a somewhat specialized field of activity, the research results

obtained are relevant and of value for foresters and other

managers of land. This is particularly so in view of the complexity

of forest ecosystems which have to be managed so that they do not

deteriorate in the course of time. Management implies controlling

and manipulating forest ecosystems on a sustained-productivity

basis to achieve explicitly stated management goals. This is

particularly important in ecologically orientated multiple-use

forestry which can be effectly planned where forest stands are

sustained as a complex of intact, albeit dynamic, ecosystems.

The book is well composed. Figure reproduction, the layout of

tables and setting of the text are clear and pleasing. There is a

useful index and most of the contributions provide comprehensive

reference lists. I can recommend the book as being of value to all

readers interested in vegetation science and forestry.

J. C. Scheepers

Vegetation Dynamics in Grasslands, Heathlands and Medi-

terranean Ligneous Formations edited by P. Poissonet, F.

Romane, M. P. Austin. E. van der Maarel & W. Schmidt. The

Hague: Junk. 1981. Pp. x + 286, 104 figures and 74 tables. Price

Dfl. 195,00/US $ 85,00.

It has been asked when the publication of a series of already

published papers as a separate book can be justified. This

question was discussed by Rutherford (1982)* and I will not

repeat his points here, but they certainly apply to this book whiclj

can be considered to be a companion volume to the book

reviewed by him, viz Volume 3 in the Advances in Vegetation

Science series. This book is Volume 4 in the same series, being the

Proceedings of the Symposium of the Working Groups for

Succession Research on Permanent Plots and Data-processing in

Phytosociology of the International Society for Vegetation

Sciences held at Montpellier, France, in September 1980. Arising

out of circumstances similar to those resulting in the appearance

of Volume 3 of this series, this is a hard-cover reprinted version of

a series of papers that appeared in Vegetatio Vols 46 and 47

(1981). It is certainly convenient to have these papers compiled

into a single hard-bound book, but one may well ask whether the

convenience justifies the additional expense.

Directing attention to the contents, it can be mentioned that

Godron's preface provides an appropriate introduction to the

papers presented at the Symposium. The first contribution is a

review-type paper on permanent quadrats by Austin. This useful

and wide-ranging coverage updates the practical implications of

some of the more modern theoretical thinking and deserves to be

widely read. Usher’s review of models of succession is cautionary

in tone and also deserves to be widely read along with Austin’s

paper. Subsequent papers also stress the dangers of simplistic

thinking when trying to unravel, comprehend and predict the

possible course of events in dynamic vegetation situations

commonly made more complicated by environmental conditions

that change spatially as well as temporally. Hard data derived

from field observation and experimentation are almost invariably

inadequate for scientific interpretation. For these reasons,

mathematical models, in which structure and function each play

their parts in computer simulations of dynamic processes, are

more promising. Subsequent papers return to this point

repeatedly although some papers do have a deceptively simplistic

look about them.

This book has much to interest the South African reader,

especially if he is interested in the dynamics of grasslands or

heathlands and other woody vegetation types, or both, perhaps

more especially if found in environments of the mediterranean

type. Predictably, some of the Australian work is of particular

interest and relevance to us where we can readily draw parallels

* Rutherford, M. C., 1982. Succession. Book review in

Bothalia 14: 168.

342

with South African situations. However, work from elsewhere is

also relevant such as work on the importance of a better

understanding of post-fire dynamics in the management of

heathland vegetation (Gimingham et al., Clement & Touffet,

Gloaguen & Guatier).

One point that does emerge clearly, and rightly so, is the

obvious but so neglected one that any real understanding of the

dynamics operating within a particular stand or type of vegetation

can only be based on scientific research. This means that any such

research must be designed on the basis that several hypotheses,

that may be helpful to explain observed features of dynamic

processes, must at least be examined and at least one hypothesis

must be thoroughly investigated and tested in any one piece of

research. For these reasons, papers such as those compiled in this

book should be prescribed reading for those involved in research

into the theory and practice of management of vegetation, which

almost invariably requires sufficient predictive understanding to

provide at least some scientific basis for manipulating vegetation

to achieve the ends considered to be desirable.

The book is durably bound and of a handy size for convenient

consultation. As one would expect in a compilation of this type,

there is no comprehensive index to the contents nor is there any

synthesis to assist the reader towards obtaining a general

overview. I found the book to be a useful and stimulating

complement to Volume 3 of the series, and I recommend it as

such. Every potential purchaser must decide whether the

convenience of such a singly bound volume of proceedings is

appropriately priced for the amount of consultation to be

expected of such a book.

J. C. SCHEEPERS

Biology and Ecology of Weeds edited by W. Holzner & M.

Numata. The Hague: Junk. 1982. Pp. ix + 461, with numerous

figures and tables. Price Dfl. 265/US $ 99,50.

Biology and Ecology of Weeds is edited by W. Holzner and N.

Numata, and is the second in the Geobotany series. In spite of the

broad title, the book concentrates on agrestal weeds which are

weeds of arable lands. Only in the preface do the editors very

briefly mention that the book deals primarily with agrestals. The

title may lead the casual reader to believe that the book has

serious omissions and that the sections dealing with non-agrestal

weeds, for example, chapter 38, are rather superficial.

The book is divided into four sections. I shall group the first two

sections and the final two sections together for review. The first

section deals with the definition of weeds, weed taxonomy,

methods for studying weeds and the control and preservation of

weeds. The second section includes various ecological aspects of

agrestal weeds, for example, their evolution, reproductive

strategies, population dynamics and competition. My first

impression of the first two sections was unfavourable, but I later

modified this view. My initial reaction was due to the presence of

some poorly written and badly structured chapters. Lack of a

good writing style from authors who do not have English as a first

language may possibly be overlooked, but poor structure and

poor annotation of diagrams cannot. For example, the diagrams

in chapter 12 (Spitters and Van den Bergh) have no labels on the

axes of the graphs and have short unexplanatory legends. Chapter

15 deals with allelopathy in weeds (Numata) and is a difficult

chapter to follow, owing to poor style and lack of structure. The

fact that paragraphs 2 and 3 are in incorrect order adds to the

reader’s confusion. It is fortunate that these chapters are

distributed among better ones and do not mar the book as a

whole.

One of the more interesting chapters is by Musselman on

parasitic weeds. The chapter is well-structured with terms clearly

and concisely defined. The aim and direction of the chapter is

clearly outlined and followed strictly. Musselman chooses five

groups of parasitic weeds to illustrate agrestal parasites, namely,

Cuscuta, Cassytha, Santalaceae, Scrophulariaceae and Oro-

banchaceae. The first two are stem parasites and the others are

root parasites. Musselman uses a common format in dealing with

each weed group, namely, description, distribution, mode of

attack, effects and currently used control measures. The text is

well illustrated by good black and white photographs showing the

parasitic weeds and their host plants.

In general, the first two sections deal with most aspects of

agrestal weed ecology adequately. One notable exception is

chapter 14 by Sagar and entitled ‘An introduction to the

population dynamics of weeds’ which is indeed an introductory

and disappointingly elementary chapter. Perhaps it would have

been less disappointing had there been the introductory chapter

followed by a more advanced complementary chapter.

The second part of the book includes section 3 which consists of

13 chapters each of which describes the weeds of different

countries, including South Africa. Each chapter is discrete and

there is little comparison between countries. In my opinion, a

final chapter that synthesized the findings from the various

countries would have concluded the section satisfactorily. The

fourth and final section consists of a selection of chapters most of

which describe the weeds of pasturelands in various countries. A

chapter on tea plantation weeds and one on aquatic weeds are

tagged on to the section.

Holzner and Immonen begin a notably well researched chapter

with an interesting history of weeds in central Europe. They then

discuss the pattern of weed distribution in Europe by comparing

the weeds of Finland, Austria and Italy. The chapter ends with a

point-form series of conclusions. In addition to the extensive list

of literature cited in the text, there is another list on the weed

floras of many European countries. Holzner and Immonen

comment informally on the relative merits of some of the

references in the latter list and note the presence of sketches and

photographs. This informative list is likely to be of great value to

readers wishing to make more detailed analyses of the weed

literature.

The book as a whole covers most aspects of agrestal weeds. Its

main flaw is that it is poorly edited. Each section contains neither

introduction nor conclusion and leaves the chapters grouped

discretely. Generally, the content of the chapters is good. The

contributors are from a wide range of countries enabling the book

to present a multilingual coverage of the literature. I feel it is a

worthwhile text for all interested in weed research.

F. M. Pressinger

Wild Flowers of Lesotho by Marthe Schmitz. Roma

(Lesotho): Essa (Pty) Ltd , Sappl. 1982. Pp. 252, map, 214 colour

plates. Price Maloti 11,50/R11,50.

This handy pocket-sized soft-cover book is a most welcome

field guide to some of the lovely flowers of Lesotho, and the first

of its kind for that picturesque land. Depicting approximately

15% of the total flora, it naturally concentrates on the larger,

more spectacular flowers but this is what is required for the

layman to make some decision on what he has seen and admired.

A larger format and more pages would have defeated the purpose

— a book to carry into the mountains or along the lowland rivers.

The book is supplied with a glossary in the form of a chapter on

scientific terminology used in the descriptions. It also has a map

for orientation, giving some of the main centres, and a three-page

introduction outlining the botanical history of the country. Two

pages of symbols with their meanings are useful in reducing the

space required for describing each plant illustrated. Scientific

names form the first index, together with the generic numbers,

but without page numbers — a small irritation. The second and

third indices (indexes) are vernacular names in English and

Sesotho.

Each colour plate is provided with genus number, generic and

specific names, family name, vernacular names if any, height and

an alphabetic and symbol-depicting description. Below this,

further information is given, sometimes elaborating the symbols

used, at other times providing further information. Many of the

photographs are clear and well produced, but a few are either

poorly reproduced or were not well-taken originally, so that it is

difficult to decide which is the plant one is dealing with. As

examples of clear and unambiguous plates one can cite Dierama

robustum (p. 57) and Papaver aculeatum (p. 77). Examples of

plates where the viewer is left bewildered, are Metalasia muricata

(p. 198) and Chrysocoma tenuifolia (p. 183). This latter

especially, shows the plants on so small a scale, they are useless

for recognition purposes. While we are quibbling about details, it

is a pity so many English spelling errors have been allowed to

creep in — kaki, popy, adress, are a few. No orthographic errors

are apparent in scientific names and a quick glance shows the

Sesotho appears entirely correct in spelling, though it lacks

diacritics. The user must remember, however, that, tragically for

343

Lesotho and its botany the author died before the book was

finally printed, and perhaps others had to see to the final stages of

production.

Probably due to its mountainous terrain, restricted European

inhabitant numbers and severe climatic conditions, Lesotho has

till now been less well treated, on the whole, botanically, than

many other parts of southern Africa. This present book helps

redress that situation for the layman, providing some basis for

Mosotho resident, casual student or visitor to appreciate the

diversity and beauty of the local plants. It may also lead, one

hopes, to better preservation of what remains of the original

vegetation. It has pioneered the way for illustrated works on the

country’s most varied flora. It would be a difficult task to provide

a workable key, so that one must search for the plant seen, and

perhaps, due to the limitations imposed, not find it, but possibly

later a more lavishly illustrated edition could include a brief key

on floral characters. One has to remember that only one (or

perhaps a few) of each genus is shown so that, for certainty, a

search through a comprehensive checklist or major herbarium is

required.

The grateful thanks of all Basotho and any others interested in

the rapidly disappearing natural vegetation of this mountain

kingdom, should go to the author, who did so much during the

many years she lived in Lesotho, to increase awareness of its

distinctive flora. The book is very well worth buying at three times

the modest price asked and no tourist especially should go into the

mountains without it.

A. Jacot Guillarmod

A Flora Checklist for Swaziland by Ellen S. Kemp.

Occasional Paper No. 2. Lobamba: Swaziland National Trust

Commission. 1983. Pp. v + 101. Price R5,00. Obtainable from

Websters, P.O. Box 292, Mbabane, Swaziland.

This checklist of the Swaziland flora comprises 2 715

indigenous species and 110 naturalized exotics belonging to 943

genera and 175 families arranged according to Schelpe (1970) and

Dyer (1975 & 1976). Synonyms are given in brackets following the

currently accepted, correct names. Only four endemic species are

known for Swaziland, namely Kniphofia umbrina, Streptocarpus

davyi, Eumorphia swaziensis and Cassipourea swaziensis.

An alphabetical index to families and genera as well as an

appendix of additional species not recorded in Compton’s An

Annotated Check List of the Flora of Swaziland ( 1966) and The

Flora of Swaziland (1976) and in Kemp’s Additions and Name

Changes for the Flora of Swaziland (1981) are provided. Each

species in the Appendix has the following information: locality

grid reference, collector’s name and specimen number, as well as

the herbarium or herbaria in which the specimen is housed. A grid

reference map of Swaziland and a list of references are also

supplied.

Mrs Kemp is to be congratulated on the production of this

useful and up-to-date checklist of the flora of Swaziland.

G. Germishuizen

Trees of Swaziland by Ellen S. Kemp. Occasional Paper No. 3.

Lobamba: Swaziland National Trust Commission. 1983. Pp. vi +

59. Price R3,00. Obtainable from Websters, P.O. Box 292,

Mbabane, Swaziland.

This small (A6 size) soft-covered booklet is a guide to the trees

of Swaziland as well as being a companion to A Flora Checklist for

Swaziland reviewed above.

There are 464 species of indigenous trees listed as well as 10

naturalized exotics — belonging to 233 genera and 79 families.

Families are arranged according to Dyer’s, The Genera of

Southern African Flowering Plants (1975 & 1976). For emphasis,

scientific names are given in capitals, italics not being available. A

useful feature of this booklet is the citation, when available, of a

synonym or synonyms under each species, which facilitates cross

reference with previous works. Tree numbers are given and are

taken from De Winter et al’s The National List of Trees (1978).

A sketch map is provided showing the geographical regions of

Swaziland, namely the Highveld, Middleveld, Lowveld and

Lubombo. These regions are given in parenthesis under each

species.

The common names of the trees have been included first in

siSwati and then in English [mainly taken from De Winter et al.

(l.c.)]. Alphabetical indices to families and genera, siSwati and

English common names are provided, as well as a list of

references used.

This booklet will undoubtedly prove most useful to all those

interested in the trees of Swaziland and, in the words of the

author, ‘although far from complete in every aspect, this is the

first time a list of trees with their distribution and common names

has been compiled for Swaziland.’ A criticism that can be levelled

at this publication is the extremely small (5 point) computer print,

which makes for difficult reading.

G. Germishuizen

Medicinal Plants and Traditional Medicine in Africa by

Abayomi Sofowara. Chichester: John Wiley & Sons. 1982. Pp.

256, 17 black and white photographs, price £14,50.

Prof. Sofowara’s approach is refreshing in that, despite his

obviously being a protagonist of traditional medicine, he does not

seek a package acceptance of it, but rather seeks to stimulate

critical appraisal via the provision of a text book.

This book is aimed primarily at students of pharmacy and

medicine in the developing countries of Africa. Its coverage

includes: terminology and definitions, history, methods and

techniques, evidence for the efficacy of some remedies,

relationships between traditional medicine and modern drugs,

methods of obtaining information on medicinal plants, screening

for bioactive agents, guidelines for research, and research trends.

He also deals with the somewhat controversial subject

‘integration or co-recognition of traditional and modern medicine’

and in assessing the advantages and disadvantages of the former,

is quick to acknowledge shortcomings such as: imprecise

diagnosis, imprecise dosage and lack of proof of efficacy.

There is also a chapter on some common medicinal plants that

deals with 10 species. Other species are mentioned elsewhere in

the text, but Prof. Sofowara makes it clear in his Preface that the

plants ‘included are only examples picked to illustrate the pattern

which a student could follow in studying other medicinal plants in

his neighbourhood’, and that he does not set out to duplicate

regional medicinal plant listings that are already available. This

approach is completely acceptable, but does not justify the accent

on medicinal plants in the title of the book.

The bibliography and the lists of references at the end of each

chapter, are useful but the bibliography is decidedly strange,

being in neither alphabetical nor time sequence, and starting with

title rather than with author. The index is better but is incomplete

e.g. it does not include latin names of fungi and other organisms

from page 228. It is also muddling to find italics used not only for

specific names, but also for some folk names e.g. ‘Ifa’ but not

‘Odu’ (page 31) that are not italicized in the text. The

photographs reproduced from WHO publications are reasonably

good, but some of herbarium specimens e.g. Fig. 15, are not.

Despite these technical drawbacks, Prof Sofowara’s book is a

distinct step forward in providing a platform for training medical

and pharmacy students — but there is not a great deal in it for the

botanist.

M. J. Wells

The Ferns and Fern Allies of Southern Africa by W. B. G.

Jacobsen. Durban: Butterworth. 1983. Pp. 542, A4, 372 figures,

186 maps. Price R65,00.

The most recent account of the Pteridophytes of southern

Africa, apart from Schelpe’s work in Flora Zambesiaca, published

in 1970, is that published by Sim in 1915, which is now outdated

and rare Africana. For this reason alone, the present book would

be welcome, as a work is needed on ferns and their allies, that is

not iniquitously expensive (although the price of the present book

probably excludes most botanists from the class of potential

purchasers). Jacobsen has set out to produce a popular guide to

southern African Pteridophyta, but has, to this reviewer’s mind,

succeeded in doing rather more than that.

The first part of the book, occupying 113 pages printed in

double columns, is filled with general information. This includes a

definition of the group dealt with, and notes on the conservation,

344

palaeobotany, morphology, anatomy and classification of the

group. There are chapters on distribution, ecology, life forms,

phytogeography and the communities in which ferns and fern

allies are found. Each community is described in a separate

section of two to twenty pages — a meaty chunk of text, in a book

meant for amateurs.

The second part of the book starts with notes on how to collect

herbarium material of ferns, classification, nomenclature and

taxonomy, a glossary and identification keys. The rest of the book

— almost 300 pages — is occupied by individual descriptions of

the families, genera and species of southern African ferns and

their allies. Three hundred black-and-white photographs, mostly

of herbarium specimens, illustrate the 338 species described here,

and their distributions are plotted on almost 200 maps. Tables

display the distribution of species with regard to rainfall,

environment, geography and altitude. An extensive bibliography

and a comprehensive index complete the volume.

For each species, the correct name and author are given, with

references to Sim, Flora Zambesiaca and other recent literature as

appropriate. The most important synonyms and literature

references to these synonyms are given. Protologues and

specimens are, however, not cited. The description of each

species is followed by notes on distribution, habitat and other

points of interest.

This major work on southern African Pteridophytes can only be

welcomed warmly, coming as it does almost 70 years after its very

rare predecessor. Sadly, one wonders whether Jacobsen will hold

the field as long as Sim did, in view of the fact that publication of

Schelpe's volume on Pteridophytes in the Flora of Southern Africa

series is believed to be imminent. This reviewer hopes that the

two works will complement rather than compete with each other.

Unfortunately, this book has a few disquieting features. Some

of the keys (e.g. those to the species of Lycopodium, Selaginella

and Alsophila) bear an uncanny resemblance to the keys in

Schelpe's account in Flora Zambesiaca, although Schelpe is not

acknowledged for these. Almost all of the keys, although of the

style called indented, are not in fact indented, making the location

of the contrasting parts of a couplet unnecessarily difficult. Three

couplets in the key to families are in fact not couplets but triplets,

but nothing in the first lead of any of these indicates the existence

of two possible alternatives. The idea of printing the keys on a

screened background is, however, a good one. They are

highlighted very effectively in this way.

One wonders whether a book of this nature is an ideal place to

publish a new species (Cheilanthes nielsii Jacobsen) formally. It

seems to this reviewer that there is a significant chance that this

new species will be lost to any index of new taxa. More serious is

the inadvertent valid publication of four new combinations (in

Cheilanthes) attributed to Schelpe, presumably to be published in

the Flora of Southern Africa, but in fact published in this book for

the first time. Jacobsen has also published two nomina nuda (one

each in Isoetes and Cheilanthes) — of plants which should be

published first in the Flora of Southern Africa. It seems that his

keenness to be ‘up-to-the-minute’ has betrayed the author of this

book in these cases.

There are also a few other criticisms that one may level at this

book. The most serious of these concerns the list of rare and

endangered species in the first part of the work. Among succulent

plants, the almost inevitable result of the publication of such a list

with any but the vaguest of localities (even ‘Africa’ may be too

precise) would be the extermination of the plants in the wild by

over-enthusiastic collectors and ‘nurserymen’ keener on a fast

buck than on conserving their resources. One hopes that fern

collectors are more amenable to self-discipline than succulent

collectors.

The text is somewhat heavier going than the general public, at

whom the book says it is aimed, may be accustomed to. For

example, one wonders what proportion of the readership of this

book, other than those versed in soil science, would be able to

form an immediate mental image of a ‘humiferous lithosol’ (the

preferred habitat of Lycopodium saururus). The glossary does not

help with this piece of jargon — it is restricted to botanical terms.

The occasional non sequitur — ‘The total number of species ... is

with 243 considerably enlarged over the 247 species listed ... in

1952’ and linguistic infelicity — ‘obnoxious (for noxious) weeds'

mar an otherwise very interesting text. Finally, one may observe

that the loss of detail caused by the normal half-tone printing

process in some of the photographs has resulted in pictures

illustrating the reason for the superiority of a good line drawing

over a photograph of low contrast, rather than illustrating any

details of the plant shown. On the other hand, many of the

pictures are excellent, particularly most of those by Mervyn

Brand and A. J. Romanowski.

This book is a major contribution, if not to the study then at

least to the popularization (which may be just as important) of the

southern African flora. It is a book of many qualities, and one

which will repay extended study. One may mention among its

minor excellences the vegetation map in the endpapers, which is

one of the most readily intelligible of its kind published in many a

long day.

H. F. Glen

Bothalia 15, 3 & 4: 345-385 (1985)

A taxonomic revision of the type section of Pelargonium L’Herit.

(Geraniaceae)

J. J. A. VAN DER WALT*

Keywords: Pelargonium (Geraniaceae), taxonomic revision, type section

ABSTRACT

Twenty four species are recognized in this taxonomic treatment of the section Pelargonium which was last

revised by Knuth in 1912. Most species occur in the south-western, southern and eastern Cape where they usually

grow in rather moist, semi-shaded habitats. A key to the identification of the species has been compiled, and at

least one illustration as well as a distribution map is presented for each species. The section is considered to be the

most primitive section of the genus with a basic chromosome number of x = 11.

INTRODUCTION

In the section Pelargonium more species than in

any other section have contributed towards the ma-

terial welfare of mankind. A very large number of

artificial hybrids, known as ‘Regal Pelargoniums’ or

‘Martha Washingtons’ in the USA, have been pro-

duced for ornamental purposes. Features of P. cu-

cullatum (L.) L’Herit. can be seen in many of these

hybrids, and this well-known species could be con-

sidered as the most important ancestor. P. graveo-

lens L’Herit., P. radens H. E. Moore and other aro-

matic species of the section are the ancestors of an-

other group of artificial ornamental hybrids known

as ‘Scented-leaved Pelargoniums’. Hybrids of P.

graveolens, especially, are grown on a commercial

scale for the production of geranium oil which is

used as a substitute for attar of roses in the perfume

trade.

When L’Heritier (1789) described the genus Pel-

argonium, he did not subdivide it. Sweet (1822) was

the first to make a subdivision: he elevated L’Heri-

tier’s genus to tribal level and his tribe Pelargonieae

included genera such as Campylia Sweet, Jenkinso-

nia Sweet and Pelargonium, which are currently con-

sidered as sections of the genus Pelargonium. His

genus Pelargonium included species of the currently

recognized section Pelargonium as well as members

of other sections such as Dibrachya Sweet.

De Candolle (1824) reduced Sweet’s tribe Pelar-

gonieae to generic level as the genus Pelargonium,

and subdivided the genus into 12 sections. One of

these sections, Pelargium (the current section Pelar-

gonium), was divided into four series. The present

species of the section Pelargonium were included in

his series Anisopetala.

Ecklon & Zeyher (1835) followed Sweet’s classifi-

cation system: they again raised the present genus

Pelargonium to tribal level as the tribe Pelargonieae,

which was divided into 15 genera. Their genus Pelar-

gonium included species of the current section Pelar-

gonium.

Harvey (1860) followed De Candolle’s system and

divided the genus Pelargonium into 15 sections. His

* Botany Department, University of Stellenbosch,

Stellenbosch 7600.

last section Pelargium, included 22 species of the

current section Pelargonium. Knuth (1912) also fol-

lowed basically the same system as De Candolle, and

divided the section Pelargonium (Pelargium) into se-

ven subsections. This subdivision was based on the

colour of the petals, leaf characters and the length of

the pedicel and hypanthium.

In 1979 Van der Walt proposed P. cucullatum as

the lectotype species of the genus Pelargonium.

The section Pelargonium is distinguished by a

combination of characters. Sweet (1812), De Can-

dolle (1824), Ecklon & Zeyher (1835), Harvey

(1860) and Knuth (1912) mentioned that the section

is characterized by a shrubby habit, five petals of

which the posterior two are larger than the anterior

three, and the presence of seven fertile stamens.

Some of these authors mentioned additional diag-

nostic features such as the shape of the leaves and

the free stipules.

The geographical distribution of the section was

discussed by Van der Walt & Vorster (1983). The 24

species occur in the south-western, southern and

eastern Cape with a few species extending north-

eastwards as far as the eastern highlands of Zim-

babwe. The two highest concentrations of species, in

the south-western and southern Cape, fall entirely

within the winter rainfall region. Many of them,

however, do not occur in close association with fyn-

bos, but rather favour primitive, moist shaded habi-

tats, often in association with forest precursors. The

species in the eastern Cape receive rain in winter as

well as summer, whereas those in the Transkei, Na-

tal, Transvaal and Zimbabwe occur in a predomi-

nantly summer rainfall region.

The section Pelargonium is considered to be the

most primitive section of the genus on account of its

rather woody, much-branched, shrubby habit, sim-

ple leaves, and five-petalled flowers with seven fer-

tile stamens. Albers & Van der Walt’s (1984) chro-

mosome study of the section supported this view.

The basic chromosome number of the section is x =

11, which is most probably also the basic number for

the genus. Furthermore, the chromosomes are rela-

tively small in comparison with those of other sec-

tions of the genus. Twelve species are diploid

(2n=22) and twelve polyploid (2n=44, 66, 88). The

346

Bothalia 15, 3 & 4 (1985)

two species occurring in the Transvaal, P. glutino-

sum (Jacq.) L’Herit, and P. graveolens, are both

polyploids, which might indicate that the section had

a southern origin.

Many species show continuous morphological va-

riation, an indication that they are probably still in

an active state of speciation. Natural hybrids be-

tween representatives of the section Pelargonium, as

well as hybrids between representatives of the sec-

tion Pelargonium and the sections Eumorpha (Eckl.

& Zeyh.) Harv. and Glaucophyllum Harv., have

been collected. A list of these hybrids with their pu-

tative parent species, is presented at the end of the

paper.

This paper represents an alpha taxonomic treat-

ment of the section. The species are phylogenetically

arranged, but the relationships of the species will be

discussed in detail in a following paper. In order to

do so, use will be made of additional features, viz

anatomy, pollen morphology and cytotaxonomy.

Section Pelargonium: DC., Prodr. 1: 658 (1824)

(series IV, Anisopetala DC. p.p.); Harv. in FI. Cap.

1: 301 (1860); Knuth in Pflanzenr. 4, 129: 320, 455

(1912) — all as Pelargium. Lectotype species: Pelar-

gonium cucullatum (L.)L’Herit. (see Van der Walt

in J1 S. Afr. Bot. 45,3: 379-380 (1979) and Van der

Walt & Vorster in Taxon 30: 307 (1981).

Genus Pelargonium: Sweet, Geran. 1: viii, 41

(1821) p.p.; Eckl. & Zeyh., Enum. 1: 78 (1835).

Erect to decumbent, branched to much-branched,

non-aromatic to aromatic, non-viscid or viscid

shrubs or subshrubs, up to 2,5 m high and 1,6 m in

diameter. Stems herbaceous when young, becoming

variously woody with age, sparsely to densely cov-

ered with different types of non-glandular and glan-

dular hairs, young stems green but becoming greyish

to brownish with age. Leaves simple, petiolate,

stipulate, indumentum variable as on stems, dark

green, green or greyish green; laminae entire to vari-

ously palmately or pinnately incised, shape variable

but often cordiform, base usually cordate seldom cu-

neate or truncate, apices of laminae/lobes/segments

obtuse or acute margins finely to coarsely dentate

or serrate, 5-250 x 3-270 mm; petioles 0-250 mm

long; stipules free, usually cordiform or triangular

and often apiculate or cuspidate, 2-20 x 1-15 mm.

Inflorescences: flowering branches in some species

profusely branched, with smaller and normal foliar

leaves or smaller foliar leaves only; peduncles un-

branched, in some species distinctly articulated at

distal and proximal ends in infructescences, 5-150

mm long, indumentum variable as on stems; invo-

lucral bracts mostly ovate or lanceolate, apiculate

to caudate, 3-10 x 1-7 mm, indumentum variable as

on stems; pseudo-umbels with 1-20 flowers each.

Pedicels 0,5-20 mm long, shorter or longer or as long

as hypanthiums. Hypanthiums 1-14 mm long, base

variably thickened, indumentum on pedicels and hy-

panthiums variable as on stems. Sepals 5, usually

lanceolate, often apiculate, green and often with a

reddish-brown tint, in some species with white mar-

gins, posterior one wider than other four, 7-20 x

2-6 mm. Petals 5, white, pink, pinkish-purple or pur-

ple; posterior two usually spathulate or obovate,

apices often emarginate, with dark purple, wine-red

or dark red feather-like markings, reflexed at less

than 90°, 90° or more than 90°, 6-35 x 3-17 mm; ante-

rior three usually spathulate or oblanceolate, with or

without narrow claws, reflexed at less than 90°, 4-28

x 2-12 mm. Fertile stamens 7 (4 long, 1 medium, 2

short), often pinkish but becoming progressively

paler towards hyaline staminal column, pollen

orange; staminodes 3. Ovary 5-lobed, usually ovoid,

densely pilose with apically directed hairs, white to

green; style 5-12 mm long, base variably hairy;

stigma with 5 recurved branches, purple to reddish;

mericarps 5, bases 3-6 mm long, tails 12-30 mm

long, plumose. 2n=22, 44, 66 or 88.

Diagnostic features

Erect to decumbent, non-aromatic to aromatic,

non-viscid or viscid subshrubs or shrubs with rather

woody stems. Leaves simple, laminae entire to vari-

ously palmately or pinnately incised. Inflorescences:

flowering branches with smaller and normal foliar

leaves or smaller foliar leaves only, pseudo-umbels

with 1-20 flowers each. Flowers white, pink, pink-

ish-purple or purple; posterior petals with dark pur-

ple, wine-red or dark red feather-like markings; fer-

tile stamens 7 (4 long, 1 medium, 2 short).

KEY TO THE SPECIES OF THE SECTION PELARGONIUM

1 Laminae without incisions or very shallowly lobed:

2 Mature laminae less than 20 mm long:

3 Laminae glabrescent to hirtellous:

4 Laminae without incisions, circular to broadly elliptic to ovate and not viscid 2. P. betulinum

4 Laminae lobed, cordiform, viscid 7. P. glutinosum

3 Laminae hispid to strigose:

5 Laminae narrowly obovate to broadly obovate, not crisped, base cuneate ... 20. P. hermanniifolium

5 Laminae reniform more or less crisped, base truncate to cordate:

6 Laminae hispid, rose-camphor-scented, posterior petals 10-18 x 4-8 mm, pseudo-umbels

2-5-flowered 17. P. englerianum

6 Laminae strigose, lemon-scented, posterior petals ca 18 x 10 mm, pseudo-umbels 1— 2(— 3)-

flowered 19. P. crispum

Bothalia 15, 3 & 4 (1985)

347

2 Mature laminae more than 20 mm long:

7 Laminae glabrescent, scabrous or at least hard to the touch:

8 Laminae glabrescent to hirtellous:

9 Laminae without incisions or lobes:

10 Laminae circular to broadly elliptic to ovate, flowering branches not profusely branched,

flowers almost actinomorphic 2. P. betulinum

10 Laminae cordiform, flowering branches profusely branched, flowers extremely zygomorphic

21. P. cordifolium

9 Laminae incised or lobed:

11 Leaves viscid, posterior and anterior petals almost equally sized 7. P. glutinosum

11 Leaves not viscid, posterior petals much larger than anterior ones:

12 Petioles and peduncles villous, young stems reddish-brown 15. P. sublignosum

12 Petioles and peduncles hirtellous to strigose, young stems green:

13 Lobes of laminae acute, flowers white to purple to pinkish-purple 12. P. scabrum

13 Lobes of laminae usually obtuse, flowers white to light pink:

14 Petioles much longer than laminae 18. P. greytonense

14 Petioles shorter than or as long as laminae 13. P. ribifolium

8 Laminae pubescent, hirsute, hispid or strigose:

15 Laminae pubescent to hirsute, flowers extremely zygomorphic 22. P. hispidum

15 Laminae strigose or hispid, flowers zygomorphic or almost actinomorphic:

16 Laminae hispid, bases truncate to cordate 17. P. englerianum

16 Laminae strigose, bases cuneate:

17 Laminae with shallow angular incisions in distal half, flowers almost actinomorphic,

posterior petals at least 15 mm long 1. P. cucullatum

17 Laminae palmatilobate, flowers zygomorphic, posterior petals usually less than 15 mm

long 12. P. scabrum

7 Laminae with long hairs or at least soft to the touch:

18 Laminae abaxially more hairy than adaxially and discolorous:

19 Laminae cordiform and not incised or lobed, flowering branches profusely branched, posterior

petals much larger than anterior ones 21. P. cordifolium

19 Laminae panduriform to cordiform, pinnatilobate, posterior petals slightly larger than anterior

ones 5. P. panduriforme

18 Laminae equally hairy adaxially and abaxially and not discolorous:

20 Laminae hood-shaped at base, posterior petals usually more than 20 mm long

1. P. cucullatum

20 Laminae not hood-shaped at base, posterior petals less than 20 mm long:

21 Pedicels up to 3 mm long, posterior petals slightly larger than anterior ones, pseudo-umbels

capitulum-like:

22 Laminae crisped, villous to densely villous, pseudo-umbels 8-20-flowered, decumbent

shrub 3. P. capitatum

22 'Laminae not crisped, pilose to villous, pseudo-umbels 3-12-flowered, erect shrub

4. P. vitifolium

21 Pedicels more than 3 mm long, posterior petals much larger than anterior ones, pseudo-umbels

not capitulum-like:

23 Young stems reddish-brown, pedicels shorter than hypanthiums, pseudo-umbels less than

10-flowered, flowering branches not profusely branched 15. P. sublignosum

23 Young stems green, pedicels longer than hypanthiums, pseudo-umbels usually more than

10-flowered, flowering branches profusely branched:

24 Laminae villous to tomentose, petals white, posterior petals with eared bases and less than

10 mm long 24. P. tomentosum

24 Laminae pubescent, sparsely villous or hirsute, petals pale pink to carmine, posterior

petals without eared bases and more than 10 mm long:

25 Laminae palmatilobate, pubescent to hirsute, posterior petals without a white

blotch 22. P. hispidum

25 Laminae shallowly lobed, sparsely villous, posterior petals with a white blotch ....

23. P. papilionaceum

1 Laminae palmatilobate to pinnatisect:

26 Margins of laminae segments revolute:

27 Laminae villous and soft to the touch, segments at least 3 mm wide 10. P. graveolens

27 Laminae hirsute and hard to the touch, segments less than 3 mm wide 11. P. radens

26 Margins of laminae segments or lobes not revolute:

348

Bothalia 15, 3 & 4 (1985)

28 Segments or lobes of laminae not incised:

29 Pedicels less than 3 mm long, posterior petals slightly larger than anterior ones:

30 Laminae glabrescent and very viscid 7. P. glutinosum

30 Laminae hairy, not viscid or somewhat viscid:

31 Laminae panduriform to cordiform, somewhat viscid, peduncles articulated, pseudo-umbels

not capitulum-like, posterior petals at least 20 mm long 5. P. panduriforme

31 Laminae cordiform, not viscid, peduncles not articulated, pseudo-umbels capitulum-like,

posterior petals less than 20 mm long:

32 Laminae crisped, villous to densely villous, pseudo-umbels 8-20-flowered, decumbent

shrub 3. P. capitatum

32 Laminae not crisped, pilose to villous, pseudo-umbels 3-12-flowered, erect shrub

4. P. vitifolium

29 Pedicels more than 3 mm long, posterior petals much larger than anterior ones:

33 Laminae villous to tomentose and soft to the touch, bases of posterior petals eared, pedicels

more than 18 mm long 24. P. tomentosum

33 Laminae pubescent, hirtellous, hirsute, hispid or villous but hard to the touch, bases of posterior

petals not eared, pedicels less than 18 mm long:

34 Laminae more or less crisped and reniform 17. P. englerianum

34 Laminae not crisped and not reniform:

35 Young stems villous and reddish-brown 15. P. sublignosum

35 Young stems hirtellous or hirsute and green:

36 Laminae palmatifid to palmatisect, hirsute to hispid, flowers pinkish-purple

14. P. citronellum

36 Laminae palmatilobate to palmatipartite, hirtellous, flowers white to light pink:

37 Petioles much longer than laminae 18. P. greytonense

37 Petioles shorter or as long as laminae 13. P. ribifolium

28 Segments or lobes of laminae incised:

38 Laminae glabrescent, sparsely strigose or strigose, pedicels usually shorter than hypanthia,

flowering branches not profusely branched:

39 Leaf segments pinnatifid to pinnatisect:

40 Leaf segments grooved adaxially and sparsely strigose 8. P. denticulatum

40 Leaf segments not grooved adaxially and strigose with a few soft hairs inbetween

6. P. quercifolium

39 Leaf segments irregularly incised:

41 Laminae viscid, posterior petals slightly larger than anterior ones, pedicels less than 3 mm long:

42 Larger veins of laminae often dark purple, laminae glabrescent to sparsely strigose,

non-aromatic, pseudo-umbels less than 3-flowered 9. P. pseudoglutinosum

42 Larger veins of laminae never dark purple, laminae glabrescent, pseudo-umbels usually

more than 3-flowered 7. P. glutinosum

41 Laminae not viscid, posterior petals much larger than anterior ones, pedicels more than 3 mm long:

43 Young stems wine-red, laminae 3-palmatisect to pinnatisect and non-aromatic

16. P. scabroide

43 Young stems green, laminae 3-palmatilobate to palmatisect, non-aromatic or lemon-

scented 12. P. scabrum

38 Laminae pubescent, hirsute or hispid, pedicels longer than hypanthia, flowering branches pro-

fusely branched:

44 Laminae sparsely hirsute to hispid and strongly lemon-scented, pseudo-umbels usually less than

6-flowered, flowers pinkish-purple 14. P. citronellum

44 Laminae pubescent to hirsute and faintly lemon-scented, pseudo-umbels usually more than

6-flowered, flowers pale pink to pink 22. P. hispidum

1. Pelargonium eucullatum (L.) L'Herit. in Ait.,

Hort. Kew. edn 1,2: 426 (1789); Volschenk, J. J. A.

v.d. Walt & Vorster in Bothalia 14: 45 (1982). Type:

Africa, without precise locality, specimen in Hort.

Cliff. 345. 17 (BM, lecto.!).

Erect, much-branched, non-aromatic shrub, up to

2 m high and 1 m in diameter. Stems herbaceous

when young but soon becoming woody, sparsely to

densely pubescent to villous and interspersed with

glandular hairs, green but becoming brownish with

age. Leaves villous or strigose and always beset with

glandular hairs, green to greyish-green; lamina cir-

cular, reniform or triangular in outline, with or with-

out shallow angular incisions, conspicuously veined

abaxially, flat to hood-shaped at base, base reni-

form, cordate or cuneate, apex obtuse to acute, mar-

gin irregularly denticulate, often entire in proximal |

of lamina, (20-) 40-55 (-110) x (20-) 50-90 (-180)

mm; petiole (8-) 20-55 (-90) mm long; stipules

ovate to narrowly ovate, acute, densely pilose, 5-10

x 3-7 mm. Inflorescence: flowering branches some-

times profusely branched, with normal and smaller

foliar leaves; peduncles 20-70 mm long, pubescent

Bothalia 15, 3 & 4 (1985)

349

or pilose and always beset with glandular hairs; in-

volucral bracts ovate to broadly ovate, acute,

sparsely pilose to villous, 5-9 x 2-4 mm; pseudo-

umbels with (1-) 3-9 (-13) flowers each. Pedicel

2-11 mm long, villous. Hypanthium 5-12 mm long.

Sepals elliptic to narrowly elliptic, villous, green to

reddish-brown, 12-20 x 2-6 mm. Petals pale pink to

dark pinkish-purple, rarely white; posterior two spa-

thulate to obovate, apices obtuse to emarginate,

with dark purple streaks and a reddish-purple tinge

at the bases, widening to reddish-purple patches, re-

flexed at less than 90°, 15-32 X 6—17 mm; anterior

three spathulate to obovate to elliptic, with reddish-

purple markings, slightly reflexed, 15-28 x 6-12

mm. 2n=22. Figs 1, 2, & 3.

Diagnostic features

Erect, much-branched, non-aromatic shrub.

Lamina circular, reniform or triangular in outline,

with or without shallow angular incisions, villous or

strigose, flat to hood-shaped at base. Pseudo-umbels

with (1-) 3-6 (-13) flowers each. Flowers relatively

large, pale pink to dark pinkish-purple, pedicel usu-

ally as long as hypanthium.

Key to the subspecies

Margin of lamina more or less angularly incised:

Lamina villous; central vascular bundle in petiole with

a fibrous column (a) subsp. cucullatum

Lamina strigose; central vascular bundle in

petiole mostly without a fibrous

column (c) subsp. strigifolium

Margin of lamina not angularly incised, lamina

villous; central vascular bundle in petiole with a

fibrous column (b) subsp. tabulare

(a) subsp. cucullatum. Volschenk, J. J. A. v.d.

Walt & Vorster in Bothalia 14:47 (1982).

Geranium cucullatum L., Sp. PI. edn 1:677 (1753) p.p., emend.

Volschenk; Burm. f., Geran. 35 (1759) p.p.; L., Sp. PI. edn 2: 946

(1763) p.p., Berg., Descr. PI. Cap. 174 (1767) p.p.; Burm. f.,

Prodr. FI. Cap. 18 (1768) p.p.; Thunb., Prodr. 114 (1794) p.p. ; FI.

Cap. 518 (1823) p.p — var . fimbriatum Burm. f., Geran. 35 (1759);

Prodr. FI. Cap. 18 (1768). Pelargonium cucullatum (L.) L'Herit.

in Ait., Hort. Kew. edn 1,2: 426 (1789) p.p.; J. J. A. v.d. Walt,

Pelarg. S. Afr. 1:12, fig. (1977) p.p.; J. J. A. v.d. Walt & Vorster,

Pelarg. S. Afr. 2:43, fig. (1981). Geraniospermum cucullatum (L.)

Kuntze, Rev. Gen. 1:93 (1891).

Geranium angulosum Mill., Gard. Diet, edn 8, no. 22 (1768).

Pelargonium angulosum (Mill.) L'Herit. in Ait., Hort. Kew. edn

1,2: 426 (1789); Salisb., Prodr. 314 (1796); Willd., Sp. PI. 3: 671

(1800); Pers., Syn. PI. 2:231 (1807); Willd., Enum. 706 (1809);

Ait. f., Hort. Kew. edn 2,4: 174 (1812); DC. , Prodr. 1:672(1824);

G. Don., Gen. Syst. 1: 740 (1831); Eckl. & Zeyh., Enum. 1:80

(1835); Harv. in FI. Cap. 1: 302 (1860); Knuth in Pflanzenr.

4,129:460 (1912); Salter in Adamson & Salter, FI. Cape Penins.

518 (1950). Geraniospermum angulosum (Mill.) Kuntze, Rev.

Gen. 1:94 (1891).

Geranium acerifolium Cav., Diss. 4:243, t. 112, fig. 2 (1787);

Thunb., Prodr. 114 (1794); FI. Cap. edn 2: 520 (1823). Iconotype:

Cav., Diss. 4: t.122, fig. 2 (1787).

Diagnostic features

Lamina slightly to strongly hood-shaped, shal-

lowly and angularly incised in distal half, base obtu-

sely subcordately to cuneately incised, villous; pe-

tiole usually shorter than lamina, with a central fi-

brous column present in the central vascular bundle.

Pseudo-umbels with 5-7 flowers each. Petals usually

dark pinkish purple, occasionally white. Fig. 1.

FIG. 1. — Pelargonium cucullatum subsp. cucullatum. a, flow-

ering branch, x 1; b, petals, x 1; c, androecium, x 2; d,

gynoecium, x 2. (From Ward-Hilhorst 4 B, collected at Bet-

ty’s Bay.)

Subsp. cucullatum has a small, continuous distri-

bution from the vicinity of Gansbaai in the east to

Gordon’s Bay in the west, with a few isolated popu-

lations in the Cape Peninsula which are separated

from the rest by False Bay (Fig. 2). The leaves of the

plants on the peninsula are less villous than else-

where. The plants occur from near the high water

mark on the narrow coastal flats and in the lower

foothills of the mountains, but always practically

within sight of the sea.

FIG. 2. — Known geographical distribution of Pelargonium cu-

cullatum. ▲ subsp. cucullatum-, ■ subsp. tabulare # subsp.

strigifolium.

350

Bothalia 15, 3 & 4 (1985)

Subsp. cucullatum is a constituent of moist coastal

fynbos. It occurs under a rainfall regime of 400-800

mm per annum, on well-drained, sandy soils which

are mainly derived from Table Mountain Sandstone.

Flowering takes place from September to Feb-

ruary. The impression was gained that individual

plants of this subspecies have a longer flowering

period than subsp. tabulare and this may be ascribed

to the moister, perhaps more favourable environ-

mental conditions experienced by subsp. cucullatum.

CAPE PROVINCE. — 3318 (Cape Town): Table Mountain,

near Constantia (-CD), Ecklon 623 (W). 3418 (Simonstown):

Kalkbaai (-AB), Zeyher s.n. (SAM); Muizenberg (-AB),

Diimmer 512 (E), Lansdell s.n. (PRE), MacOwan 79 (BOL;

PRE; SAM; W). Marloth 35 (PRE), Penther 2171 (W), Scott- El-

liot 197 (E); Silver Mine Valley (-AB), Pillans 2708 (BOL); Sim-

onstown (-AB), Marloth 3615 (PRE; STE), Volschenk 22

(STEU); Gordon's Bay (-BB), Davis s.n. (SAM), Duthie 306

(STEU), Volschenk 40 (STEU); between Gordon’s Bay and Bet-

ty’s Bay (-BD), Dyer 5761 (PRE); between Gordon’s Bay and

Koelbaai (-BD), Ihlenfeldt 1673 (PRE); Betty’s Bay (-BD), Rv-

croft 2722, 2775, 3155 (NBG); Van der Schijff 7424 (PRE), Van

der Walt 433 (STEU), Volschenk 23, 49 (STEU), Werdermann

& Oberdieck 711 (PRE); Hangklip (-BD), Rodin 3126 (BOL),

Taylor 5866 (NBG); Harold Porter Reserve (-BD), Ebersohn

62168 (NBG), Van der Walt 649 (STEU); Koelbaai (-BD), Bou-

cher 473 (PRE). 3419 (Caledon): near Hermanus (-AC), Gillett

28 (STE), Jordaan 896 (STE), Rogers 26600 (PRE), Van der Walt

492 (STEU); Kleinmond (-AC), Pillans 8234 (BOL), Strey 2964

(PRE), Van Breda 1432 (PRE), Volschenk 1 , 2, 3, 4 (STEU);

Voelklip (-AC), Barker 1613 (NBG); Kleinrivier Mountains

(-AD), Ecklon 625 (W); Mosselrivier (-AD), Potts 5011 (SAM),

Verkiis s.n. (STE); near Stanford (-AD), Radloff s.n. (PRE); be-

tween Stanford and Hermanus (-AD), Gillett 4i75 (BOL), Marsh

872 (STE).

(b) subsp. tabulare Volschenk in Bothalia 14: 49

(1982). Type: Africa, without precise locality, speci-

men in Hort. Cliff. 345. 17 °c (BM, holo!).

Geranium cucullatum L., Sp. PI. edn 1: 677 (1753) p.p.; Burm.

f., Geran. 35 (1759) p.p.; L., Sp. PI. edn 2: 946 (1763) p.p.; Berg.,

Descr. PI. Cap. 174 (1767) p.p.; Burm. f.. Prodr. FI. Cap. 18

(1768) p.p.; Cav., Diss. 4:241, t. 106, fig. 1 (1787); Thunb.,

Prodr. 114 (1794) p.p.; FI. Cap. 518 (1823) p.p.. Pelargonium cu-

cullatum (L.) L’Herit. in Ait., Hort, Kew. edn 1,2: 426 (1789)

p.p.; Willd. , Sp. PI. 3: 670 (1800); Enum. 706 (1809); Pers., Syn.

PI. 2: 231 (1807); Ait. f., Hort. Kew. edn 2,4: 174 (1812); DC.,

Prodr. 1: 671 (1824); Spreng., Syst. Veg. 3: 58 (1826); G. Don,

Gen. Syst. 1: 740 (1831); Eckl. & Zeyh., Enum. 1: 80 (1835);

Harv. in FI. Cap. 1: 302 (1860); Marloth. FI. Kapl. 116, fig. 30

(1908); Knuth in Pflanzenr. 4,129: 466 (1912); Marloth, FI. S.

Afr. 2: 90 (1925); Salter in Adamson & Salter, FI. Cape Penins.

518 (1950); Mason, Western Cape Sandveld Flow. 134, t. 56

(1972); J. J. A. v.d. Walt, Pelarg. S. Afr. 1: 12, fig. (1977) p.p.

(pro icon.).

Diagnostic features

Lamina hood-shaped, circular in outline, base re-

niformly to cordately incised, villous; petiole equal-

ling or exceeding lamina in length, with a central fi-

brous column present in the central vascular bundle.

Pseudo-umbels with (5-) 6-9 (-13) flowers each. Pe-

tals usually dark pinkish purple. Fig. 3.

Subsp. tabulare occurs in two relatively restricted

areas separated from each other by a distance of

about 90 km, namely on the Cape Peninsula where it

ranges from Lion’s Head and Table Mountain south-

wards to Cape Point and in the vicinity of Saldanha

Bay (Fig. 2). On the Cape Peninsula it occurs as a

component of fynbos, from sea level to about 500 m

against the mountain slopes, but always practically

within sight of the sea. It grows on well-drained, usu-

FIG. 3. — Pelargonium cucullatum subsp. tabulare. a, flowering

branch, x 1; b, petals, x 1; c, androecium, x 1,5; d, gynoe-

cium, x 2; e, pedicel, hypanthium and sepals, x 1. (From

Ward-Hilhorst 35, collected at Scarborough.)

ally stony and often sandy soils derived from Table

Mountain Sandstone, shale and tillite. It receives an

annual rainfall of 400-1 000 mm on the Cape Penin-

sula. In the Saldanha Bay area it occurs in fynbos

vegetation in sheltered ravines, under conditions

somewhat reminiscent of those on the Cape Penin-

sula, but receiving only about 200 mm of rain per

annum.

It should be noted that the Saldanha Bay plants

are smaller than those on the Cape Peninsula, espec-

ially in respect of the leaves and flowers, but this ap-

pears to be a result of less favourable environmental

conditions and is not of taxonomic significance.

The flowering period extends from late September

to February , with a peak in October and November.

CAPE PROVINCE. — 3318 (Cape Town): Langebaan (-AA),

Axelson 3576 (NBG), Taylor 3773 (NBG); Postberg Nature Re-

serve (-AA), Volschenk 28, 29 (STEU); peninsula west of Sal-

danha Bay (-AA), Pillans 6883 (BOL), Salter 3925 (BOL);

Camp’s Bay (-CD), Ecklon & Zeyher 4827 (SAM), Marloth 9356

(PRE), Volschenk 9,10 (STEU); Cape Town (-CD), Pegler s.n.

(PRE), Young s.n. (PRE); Devil’s Peak (-CD), Thompson 31

(PRE), Wolley-Dod 2469 (BOL); Kirstenbosch (-CD), Grant

2639 (PRE); north of Window Stream (-CD), Esterhuysen 501

(BOL; PRE); Lion’s Head (-CD), Van der Walt 470 (STEU);

Sea Point (-CD), Willms 3074 (Z); Signal Hill (-CD), Volschenk

6, 7, 8 (STEU); Table Mountain (-CD), Bolus 2729 (BOL),

Diimmer 332 (E), Ecklon 608 (W), Ecklon & Zeyher 621 (SAM;

W), Flanagan 2415 (PRE), Guthrie 2394 (NBG; PRE), MacOwan

6472 (SAM), Rodin 3221 (BOL; PRE), Schenk 551 (Z), Thode

A5 (PRE), Werdermann & Oberdieck 65 (PRE). 3418 (Simons-

town): farm Bergvliet (-AB), Purcell 237 (SAM); s.n. (BOL);

south of Houtbaai (-AB), Smuts s.n. (STE); Muizenberg (-AB),

Pillans s.n. (PRE); Oceanview near Kommetjie (-AB), Van der

Walt 664 (STEU); Simonstown, Red Hill (-AB), Leighton 3062

Bothalia 15, 3 & 4 (1985)

351

(BOL), Pillans 2194 (BOL), Rogers 11282 (PRE); Wynberg Hill

(-AB), Salter 6393 B (BOL), Verdoorn s.n. (PRE); Cape of Good

Hope Nature Reserve (-AD), Dahlstrand 1080 (PRE); Cape

Point (-AD), Herre s.n. (STE); Buffelsbaai (-AD), Gillett 768

(STE); Olifantsbos (-AD), Leighton 3055, 3056, 3057 (BOL),

Rycroft 2189 (NBG); near Vasco da Gama Hill (-AD), Hutchin-

son 652 (BOL; PRE),

(c) subsp. strigifolium Volschenk in Bothalia 14:

50 (1982). Type: as for P. acerifolium L’Herit. be-

low.

Pelargonium acerifolium L’Herit. in Ait., Hort. Kew. edn 1,2:

427 (1789); Geran. t. 21 (1792); Salisb., Prodr. 315 (1796); Pers.,

Syn. PI. 2: 231 (1807); Willd., Enum. 706 (1809); Ait. f., Hort.

Kew. edn 2,4: 174 (1812); DC., Prodr. 1: 672 (1824); Spreng. ,

Syst. Veg. 3; 61 (1826); Eckl. & Zeyh., Enum. 1: 80 (1835); Knuth

in Pflanzenreich 4, 129:461 (1912); non Geranium acerifolium

Cav. (1787), Pelargonium angulosum (Mill.) L'Herit. var. acerifo-

lium (L'Herit.) Harv. in FI. Cap. 1: 303 (1860). Iconotype:

L’Herit., Geran. t. 21 (1792).

Diagnostic features

Lamina flat to somewhat hood-shaped, shallowly

and angularly incised in distal half, base cuneately

incised, strigose and with glandular hairs longer than

ordinary hairs; petiole usually shorter than lamina,

with a central fibrous column absent (rarely present)

in the central vascular bundle. Pseudo-umbels with

3-5 flowers each. Petals usually light pink to pinkish

purple. Fig. 4.

Subsp. strigifolium has a montane distribution in

the south-western Cape Province, being known from

the mountains around Bainskloof in the north to

FIG. 4. — Pelargonium cucullatum subsp. strigifolium. a, flow-

ering branch, x 1; b, petals, x 1; c, androecium, x 2; d,

gynoecium, x2; e, pedicel, hypanthium, sepals and stamens,

x 1. (From Van der Walt 463, cultivated in Stellenbosch.)

Baardskeerdersbos in the south, and the southern

Hottentotsholland Mountains in the west to the

Kleinrivier Mountains near Caledon in the east, at

altitudes of 300-900 m (Fig. 2). In contrast to the

subspecies cucullatum and tabulare, this subspecies

does not occur in close proximity to the sea.

Like the other two subspecies, it is a constituent of

fynbos. Few precise rainfall figures are available for

its montane habitats, but the rainfall over the

general area varies from about 600 to 1 000 mm per

annum. It grows on soils derived from sandstone,

shale and tillite, and in the vicinity of Paarl and

Jonkershoek also on soils derived from granite and

granofir. These soils can be quite heavy.

Plants from the Swartberg near Caledon are

smaller than elsewhere, probably occurring under

less than optimal conditions at the eastern border of

the area which is climatically suitable for this subspe-

cies.

Flowering takes place from September to January,

with a peak in October and November.

CAPE PROVINCE. — 3318 (Cape Town): Paarl Mountain

(-DB), Marais s.n. (STEU), Marloth 3480 (PRE), Van der Walt

652 (STEU), Volschenk 12, 33 (STEU); Wellington (-DB), Kno-

bel s.n. (PRE); Jonkershoek (-DD), B or char dt 301 (STE), Ma-

rais s.n. (STEU), Smith 306b (STEU), Van der Walt 428, 506, 641

(STEU), Volschenk 5 (STEU); Jonkershoek, Assegaaiboskloof

(-DD), Van der Walt 517 (STEU); Jakkalsvlei (-DD), Taylor

5153, 6956 (PRE); Swartboskloof (-DD), Van der Merwe 2126

(STE), Van der Walt 418 (STEU), Van Rensburg 2074 (PRE),

Walgate 987 (STE); Stellenbosch (-DD), Wawra 123 (W); Stel-

lenbosch, Botmaskop (-DD), Van der Walt 537 (STEU), Van

Rensburg 371, 374 (STE); near Paradise (-DD), Herre s.n.

(STE). 3319 (Worcester): Bainskloof (-CA), Gillett 788 (BOL),

Kies 43 (NBG); Du Toitskloof, western entrance (-CC), Pillans

8461 (BOL); Franschhoek (-CC), Boucher 2353 (PRE; STE),

Phillips 1058 (SAM), R.G. 5136 (PRE), Thode A2187 (PRE),

Volschenk 35 (STEU); Franschhoek Forest Reserve (-CC), Ester-

huysen s.n. (BOL); Klein Drakenstein Mountains, Kasteelkloof

(-CC), Kruger 1465 (STE); mountains between Franschhoek and

Villiersdorp (-CC), Bolus 5136 (BOL); Wemmershoek Moun-

tains (-CC), Drijfhout 463 (STEU), Esterhuysen 4044 (PRE);

17718 (BOL; PRE). 3418 (Simonstown): mountains south of Gor-

don’s Bay (-BB), Marloth 10119 (PRE), Smith 306a (STEU);

Hottentotshollandkloof (-BB), Ecklon & Zeyher 627 (SAM; W);

Koelberg Forest Reserve (-BB), Rycroft 1438 (BOL; NBG;

PRE); Sir Lowry’s Pass (-BB), Volschenk 39 (STEU); Steenbras

Dam (-BB), De la Bat s.n. (STE; STEU), Salter 6518 (BOL);

Arieskraal near Palmiet River (-BD), Barker 3335 (NBG),

Leighton 784 (BOL); between Betty’s Bay and Cape Hangklip

(-BD), lhlenfeldt 1718 (PRE); Kleinmond (-BD), Compton

12370 (NBG). 3419 (Caledon): between Villiersdorp and Elgin

(-AA), Bayliss 707 (PRE); Elgin (-AA), Van Breda 563 (PRE);

Grabouw (-AA), Bolus 4112b (BOL); 4112c (BOL); Lebanon

River (-AA), Van der Zel Z19; Z65 (PRE); Houhoek Pass, 5 km

west of Botrivier (-AA), Barker 8804 (NBG), Volschenk 38

(STEU); Palmiet River (-AA), Penther 2153 (W); Caledon

(-AB), Penther 2136 (W); Swartberg (-AB), Galpin 3820 (PRE),

Volschenk 36 (STEU), Zeyher s.n. (PRE; SAM); Paardeberg at

Palmiet River Mouth (-AC), Grobler 29286 (PRE; STE); Cale-

don, Hartebees River (-BC), Elbrecht 22136 (PRE); near Strand-

kloof (-CB), Van der Walt 603 (STEU); Baardskeerdersbos, on

the hills (-DA), Volschenk 44 (STEU).

2. Pelargonium betulinum (L.) L'Herit. in Ait.,

Hort. Kew. edn 1,2: 429 (1789); Curtis in Curtis’s

bot. Mag. 5: t. 148 (1792); Willd., Sp. PI. 3: 665

(1800); Pers., Syn. PI. 2: 230 (1806); Ait. f., Hort.

Kew. edn 2,4: 170 (1812); DC., Prodr. 1: 699 (1824);

Spreng., Syst. Veg. 3: 57 (1826); Eckl. & Zeyh.,

Enum" 1: 78 (1835); Harv. in FI. Cap. 1: 301 (1860);

Szyszyl., Polypet. Thalarn. Rehm.: 14 (1887); Knuth

in Pflanzenr.' 4,129: 457 (1912); Adamson & Salter,

352

Bothalia 15, 3 & 4 (1985)

FI. Cape Penins.: 519 (1950); Levyns, Guide Flow.

Cape Penins.: 173, fig. 84 (1866); Kidd, Wild Flow.

Cape Penins. edn 2: t. 62 (1973); J.J.A. v.d. Walt,

Pelarg. S. Afr. 1; 49, fig. (1977). Lectotype; ‘Habitat

in Africa’, LINN 858.9!

Geranium betulinum L.. Sp. PI. edn 1: 679 (1753); Burm. f.,

Geran.: 33 (1759); L., Sp. PI. edn 2,2: 946 (1763); Berg., Descr.

PI. Cap.; 175 (1767); Burm. f.. Prodr. FI. Cap.: 18 (1768); Mill.,

Gard. Diet, edn 8, no. 43 (1768); Murray, Syst. Veg. edn 14; 613

(1784); Cav.. Diss. 4: 238 (1787); Thunb.,' ProdC: 113 (1794);

Thunb., FI. Cap.: 516 (1823).

Pelargonium betulaefolium Salisb., Prodr.: 316 (1796), nom.

superfl.

P. penicillaium Willd . , Hort. Berol.: 37 (1805). Type: 'Habitat

ad Caput, bonae spei'. Willd. Herb. B12451 (B. holo.!).

P georgense Knuth in Reprium nov. Spec. Regni veg. 28: 92

(1930). Type: Cape Province, Ruigtevlei near Swart River,

George, Fourcade 1542 (BOL. holo.!; PRE!; STE!).

Erect to decumbent, branched to much-branched,

non-aromatic shrub or subshrub, up to 1,5 m high

and 1,5 m in diameter. Stems herbaceous when

young but soon becoming woody, sparsely to

densely pubescent and with a few glandular hairs in

between, green but becoming brownish with age.

Leaves glabrescent to hirtellous and with a few scat-

tered glandular hairs, green; lamina circular to

broadly elliptic to ovate, flat, veins depressed, base

shallowly cordate to cuneate, apex obtuse, margin

usually rather coarsely dentate to serrate with the

teeth often red-tipped, (8-) 20 (-40) x (5-) 15 (-30)

mm; petiole (4-) 10 (-20) mm long; stipules triangu-

lar, usually acuminate, pubescent and with a few

glandular hairs in between, 3-7 x 1^1 mm. Inflores-

cence: flowering branches with normal and smaller

foliar leaves; peduncles 15-90 mm long, pubescent

FIG. 5. — Pelargonium betulinum. a, flowering branches, X 1; b,

petals, x 1 ; c, androecium. x 1 ,5; d, gynoecium, x 3. (From

Ward-Hilhorst 5, collected at Leeukoppie.)

to villous and with glandular hairs interspersed; in-

volucral bracts lanceolate, apiculate, indumentum as

on stipules, 6-8 x 2-3 mm; pseudo-umbels with 2-6

flowers each. Pedicel 4-15 mm long, pubescent to

pilose and with glandular hairs interspersed. Hypan-

thium 3-8 mm long, indumentum as on pedicel. Se-

pals lanceolate, indumentum abaxially as on pedicel,

green to reddish-brown with a white margin, ca 12 x

2-4 mm. Petals pale pink to purplish, exceptionally

almost white; posterior two broadly spathulate to

obovate, with feather-like reddish-purple markings,

re flexed at less than 90°, ca 25 x 15 mm; anterior

three spathulate with rather broad claws, slightly re-

flexed, ca 22 x 7 mm. 2n=22. Fig. 5.

Diagnostic features

Erect to decumbent, branched to much-branched,

non-aromatic shrub or subshrub. Lamina circular to

broadly elliptic to ovate, flat, veins depressed, mar-

gin rather coarsely dentate to serrate. Pseudo-um-

bels with 2-6 relatively large, pink to purplish flow-

ers each, pedicel usually longer than hypanthium.

P. betulinum occurs from the vicinity of Yzerfon-

tein in the south-western Cape along the coast to

Knysna in the southern Cape (Fig. 6). The plants

usually grow on sandy dunes or flats, mostly in close

proximity to the coast. Its distribution area falls en-

tirely in the winter rainfall region. Temperatures are

high during summer and the winters are practically

frost free.

P. betulinum flowers from August to January with

a peak during September to November. Odd flowers

are found throughout the year.

There is a resemblance between the flowers of P.

betulinum and P. cucullatum (L.) L’Herit., but their

leaves differ markedly, those of the latter species be-

ing larger, hairy and more or less hood-shaped. A

natural hybrid between these two species has been

recorded (Van der Walt & Vorster, Pelarg. S. Afr.

2:45 (1981). Superficially, the flowers and leaves of

P. betulinum resemble those of P. elegans (Andr.)

Willd. of the section Campylia , but the habit of these

two species differs completely.

Whitish flowers of P. betulinum have been re-

corded from plants occurring in Llandudno on the

Cape Peninsula, and also from those growing on the

hard, lime-rich dunes of De Hoop Nature Reserve

near Bredasdorp. Plants collected in the Mossel

Bay-Knysna areas, show a much coarser leaf margin

than those from the south-western Cape.

CAPE PROVINCE. — 3318 (Cape Town): Slangkop near

Yzerfontein (-AD). Leighton 3058 (BOL), Sidey 2148 (MO);

Rondebosch (-CD), Wolley Dod 56, 57 (BOL); Kirstenbosch

(-CD), Ecklon & Zeyher 1848 (MO); near Cape Town (-CD),

Laubner 191 (Z), Young 26418 (PRE). Zeyher 14586 (SAM);

Blackheath (-DC), Bayliss 4367, 2428 (NBG), Van der Walt 438

(STEU); Vygekraal (-DC), Pillans 2784 (BOL); Durbanville

(-DC), Barker 1652 (NBG). 3321 (Ladismith): Huisrivier Pass

(-CB), Bayliss 4367 (MO). 3322 (Oudtshoorn): Wilderness

(-DC), Martin 130 (NBG); Ruigtevlei near George (-DD), Four-

cade 1542 (BOL; K; PRE). 3418 (Simonstown): Llandudno

(-AB), Van der Walt 486 (STEU); Klein Leeukop (-AB), Vol-

schenk 18 (STEU); Retreat (-AB), Dtimmer 379, 436 (E); Wyn-

berg (-AB). Bolus 513 (BOL; PRE). 3262 (BOL; SAM), Mac-

Owun 513 (BOL; PRE), 2703 (MEL; SAM); Wynberg Hill

(-AB), Marloth 202 (PRE); Hout Bay (-AB), Bond 443 (NBG),

Bothalia 15, 3 & 4 (1985)

353

FIG. 6. — Known geographical distribution of Pelargonium betu-

linum.

Fairall 180 (NBG), Prior s.n. (Z); Muizenberg (-AB), Brain 5865

(SRGH); Chapman’s Bay (-AB), Salter 7812 (BOL); Kalk Bay

(-AB). Scot! Elliot 1157 (E); Karbonkelberg (-AB), Barker 1669

(NBG), Esterhuysen 26245 (BOL); Kommetjie (-AB), Kies 139

(NBG); Bergvliet Farm (-AB), Purcell 239 (SAM); Fish Hoek

(-AB), Young 263 (PRE); Olifantsbos (-AD), White 5691

(PRE); between Eersterivier and Swartklip (-BA). Pillans 9210,

9248 (BOL); Cape Flats (-BA), Marloth 5322, 6005 (PRE), Reh-

mann 2170 (Z). Thompson 1883 (PRE). Werdermann & Ober-

dieck 231 (PRE); Strand (-BB), Parker 3922 (BOL; NBG), Strey

720 (PRE); Betty’s Bay (-BD), Ebersohn s.n. (NBG), Taylor

4765 (NBG), Topper 126 (NBG), Low der Walt 435 (STEU), Vol-

schenk 25, 51 (STEU); Hangklip (-BD), Boucher 521 (PRE),

Lavranos 15157 A (PRE), Taylor 5844 (NBG); Kogel Bay (-BD),

Werdermann & Oberdieck 284 (PRE). 3419 (Caledon): near Stan-

ford (-AD), Gillett 4403 (BOL; PRE). Van der Walt 601 (STEU),

Volschenk 48 (STEU); Kleinrivier Mountains (-AD), Ecklon &

Zeyher 30280 (SAM); near Uilenkraalmond (-BC), Volschenk 45

(STEU); Die Kelders (-CB), Taylor 6914 (PRE), Volschenk 47

(STEU); Van Dyksbaai (-CB), Volschenk 46 (STEU); Gansbaai

(-CB), Rabinowitz s.n. (SAM), Schonken 158 (STEU); Strand-

kloof (-CB). Compton 18215 (NBG), Leighton 1930 (BOL), Van

der Walt 607, 1048 (STEU); Baardskeerdersbos (-DA). Hugo 625

(PRE); Heuningrug (-DB), Fischer 281 (STEU); Mierkraal

(-DB), Fischer 274 (STEU); Poort near Bredasdorp (-DB), Bar-

ker 2529 (NBG), Compton 9023, 23198 (NBG), Esterhuysen 2998

(PRE), Leighton 21083 (BOL). 3420 (Bredasdorp): Kathoek near

Bredasdorp (-AD), Pillans 9378 (BOL); De Hoop Nature Re-

serve (-AD), Van der Walt 1330 (STEU); near Potberg (-BC),

Esterhuysen 23243 (BOL); Bredasdorp Hills (-CA), Bond 455

(NBG); Farm Kleiheuwel (-CA), Albertyn s.n. (STEU); Cape

Agulhas (-CC). Fischer 289 (STEU), Pillans 8165 (BOL), Venter

7452 (STEU). 3421 (Riversdale): Corente River (-AA), Muir

5061 (PRE); Melkhoutfontein (-AB), Galpin 3822 (PRE); Re-

nosterkop (-AB), Smith 4953 (PRE); Riversdale (-AB). Muir

1069 (BOL; PRE); S of Riversdale (-AC), Oliver 5993 (PRE);

Stilbaai (-AD), Johnson 109 (NBG), Morris 262 (NBG); between

Riversdale and Stilbaai (-AD), Goldblatt 4146 (MO), Wurts 1563

(NBG); near Albertinia (-BA), Bayliss 3658 (NBG; SRGH; Z);

between Gouritsmond and Stilbaai (-BD), Rycroft 3121 (NBG).

3422 (Mossel Bay): Fransmanshoek (-AA), Van der Walt 678

(STEU); Mossel Bay (-AA), Morau s.n. (BOL), Rogers 4180

(GRA; NBG); Groot-Brakrivier (-AA), Sidey 1730 (MO; PRE);

5 km W of Groot-Brakrivier (-AA), Thompson 556 (PRE); Gou-

kamma (-BB). Heinecken 272 (PRE). 3423 (Knysna): Sedgefield

(-AA), Middlemost 2051 (NBG); Knysna (-AA), Keet 847

(GRA; PRE), Leighton 3334 (BOL), Marloth 7533 (PRE).

3. Pelargonium capitatum (L.) L'Herit. in Ait.,

Hort. Kew. edn 1,2: 425 (1789); Salisb., Prodr.: 314

(1796); Willd., Sp. PI. 3: 676 (1800); Pers., Syn. PI.

2: 232 (1806); Willd., Enum.: 707 (1809); Ait. f.,

Hort. Kew. edn 2,4: 176 (1812); DC., Prodr. 1: 676

(1824); Spreng., Syst. Veg. 3: 61 (1826); Eckl. &

Zeyh., Enum. 1: 80 (1835); Harv. in FI. Cap. 1: 303

(1860); Hanks & Small, N. Amer. FI. 25: 24 (1907);

Knuth in Pflanzenr. 4,129: 467 (1912); Phillips in

Rep. S. Afr. Ass. Advmt. Sci.: 456 (1918); Adam-

son & Salter, FI. Cape Penins.: 519 (1950); Batten &

Bokelmann, Wild Flow. E. Cape: 90, t. 75: 5 (1966);

Courtenay-Latimer & Smith, Flow. PI. Tsitsi-

kamma: t. 39 (1967); Mason, West. Cape Sand.

Flow.: 134, t. 56,5 (1972); J.J.A. v.d. Walt, Pelarg.

S. Afr. 1: 7, fig. (1977); Rourke, Fairall & Snyman

in J1 S. Afr. Bot. 47: 562 (1981). Lectotype: ‘Habitat

in Africa’, LINN 858.17!

Geranium capitatum L., Sp. PI. edn 1,2: 678 (1753); Burm. f.,

Geran.: 35 (1759); L., Sp. PI. edn 2,2; 947 (1763); Burm. f..

Prodr. FI. Cap.: 18 (1768); Mill., Card. Diet, edn 8, no. 25

(1768); L., Mant.: 432 (1771); Murray, Syst. Veg. 13; 511 (1774);

Cav., Diss. 4: 249, t. 105, fig. 1 (1787); Murray, Syst. Veg. 15: 652

(1797); Thunb.. Prodr. 2:^114 (1800); Andr.. Geran7l: C. ic

(1805); Desf., Arb. 1: 461 (1809). Geraniospermum capitatum

(L.) Kuntze, Rev. Gen. 1: 94 (1891).

Pelargonium drummondii Hook. f. in Curtis’s bot. Mag. 120: t.

7346 (1894) ex descr. & icon., non Turcz. in Bull. Soc. Nat. Mos-

cow 31: 421 (1858).

Decumbent, much-branched, rose-scented sub-

shrub, up to lm high and 1,6m in diameter. Stems

herbaceous, becoming somewhat woody with age,

pilose to densely villous and with glandular hairs in-

terspersed, green but becoming brownish with age.

Leaves villous to densely villous and densely inter-

spersed with glandular hairs, green to greyish-green;

lamina 3-5-palmatilobate to 3-5-palmatipartite,

crisped, base cordate, lobes sometimes shallowly

FIG. 7. — Pelargonium capitatum. a, flowering branch, x 1; b.

petals, x 1; c, androecium, x 2; d, gynoecium, x 2; e, pedi-

cel, hypanthium and sepals, x 1. (From Ward-Hilhorst 3; col-

lected at Muizenberg.)

354

Bothalia 15, 3 & 4 (1985)

lobed. apices of lobes obtuse, margins irregularly

crenate-dentate, (20-) 40 (-115) x (30-) 50 (-160)

mm; petiole (5-) 20 (-55) mm long; stipules broadly

ovate to cordiform, often apiculate, 7-12 x 5—14

mm. Inflorescence: flowering branches with normal

and smaller foliar leaves; peduncles 30-150 mm

long, villous to densely villous and densely inter-

spersed with glandular hairs; involucral bracts

broadly lanceolate, apiculate to cuspidate, indumen-

tum as on leaves, 8-10 x 3-5 mm; capitulum-like

pseudo-umbels with 8-20 flowers each. Pedicel 0,5-2

mm long. Hypanthium 3-8 mm long, densely villous

and with glandular hairs interspersed. Sepals lanceo-

late, apiculate, indumentum as on hypanthium,

green but apices often reddish, ca 10 x 3-4 mm. Pe-

tals pale pink to dark pink-purple; posterior two spa-

thulate, apices obtuse, with dark purple markings,

reflexed at ca 90°, ca 18 x 6 mm; anterior three spa-

thulate with narrow claws, slightly reflexed, ca 16 x

6 mm. 2n=66. Fig. 7.

Diagnostic features

Decumbent, much-branched, rose-scented sub-

shrub. Lamina 3-5-palmatilobate to 3-5-palmatipar-

tite, crisped, villous to densely villous. Pseudo-um-

bels capitulum-like, with 8-20 flowers each. Flowers

pale pink to dark pink-purple, pedicels much shorter

than hypanthium.

P. capitation occurs from Lambert’s Bay in the

west along the coast through the Transkei to Zulu-

land (Fig. 8). The plants usually grow on sandy

dunes or flats, mostly in close proximity to the coast.

In the interior, it is usually found in disturbed habi-

tats. Populations in the south-western part of its dis-

tribution area receive winter rains, those in the

southern Cape winter as well as summer rains,

whereas the eastern Cape, Transkei and Natal popu-

lations receive rain predominantly during the sum-

mer months. Temperatures are high to very high

during the summer and the winters frost free.

P. capitation has a definite flowering peak during

spring, especially in September and October, but

odd flowers may be found throughout the year.

P. capitation is closely related to P. vitifolium

(L.)L’Herit. and the differences between these two

species are tabulated under P. vitifolium. The villous

leaves of P. capitation can also be confused with

those of P. tomentosum Jacq., but they have a com-

pletely different odour, the leaves of P. capitation

being rose scented and those of P. tomentosum hav-

ing a peppermint scent. Furthermore, the inflores-

cences and flowers of these two species differ mark-

edly.

Two specimens of P. capitation, LINN 858.17 and

Hort. Cliff. 345.15 (BM), both with the epithet in

the handwriting of Linnaeus, are available as pos-

sible lectotypes. I decided to choose the specimen in

LINN as the lectotype, because Linnaeus’s first re-

ference in Species Plantarum (1753) is to Hort. Ups.

and not to Hort. Cliff.

NATAL. — 2831 (Nkandla): Nkandla Forest Reserve (-CA).

Moffett 3003 (STEU); Ngoye (-DC). Wood 9360 (BOL; MEL;

PRE; SAM). 2930 (Pietermaritzburg): Kranzkloof (-DD), Hay-

garth 22337 (PRE), Rogers 17177 (Z). 3030 (Port Shcpstone); St.

LTG. 8. — Known geographical distribution of Pelargonium capi-

tatum.

Michaels (-AB), Prosser 1458 (PRE); Umkomaas (-BB), Dyer

3122 (PRE), Forbes 1227 (NH), Theiler 10027 (PRE), Wood 7177

(E: NH); Pennington (-BC). Repton 1282 (PRE); Ifafa (-BC),

Huntley 216 (PRE); Clansthal, Umzinto (-BC). Gallway 102

(MO); Scottburgh (-BD), Crookes 3 (PRE); Uvongo (-CB),

Mogg 13198 (SRGH); Umtamvuna (-CC), Codd 10696 (PRE),

Strey 5842 (NH: PRE); Port Shepstone (-CD). Wood 9759

(NBG); Southbroom (-CD), Whellan 1105 (PRE; SRGH). 3130

(Port Edward): Port Edward (-AA), Strey 4344, 5842 (PRE);

Craigadour Farm near Port Edward (-AA), Nicholson 1605

(PRE).

TRANSKEI. — 3128 (Umtata): Mtentu (-CD). Strey 8647

(PRE; SRGH). 3129 (Port St. Johns): Lusikisiki Magna Waterfall

(-BC), Venter & Vorster 16 (PRE); Mateku Waterfall (-BC),

Strey 10167 (E; PRE); Mkambati Leper Hospital (-BD), Marais

1196 (PRE); Ndininindi (-CB), Venter 932 (PRE); Umtata River

Mouth (-CC), Acocks 13566 (PRE). 3130 (Port Edward): Llm-

tentu River Mouth (-AC), Marais 1196 (NBG; PRE). 3228 (But-

terworth): Mazeppa Bay (-BC). Theron 1201 (BOL; PRE); Ken-

tani Coast (-CB). Pegler 120 (BOL; SAM); Kobongaba (-DA).

Taylor 3538 (NBG).

CAPE PROVINCE. — 3218 (Clanwilliam): Lamberts Bay

(-AB). Hardy & Bayliss 1034 (PRE); 20 km from Elandsbaai on

Redelinghuys Road (-AD), Van der Walt 598 ( ST Ed); between

Darling and Ysterfontein (-BA), Rycroft 1818 (NBG); near

Paleisheuwel (-BC), Schonken 89 (STEU); Het Kruis (-DA),

Zinn s.n. (SAM); Piketberg (-DD), Schonken 88 (STEU). 3219

(Wuppertal): Theerivier, Citrusdal (-CA), Hanekom 1143 (PRE;

SRGH). 3228 (Butterworth): Kei Mouth (-CB). Flanagan 163

(BOL; PRE), Van der Walt 685 (STEU). 3318 (Cape Town):

Hopefield (-AB), Bachmann 1534 (Z), Letty 114 (PRE); near

Darling (-AD), Bolus 12629 (BOL), Winkler 115 (NBG); Melk-

bcsstrand (-CB), Dahlstrand 1055 (PRE); Robben Island (-CD),

Walgate 463 (NBG); between Blouberg and Melkbos (-CD), Ax-

elson 357 (NBG); Devils Peak (-CD), Thompson s.n. (PRE);

Lions Head (-CD), Van der Walt 469 (STEU); near Bolus Her-

barium (-CD), Leighton 3206, 3207, 3208 (BOL); Kirstenbosch

(-CD). Barker 225 (BOL), Zeyher 32 (SAM); Campsbay (-CD),

Elbrecht 22135 (PRE), Hutchinson 147 (BOL; PRE). Rehmann

1603 (Z); Kenilworth Race Course (-CD), Van der Walt 545

(STEU); Rondebosch (-CD). Diimmer 269 (E); Clifton (-CD),

Phillips s.n. (PRE); Cape Town (-CD), Bolus 2728 (PRE),

Schinz s.n. (Z); Dieprivier (-DA), Ecklon & Zeyher 7 (PRE);

Groot Springfontein (-DC), Wasserfall 248 (NBG); Joostenberg

(-DC), Barker 9603 (NBG): Blackheath (-DC), Van der Walt 443

(STEU); W of Koelenhof (-DD), Thompson 2648 (PRE); Klap-

muts (-DD), Ecklon & Zeyher 631 (MO; SAM); Lyndoch

(-DD), Taylor 5796 (NBG). 3319 (Worcester): 18 km from Wor-

cester on Bothashalt Road (-CB). Van Breda 554 (PRE); Bothas-

halt (-CB). Van Breda 912 (PRE; SRGH); near Franschhoek

(-CC), Van der Walt 453 (STEU), Volschenk 15 (STEU);

Wemmershoek Mountains (-CC), Van der Walt 465 (STEU);

Franschhoek (-CC), Phillips 1055 (SAM), Scltlechter 9220

(BOL), Schlechter 9253 (MO), Thode A2188 (PRE). 3322

(Oudtshoorn): Robinson Pass (-CC), Heginbotham 220 (NBG),

Bothalia 15, 3 & 4 (1985)

355

Van der Walt 1125 (STEU); Ruitersbos (-CC), Van Niekerk 39

(BOL); Montagu Pass (-CD), Van der Walt 423(a), 1138, 1307

(STEU); 10 km E of George (-CD), Schonken 186A (STEU);

Wilderness (-DC), Heginbotham 196 (NBG); near Rondevlei,

Sedgefield (-DC), Bayliss s.n. (PRE); Rheenendal near Knysna

(-DD), Neser s.n. (STEU); Ruigtevlei (-DD), Martin 4540

(PRE). 3323 (Willowmore): Prince Alfred’s Pass (-CC), Van der

Walt 721 (STEU); Nature’s Valley (-DC), Immelman 78 (PRE);

near Nature’s Valley (-DC), Von Teichman 355 (PRE). 3325

(Port Elizabeth): Van Stadens Pass (-CC), Theron 569 (PRE);

near Uitenhage (-CD), Bolus 757 (BOL). 3326 (Grahamstown):

Hoffman’s Bosch (-AC), Britten 1297 (PRE); Komgha (-AC),

Flanagan 1525 (SAM); Assegaaibos (-AD), Breyer 23593 (PRE);

Farm Rosslyn, Bloukrans (-BC), Bayliss 5328 (Z), Bayliss 5374

(MO; WAG); Grahamstown (-BC). MacOwan 677 (SAM), Rog-

ers 27417 (Z); Woest Hill (-BC), Bayliss 4501 (MO); between

Bathurst and Port Alfred (-BD), Barker 10501 (NBG);

Boesmansriviermond (-DA), Van der Walt 693 (STEU). 3327

(Peddie): East London (-BB), Batten s.n. (NBG), Galpin 5841

(PRE), Leifert 79 (Z), Nanni 135 (PRE). 3418 (Simonstown):

Kalk Bay (-AB), Bolus 2728 (BOL), Goldblatt 1405 (MO); Wyn-

berg (-AB), Schenck 610 (Z), Zeyher s.n. (SAM); Wynberg Hill

(-AB), Pillans 10163 (MO); near Wynberg (-AB), Goldblatt 2680

(MO); Llandudno (-AB), Van der Walt 490 (STEU), Willems 69

(NBG), Young 217 (PRE); Shelly Beach (-AB), Letty s.n (PRE);

Hout Bay (-AB), Galpin 3806 (PRE), Marloth 428 (PRE); Kom-

metjie (-AB), Young 266 (PRE); between Retreat and Muizen-

berg (-AB), Schlechter 644 (Z); Muizenberg (-AB), Pillans 2857

(BOL); Silvermine (-AB), Oberdieck 112 (PRE), Werdermann &

Oberdieck 116 (PRE); road to Constantia Neck (-AB), Guthrie

1000 (NBG); Klaasjagersberg (-AB), Sidey 2149 (MO): Bergvliet

(-AB), Purcell 236 (SAM); Simonstown (-AB), Brown s.n. (E),

Duruo s.n. (E), Marloth 3619 (PRE); between Simonstown and

Miller's Point (-AB), Phillips 379 (NBG); Cape Point Nature Re-

serve (-AD), Leighton 3054 (BOL), Ry croft 2192 (NBG); Isoetes

Vlei, Cape Flats (-BA), Mathews 27 (NBG), Rowe 26 (NBG);

Strandfontein (-BA), Van Zinderen-Bakker 23 (NBG), White

5107 (PRE); Macassar (-BA), Taylor 3185 (PRE), Werdermann

& Oberdieck 200 (PRE; WAG); Cape Flats (-BA), Phillips s.n.

(Z), Rehmann 2177 (Z), Rogers 27217 (Z); Faure (-BB), Van der

Walt 516 (STEU); Sir Lowry's Pass (-BB), Caplan s.n. (Z), Guth-

rie 2991 (NBG); Strand (-BB), Parker 4925 (BOL; NBG), Strey

742 (PRE); Groot Hangklip (-BD), Boucher 775 (PRE); Kogei-

baai (-BD), Boucher 474 (PRE; SRGH); Betty’s Bay (-BD),

Topper 147 (NBG), Van der Walt 432 (STEU). 3419 (Caledon):

Houwhoek Pass (-AA), Van der Walt 794 (STEU); Caledon

(-AB), Gillett 4484A (PRE); Onrusrivier (-AC), Van Niekerk

340 (BOL; PRE); Hermanus (-AC), Gillett 4483 (PRE); Ver-

mont (-AC), Gillett 4483 (BOL; PRE); Kleinmond (-AC), De

Vos 775 (PRE), Van der Merwe s.n. (STEU); near Brakfontein,

Riviersonderend (-BB), Fischer 306 (STEU); Uilenkraal (-BC),

Taylor 1555 (SAM); Oubos near Greyton (-BD), Van der Walt

709 (STEU); Baviaansfontein, Gansbaai (-CB), Stokoe 7400

(BOL); 16 km E of Gansbaai (-CB), Taylor 4903 (MO; NBG).

3420 (Bredasdorp): Ratelrivier (-AA), Van Breda 827 (PRE);

Stormsvlei (-AA), Fischer 290 (STEU); Farm Elandspoort

(-BC), Pillans 9411 (BOL); Cape Infanta (-BD), Blum s.n. (E);

between Elim and Bredasdorp (-CA), Marsh 925 (PRE); Farm

Kleiheuwel (-CA), Van der Walt 609 (STEU); Cape Agulhas

(-CC), Venter 7455 (STEU). 3421 (Riversdale): near Albertinia

(-BA), Muir 1802 (BOL; PRE). 3422 (Mossel Bay): Little Brak

River (-AA). Sidney 1731 (PRE); 19 km from Mossel Bay to

George (-AA), Acocks 15393 (PRE), Story 3102 (PRE); Rhee-

bok (-AA), Van der Walt 673 (STEU); 5 km from Great Brak

River (-AA), Marsh 567 (PRE); Great Brak River (-AA),

Thorne s.n. (SAM); near Mossel Bay (-AA), Lewis 3910 (SAM),

Rogers 4161 (NBG; Z); near Bottelierskop (-AA), Van der Walt

1130, 1134 (STEU); Goukamma (-BB), Schlieben & Ellis 12334

(PRE). 3423 (Knysna): Knysna (-AA), Bayliss 1378 (Z), Breyer

23974 (PRE), Galpin 3819 (PRE), Rehmann 463 (Z), Rogers

22321 (PRE), Rogers 27156 (Z); Buffalo Bay (-AA), Keet 846

(PRE); Robbeberg (-AA), Werdermann Oberdieck 962

(PRE); Plettenberg Bay (-AB), Ecklon & Zeyher s.n. (PRE),

Lavranos 12939 (PRE), Rogers 28396 (Z), Rogers & Smart 24656

(PRE), Theron 1780 (PRE); Keurboomsrivier (-AB), Taylor

6010 (NBG); Tsitsikamma National Park (-BB). Botha 126

(PRE), Liebenberg 7876 (PRE). 3424 (Humansdorp): Witelsbos

(-AA), Leighton 3070 (BOL); Clarkson (-AB), Thode A756

(PRE); near Kareedouw (-AB), Fourcade 4413 (PRE), Thode

A2590 (PRE); Klipdrift (-BA), Thode A2467 (PRE); Cape St.

Francis (-BB). Bayliss 6797 (MO; Z); 32 km from Humansdorp

to Cape St. Francis (-BB), Wells 2966 (PRE); 10 km NE of Jef-

frey’s Bay (-BB), Hutchinson 1465 (BOL). 3425 (Skoen-

makerskop): Skoenmakerskop (-BA), Van der Walt 564 (STEU).

4. Pelargonium vitifolium (L.) L’Herit. in Ait.,

Hort. Kew. edn 1,2: 425 (1789); Salisb., Prodr.: 314

(1796); Willd., Sp. PI. 3: 675 (1800); Pers., Syn. PI.

2: 232 (1806); Willd., Enum.: 707 (1809); Ait. f.,

Hort. Kew. edn 2,4: 176 (1812); DC., Prodr. 1: 674

(1824); Spreng., Syst. Veg. 3: 60 (1826); G. Don,

Gen. Syst. 1: 740 (1831); Harv. in FI. Cap. 1: 303

(1860); Knuth in Pflanzenr. 4,129: 468 (1912); Ad-

amson & Salter, FI. Cape Penins.: 519 (1950); J.J.A.

v.d. Walt, Pelarg. S. Afr. 1: 49, fig. (1977). Lecto-

type: Habitat in Africa, Hort. Ups., 858.15 (LINN).

Geranium vitifolium L., Sp. PI. edn 1,2: 678 (1753); Burm. f.,

Geran.: 34 (1759); L., Sp. PI. edn 2,2: 947 (1763); Burm. f.,

Prodr. FI. Cap.: 18 (1768); Mill., Gard. Diet, edn 8, no. 26

(1768); Cav., Diss. 4: 245, t. Ill, fig. 2 (1787); Thunb., Prodr. 2:

114 (1800); Thunb., FI. Cap. edn 2: 521 (1823).

G. ribisioides Burm. f., Geran.: 34 (1759). Type: ‘Habitat in

Cap. Bon. Spei.’, Burm. Herb. 3771/76 (G).

Erect, branched, strongly aromatic shrub, up to 1

m high and 0,75 m in diameter. Stems herbaceous

when young but soon becoming woody, villous and

interspersed with glandular hairs. Leaves pilose to

villous and densely interspersed with glandular

hairs, green; lamina cordiform in outline, shallowly

3 (-5)-lobed, base cordate, apices of lobes obtuse,

margins irregularly crenate, (25-) 45-60 (-87) x

(25-) 60-80 (-115) mm; petiole (25-) 50 (-145) mm

long; stipules broadly ovate to triangular or narrowly

triangular, often apiculate, 8-15 x 5-15 mm. Inflo-

rescence: flowering branches with normal and

smaller foliar leaves; peduncles 15-100 mm long, pi-

lose to villous and densely interspersed with glandu-

lar hairs; involucral bracts ovate to lanceolate, acute

to caudate, indumentum as on stipules, 5-10 x 2-5

mm; capitulum-like pseudo-umbels with 3-12 flow-

ers each. Pedicel 0,5-3 mm long, indumentum as on

stipules. Hypanthium 2-4 mm long, indumentum as

on stipules. Sepals lanceolate, abaxially pilose to hir-

sute and with glandular hairs interspersed, green

with a reddish-brown tint, ca 10 x 2,5-4 mm. Petals

pale pink to light pinkish-purple; posterior two spa-

thulate, apices obtuse to retuse, with dark purple

markings, reflexed at ca 90°, ca 18 x 7 mm; anterior

three spathulate with narrow claws, slightly reflexed,

ca 16 x 5 mm. 2n=88. Fig. 9.

Diagnostic features

Erect, branched, strongly aromatic shrub. Lamina

cordiform in outline, shallowly 3- (5)-lobed, pilose

to villous. Capitulum-like pseudo-umbels with 3-12

flowers each. Flowers pale pink to light pinkish-

purple, pedicel and hypanthium relatively short,

pedicel usually slightly shorter than hypanthium.

P. vitifolium occurs from the Cape Peninsula in

the south-western Cape eastwards along the coast to

Knysna in the southern Cape (Fig. 10). Its distribu-

tion area falls entirely in the winter rainfall region.

Although consistently associated with mountains, it

is confined to the lower slopes where it occurs in

shady ravines, usually near streams between bould-

ers or on scree. The soil in these situations is usually

sandy, often with a high percentage of organic mat-

356

Bothalia 15, 3 & 4 (1985)

FIG. 9. — Pelargonium vitifolium. a, flowering branch, x 1; b,

petals, x 2; c, androecium, x 2; d, gynoecium, x 2. (From

W ard-Hilhorst 88, collected at Helderberg Nature Reserve.)

ter. Temperatures are high during summer and the

winters are practically frost free.

P. vitifolium flowers from August to January with

a peak in October-November. Except for the mid-

winter months, odd flowers can be found throughout

the year.

P. vitifolium is closely related to P. capitatum (L.)

L’Herit. The following characters may be used to

distinguish between these two species:

P. vitifolium

1. Erect shrub with an

unpleasant scent.

2. Leaves pilose to

villous and

not crisped.

3. Pseudo-umbels

with 3-12

flowers

each.

P. capitatum

1 . Decumbent rose-

scented subshrub.

2. Leaves villous

to densely villous

and crisped.

3. Pseudo-umbels

with 8-20

flowers

each.

Two specimens of P. vitifolium , LINN 858. 15 and

Hort. Cliff. 345. 15 (BM), both with the epithet in

the handwriting of Linnaeus, are available as pos-

sible lectotypes. I decided to choose the specimen in

LINN as the lectotype, because Linnaeus referred to

Hort. Ups., and not to Hort. Cliff., in his Species

Plantarum (1753).

(-CD), Salter 6392 (BOL); Paarlberg (-DB), Drege s.n. (PRE);

Dal Josafat near Paarl (-DB), Tyson 940 (SAM); Jonkershoek

near Stellenbosch (-DD), Page 14201 (BOL), Van der Merwe

23-59 (SRGH), Van der Walt 515 (STEU), Van der Wall

Vorster 1324 (STEU); Swartboskloof, Jonkershoek (-DD), Van

Rensburg 2188 (PRE); Guardian Peak, Jonkershoek (-DD). Es-

terhuysen 24127 (BOL); Bothmaskop, Stellenbosch (-DD), Van

der Walt 523 (STEU); Banhoek near Stellenbosch (-DD), Mart-

ley s.n. (BOL), Thorne s.n. (SAM); Vergenoeg, Stellenbosch

(-DD), Van der Walt 499 (STEU). 3319 (Worcester); Du Toits-

kloof (-CA), Pillans 8492 (BOL); French Hoek (-CC), Schlechter

9220 (Z), 9253 (E; BOL; PRE; Z), 1574a (PRE). 3322

(Oudtshoorn): Robinson Pass near Mossel Bay (-CC), Esterhuy-

sen 19419 (BOL; PRE). 3418 (Simonstown); Wynberg Park

(-AB). Rogers s.n. (PRE); Wynberg Hill (-AB), Salter s.n.

(PRE), Wolley Dod 1977 (BOL); Helderberg, Somerset West

(-BB), Esterhuysen 14639 (BOL; MO); Helderberg Nature Re-

serve, Somerset West (-BB). Van der Walt & Vorster 1321

(STEU); Hottentot Holland (-BB), Zeyher 221 (MEL). 3419

(Caledon); Elim near Bredasdorp (-DA), Bolus 8527 (BOL), s.n.

(NBG). 3423 (Knysna): Portland near Knysna (-AA), Duthie

1161 (BOL; STE).

FIG. 10. — Known geographical distribution of Pelargonium viti-

folium.

5. Pelargonium panduriforme (sphalm. pandu-

raeforme) Eckl. & Zeyh., Enum. 1: 82 (1835);

Knuth in Pflanzenr. 4,129: 473 (1912). Type: Cape

Province, ‘. . . in lateribus montium “Winterhoeck-

berge” occidentem spectantibes in deserto “Karro”

dicto (Uitenhage)’, Ecklon & Zeyher 640 (S, lecto.!;

M!; MO!; OXF!; SAM!).

Geranium terebinthinaceum Murray in Comm. Soc. Gott. 7: 88,

t. 4 (1786). Iconotype: Murray in Comm. Soc. Gott. 7: 88, t. 4

(1786).

Geranium quercifolium sensu Cav., Diss. 4: 246, t. 119, fig. 1

(1787).

Pelargonium quercifolium sensu L'Herit. in Ait.. Hort. Kew.

edn 1,2: 422 (1789) excl. var. 6; sensu L'Herit., Geran. t. 14

(1792); sensu Salisb . , Prodr.: 313 (1796) p.p.; sensu Willd., Sp.

PI. 3: 678 (1800) p.p.; sensu Per's. Syn. PL: 232 (1806) p.p.; sensu

Dietr., Lex. Gart. 7: 49 (1807) p.p.; sensu Willd., Enum. 2: 708

(1809) p.p.; sensu DC.,, Prodr. 1: 678 (1824) p.p.; sensu Eckl. &

Zeyh., Enum. 1: 82 (1835); sensu Steud., Norn. Bot. edn 2,2: 289

(1841) p.p.; sensu Harv. in FI. Cap. 1: 306 (1860) p.p.; sensu

Knuth in Pflanzenr. 4,129: 472 (1912) p.p.; omnes non P. querci-

folium (L.f.) L'Herit (1789).

CAPE PROVINCE. — 3318 (Cape Town): Devil’s Peak,

Table Mountain (-CD). Zeyher s.n. (BOL); Ferry Gully, Table

Mountain, Esterhuysen 18330 (BOL); Kirstenbosch (-CD), Ester-

huysen 30839 (BOL); U.C.T., Rondebosch (-CD). Esterhuysen

23545 (BOL; PRE); near Rhodes Memorial, Table Mountain

Erect, branched, strongly balm-scented, some-

what viscid shrub, up to 1,75 m high and 0,5 m in

diameter. Stems herbaceous when young but soon

becoming woody, villous and with glandular hairs in-

Bothalia 15, 3 & 4 (1985)

357

terspersed, green but becoming brownish with age.

Leaves usually soft to the touch, indumentum ex-

tremely variable but always with glandular hairs;

lamina panduriform to cordiform in outline, pinnati-

lobate to pinnatipartite with the segments occasion-

ally shallowly lobed, adaxially glabrate to villous and

green, abaxially sparsely villous to lanate and grey-

ish-green, base cordate, apices of lobes obtuse, mar-

gins crenate, (20-) 35 (-60) x (15-) 25 (-75) mm;

petiole (7-) 15 (-30) mm long, sparsely villous to vil-

lous; stipules triangular to cordiform, apices often 2-

or more dentate, ca 6 x 5 mm. Inflorescence: flow-

ering branches with normal and smaller foliar leaves;

peduncles 20-80 mm long, distinctly articulated at

distal and proximal ends in infructescences, villous

to densely villous and interspersed with glandular

hairs; involucral bracts lanceolate to ovate, often

cuspidate, indumentum as on peduncles, 6-10 x 4-6

mm; pseudo-umbels with 2-20 relatively large flow-

ers. Pedicel 1-3 mm long. Hypanthium 6-13 mm

long, villous and with glandular hairs interspersed.

Sepals lanceolate, apiculate, indumentum as on hy-

panthium, green to reddish-brown with white mar-

gins, 8-14 x 3-5 mm. Petals pale pink to pink; post-

erior two spathulate, apices obtuse to emarginate,

with dark purple markings, reflexed at less than 90°,

20-35 x 6—13 mm; anterior three spathulate with

long narrow claws, slightly reflexed, 15-28 x 5-10

mm. 2n=44. Fig. 1 1 .

yy

FIG. 11. — Pelargonium panduriforme. a, lowering branch, x 1;

b, petals, x 1; c, androecium. x 2; d, gynoecium, x 3; e,

pedicel, hypanthium and sepals, x 1. (From Van der Walt

890, cultivated in Stellenbosch.)

Diagnostic features

Erect, branched, strongly balm-scented, some-

what viscid shrub. Lamina panduriform to cordiform

in outline, pinnatilobate to pinnatipartite, abaxially

sparsely villous to lanate and greyish-green. Pedun-

cles distinctly articulated, pseudo-umbels with 2-20

relatively large flowers each. Flowers pale pink to

pink, pedicel shorter than hypanthium.

P. panduriforme occurs from Antoniesberg in the

vicinity of Willowmore eastwards to near Riebeeck

East, and it has also been collected near Engcobo in

the Transkei (Fig. 12). It is extremely plentiful on

the Baviaanskloof and Kouga Mountains where it

grows on the lower foothills or ravines, often close to

streams. The annual rainfall of the area in which it

occurs is relatively low and it is spread throughout

the year. The summers are very hot and frost occurs

frequently during the winter months.

P. panduriforme flowers from August to January

but the odd flower may be found throughout the

year.

P. panduriforme is closely related to P. quercifo-

lium (L.f.) L’Herit. and the differences between

these two species are tabulated under P. quercifo-

lium. There is evidence that hybridization takes

place between sympatric populations.

FIG. 12. — Known geographical distribution of Pelargonium pan-

duriforme.

TRANSKEI. — 3127 (Lady Frere): near Engcobo (-DB),

Acocks 13817 (PRE), Esterhuysen 29143 (BOL).

CAPE PROVINCE. — 3323 (Willowmore): Antoniesberg

(-AD), Esterhuysen 24970 , 24984 (BOL); Baviaanskloof (-BC),

Bayliss 5586. 7707 (MO), s.n. (PRE), Esterhuysen 24986 (BOL;

PRE); Bo Kouga (-CB), Bayliss s.n. PRE); Kouga Mountains

(-CB), Compton 10529 (NBG), Esterhuysen 4676 (BOL; GRA);

near Misgund (-CD), Compton 7501 (BOL; NBG); Braamrivier

(-DB), Esterhuysen 16307 (BOL); near Studtis (-DB), Olivier

1645 (STEU), Van der Walt & Vorster 1382 (STEU); Bruinklip

(-DC), Van Breda 554 (PRE); Louterwater (-DC), Compton

4241 (BOL); Joubertina (-DD), Esterhuysen 6969. 21223.

(BOL), 22795, 24240 (BOL; PRE), Long 1128 (PRE), Prosser

3323 (GRA), Van der Walt 682, 859 (STEU); Twee Riviere

(-DD), Esterhuysen 7055 (BOL); Kromrivier (-DD), Ecklon &

Zeyher 641 (MO; SAM). 3324 (Steytlerville): Cockscomb (-BD).

Archibald s.n. (BOL); Kouga (-CB), Esterhuysen 4677, 6708

(BOL); Kougarivier (-CC), Compton 5296 (BOL); Assegaai-

bosch (-CD), Esterhuysen 6657 (BOL); Boplaas (-CD), Acocks

21281 (PRE); 40 km W of Patensie (-DA), Wurts 2105 (NBG);

358

Bothalia 15, 3 & 4 (1985)

Melkhoutboom (-DB), Long 1381 (PRE); De Mistkraal (-DB),

Compton 23438, 24067 (NBG); near Combrink (-DD), Acocks

13695 (PRE). 3325 (Port Elizabeth): Ann's Villa (-BB), Bayliss

2948 (NBG). Taylor 9429 (PRE). Van der Walt 890 (STEU);

Palmrietrivier Catchment (-CA), Scharf 1739 (PRE); Winter-

hoek Mountains (-CA), Ecklon & Zeyher 640 (M; MO; OXF; S;

SAM). 3326 (Grahamstown): Mitford Park (-AA), Brink 98

(GRA).

6. Pelargonium quercifolium (L.f.) L'Herit. in

Ait., Hort. Kew. edn 1,2: 422 (1789), excl. var. a;

Salisb., Prodr.: 313 (1796) excl. L’Herit., Geran. t.

14; Willd., Sp. PI. 3: 678 (1800) p.p.; Pers. Syn. PI.

2: 232 (1806) p.p.; Dietr., Lex. Gart. 7: 49 (1807)

p.p.; Willd., Enum. 2: 708 (1809) p.p.; Desf., Arb.

1: 462 (1809) p.p.; Ait. f., Hort. Kew. edn 2,4: 177

(1812) p.p.; DC., Prodr. 1: 678 (1824) p.p.; Steud.,

Norn. Bot. edn 2,2: 289 (1841) p.p.; Harv. in FI.

Cap. 1: 306 (1860) p.p.; Knuth in Pflanzenr. 4,129:

472 (1912) p.p. Type: Cape Province, 'Habitat in

Cap. bonae Spei', Thunberg s.n. (UPS, holo.!).

Geranium quercifolium L.f., Suppl.: 306 (1781); Murray, Syst.

Veg. 14: 619 (1784); Thunb., Prodr. 2: 115 (1800); Thunb., FI.

Cap. edn 2; 521 (1823) excl. ref. Cavanilles.

Pelargonium karrooense Knuth in Reprium nov. Spec. Regni

veg. 19: 231 (1932). Type: Cape Province, ‘Grosse Karoo, Klip-

drift', Schlechter 2278 (Bt, holo.; Z, lecto.l).

P. quercifolium (L.f.) L'Herit. var. pinnatifidum L'Herit., Ge-

ran.: t. 15 (1792), name and illustration (description: L’Herit. in

Ait., Hort. Kew. edn 1,2: 322 (1789). Iconotype: L’Herit., Ge-

ran.: t. 15 (1792).

Erect, much-branched, strongly balm-scented,

viscid shrub, up to 1,75 m high and 0,75 m in di-

ameter. Stems herbaceous when young but soon be-

coming woody, villous and densely interspersed with

long glandular hairs, green but becoming brownish

with age. Leaves hard to the touch (almost sca-

brous), strigose with a few long soft hairs in between

and densely interspersed with long glandular hairs,

green; lamina cordiform to triangular in outline, 3-

palmatisect to pinnatisect, with the segments irregu-

larly pinnatifid to pinnatisect, apices of segments

acute to obtuse, margins irregularly dentate-serrate,

(10-) 20 (-80) x (12-) 25 (-70) mm; petiole (7-) 15

(-30) mm long, indumentum as on stems but with

sharp, appressed stiff hairs in between; stipules tri-

angular to cordiform, ca 4x5 mm. Inflorescence:

flowering branches with normal and smaller foliar

leaves; peduncles 20-80 mm, distinctly articulated at

distal and proximal ends in infructescences, villous

and densely interspersed with long glandular hairs;

involucral bracts broadly ovate, caudate, villous and

with glandular hairs interspersed, 5-9 X 5-7 mm;

pseudo-umbels with 2-6 flowers each. Pedicel 1-2

mm long. Hypanthium 6-11 mm long, villous and

densely interspersed with glandular hairs. Sepals lan-

ceolate, apiculate, indumentum abaxially as on hy-

panthium, green to reddish-brown with white mar-

gins, ca 10 x 3-5 mm. Petals pale pink to dark pink-

ish-purple; posterior two spathulate, apices obtuse

to emarginate, with dark purple markings reflexed at

less than 9u°, 18-25 x 5-8 mm; anterior three spa-

thulate with long narrow claws, slightly reflexed, ca

20 x 4-6 mm. 2n=44. Fig. 13.

Diagnostic features

Erect, much-branched, strongly balm-scented,

viscid shrub. Lamina 3-palmatisect to pinnatisect

FIG. 13. — Pelargonium quercifolium. a, flowering branch, x 1;

b, petals, x 1.5; c, androecium, x 2; d, gynoecium, x 2;

pedicel, hypanthium, sepals and stamens, x 1. (From Van

der Walt 675, cultivated in Stellenbosch.)

with the segments irregularly pinnatifid to pinnati-

sect, almost scabrous. Peduncles distinctly articu-

lated, pseudo-umbels with 2-6 flowers each. Flowers

pale pink to dark pinkish-purple, relatively large,

pedicel much shorter than hypanthium.

P. quercifolium is confined to the one degree

squares which include the towns of Oudtshoorn and

Willowmore (Fig. 14). The annual rainfall in these

Karoo regions is relatively low and it is spread

FIG. 14. — Known geographical distribution of Pelargonium

quercifolium.

Bothalia 15, 3 & 4 (1985)

359

throughout the year. The summer temperatures are

very high and frost occurs frequently during the mid-

winter months. P. quercifolium is usually associated

with fynbos vegetation and it is often found on rocky

hills or mountain slopes in rhenosterbosveld. It is

common along the road between Oudtshoorn and

Willowmore where it grows on disturbed soil.

P. quercifolium flowers from August to January

but the odd flower may be found throughout the

year.

P. quercifolium and P. panduriforme Eckl. &

Zeyh. are closely related species and the following

characters can be used to distinguish between them:

P. quercifolium

1. Leaves 3-palmatisect to

pinnatisect with the

segments irregularly

pinnatifid to pinnatisect.

2. Leaves almost scabrous and

abaxially sometimes villous

but never lanate.

3. Leaves usually viscid. 3.

4. Pseudo-umbels with 2-6 4.

flowers each.

5. Posterior petals up to 25 5.

mm long.

P. panduriforme

1. Leaves pinnatilobate to

pinnatipartite with the

segments occasionally

shallowly lobed.

Leaves usually soft to the

touch and abaxially villous

to lanate.

Leaves occasionally some-

what viscid.

Pseudo-umbels with 2-20

flowers each.

Posterior petals up to 35

mm long.

CAPE PROVINCE. — 3322 (Oudtshoorn): Fonteinplaas,

Kango (-AC), Moffett 230 (STEU); Swartberg Pass (-AC),

Schonland 272 (Z); near De Rust (-BC), Acocks 18284 (GRA;

PRE); near Meiringspoort (-BC), Marloth 11325 (PRE); Rooi-

krantz near Oudtshoorn (-CA), Hops 2 (BOL); 18 km E of

Oudtshoorn (-CB), Schonken 144 (STEU); near Camfer (-CD),

Acocks 22859 (PRE), Esterhuysen 7112 (BOL), Schonken 204

(STEU), Van der Walt 675, 1139, 1308 (STEU); Klipdrift (-CD),

Schlechter 2278 (Z); between Oudtshoorn and Montagu Pass

(-CD), Bolus 14499 (BOL); near junction Oudtshoorn-George

Road (-CD). Bayliss s.n. (PRE); Aangenaam (-DA), Van Nie-

kerk 468 (BOL); below Kammanassieberg (-DB), Coppejans

1377 (WAG). Thompson 1378 (PRE); between Noll and Union-

dale near Keurboomsrivier (-DD), Nesers.n. (STEU). 3323 (Wil-

lowmore): Willowmore (-AD). West 219 (BOL); Antoniesberg

(-AD), Esterhuysen 24963 (BOL); Vaalwater (-AD), Van der

Walt 725 (STEU); Georgida (-AD), Esterhuysen 6392 (BOL;

PRE); 13 km from Willowmore to Uniondale (-AD), Van der

Walt 702 (STEU); Baviaanskloof (-BC), Bayliss 7110, 7492

(MO), Boucher 31 (STEU), Gill s.n. (BOL); near Uniondale

(-CA). Bolus 2278 (BOL; SAM), Marloth 10975 (PRE), Ryder 17

(BOL), Zinn s.n. (SAM); Uniondale Kloof (-CA), Esterhuysen

4699 (BOL); Potjieshoogte (-CA), Thompson 961 (PRE); 10 km

W of Uniondale (-CA), Schonken 140 (STEU); 7 km from

Uniondale to Willowmore (-CA), Van der Walt 723 (STEU);

near Avontuur (-CA), Van der Walt 854 (STEU), Wells 2843

(GRA; PRE); 16 km W of Avontuur (-C A), Theron 1705 (PRE);

Bo Kouga (-CB), Bayliss 7170 (MO); Onder Kouga (-CD), Bay-

liss 6032 (MO; NBG); Nuwekloof (-DA), Acocks 19909 (PRE).

7. Pelargonium glutinosum (Jacq.) L'Herit. in

Ait., Hort. Kew. edn 1,2: 426 (1789); Curtis in Cur-

tis’s bot. Mag. 3: t. 95 (1791); L'Herit., Geran.: t. 20

(1792); Salisb., Prodr.: 314 (1796); Willd., Sp. PI. 3:

676 (1800); Willd., Enum.: 707 (1809); Desf., Arb.

1: 462 (1809); Ait. f., Hort. Kew. edn 2,4: 176

(1812); DC., Prodr. 1: 679 (1824); Eckl. & Zeyh.,

Enum. 1:82 (1835); Harv. in FI. Cap. 1: 306 (1860);

Knuth in Pflanzenr. 4,129: 475 (1912). Type: 'Ex

Africa est.’, (W, holo.!, specimen with Jacquin’s

handwriting).

Geranium glutinosum Jacq., Coll. 1: 85 (1787); Jacq., Icon. PI.

Rar. 1: t. 131 (1787).

G. viscosum Scop., Del. FI. Faun. 2: 27, t. 14 (1786); Cav.,

Diss. 4; 246, t. 108, fig. 2 (1787); Thunb., Prodr. 2: 115 (1800).

Iconotype: Scop., Del. FI. Faun. 2: t. 14 (1786).

G. crataegi folium Roth, Bot. Abh.: 50, t. 9 (1787). Iconotype:

Roth, Bot. Abh.: t. 9 (1787).

Pelargonium erectum Knuth in Bot. Jb. 44: 30 (1910); Knuth in

Pflanzenr. 4, 129: 474 (1912). Type: Cape Province, Worcester,

MacOwan 1705 (Bt, holo.; SAM, lecto.l; Gl; GRA!; K!; Wl;

Z!).

Erect, much-branched, strongly balm-scented,

viscid shrub, up to 1 ,8 m high and 1 m in diameter.

Stems herbaceous when young but soon becoming

woody, glabrous to villous but always with numerous

glandular hairs, green but becoming brownish with

age. Leaves glabrescent but with numerous glandu-

lar hairs, green to dark green; lamina cordiform in

outline; 3-palmatilobate to pinnatisect, basal seg-

ments often irregularly incised, base cordate, apices

of segments acute, margins finely to coarsely den-

tate-serrate and with short sharp-pointed hairs,

(15-) 50 (-120) x (12-) 55 (-130) mm; petiole (5-)

30 (-80) mm long, indumentum as on stems; stipules

narrowly triangular to triangular, cuspidate, indu-

mentum as on stems, ca 7 x 5 mm. Inflorescence:

flowering branches with normal and smaller foliar

leaves; peduncles 15-80 mm long, distinctly articu-

lated at distal and proximal ends in infructescences,

glabrous to sparsely villous but always with numer-

ous glandular hairs; involucral bracts ovate, cuspi-

date, villous and densely interspersed with glandular

hairs, 5-8 x 2-A mm; pseudo-umbels with 1-8 flow-

ers each. Pedicel 1 mm long. Hypanthium 3-10 mm

long, villous and densely interspersed with glandular

hairs. Sepals narrowly lanceolate to broadly spathu-

late, apiculate, indumentum abaxially as on hypan-

thium, green to reddish-brown with a white margin,

ca 12 x 3-6 mm. Petals pale pink to dark pink; post-

erior two spathulate, apices obtuse, with dark pink-

purplish markings, reflexed at less than 90°, 12-25 x

5-8 mm; anterior three spathulate with long narrow

claws, slightly reflexed, 12-25 x 4—8 mm. 2n=44.

Figs 15, 16 & 17.

Diagnostic features

Erect, much-branched, strongly balm-scented,

viscid shrub. Lamina 3-palmatilobate to pinnatisect,

glabrescent. Peduncles distinctly articulated,

pseudo-umbels with 1-8 flowers each. Flowers pale

pink to dark pink, pedicel shorter than hypanthium .

P. glutinosum occurs from Piquetberg in the

south-western Cape to the Kei River in the eastern

Cape. It is also known from a single locality in the

Soutpansberg, northern Transvaal (Fig. 18). The

distribution pattern is largely correlated to mountain

ranges. It grows on well-drained soil in relatively

moist habitats, often in close proximity to running

water. In the Karoo regions of its distribution area

with a low annual rainfall, it is confined to a moun-

tainous habitat where the rainfall is considerably

higher than in the surrounding lower-lying areas.

The rainfall pattern seems to be of little significance

as it occurs in winter rainfall regions, summer rain-

fall regions or in regions receiving rain throughout

the year. Temperatures are high during the summer

and frost is possible during the mid-winter in some of

the inland localities.

360

Bothalia 15, 3 & 4 (1985)

P. glutinosum flowers sporadically throughout the

year, but with a definite peak during the spring

months of September and November.

It is an extremely variable species, both as far as

the leaves and flowers are concerned. Three forms

can be distinguished:

(a) Plants from the western part of the distribution

area have deeply incised leaves and pale pink flow-

ers. The flower buds are distinctly pear-shaped (Fig.

15). This form was described as P. erectum by Knuth

in 1910. Plants from northern Transvaal show a

strong resemblance to this form, although the flow-

ers are smaller and the hypanthium, especially, is

much shorter.

(b) Plants from the Karoo regions have less in-

cised leaves and pink flowers (Fig. 16).

(c) Plants from the eastern part of the distribution

area have shallowly incised leaves with dark pink

flowers (Fig. 17).

The distribution of these three forms is by no

means disjunct and a continuous variation pattern of

leaf and floral characters occurs. It is therefore not

possible to recognize infraspecific taxa although con-

siderable differences exist between the extreme

forms.

The leaves of P. denticulatum Jacq., P. panduri-

forme Eckl. & Zeyh. and P. quercifolium (L.f.)

L’Herit. are also viscid and emit the same balm scent

as P. glutinosum. Inflorescence and floral characters

indicate a very close relationship between P. glutino-

sum and P. denticulatum.

FIG. 15. — Pelargonium glutinosum. a, flowering branch, X 1; b,

petals, x 2; c, androecium, x 2; d, gynoecium, x 3; e, hy-

panthium and sepals, x 1,5; f, sepals, x 1,5. (From Van der

Walt & Vorster 1363, cultivated in Stellenbosch.)

FIG. 16. — Pelargonium glutinosum. a, flowering branch, x 1; b,

petals, x 1,5; c, androecium, x 2; d, gynoecium, x 3; e,

pedicel hypanthium and sepals, x 1. (From Joubert s.n. (sub

STEU 1007), cultivated in Stellenbosch.)

FIG. 17. — Pelargonium glutinosum. a, flowering branch, x 1; b,

petals, x 1; c, androecium, x 2; d, gynoecium, x 3; e, flower

with petals removed. (From Schonken 139, cultivated in Stel-

lenbosch.)

Bothalia 15, 3 & 4 (1985)

361

Geranium viscosum Cav. (1786) is an older name

than G. glutinosum Jacq. (1787). However, this

name cannot be applied to this species, because it

represents a later homonym of G. viscosum Mill.

(1768).

TRANSVAAL. —2329 (Pietersburg): Happy Rest Nature Re-

serve (-BA), Van der Walt & Vorster 1363 (STEU).

CAPE PROVINCE. —3123 (Victoria West): Hoekplaas, Mur-

raysburg (-DD), Acocks 23527 (PRE). 3124 (Hanover): Wapads-

berg Pass (-DD), Acocks 16215 (PRE), Hall 263 (NBG). 3126

(Queenstown): Rockwood, Queenstown (-DD), Galpin 2494

(PRE). 3218 (Clanwilliam): Moutonsvlei, Piquetberg (-DA), Pil-

lans 7290 (BOL). 3224 (Graaff Reinet): farm Rietvlei, Graaff

Reinet (-AB), Galpin 10006 (PRE); 40 km NW of Graaff Reinet

(-AB), Acocks 17548 (PRE); Graaff Reinet (-BC), Bolus 40

(MEL; MO; PRE), Bolus 152 (SAM), Page 14392 (BOL); Pla-

teau of Desolation (-BC), Francis 28 (BOL); Naudesberg Pass

(-BC), Moffett 1020 (STEU); farm De Nek (-BC), Olivier 1672

(STEU); Oudeberg (-DD), Bolus s.n. (MO). 3225 (Somerset

East): Buffelshoek Pass (-AC), Acocks 11973 (PRE). Mountain

Zebra National Park (-AD), Liebenberg 7125 (PRE), Muller 605

(PRE; SRGH); Karreebosch, Cradock (-BA), Long 761 (PRE);

Bruintjieshoogte (-CB), Scott Elliot 557 (E); Boschberg, So-

merset East (-DA), MacOwan 1728 (MEL; SAM); Ongegund

(-DA), P.T. van der Walt 334 (PRE). 3227 (Stutterheim): Wind-

voelberg, Cathcart (-AC), Roberts 1780 (PRE). 3228 (Butter-

worth): Kei River (-CA), Ecklon & Zeyher 643 (SAM). 3319

(Worcester): Hex River Pass (-BD), Grant s.n. (MO; PRE);

Worcester (-CB), MacOwan 1705 (G; GRA; K; SAM; Z), Mar-

loth 2304 (PRE); Keeromsberg (-DA), Esterhuysen 26595, 28160

(BOL). 3320 (Montagu): Kogmanskloof (-CC), Acocks 20344

(PRE), Esterhuysen 23809 (BOL; PRE), Kuntze s.n. (Z); Don-

kerkloof, Montagu (-CC), Compton 18471 (NBG); Montagu

(-CC), Kensit 157 (BOL), Page 15644 (BOL); Tradouw Pass

(-DC), Barnard s.n. (SAM), Marsh 693 (PRE), Van der Walt

826, 1294 (STEU); Rooihoogte Pass (-DC), Boucher 97 (STEU).

3321 (Ladismith): Buffelskloof near Ladismith (-AC), Esterhuy-

sen 14009 (BOL); Seven Weeks Poort (-AD), Compton 7368

(NBG), Neser s.n. (STEU), Schonken 148 (STEU), Van der Walt

1113 (STEU); Toringberg (-AD), Wurts 1035 (NBG); Water-

kloof near Ladismith (-BD), Van der Walt 626, 1119 (STEU);

Kliphuisvlei (-BD), Oliver 5516 (PRE); 51 km from Muiskraal to

Calitzdorp (-CB), Admiraal 207 (PRE); near Van Wyksdorp

(-CB), Van Breda 746 (PRE); Garcia’s Pass (-CC ), Acocks 15437

(PRE), Galpin 3811 (GRA; PRE), Johnson 131 (NBG), Van der

Walt 623, 1295 (STEU); Gamka Reserve near Calitzdorp (-DA),

Esterhuysen 33855 (BOL); Rooiberg Pass (-DA), Compton 3970

(BOL),' Oliver 5310 (STE), Van der Walt 1302 (STEU). 3322

(Oudsthoorn): Swartberg Pass near Prince Albert (-AA), Bolus

11456 (BOL), Hafstrom & Acocks 1973 (PRE); Swartberg Pass

(-AC), Hutchinson 1172 (PRE), Van der Walt 728, 1149, 1312

(STEU); Boomplaas, Kango (-AC), Hugo 32 (STEU); Schoe-

manskloof (-AD), Venter 7445, 7449 (STEU); Meiringspoort

FIG. 18. — Known geographical distribution of Pelargonium glu-

tinosum.

(-BC), Hafstrom & Acocks 751 (PRE), Van der Walt 1140, 1144

(STEU); Blesberg (-BC), Esterhuysen 24896 (BOL); Oudtshoorn

(-CA), Hops 2 (BOL); Groenkloof near Oudtshoorn (-CA),

Starke s.n. (BOL). 3323 (Willowmore): Blydeberg near Willow-

more (-AA), Andreae 944 (PRE); Slypsteenberg (-AC), Ester-

huysen 6319 (BOL), Schonken 139 (STEU); Hot Springs (-AC),

Fourcade 6088 (BOL); Swanepoelspoortberg (-BB), Marloth

4133 (PRE). 3325 (Port Elizabeth): Gamtoosrivier (-CC), Ecklon

& Zeyher 642 (SAM).

8. Pelargonium denticulatum Jacq., Hort.

Schoenbr. 2: 5, t. 135 (1797); Willd., Sp. PI. 3; 680

(1800); Pers., Syn. PI. 2; 232 (1806); DC., Prodr. 1:

674 (1824); Ait., Hort. Kew. edn 2,4: 179 (1812);

Sweet, Geran. 2: 109 (1822); Eckl. & Zeyh., Enum.

1: 83 (1835); Harv. in FI. Cap. 1: 307 (1860); Knuth

in Pflanzenr. 4,129: 476 (1912); Compton in Trans.

R. Soc. S. Afr. 19: 294 (1931); J.J.A. v.d. Walt &

Vorster, Pelarg. S. Afr. 2: 51, fig. (1981). Iconotype:

Jacq., Hort. Schoenbr. 2: t. 135 (1797).

Geranium denticulatum (Jacq.) Poir., Encycl. Suppl. 2: 755

(1811). Geraniospermum denticulatum (Jacq.) Kuntze, Rev. Gen.

1: 94 (1891).

Erect, much-branched, balm-scented, viscid

shrub, up to 2 m high and 1 m in diameter. Stems

herbaceous when young but soon becoming woody,

villous to hirsute and with numerous glandular hairs

interspersed, dark green and sometimes flushed with

purple, becoming brown with age. Leaves sparsely

strigose but densely beset with glandular hairs,

green; lamina pinnatisect to palmatisect with pinna-

tisect segments, base cordate, (40-) 60-80 (-100) x

(45-) 70-90 (-110) mm, segments narrow and adax-

ially grooved, apices acute, margins irregularly den-

ticulate, apices of teeth acute; petiole (25-) 50-60

(-90) mm long; stipules asymmetric-triangular, ca 6

x 2-5 mm. Inflorescence: flowering branches with

normal and smaller foliar leaves; peduncles 20-60

mm long, distinctly articulated at distal and proximal

ends in infructescences, sparsely villous to villous

and with numerous glandular hairs interspersed; in-

volucral bracts lanceolate to ovate, acuminate, indu-

mentum as on peduncles, 4-7 x 2-3 mm; pseudo-

umbels with 3-7 (-9) flowers each. Pedicel 1-2 mm

long. Hypanthium 4—9 mm long, indumentum as on

peduncles. Sepals lanceolate, acuminate, indumen-

tum abaxially as on peduncles, green with a reddish-

brown tint, 8-10 x 2-4 mm. Petals pinkish-purple;

posterior two spathulate to obovate, apices obtuse to

emarginate, with dark red to purple markings, re-

flexed at ca 90°, ca 20 x 6 mm; anterior three spathu-

late with narrow claws, reflexed at less than 90°, ca

14 x 4 mm. 2n=44. Figs 19 & 20.

Diagnostic features

Erect, much-branched, balm-scented, viscid

shrub. Lamina pinnatisect to palmatisect with pinna-

tisect segments, segments narrow and adaxially

grooved, apices acute, apices of teeth acute, sparsely

strigose. Peduncles distinctly articulated, pseudo-

umbels with 3-9 flowers each. Flowers relatively

small, pinkish-purple, pedicel shorter than hypan-

thium.

P. denticulatum occurs in the south-western and

southern Cape and it is known from Algeria near

Clanwilliam south-eastwards to Baviaanskloof near

362

Bothalia 15, 3 & 4 (1985)

FIG. 19. — Pelargonium denticulatum. a, flowering branch, x 1;

b, petals, X 2; c, androecium, X 3; d, gynoecium, X 4; e,

sepals, x 2. (From Van der Walt 728. cultivated in Stellen-

bosch.)

FIG. 20. — Pelargonium denticulatum. a, flowering branch, x 1;

b, petals, x 1,5; c, androecium, x 2; d, gynoecium, x 4; e,

schizocarp with mericarps, x 2. (From Van der WaltllS. cul-

tivated in Stellenbosch.)

FIG. 21 . — Known geographical distribution of Pelargonium den-

ticulatum.

The leaves of P. denticulatum could also be con-

fused with those of P. radens H.E. Moore. The dif-

ferences between the leaves of these two species are

tabulated under P. radens.

P. denticulatum was introduced into England from

the Cape as early as 1789 by Masson, who was sent

there to collect live plants for the Royal Botanic

Gardens at Kew. According to Sweet, who pub-

lished a description and a beautiful coloured illustra-

tion in his Geraniaceae 2: t. 109 (1822), it was culti-

vated in Kew until at least 1821 but there is no evi-

dence that Jacquin obtained the material which he

used to describe the species in 1797, from Kew. In

the herbarium of the Naturhistorisches Museum in

Vienna there are several specimens of P. denticula-

tum, but none has any inscription in Jacquin’s hand-

writing, and there is also no evidence that he ever

used any of these for his original description. It

would be presumptuous, therefore, to consider any

of them as the type specimen of P. denticulatum. For

this reason, we consider it more prudent to recog-

nize the illustration which accompanied the original

description, as an iconotype.

Willowmore (Fig. 21). As in the case of so many

other representatives of the section Pelargonium,

this species is confined to mountains or their im-

mediate vicinity, where the microclimate is consider-

ably moister than in the neighbouring lower-lying

areas. It is usually found in ravines and near streams.

Its distribution area receives most of the rainfall dur-

ing the winter months. Exact rainfall and tempera-

ture records for these mountainous areas are not yet

available.

P. denticulatum flowers between April and

November with a peak in September, and occasion-

ally as late as the following January. Inflorescence

and floral characters indicate a very close relation-

ship between P. denticulatum and P. glutinosum

(Jacq.) L'Herit. The leaves of these two species are

both viscid and they have the same balm scent. How-

ever, the leaves of P. glutinosum are less divided

(pinnatisect) and the leaf segments consequently

much wider than those of P. denticulatum.

Bothalia 15, 3 & 4 (1985)

363

CAPE PROVINCE. — 3219 (Wuppertal): Algeria (-AC), Van

der Walt 743 (STEU); Middelberg, Cedarberg (-AC), Kerfoot

6108 (NBG). 3222 (Beaufort West): farm Donkerhoek, Nieuvels

Mountains (-BA), Liebenberg 327 (PRE). 3319 (Worcester): Ka-

roopoort (-BA). Hutchinson 449 (BOL). 3320 (Montagu): Witte-

berge, 1ft km W of Laingsburg (-BA), Fischer 240 (STEU);

Witteberg, Matjiesfontein )-BC), Compton 2568 (BOL), Ester-

huysen 28880 (BOL). Marloth 2967 (BOL); Tradouw Pass (-DC),

Taylor 394 (BOL). 3321 (Ladismith): Seven Weeks Poort (-AD),

Barker 5457 (BOL; NBG), Bayliss 2438 (NBG; Z), Esterhuysen

24746 (BOL), Marloth 2947 (PRE), Phillips 1414 (PRE; SAM),

Rvcroft 2706 (NBG), Salter 3118 (BOL), Taylor 6966 (PRE),

Wells 3758 (GRA; PRE), Wisura 1789 (NBG), Warts 1009

(NBG); Waterkloof, Ladismith (-AD), Stokoe 1896 (PRE);

Rooiberg (-CB), Compton 3916 (BOL); farm Langberg (-DC),

Van der Walt 718 (STEU); N of Cloete’s Pass (-DD), Goldblatt

4162 (MO; PRE); Cloete’s Pass (-DD), Muir 1070 (BOL; PRE),

Schlieben & Ellis 12357 (PRE; WAG). 3322 (Oudtshoorn): near

Prince Albert (-AA), Bayliss 1945 (NBG; Z), Marloth 4473

(PRE); Schoemanspoort (-AC), Britten 1635 (GRA; PRE);

Oudtshoorn (-CA). Britten 54 (GRA); Robinson Pass (-CC),

Middlemost 2022 (NBG). 3323 (Willowmore): Baviaanskloof

(-BC), Bayliss 5747 (MO; WAG; Z).

9. Pelargonium pseudoglutinosum Knuth in Re-

prium nov. Spec. Regni veg. 45: 64 (1938). Type:

Cape Province, ‘Distr. Uniondale, Long Kloof am

Keurbooms River, 510 m", Fourcade 3388 (Bt,

holo.; BOL (Fourcade colln), lecto.!).

P. uniondalense Knuth in Reprium nov. Spec. Regni veg. 28: 91

(1930); non Knuth loc. cit. 18: 135 (1918).

P. uniondalense Knuth var. scabridum Knuth in Reprium nov.

Spec. Regni veg. 28:91 (1930). P. pseudoglutinosum Knuth var.

scabridum (Knuth) Knuth in Reprium nov. Spec. Regni veg. 45:

64 (1938). Type: Cape Province, 'Distr. Uniondale, zwischen

Uniondale und Avontuur, 690 m\ Fourcade 2922 (Bt, holo.;

BOL (Fourcade colln). lecto.!).

Erect to decumbent, much-branched, non-aro-

matic, viscid shrub, up to 1 m high and 1,25 m in

diameter. Stems herbaceous when young but soon

becoming woody, with numerous glandular hairs

and long soft hairs in between, green but becoming

wine-red and eventually brown with age. Leaves gla-

brescent to sparsely strigose and with glandular

hairs, green but the larger veins adaxially often dark

purple; lamina ovate in outline, pinnatipartite to

pinnatisect with the segments irregularly incised,

base cuneate to obtuse, apices of segments acute,

margins coarsely dentate and with short conical

hairs, (20-) 35 (-50) x (15-) 25 (^10) mm; petiole

10-20 mm long; stipules triangular, cuspidate to sub-

ulate, sparsely hispid with glandular hairs inter-

spersed, 6-8 x 2-4 mm. Inflorescence: flowering

branches with normal and smaller foliar leaves; pe-

duncles 5-25 mm long, distinctly articulated at distal

and proximal ends in infructescences, with grandular

hairs and long soft hairs in between; involucral

bracts ovate, apiculate, with glandular hairs and long

soft hairs on the margins, 3-5 x 3^4 mm; pseudo-

umbels with 1-2 flowers each. Pedicel 1-2 mm long.

Hypanthium 6-10 mm long, with many glandular

hairs. Sepals lanceolate, cuspidate, abaxially with

many glandular hairs, green to wine-red, ca 10 x 4-5

mm. Petals pale pink to pink; posterior two spathu-

late, apices obtuse to emarginate, with a dark pink

blotch and purple markings, reflexed at ca 90°, 15-25

x 5-8 mm; anterior three spathulate with long nar-

row claws, slightly reflexed, ca 12-20 x 4-6 mm.

2n=44. Fig. 22.

FIG. 22. — Pelargonium pseudoglutinosum. a, flowering branch,

x 1; b, petals, x 1,5; c, androecium, x 3; d, gynoecium, x 3.

(From Van der Walt 855, cultivated in Stellenbosch.)

Diagnostic features

Erect to decumbent, much-branched, non-aro-

matic, viscid shrub. Lamina pinnatipartite to pinna-

tisect, glabrescent to sparsely strigose, larger veins

adaxially often dark purple. Peduncles distinctly ar-

ticulated, pseudo-umbels with 1-2 flowers each.

Flowers pale pink to pink, pedicel shorter than hy-

panthium.

P. pseudoglutinosum has a restricted distribution

in the southern Cape, and it is only known from

Mannet jiesberg, Keurboomsrivier, Uniondale and

Prince Alfred’s Pass (Fig. 23). It grows in ravines

near streams on well-drained soil. Its distribution

area receives an annual rainfall of approximately 200

mm which is evenly spread throughout the year.

Temperatures are high during summer and frost oc-

curs regularly during the winter.

P. pseudoglutinosum flowers from September to

January. The odd flower may be found throughout

the year.

P. denticulatum Jacq., P. glutinosum (Jacq.)

L’Herit., P. pseudoglutinosum and P. quercifolium

(L.f.) L’Herit. are the only species of the section

with viscid leaves. The leaves of P. pseudoglutino-

sum could be confused with those of P. glutinosum.

P. glutinosum , however, has strongly aromatic

leaves and lacks the purple veins which are often

found in P. pseudoglutinosum. The 1-2-flowered

pseudo-umbels are the most reliable feature to dis-

364

Bothalia 15, 3 & 4 (1985)

FIG. 23. — Known geographical distribution of Pelargonium

pseudoglutinosum.

tinguish P. pseudoglutinosum from P. glutinosum

which has more than two flowers per pseudo-umbel.

CAPE PROVINCE. — 3322 (Oudtshoorn): Mannetjiesberg

(-DB), Esterhuysen 6490 (BOL); Keurboomsrivier (-DD), Four-

cade 3388 (BOL). 3323 (Willowmore): Uniondale (-CA), Ester-

huysen 16417 (BOL); 8 km S of Uniondale (-CA), Theron 1372

(PRE); Uniondale Poort (-CA), Van der Walt 855 (STEU); be-

tween Avontuur and Uniondale (-CA), Fourcade 2922 (BOL);

Prince Alfred's Pass (-CC), Schonken 136 (STEU).

fO. Pelargonium graveolens L'Herit. in Ait.,

Hort. Kew. edn f ,2; 423 (1789); L'Herit., Geran.: t.

17 (1792); Willd. , Sp. PI. 3: 678 (1800); Pers., Syn.

PI. 2; 232 (1806); Willd., Enurn. 2: 708 (1809); Ait.

f., Hort. Kew. edn 2,4: 178 (1812); DC., Prodr. 1:

678 (1824); Eckl. & Zeyh., Enum. 1: 82 (1835);

Harv. in FI. Cap. 1; 306 (1860); Knuth in Pflanzenr.

4,129: 475 (1912); Burtt Davy, FI. Transv.: 190

(1926); H.E. Moore in Baileya 3: 15 (1955); Muller

in FI. Zamb. 2,1: 147, t. 22 (1963); Batten & Bokel-

mann. Wild Flow. E. Cape: 88, fig. 74.3 (1966). Ico-

notype: L'Heritier, Geran.: t. 17 (1792).

Geranium graveolens (L'Herit.) Thunb., Prodr. 2: 115 (1800);

Thunb., FI. Cap. edn 2; 522 (1823).

G. radula Roth, Bot. Abh.: 51, t. 10 (1787), non Cav. Pelargo-

nium asperum Ehrh. ex Willd., Sp. PI. 3: 678 (1800); Pers., Syn.

PI. 2: 232 (1806); Ait. f., Hort. Kew. edn 2.4; 178 (1812); DC.,

Prodr. 1: 679 (1824). Geranium asperum (Ehrh. ex Willd.) Poir.,

Encycl. Suppl. 2: 759 (1811). Iconotype; Roth, Bot. Abh.; t. 10

(1787).

G. terebinthinaceum Cav., Diss. 4; 250, t. 114, fig. 1 (1787), non

Murray. Geraniospermum terebinthinaceum (Cav.) Kuntze, Rev.

Gen. 1: 94 (1891). Pelargonium terebinthinaceum (Cav.) Desf.,

Arb. 1; 462 (1809); Hanks & Small, N. Amer. FI. 25: 24 (1907).

Iconotype: Cavanilles, Diss. 4: t. 114, fig. 1 (1787).

P. intermedium Knuth in Bot. Jb. 44: 27 (1910). Syntypes: Cape

Province, ‘Howison's Poort’, Schonland 614 (GRA, lecto.l; NH!;

Z! ) ; ‘Kap der guten Hoffnung', Krebs s.n. ( Bt) -

Erect, much-branched, strongly rose-scented

shrub, up to 1,3 m high and 1 m in diameter. Stems

herbaceous when young, becoming woody with age,

villous to densely villous and densely interspersed

with glandular hairs, green but becoming brownish

with age. Leaves soft to the touch, indumentum vari-

able but always with numerous glandular hairs;

lamina cordiform in outline, palmatipartite to pinna-

tisect with the segments irregularly pinnatipartite to

pinnatisect, adaxially sparsely villous to villous and

green, abaxially villous to densely villous or almost

lanate and greyish-green, base cordate, segments ad-

axially grooved, apices obtuse to acute, margins ir-

regularly serrate and revolute, (20—) 40 (-55) x

(30-) 60 (-100) mm; petiole (10-) 30 (-80) mm long;

stipules asymmetric-triangular, cuspidate, ca 6 x 4

mm. Inflorescence: flowering branches with normal

and smaller foliar leaves; peduncles 15-40 mm long,

villous to densely villous and densely interspersed

with glandular hairs; involucral bracts ovate to lan-

ceolate, cuspidate, indumentum as on peduncles,

4—6 x 2-3 mm; pseudo-umbels with 3-5 (-7) flowers

each. Pedicel 1-7 mm long. Hypanthium 4-15 mm

long, indumentum as on peduncles. Sepals lanceo-

late, indumentum adaxially as on peduncles, green

to reddish-brown with a white margin, 8-11 x 2-3,5

mm. Petals white to pinkish-purple; posterior two

spathulate, apices obtuse to emarginate, with wine-

red feather-like markings, reflexed at ca 90°, 20-30 x

6-8 mm; anterior three spathulate to oblanceolate

with narrow claws, reflexed at ca 45°, 17-22 x 4—5

mm. 2n=88. Fig. 24.

FIG. 24. — Pelargonium graveolens. a, flowering branch, x 1; b,

petals, x 1; c, androecium, x 2; d, gynoecium, x 3. (From

Van der Walt 690, cultivated in Stellenbosch.)

Diagnostic features

Erect, much-branched, strongly rose-scented

shrub. Lamina palmatipartite to pinnatisect with ir-

regularly pinnatipartite to pinnatisect segments, soft

to the touch, margins of segments revolute. Pseudo-

umbels with 3-5 (-7) white to pinkish-purple flowers

each, pedicel usually shorter than hypanthium.

Bothalia 15, 3 & 4 (1985)

365

P. graveolens has two separate areas of distribu-

tion in southern Africa, one in the northern Trans-

vaal and one in the south-eastern part of the Cape

Province. It is also recorded from Zimbabwe and

Mozambique (Fig. 25). In northern Transvaal it

grows on mountains and occurs from Blouberg in the

west to Wolkberg near Pilgrim’s Rest in the east. In

the Cape Province it occurs from near George in the

west to Grahamstown in the east.

It is usually found in relatively moist habitats in

semi-shaded situations. The populations in the

northern Transvaal receive summer rain and those in

the Cape Province rain throughout the year. The

summers in both areas are hot to very hot and the

winters mild.

One can only speculate that it had a continuous

distribution in earlier times along mountain ranges

from northern Transvaal to the Cape Province.

FIG. 25. — Known geographical distribution of Pelargonium

graveolens.

P. graveolens flowers from August to January with

a peak in September and October. However, the

odd flower may be found throughout the year.

P. graveolens and P. radens H.E. Moore are two

closely related species. Their floral structure is very

similar and they have the same fragrance. They can

only be distinguished on leaf characteristics, and

these differences can be tabulated as follows:

P. graveolens

1. Leaves villous and

soft to the touch.

2. Leaf segments at

least 3 mm wide.

3. Leaf margins less

revolute.

P. radens

1 . Leaves hirsute and

almost scabrous.

2. Leaf segments less

than 3 mm wide.

3. Leaf margins more

revolute.

TRANSVAAL. — 2229 (Waterpoort): 8 km W of Wyllie’s

Poort (-DD), Hutchinson & Gillen 4384 (BM). 2329 (Peters-

burg): Blouberg (-AA), Esterhuysen 21440 (BOL; MO; PRE),

Meeuse 10337 )PRE; SRGH), Strey & Schlieben 8507 (PRE;

SRGH), Van der Schijff 5398 (PRE); Letsjume (-AB), Venter

6206 (PRE); Hanglip (-BB), Meeuse 10193 (PRE); 18 km W of

Louis Trichardt (-BB), Prosser 2022 (PRE); Wolkberg (-CC),

Leighton 3236 (BOL). 2427 (Thabazimbi): Krantzberg (-BC),

Codd 4809 (PRE), Van der Merwe 2030 (PRE); farm Groothoek

(-BC), Codd 3969 (PRE). 2429 (Zebediela): Makapaansberg

(-AA), Moffett 1741 (STEU). 2430 (Pilgrim’s Rest): Wolkberg

(-AA), Muller & Scheepers 153 (PRE).

CAPE PROVINCE. — 3322 (Oudtshoorn): Swartrivier be-

tween George and Wilderness (-DC), Oosthuizen s.n. (STEU).

3325 (Port Elizabeth): Enon (-BC), Thode A2614 (NH; PRE);

Loerie Plantation (-CC), Dix 4 (BOL). 3326 (Grahamstown):

Riebeek East (-AA), Bayliss 8067 (MO); near Alicedale (-AC),

Bayliss 8098 (MO); Howieson’s Poort (-AD), Barker 10612

(NBG), Britten 2553 (GRA), Leighton 3071 (BOL), Schonland

614 (GRA; NH; Z), Van der Walt 690 (STEU); Grahamstown

(-BC), Bokelmann 3-PL13 (NBG), Daly & Sole 2709 (PRE),

Galpin s.n. (GRA; PRE), MacOwan 813 (BOL; GRA; Z), Mac-

Owan s.n. (BOL; MEL; SAM).

11. Pelargonium radens H.E. Moore in Baileya

3,1: 22 (1955); J.J.A. v.d. Walt, Pelarg. S. Afr. 1:

38, fig. (1977). Iconotype: Cavanilles, Diss. 4: t. 101,

fig. 1 (1787).

Geranium radula Cav., Diss. 4: 262, t. 101, fig. 1 (1787); ?Cur-

tis in Curtis’s bot. Mag. 3: t. 95 (1789); Thunb.,, Prodr. 2: 116

(1800); Thunb., FI. Cap. edn 2: 529 (1823); non Roth., Bot.

Abh.: 51, t. 10 (1787). Pelargonium radula (Cav.) L’Herit. in

Ait., Hort. Kew. edn 1,2: 423 (1789); Willd., Sp. PI. 3: 679

(1800); Pers., Syn. PI. 2: 232 (1806); Willd., Enum. 2: 709 (1809);

Desf., Arb. 1: 462(1809); Ait. f.. Hort. Kew. edn 2,4: 178(1812);

DC., Prodr. 1: 679 (1824); Eckl. & Zeyh.. Enum. 1: 82 (1835);

Steud., Norn. Bot. edn 2,2: 289 (1841); Harv. in FI. Cap. 1: 307

(1860); Knuth in Pflanzenr. 4.129: 477 (1912). Geraniospermum

radula (Cav.) Kuntze, Rev. Gen. 1: 95 (1891).

Geranium revolutum Jacq., Coll. 1: 84 (1787); Icon. PI. Rar. 1:

t. 133 (1787); non Andr., Bot. Rep. 5: t. 354 (1804). Iconotype:

Jacquin, Icon. PI. Rar. 1: t. 133 (1787).

Pelargonium multifidum Salisb., Prodr.: 313 (1796). nom. il-

legit.

Erect, much-branched, rose-scented shrub, up to

1,5 m high and 0,75 m in diameter. Stems herba-

ceous when young but soon becoming woody,

sparsely to densely hirtellous and densely inter-

spersed with glandular hairs, green but becoming

brownish with age. Leaves hard to the touch (almost

scabrous), hirsute and with glandular hairs inter-

spersed, greyish-green to green; lamina palmatisect

to pinnatisect with narrow, irregularly pinnatisect

segments, segments adaxially grooved, apices usu-

ally obtuse, margins revolute, (20—) 45 (-85) x (25-)

70 (-150) mm; petiole (10-) 40 (-150) mm long;

stipules asymmetric and narrowly triangular, hirtel-

lous to villous and densely interspersed with glandu-

lar hairs, 4-6 x 2-3 mm. Inflorescence: flowering

branches with normal and smaller foliar leaves; pe-

duncles 12-20 mm long, sparsely to densely hirtel-

lous and densely interspersed with glandular hairs;

involucral bracts lanceolate, cuspidate, indumentum

as on stipules, 4-6 x 2-3 mm; pseudo-umbels with

3-5 (-8) flowers each. Pedicel 2-10 mm long, indu-

mentum as on stipules. Hypanthium 2-8 mm long.

Sepals lanceolate, indumentum adaxially as on stipu-

les, green to reddish-brown with white margins, ca

8 x 1,5-3, 5 mm. Petals pale purple to pinkish-pur-

ple; posterior two spathulate, apices obtuse to emar-

ginate, with dark wine-red feather-like markings, re-

flexed at ca 90°, ca 20 x 5 mm; anterior three spathu-

late with narrow claws, reflexed at ca45°, ca 15 x 3,5

mm. 2n=88. Fig. 26.

Diagnostic features

Erect, much-branched, rose-scented shrub.

Lamina palmatisect to pinnatisect, segments pinnati-

sect, narrow, apices usually obtuse, margins revo-

366

Bothalia 15, 3 & 4 (1985)

FIG. 26. — Pelargonium radens. a, flowering branch, x l;b, pe-

tals, x 1,5; c, androecium, x 2; d, gynoecium, X 3. (From

Ward-Hilhorst s.n. (sub STEU 1369), cultivated in Stellen-

bosch.)

lute, almost scabrous. Pseudo-umbels with 3-8 flow-

ers each. Flowers pale purple to pinkish-purple, ped-

icel as long as hypanthium.

P. radens occurs from Tulbagh in the south-west-

ern Cape, eastwards along the coast to Engcobo in

the Transkei (Fig. 27). It is particularly common in

the coastal regions of the southern Cape. In the

south-western Cape it receives winter rains, in the

southern Cape winter as well as summer rains, and

in the eastern Cape and Transkei predominantly

summer rains. It is usually found in mountainous,

rather moist habitats, but apart from the water re-

quirements, it is apparently adapted to a wide range

of climatic conditions especially as far as tempera-

tures are concerned.

P. radens flowers from August to January with a

peak during the spring months (September to Oc-

tober).

The leaves of P. radens resemble those of P. denti-

culatum Jacq. and the differences between the leaves

of these two species can be tabulated as follows:

P. radens

1 . Leaf segment margins

revolute.

2. Apices of leaf

segments usually

obtuse.

3. Leaves not viscid.

P. denticulatum

1. Leaf segment margins

not revolute.

2. Apices of leaf

segments acute.

3. Leaves viscid.

Moore (1955) explained in detail why it was

necessary to replace the old established name P. ra-

FIG. 27. — Known geographical distribution of Pelargonium ra-

dens.

data with P. radens. The name Geranium radida

Cav. is illegitimate since it was superfluous when

published.

TRANSKEI. — 3127 (Lady Frere): near Cala (-DA), Flana-

gan 2628 (PRE; SAM), Bolus 8814 (BOL); Engcobo (-DB), Es-

terhuysen 29218 (BOL); Flanagan 2729 (PRE).

CAPE PROVINCE. — 3319 (Worcester): Waterfall, Tulbagh

(-AC), Bolus s.n. (BOL), Ecklon & Zeyher 644 (SAM; W). 3320

(Montagu): Tradouw Pass (-DC), Hafstrom & Acocks 1981

(PRE), Wall s.n. (MO). 3322 (Oudtshoorn): Robinson Pass

(-CC), Van der Walt 1153 (STEU); 10 km E of George (-CD),

Schonken 187 (STEU); Saasveld (-DC), Oosthuizen s.n.

(STEU), Schonken 188 (STEU); Kaaimansgat (-DC), Compton

15799 (NBG). 3323 (Willowmore): Keurboomsrivier (-CC), Es-

terhuysen 27351 (BOL); Prince Alfred’s Pass (-CC), Horn s.n.

(PRE); De Vlug (-CC), Van der Walt 845 (STEU); Braamrivier,

Uniondale (-DB), Esterhuysen 16318 (BOL; NBG); Die Hoek

near Joubertina (-DD), Esterhuysen 22791 (BOL). 3325 (Port

Elizabeth): Suurberg Pass (-BB), Taylor 9426 (PRE);

Olifantskop (-BC), Olivier 1700 (STEU); Farm Deyselskraal,

Llitenhage (-CA), Scharf 1869 (PRE); Groendal, Wilderness

(-CA), Scharf 1641 (PRE); Van Stadensrivier (-CC), Ecklon &

Zeyher 645 (S; SAM; W); Van Stadensberg (-CC), Ecklon &

Zeyher 2096 (S), Elliot 255 (E); 49 km from Port Elizabeth to

Humansdorp (-CC), Stirton 6348 (MO; PRE); Bulkrivier Reser-

voir (-CC), Holland 3659 (BOL); Uitenhage (-CD), Zeyher 301

(SAM), Zeyher 319 (MEL); 24 km from Uitenhage on Elandsri-

vier road (-CD), Acocks 21073 (PRE); Kameahs (-CD), West 316

(BOL); Swartkopsrivier (-DC), Drege s.n. (SAM), Drege 7444

(E; K; MO; PRE; S: W); Walmer (-DC), Paterson 129 (Z). 3326

(Grahamstown): near Sidbury (-AD), Dyer 2233 (GRA); Cold-

spring (-AD), Rogers 3988 (Z); between Grahamstown and Asse-

gaaibos (-AD), Ecklon & Zeyher 199 (PRE; SAM); Bothasberg

(-BA), Ecklon & Zeyher 646 (K; MEL; MO; PRE; S; W; Z);

Grahamstown (-BC), Bayliss 5275, 7507 (GRA; MO), Ecklon &

Zevher 2097 (GRA; W), MacOwan s.n. (BOL; GRA; MEL).

MacOwan 460 (GRA; SAM; Z), Rogers 29470 (SAM; Z); 8 km

from Grahamstown to Fort Beaufort (-BC), Stirton 6207 B

(PRE); Mountain Drive, Grahamstown (-BC), Bond 1245

(NBG); Farm Earls Rust near Grahamstown (-BC), Dyer 604

(PRE); Collingham Tower near Grahamstown (-BC), Leighton

3072 (BOL); Fraser’s Camp Reserve near Grahamstown (-BD),

Rhumane P. 29 (GRA); Longvale, De Kol (-DB), Grant 24

(GRA). 3423 (Knysna): Plettenberg Bay (-AB), Kapp 117

(PRE), Keet 1014 (GRA; PRE), Rogers 22320 (PRE), Rogers

26776 (Z), Taylor 6039 (NBG). 3424 (Humansdorp): Clarkson

(-AB), Thode A758 (MO; NH; PRE).

12. Pelargonium scabrum (L.) L'Herit. in Ait.,

Hort. Kew. edn 1,2: 430 (1789); Jacq., Icon. PI. Rar.

3: 11, t. 542 (1789); Salisb., Prodr.: 313 (1796);

Bothalia 15, 3 & 4 (1985)

Willd., Sp. PI. 3: 681 (1800); Pers., Syn. PI. 2: 233

(1806); Desf., Arb. 1: 462 (1809); Ait. f., Hort.

Kew. edn 2,4: 179 (1812); DC., Prodr. 1: 677 (1824);

Spreng., Syst. Veg. 3: 60 (1826); Eckl. & Zeyh.,

Enum. 1: 81 (1835); Steud., Nom. Bot. edn 2,2: 289

(1841); Harv. in FI. Cap. 1: 304 (1860); Knuth in

Pflanzenr. 4,129: 462 (1912); J. J. A. v.d. Walt, Pel-

arg. S. Afr. 1: 42, fig. (1977). Lectotype: Specimen

LINN 858.8 (LINN!), specific epithet in Linnaeus’s

handwriting.

Geranium scabrum L., Cent. PI. 1: 281 (1755); L., Syst. Nat. 2:

1142 (1759); L., Amoen. 4: 281 (1759); Burnt, f., Geran.: 34, t. 1

(1759); L., Sp. PI. edn 2,2: 946 (1763); Burnt, f.. Prodr. FI. Cap.:

18 (1768); Murray, Syst. Veg. 14: 613 (1784); Cav., Diss. 4: 247, t.

108, fig. 1 (1787); Jacq., Coll. 2: 327 (1789); Thunb. , Prodr. 2: 115

(1800); Andrews, Geran. 1: C, ic (1805); Thunb. , FI. Cap. edn 2:

523 (1823). Geraniospennum scabrum (L.) Kuntze, Rev. Gen. 1:

95 (1891).

Pelargonium balsameum Jacq., Icon. PI. Rar. 3: t. 543 (1794);

Jacq., Coll. 5: 136 (1797); Willd., Sp. PI. 3: 679 (1800); Pers.,

Syn. PI. 2: 232 (1806); Willd., Enum. 2: 709 (1809); Ait. f., Hort.

Kew. edn 2,4: 178 (1812); DC., Prodr. 1: 679 (1824). Geranium

balsameum (Jacq.) Poir., Encycl. Suppl. 2: 754 (1811). Pelargo-

nium scabrum (L.) L'Herit. var. balsameum (Jacq.) Harv. in FI.

Cap. 1: 304 (1860); Knuth in Pflanzenr. 4,129: 462 (1912). Icono-

type: Jacq., Icon. PI. Rar. 3: t. 543 (1794).

Erect, often many-stemmed, soboliferous, usually

much-branched, lemon-scented or non-aromatic

shrub, up to 2,5 m high and 1,5 m in diameter. Stems

herbaceous when young but soon becoming woody,

strigose and with glandular hairs interspersed, green

but becoming brownish with age. Leaves hard to the

touch, scabrous, indumentum variable but always

with glandular hairs, green; lamina rhomboidal in

outline, 3-palmatilobate to 3-palmatisect with the

segments irregularly incised, adaxially glabrescent to

sparsely strigose, abaxially strigose, base cuneate,

lateral segments often bifid and terminal segment of-

ten trifid, apices acute, margins finely to coarsely

dentate-serrate, (20-) 40 (-100) x (10-) 40 (-110)

mm; petiole (5-) 15 (-70) mm long, indumentum as

on stems; stipules asymmetric-triangular, usually

cuspidate, margins strigose, 5-10 x 3-8 mm. Inflo-

rescence: flowering branches with normal and

smaller foliar leaves; peduncles 6-50 mm long,

sparsely strigose to strigose with glandular hairs in-

terspersed; involucral bracts lanceolate to ovate,

cuspidate, densely beset with glandular hairs and

margins strigose, 3-10 x 1-5 mm; pseudo-umbels

with 2-6 flowers each. Pedicel 3-12 mm long,

sparsely strigose but densely beset with glandular

hairs. Hypanthium 3-12 mm long, indumentum as

on pedicel, purplish. Sepals lanceolate to spathulate,

apiculate, indumentum abaxially as on pedicel,

green to purplish, 7-12 x 2-4 mm. Petals white to

purple to pinkish-purple; posterior two spathulate,

sometimes emarginate, usually with feather-like

dark purple markings, reflexed at ca 90°, 6-15 x 3-7

mm; anterior three narrowly oblanceolate to almost

linear with narrow claws, slightly reflexed, 4-12 x

1,5-4 mm. 2n=22. Figs 28, 29 & 30.

Diagnostic features

Erect, soboliferous, lemon-scented or non-aro-

matic shrub. Lamina 3-palmatilobate to 3-palmati-

sect with the segments irregularly incised, scabrous.

Pseudo-umbels with 2-6 flowers each. Flowers white

367

FIG. 28. — Pelargonium scabrum. a, flowering branch, x 1; b,

petals, x 2; c, androecium, x 2; d, gynoecium, x 4; e, meri-

carp, x 2. (From Fischer s.n. (sub STEU 1925), cultivated in

Stellenbosch.)

FIG. 29. — Pelargonium scabrum. a, flowering branch, x 1; b.

petals, x 2; c, androecium, x 3; d, gynoecium, x 4; e, meri-

carp, x 3. (From Van der Walt 552, cultivated in Stellen-

bosch.)

368

Bothalia 15, 3 & 4 (1985)

FIG. 30. — Pelargonium scabrum. a, flowering branch, x 1; b,

petals, x 2; c, androecium, x 2; d, gynoecium, x 2; e, flower

with petals removed, x 1. (From Van der Walt 511, culti-

vated in Stellenbosch.)

to purple to pinkish-purple, anterior three petals

much narrower than two posterior ones, pedicel as

long as hypanthium.

P. scabrum has a relatively large distribution area.

It occurs from Steinkopf in the north-western Cape

southwards along the west coast to the south-west-

ern Cape and from here eastwards to the district of

Willowmore in the southern Cape. It usually grows

in mountainous habitats on well-drained, sandy soil

(Fig. 31). The western part of its distribution area

receives winter rains while the eastern part receives

rain throughout the year. Temperatures are high

during the summer months and some of its habitats

are covered with snow during the winter.

P. scabrum flowers from August to January al-

though the odd flower may be found throughout the

year.

P. scabrum is closely related to P. ribifolium Jacq.

and the differences between them are discussed

under the latter species. The leaves of P. scabrum

resemble those of P. citronellum J. J. A. v.d. Walt,

but the flowers of these two species differ markedly.

P. scabrum is a very variable species, especially as

far as the structure of the leaves is concerned. The

leaf segments of plants in the western Cape (Fig. 28)

are much longer and narrower than those in the

south-western and southern Cape (Fig. 30). The

form with extremely narrow leaf segments (Fig. 29)

was described as P. balsameum by Jacquin (1794),

and Harvey (1860) considered it as a variety of P.

scabrum. Plants in the western Cape have purplish

flowers, those in the south-western Cape pinkish-

purple flowers and those in the southern Cape white

flowers.

The extreme forms of P. scabrum differ to such an

extent that one is tempted to distinguish infraspecific

taxa. However, a detailed study of a large number of

specimens from the entire distribution area revealed

a continuous variation pattern of leaf and floral

characters. This makes a delimitation of infraspecific

taxa impossible.

P. scabrum hybridizes frequently in nature with

other species of the section Pelargonium, and even

with species of other sections (see list of hybrids at

the end of this paper).

Linnaeus originally described Geranium scabrum

in Centuria (1755). In this description the scabrous

nature of P. scabrum is not mentioned, but there is

nothing in the description which is not applicable to

P. scabrum. Linnaeus considered Geranium afri-

canum ... of Plukenet ( Phytographia : 169, t. 186,

fig. 5 (1692)) as a possible synonym of G. scabrum.

Unfortunately the identity of G. africanum . . . can-

not be determined from the illustration, but it does

not resemble P. scabrum.

It is strange that Linnaeus did not refer to the Cen-

turia publication in his description of G. scabrum in

Systema Naturae (1759). In this description the sca-

brous nature of the leaves is mentioned, leaving no

doubt about Linnaeus’s concept of the species.

FIG. 31 . — Known geographical distribution of Pelargonium sca-

brum.

CAPE PROVINCE. — 2917 (Springbok): 5 km S of Steinkopf

(-BC), Taylor 1203 (BOL); 25 km W of Springbok (-DB),

Acocks 19583 (PRE); near Nababiep (-DB), Van der Walt 1402

(STEU). 3017 (Hondeklipbaai): 6 km N of Bowesdorp (-BB),

Van der Walt 410 (STEU); Bowesdorp (-BB). Van der Walt 557

(STEU); Kamieskroon (-BB), Thorne s.n. (SAM); 4 km E of Ka-

mieskroon (-BB), Fischer 124 (STEU); 35 km S of Kamieskroon

(-BD), Hardy & Bayliss 1094 (PRE; Z). 3018 (Kamiesberg): 10

km E of Kamieskroon (-AA), Lavranos 15162 (PRE); Ellenboog

near Kamieskroon (-AA), Esterhuysen 23611 (BOL); Rooifon-

tein. Kamiesberg (-AB), Adamson 1518 (PRE); Kamiesberg

(-AC), Thorne s.n. (NBG); near Garies (-CA), Rodin 1430

(MO). 3118 ( Vanrhynsdorp): Vanrhynsdorp (-DA), Van Breda

Bothalia 15, 3 & 4 (1985)

369

1394 (PRE), Van der Wall 552 (STEU); Gifberg (-DC), Bayliss

601 (PRE), Esterhuysen 21957 (BOL; PRE), Phillips 7594 (BOL;

SAM), Werdermann & Oberdieck 523 (MO; PRE); Nardouw

Mountains (-DC), Barker 4752, 7444 (NBG); Bulshoek (-DD),

Compton 18841 (NBG), Leighton 2383 (BOL). 3119 (Calvinia):

Nieuwoudtville (-AC), Galpin 11130 (PRE), Van Son s.n.

(PRE); Vanrhyns Pass (-AC), Hall 962 (NBG), Hardy 77S(PRE;

SRGH), Taylor 2869 (NBG); Glen Ridge (-AC), Barker 9638

(NBG); Calvinia (-BD), Theiler 79 (PRE); Lokenburg (-CA),

Acocks 17232 (PRE), Story 4285 (GRA; PRE); Botterkloof

(-CD), Barker 10726 (MO; NBG), Salter 1657 (BOL). 3218

(Clanwilliam): Graafwater (-BA), Thode A1970 (MO); Clanwil-

liam (-BB), Rogers 16790 (Z); Pakhuis Pass (-BB). Schlieben &

Breda 9883 (SRGH); Boschkloof (-BB), Bond 530 (NBG);

Lambertshoekberg (-BC), Maguire 417 (NBG); Piquetberg

(-DA), Grant 3439 (MO), Guthrie 2560 (NBG), Martin 899

(NBG), Taylor 3905 (PRE), Theiler 5 (PRE); Mouton’s Vlei

(-DA), Marloth 11489 (PRE), Pillans 7284, 7334 (BOL); De

Hoek (-DA), Barker 2565 (NBG), Compton 19953 (NBG); Pie-

kenierskloof (-DB), Schlechter 4965 (BOL; Z); Grey’s Pass

(-DB), Ihlenfeldt 1037 (PRE), Steyn 366 (NBG); Versfeld Pass

(-DD). Drijfhout 1427 (STEU), Liebenberg 8354 (MO; PRE;

SRGH). 3219 (Wuppertal); Krakadouwsberg (-AA), Ecklon &

Zeyher s.n. (SAM), Middlemost 1859 (NBG); Algeria (-AC),

Story 2953 (PRE); Warm Baths near Citrusdal (-CA), Pearson &

Glover 7061 (PRE); near Citrusdal (-CA), Grant s.n. (MO); Kee-

rom (-CC), Pillans 8730 (PRE); Waboomsrivier (-CD), Ester-

huysen 8940 (BOL). 3318 (Cape Town): Paarlberg (-DB), Bolus

2602 (BOL; SAM), Volschenk 31 (STEU); Bothmaskop, Stellen-

bosch (-DD), Van der Walt 511 (STEU); Stellenbosch (-DD),

Prior s.n. (PRE; Z). 3319 (Worcester): Michell’s Pass (-AD), Bo-

lus 2602 (BOL), Esterhuysen 20722 (BOL), Guthrie 2403 (NBG),

Rehmann 2347 (Z), Van der Walt 1284 (STEU), Walgate 369

(NBG); Mostertshoek (-AD), Wasserfall 836 (NBG); Waaihoek

(-AD), Esterhuysen 22243 (BOL); Matroosberg (-BC), Phillips

1948 ( SAM); Hex River (-BD), Esterhuysen 21590 (BOL), Lamb

1299 (SAM), Tyson 7216 (PRE; SAM); Worcester (-CB), Mac

Owan s.n. (SAM), Marloth 2305 (PRE), Rehmann 2457 ( Z); Mal-

kopskloof (-CB), Esterhuysen 3379 (BOL; PRE); Brandwacht

Mountains (-CB), Acocks 15277 (PRE), Van Breda 751 (PRE);

Audensberg (-CB), Compton 9790 (NBG); Groot Drakenstein

Mountains (-CC), Esterhuysen 24030 (BOL); Sanddrift (-DA),

Schlieben & Van Breda 9930 (PRE; SRGH), Van der Walt 535

(STEU); Sandhills (-DA), Van der Walt 634 (STEU). 3320 (Mon-

tagu); Kogmanskloof (-CC). Ecklon & Zeyher 634 (SAM), Ester-

huysen 23808 (BOL); Montagu Baths (-CC), Compton 18360

(NBG), Lewis 1772 (SAM), Page 117 (PRE), Page s.n. (BOL).

3321 (Ladismith): Seven Weeks Poort (-AD), Barker 20597

(BOL), Bayliss 2442 (NBG; Z), Gillett 1701 (BOL). Phillips 1413

(SAM; Z), Schonken 147 (STEU), Van der Wall 630 (STEU),

Wells 3757 (GRA; PRE); near Hoeko (-AD), Wurts 1481

(NBG); Rooiberg Pass (-DA), Oliver 5302 (PRE); farm Lang-

berg (-DC), Van der Walt 837 (STEU); Cloete's Pass (-DC),

Goldblatt 4163 (MO; PRE). 3322 (Oudtshoorn): Prince Albert

(-AA), Bolus 10238 (NH), Marloth 4472 (PRE; STE), Tugwell 4

(BOL); Boomplaas (-AC), Van der Walt 403 (STEU); Swartberg

Pass (-AC), Hutchinson 1174 (BOL). Moffett 700 (STEU), Van

der Walt 727, 1311 (STEU); Schoemanshoek (-AD), Barker 67

(BOL), Hops 5 (BOL); Schoemanskloof (-AD), Venter 7443

(STEU); Meiringspoort (-BC), Esterhuysen 24885 (BOL),

Hafstrom & Acocks 750 (PRE), Van der Walt 704, 1141 (STEU);

Kamanassie Valley (-CA), Fourcade 3597 (PRE); 18 km E of

Oudtshoorn (-CB), Schonken 143 (STEU); Robinson Pass

(-CC), Van der Walt 1154 (STEU); Outeniqua Pass (-CD), Lewis

4370 (SAM); Klipdrift (-CD), Schlechter 2253 (BOL; Z); Camfer

(-CD), Schonken 206 (STEU), Van der Walt 676, 1309 (STEU);

Mannetjiesberg (-DB). Esterhuysen 6488 (BOL). 3323 (Willow-

more): Slypsteenberg (-AC), Schonken 138 (STELO; Georgida

(-AD). Esterhuysen 6377 (BOL; PRE); Modderfontein (-AD),

Andreae 1045 (PRE; STE); Avontuur (-CA), Bolus 2279 (BOL),

Fourcade 5851 (PRE); 8 km from Uniondale (-CA), Theron 1374

(BOL; PRE); Haarlem (-CB). Schonland 3083 (GRA; PRE);

Prince Alfred’s Pass (-CC), Schonken 135 (STEU). 3324 (Steyt-

lerville): Suuranysberg (-CD), Wurts 2062 (NBG). 3419 (Cale-

don): Botrivier (-AA), Welman 757 (BOL; GRA); near Ge-

nadendal (-BA), Baur 6607 (PRE). 3421 (Riversdale): Gourits

River Bridge (-BB), Galpin 3813 (GRA; PRE).

13. Pelargonium ribifolium Jacq., Icon. PI. Rar.

3: 11, t. 538 (1794); Jacq., Coll. Suppl. 5: 141 (1797);

Willd., Sp. PI. 3: 677 (1800); Pers., Syn. PI. 2: 232

(1806); Desf., Arb. 1: 460 (1809); Willd., Enum. 2:

708 (1809); Ait. f., Hort. Kew. edn 2,4: 177 (1812);

DC., Prodr. 1: 671 (1824); Spreng., Syst. Veg. 3: 61

(1826); Eckl. & Zeyh., Enum. 1: 79 (1835); Steud.,

Nom. Bot. edn 2,2: 289 (1841); Harv. in FI. Cap. 1:

305 (1860); Knuth in Pflanzenr. 4: 129: 459 (1912); J.

J. A. v.d. Walt & Vorster in J1 S. Afr. Bot. 46: 289

(1980); J. J. A. v.d. Walt & Vorster, Pelarg. S. Afr.

2: 121, fig. (1981). Type: Cape Province, ‘Ex Pro-

montorio bonae Spei,’ (W, holo.!, specimen with

Jacquin’s handwriting).

Geranium ribifolium (Jacq.) Poir., Encycl. Suppl. 2: 754

(1812). Geraniospermum ribifolium (Jacq.) Kuntze, Rev. Gen. 1:

95 (1891).

Pelargonium populifolium Eckl. & Zeyh., Enum. 1: 81 (1835);

Knuth in Pflanzenr. 4,129: 459 (1912). Type: Cape Province, '. . .

laterum montium “Zuurberge” prope “Enon" (Uitenhage) eo-

rumque ad “Langkloof” (Georg)’ Ecklon & Zeyher 632 (SI;

SAMI).

P. schonlandii Knuth in Bot. Jb. 44: 28 (1909); Knuth in Pflan-

zenr. 4,129: 470 (1912). Type: Cape Province, ‘Grahamstown

Kloof bei Grahamstown’, Schonland 568 (GRA!; NH!; PRE!;

Z!).

Erect, much-branched, aromatic shrub, up to 2 m

high and 1 m in diameter. Stems herbaceous when

young but soon becoming woody, hirtellous and

densely interspersed with glandular hairs, green.

Leaves sparsely hirtellous to hirtellous and densely

interspersed with glandular hairs, light green; lamina

3 (-5)-palmatilobate to 3 (-5)-palmatifid, base

mostly cordate, apices of lobes mostly obtuse (rarely

acute), margins dentate, (25-) 40-50 (-90) x (30-)

50-60 (-100) mm; petiole (20—) 40 (-100) mm long;

FIG. 32. — Pelargonium ribifolium. a, flowering branch, x 1; b,

petals, x 2; c, androecium, x 2; d, gynoecium. x 3; e, flower

with petals removed, x 2; f, mericarp, x 2. (From Ward-Hil-

horst 207, cultivated in Kirstenbosch.)

370

Bothalia 15, 3 & 4 (1985)

stipules triangular to cordiform, acuminate, 5-11 x

2-8 mm. Inflorescence: flowering branches profusely

branched, with normal and smaller foliar leaves; pe-

duncles 30-40 mm long, hirtellous and densely inter-

spersed with glandular hairs; involucral bracts lan-

ceolate, acuminate, indumentum as on peduncles,

5-7 x 1-2 mm; pseudo-umbels with 6-12 flowers

each. Pedicel 7-10 mm long, indumentum as on pe-

duncles. Hypanthium 6-8 mm long, prominently

thickened at the base. Sepals lanceolate, indumen-

tum abaxially as on peduncles, green with a reddish

tint, ca 10 x 2-5 mm. Petals white, posterior two

obovate to obcordiform, apices obtuse to emargi-

nate, with wine-red markings, reflexed at more than

90°, ca 17 x 11 mm; anterior three narrowly spathu-

late with narrow claws, occasionally with thin wine-

red lines near bases, reflexed at less than 90°, ca 13 x

3 mm. 2n=22. Fig. 32.

Diagnostic features

Erect, much-branched, aromatic shrub. Lamina 3

(-5)-palmatilobate to 3 (-5)-palmatifid, hirtellous,

apices of lobes mostly obtuse. Flowering branches

profusely branched. Pseudo-umbels with 6-12 flow-

ers each. Flowers white, anterior three petals much

narrower than posterior two, hypanthium promi-

nently thickened at base, pedicel as long as hypan-

thium.

P. ribifolium occurs in the eastern Cape Province,

where it is particularly common, as well as in the

southern Cape. It is found for more than 400 km

from the Swartberg Pass in the Oudtshoorn district

eastwards as far as the Katberg in the Fort Beaufort

district (Fig. 33). The distribution pattern is largely

correlated to the mountain ranges of the southern

and eastern Cape. P. ribifolium occurs in fairly moist

habitats, and this explains its distribution pattern as

the rainfall in the mountains is higher than in the

neighbouring lower-lying areas. A component of

moist ravine vegetation, which usually consists of

sclerophyllous shrub, the species is often associated

with the margins of ravine forest. It grows in direct

sunlight but is tolerant of shade and is frequently

found in partially or even completely shady situa-

FIG. 33. — Known geographical distribution of Pelargonium ribi-

folium.

tions. The soil is usually well-drained, sandy and

acid, often with a considerable amount of organic

matter. The plants probably have a fairly constant

supply of moisture throughout the year.

P. ribifolium flowers sporadically throughout the

year, but especially during the spring months of Sep-

tember to November, with a peak in October.

P. ribifolium resembles P. scabrum (L.) L’Herit.

in its habit, and there is some resemblance in the

shape of the leaves. The flowers of P. scabrum, how-

ever, are seldom white, and the thickening at the

base of the hypanthium is less conspicuous. Fur-

thermore, the lobes of the leaves are acute instead of

obtuse, as is mostly the case in P. ribifolium.

CAPE PROVINCE. — 3225 (Somerset East): Elim (-BB),

Schlechter 7736 (GRA; PRE; Z); Boschberg (-DC). Bolus 1752

(BOL), MacOwan 2214 (MEL; SAM). 3226 (Fort Beaufort):

Katberg (-DA), Dyer 791 (GRA; PRE). 3322 (Oudtshoorn):

Swartberg Pass, Wisura 1805 (NBG); Meiringspoort (-BC),

Haf strom & Acocks 750 (PRE). 3323 (Willowmore): Kouga

Mountains (-CB), Esterhuysen 4675 (BOL); Tsitsikamma Moun-

tains (-DC), Esterhuysen 22817 (BOL). 3324 (Steytlerville): Ba-

viaanskloof (-CA), Bayliss s.n. (PRE), Esterhuysen 24985 (BOL;

NBG; PRE); Baviaanskloof Mountains near Smitskraal (-CB),

Thompson 1934 (STE); Kouga River Poort (-CC), Fourcade 3080

(STE); Suuranys (-CD), Bayliss s.n. (MO; PRE; WAG); 5 km

from Kareedouw (-CD), Acocks 20036 (PRE); Cambria (-DA),

Fourcade 5191 (PRE); between Cambria and Andrieskraal

(-DA), Fourcade 2755 (MO; PRE); 58 km from Andrieskraal to

Billson (-DC), Marsh 1392 (PRE; STE); Hankey (-DD), Long

1351 (PRE). 3325 (Port Elizabeth): between Ann’s Villa and

Zuurberg Inn (-BB), Van der Walt 701 (STEU); Zuurberg (-BC),

Barker 4954 (NBG), Compton 20275 (BOL; NBG). Dr'ege 7447

(S), Ecklon & Zeyher 632 (S; SAM), Van der Walt 886 (STEU);

Sandrivier Reservation (-CA), Holland 3675 (BOL); Springfields

(-CB), Paterson 930 (GRA); Otterford (-CC), Schonken 126

(STEU); Van Staden’s Pass (-CC), Dahlstrand 2975 (GRA;

PRE), West 387 (BOL); Witteklip (-CC), Rodin 1041 (BOL;

MO); Loerie Plantation near Humansdorp (-CC), Dix 121 (BOL;

GRA); Longmore Forest Reserve (-CC), Long 1035 (GRA;

PRE); near Uitenhage (-CD), Ecklon & Zeyher 633 (S; SAM);

Groendal near Uitenhage (-CD), Long 1158 (PRE); Baakensri-

vier. Fern Glen (-DC), Olivier 1228 (MO; NBG). 3326 (Gra-

hamstown): Riebeek-Oos (-AA). Bayliss 3587 (NBG; Z). Swart-

watersberg (-AA), Acocks 12115 (PRE); Howieson’s Poort

(-AD), Britten 990 (GRA), Van der Walt 689 (STEU); Paradise

kloof near Grahamstown (-AD), Britten 2629 (GRA); between

Grahamstown and Assegaaibosch (-AD), Ecklon & Zeyher 206

(BOL; PRE; SAM); Grahamstown (-BC), Britten 5131, 5166

(Z), Daly & Sole 330 (Z), Dyer 2150 (GRA; PRE), Galpin 158

(GRA; PRE), Rogers 27315 (BOL; GRA; SAM; Z), Rogers

28676 (Z), Schonland 568 (GRA; NH; PRE; Z), Sidey 1362, 3056

(PRE), Van Dam 22094 (PRE). Van der Walt 978 (STEU). 3424

(Humansdorp): Mistkraal (-BB), Compton 23436 (NBG); Cape

St Francis (-BB), Bayliss s.n. (PRE).

14. Pelargonium citronellum, J.J.A. v.d. Walt in

S. Afr. J. Bot. 2,1: 76, fig. 5 (1983). Type: Cape

Province, between Riversdale and Herbertsdale

near farm Langberg, Britz s.n. sub STEU 1082

(PRE, holo.!; K!; NBG!; STEU!).

Erect, much-branched, strongly lemon-scented

shrub, up to 2m high and 1 m in diameter. Stems her-

baceous when young but soon becoming woody, hir-

sute and with numerous glandular hairs, green but

becoming brownish with age. Leaves sparsely hirsute

to hispid and with numerous glandular hairs inter-

spersed, green; lamina palmatifid to palmatisect

with segments sometimes irregularly incised, con-

spicuously veined abaxially, base cuneate to cordate,

apices of lobes acute, margins irregularly serrate-

dentate, (35-) 50 (-110) x (25-) 55 (-130) mm; pe-

Bothalia 15, 3 & 4 (1985)

371

tiole (15-) 35 (-60) mm long; stipules narrowly tri-

angular to triangular, 6-10 x 3-6 mm. Inflorescence:

flowering branches profusely branched, with normal

and smaller foliar leaves; peduncles 20-50 mm long,

hirsute and with glandular hairs; involucral bracts

narrowly ovate to ovate, apiculate, indumentum as

on peduncles, 6-8 x 3-4 mm; pseudo-umbels with

5-6 (-8) flowers each. Pedicel 5-15 mm long, indu-

mentum as on peduncles. Hypanthium 3-8 mm long.

Sepals lanceolate, indumentum abaxially as on pe-

duncles, green with a reddish-brown tint, ca 10 x

2-4 mm. Petals pinkish-purple; posterior two spa-

thulate to obovate, apices obtuse to emarginate,

with dark purple markings, reflexed at ca 90°, ca 20

x 8 mm; anterior three spathulate with narrow

claws, reflexed at less than 90°, ca 20 x 4 mm.

2n=22. Fig. 34.

FIG. 34. — Pelargonium citronellum. a, flowering branch, x 1; b,

petals, x 1,5; c, androecium, x 1,5; d, gynoecium, x 2.

(From Van der Walt 1296, cultivated in Stellenbosch.)

Diagnostic features

Erect, much-branched, strongly lemon-scented

shrub. Lamina palmatifid to palmatisect, sparsely

hirsute to hispid. Flowering branches profusely

branched. Pseudo-umbels with 5-8 pinkish-purple

flowers each, pedicel usually longer than hypan-

thium.

P. citronellum is apparently confined to the one

degree square which includes the town of Ladismith

in the southern Cape (Fig. 35). It is common on the

northern foothills of the Langeberg Range between

Muiskraal and Flerbertsdale. As the case with many

other representatives of the section Pelargonium, it

is usually found near streams in well-drained sandy

soil. The distribution area of P. citronellum receives

most of its rainfall during the winter months.

P. citronellum flowers between August and Janu-

ary with a peak in September and October.

P. citronellum exhibits many leaf characters which

are intermediate between those of P. scabrum (L.)

L’Herit, and P. hispidum (L.f.) Willd. The leaves of

P. citronellum are less scabrous and more strongly

lemon-scented than those of P. scabrum. Those of

P. hispidum are aromatic but not lemon-scented,

and less scabrous, than those of P. citronellum.

Most of the herbarium specimens of P. citronel-

lum collected by previous workers, have been identi-

fied as P. scabrum. This error could be ascribed to

the similarity which exists between the leaves of the

two species. However, the flowers of these two

species differ considerably. The flowers of P. citro-

nellum are much larger and darker in colour than

those of P. scabrum. The latter is a very variable

species with a large distributional range stretching

from the north-western Cape to the southern Cape.

The southern Cape form of P. scabrum resembles P.

citronellum much more than the western Cape form,

and it is possible that P. citronellum evolved from

this form. P. scabrum is also known as a species

which hybridizes easily with other species, and it is

therefore also possible that P. citronellum originated

as a natural hybrid between P. scabrum and P. hispi-

dum. A natural hybrid between P. citronellum and

P. hispidum has been recorded (Van der Walt) 1124

in STEU).

FIG. 35. — Known geographical distribution of Pelargonium cit-

ronellum.

CAPE PROVINCE. — 3321 (Ladismith): Buffelskloof near

Ladismith (-AC), Esterhuysen 14008, 26833 (BOL); Seven

Weeks Poort (-AD), Esterhuysen 24745 (BOL), Liebenberg 714

(PRE), Wells 3759 (PRE); Waterkloof near Ladismith (-AD),

Hutchinson 1113 (BOL; K), Van der Walt 1120 (STEU); Muis-

kraal (-CC), Muir 2730 (PRE), Van der Walt 1296 (STEU); 12

km E of Muiskraal (-CD), Van der Walt 716 (STEU); Farm Lang-

berg (-DC), Britz s.n. (K; NBG; PRE; STEU), Van der Walt 835

(STEU).

372

Bothalia 15, 3 & 4 (1985)

15. Pelargonium sublignosum Knuth in Pflan-

zenr. 4,129: 460 (1912). Type: ‘Sudwestliche Kap-

provinz, Michell’s Pass 660 m’, Schlechter 9976 (Bt,

holo.; Z, lecto.!; BM!; BOL!; G!; GRA!; K!; P!;

PRE!).

Erect, branched, strongly sweet-scented or non-

aromatic shrub, up to 1 m high and 0,5 m in di-

ameter. Stems herbaceous when young, becoming

woody with age, sparsely to densely villous inter-

spersed with glandular hairs and often shorter

thicker hairs, green but becoming reddish-brown

with age. Leaves sparsely hirtellous to villous (long

soft hairs predominantly on veins of lamina) inter-

spersed with glandular hairs, green to dull green;

lamina cordiform in outline, 3-(5)-lobed, base cor-

date, apices of lobes obtuse to acute, margins den-

tate to coarsely dentate and often reddish, (20-) 35

(-60) x (25-) 40 (-80) mm; petiole (10-) 20 (-60)

mm long; stipules ovate to cordiform, often apicu-

late, 5-10 x 3-10 mm. Inflorescence: flowering

branches with normal and smaller foliar leaves; pe-

duncles 20-40 mm long, sparsely villous to villous

interspersed with glandular hairs and often shorter

thicker hairs; involucral bracts lanceolate to cordi-

form, apiculate, glabrescent but margins with long

hairs, 5-6 x 3^1 mm; pseudo-umbels with 3-9 flow-

ers each. Pedicel 4-10 mm long, with glandular

hairs, short hairs and sparsely interspersed with long

hairs. Hypanthium 5-14 mm long, prominently

thickened at base, indumentum as on pedicel. Sepals

lanceolate, apiculate, indumentum abaxially as on

pedicel, green with white margins, ca 9 x 3^1 mm.

rr\.

FIG. 36. — Pelargonium sublignosum. a, flowering branch, x 1;

b, petals, x 1,5; c, androecium, x 2; d, gynoecium, x 3.

(From Van der Wall 591, cultivated in Stellenbosch.)

Petals white, light pink or pink; posterior two spa-

thulate to obovate, with feather-like pinkish-purple

markings, reflexed at ca 90°, ca 18 x 4-7 mm; ante-

rior three narrowly spathulate with short narrow

claws, slightly reflexed, ca 15 x 2 mm. 2n=22. Fig.

36.

Diagnostic features

Erect, branched, sometimes strongly sweet-

scented shrub. Stems villous becoming reddish-

brown with age. Lamina cordiform in outline, 3- (5)-

lobed, sparsely hirtellous to villous, margins often

reddish. Pseudo-umbels with 3-9 flowers. Flowers

white, light pink or pink, pedicel usually slightly

shorter than hypanthium.

P. sublignosum occurs in a relatively small area in

the south-western Cape. It is confined to mountain-

ous habitats and so far it has only been collected on

Piquetberg and the mountain ranges between Por-

terville and Ceres (Fig. 37). This area falls entirely in

the winter rainfall region. It grows in ravines or in

other moist habitats, often between rocks. The soil

in these situations consists of rather coarse, white

sand. It is adapted to grow in a wide range of tem-

peratures. Very high temperatures are experienced

during summer and snow often occurs during winter.

P. sublignosum flowers from October to February

with a peak in November and December. The odd

flower is found throughout the year.

P. sublignosum exhibits a high degree of variation

as far as fragrance and indumentum are concerned,

especially of the stems and leaves, and the structure,

colour and size of the flowers. This leaves the im-

pression that it is possibly a relatively young species

in the final stages of speciation. The leaves of P. sub-

lignosum resemble those of P. scabrum (L.)

L’Herit., and it is possible that it evolved from this

species. Natural hybrids between P. sublignosum

and P. scabrum have been observed on the farm

Grootfontein near Porterville. Hybrids between P.

sublignosum and P. grandiflorum (Andr.) Willd.

have also been reported from this area. The habit,

leaves and indumentum of P. sublignosum show

some resemblance to those of P. scabroide Knuth

FIG. 37. — Known geographical distribution of Pelargonium sub-

lignosum.

Bothalia 15, 3 & 4 (1985)

373

and P. englerianum Knuth, which may be indicative

of a relationship between these species.

CAPE PROVINCE. — 3218 (Clanwilliam): Piquetberg

(-DA), Edwards 215 (Z). 3219 (Wuppertal): Elandskloof (-CA),

Smuts & Gillett 3479 (BOL); Hexberg (-CA), Esterhuysen 18430

(BOL); Olifants River Mountains (-CC), Esterhuysen 15339

(BOL); Twenty Four Rivers Mountains (-CC), Esterhuysen

21906 (BOL; K; PRE); Farm Grootfontein near Porterville

(-CC), Van der Walt 591, 905 (STEU). 3319 (Worcester): Neeth-

lingsberg, Witzenberg (-AC), Esterhuysen 22513 (BOL; PRE);

Michell’s Pass near Ceres (-AD), Compton 10061 (NBG), Ester-

huysen 14688 (BOL; PRE), Schlechter 9976 (BM; BOL; G;

GRA; K; P; PRE; Z), Thode A2230 (PRE), Van der Walt 1285

(STEU); Ceres (-AD), Barker 8107 (NBG), Esterhuysen 28416

(BOL), Rogers 17600 (Z), Van der Walt 581, 645 (STEU).

16. Pelargonium scabroide Knuth in Reprium

nov. Spec. Regni veg. 18: 293 (1922). Syntypes:

Cape Province, ‘Gydow bei Ceres, auf den

Abhangen der Skurfdeberg’, Bolus 1115 (BOL,

lecto!; G!; SAM!; UPS!; W!); Schlechter 9988 (BM!;

BOL! p.p.; G! p.p.; K!; PRE! p.p. ; S! p.p.; Z! p.p.);

‘Constantiaberg’, Schlechter 877 (G!; P!; WU!; Z!).

Erect to decumbent, branched, non-aromatic sub-

shrub, up to 0,75 m high and 0,5 m in diameter.

Stems herbaceous when young, becoming woody

with age, villous to hispid and with small glandular

hairs in between, wine-red but becoming greyish-

brown with age. Leaves hard to the touch (almost

scabrous), green; lamina cordiform in outline, 3-pal-

matisect to pinnatisect with the segments irregularly

incised, adaxially glabrescent, abaxially strigose to

hispid and with small glandular hairs and sometimes

long soft hairs in between, base cordate, apices of

segments acute, margins coarsely serrate-dentate

and wine-red, (12-) 25 (-45) x (15-) 30 (-55) mm;

petiole ( 10 — ) 30 (-70) mm long, indumentum as on

stems, wine-red; stipules triangular to cordiform,

usually caudate, glabrescent but margins ciliate, ca 5

x 4 mm. Inflorescence: flowering branches with nor-

mal and smaller foliar leaves; peduncles 10-35 mm

long, glabrescent to sparsely hispid with small glan-

dular hairs and sometimes long soft hairs in be-

tween; involucral bracts lanceolate to cordiform,

acuminate, glabrescent to hispid but margins ciliate,

4-7 x 2-3 mm; pseudo-umbels with 3-5 flowers

each. Pedicel 4—10 mm long, sparsely hispid with

small glandular hairs and sometimes long soft hairs

in between. Hypanthium 6-10 mm long, promi-

nently thickened at base, indumentum as on pedicel.

Sepals lanceolate, apiculate, indumentum abaxially

as on pedicel, green but becoming wine-red with

age, ca 7 x 1,5-3 mm. Petals white to pale pink-pur-

ple; posterior two spathulate, with feather-like dark

pink to wine-red markings, reflexed at ca 90°, ca 17

x 5 mm; anterior three narrowly spathulate with

narrow claws, with a pink streak, reflexed at ca 45°,

ca 12 x 2 mm. 2n=22. Fig. 38.

Diagnostic features

Erect to decumbent, branched, non-aromatic sub-

shrub, young stems wine-red, villous to hispid.

Lamina 3-palmatisect to pinnatisect with the seg-

ments irregularly incised, margins coarsely serrate-

dentate and wine-red, almost scabrous. Pseudo-um-

bels with 3-5 flowers each. Flowers white to pale

pink-purple with darker markings, pedicel as long as

hypanthium.

FIG. 38. — Pelargonium scabroide. a, flowering branch, x 1; b,

petals, x 2; c, androecium, x 2,5; d, gynoecium, x 3. (From

Ward-Hilhorst 116, collected at Gydouw Pass.)

P. scabroide has a relatively small distribution

area. So far it has only been collected on the moun-

tains between Porterville and Touws River (Fig. 39).

It grows on coarse sandy soil in the shade of over-

hanging sandstone boulders which create rather

moist habitats. The distribution area is situated en-

tirely in the winter rainfall region. Snow falls occur

regularly during the winter and the summers are

very hot and dry.

Contrary to most species of the section Pelargo-

nium which usually flower from spring to early sum-

mer, P. scabroide flowers during the hot summer

months. Its flowering period has been recorded as

from November to February.

The wine-red young stems with long hairs and

typical leaf structure of P. scabroide distinguish it

from other species of the section. P. scabroide has

many characters in common with P. sublignosum

Knuth and P. englerianum Knuth, and they un-

doubtedly are closely related. It is, however, easy to

distinguish between these three species by leaf

characters.

Nine specimens of Schlechter 9988 (cited by Knuth

as a type) have been studied. Two of these speci-

mens consist of P. scabroide material only, and five

consist of a mixture of P. scabroide and P. engleria-

num. The last two specimens, housed in MO and W,

consist entirely of P. englerianum material.

The locality of Schlechter 877 (also cited by Knuth

as a type), is given as ‘Constantiaberg’. This locality

on the Cape Peninsula is doubtful, and it can most

374

Bothalia 15, 3 & 4 (1985)

FIG. 39. — Known geographical distribution of Pelargonium sca-

broide.

probably be ascribed to an error on the part of Sch-

lechter: apparently Schlechter did not keep a regis-

ter for specimens below No. 1700 (Jessop, 1964).

CAPE PROVINCE. — 3319 (Worcester): farm Suurvlakte

near Porterville (-AA). Van der Walt 1071 (STEU); Groot Win-

terhoek (-AA). Van der Walt 1077 (STEU); between Rosendal-

fontein and Visgat (-AA), Pillans 9664 (BOL); Visgat near Wit-

zenberg (-AA), Esterhuysen 13433 (BOL); east slopes of Witzen-

berg (-AA), Esterhuysen 28405 (BOL), Pillans 9634 (BOL);

Skurfteberg (-AB), Taylor 8670 (PRE); Gydo (-AB), Bolus 1115

(BOL; G; S; SAM; UPS; W), Schlechter 9988 (BM; BOLp.p.; G

p.p.; K; PRE p. p. ; S p.p.; Z p.p.). Van der Walt 1288 (STEU);

between Gydo and Witzenberg (-AB), Isaac s.n. (BOL), Van der

Walt 578 ( STEU); Bokkerivier Farms (-BD). Booysen62 (NBG).

17. Pelargonium englerianum Knuth in Pflan-

zenr. 4, 129: 470 (1912). Type: Cape Province,

Prince Albert, Swartberg Pass, Bolus 11452 (BOL,

holo.!).

Erect to decumbent, much-branched, rose-cam-

phor-scented subshrub, up to 1 m high and 0,75 m in

diameter. Stems herbaceous when young, becoming

woody with age, often long and slender, villous and

densely interspersed with glandular hairs, green but

soon becoming purplish and eventually brown.

Leaves hispid and densely interspersed with glandu-

lar hairs, green; lamina reniform, sometimes 3-pal-

matilobate to 3-palmatisect, more or less crisped,

base truncate to cordate, apices of lobes obtuse to

acute, margin coarsely dentate, (5-) 15 (-35) x (5-)

20 (-40) mm; petiole (3-) 15 (-40) mm long, indu-

mentum as on stems; stipules cordiform, often api-

culate, ca 4 x 4 mm. Inflorescence: flowering

branches with normal and smaller foliar leaves; pe-

duncles 10-40 mm long, villous to hispid with glan-

dular hairs interspersed; involucral bracts ovate to

cordiform, apiculate, 4-5 x 3-4 mm; pseudo-umbels

with 2-5 flowers each. Pedicel 6-20 mm long,

sparsely hispid to hispid and with glandular hairs in-

terspersed. Hypanthium 1-10 mm long, prominently

thickened at the base. Sepals narrowly ovate to lan-

ceolate, indumentum abaxially as on pedicel, green

with white margins, ca 6-10 x 2-4 mm. Petals white

to pinkish-purple; posterior two spathulate, with

feather-like dark purple markings, reflexed at ca 90°,

FIG. 40. — Pelargonium englerianum. a, flowering branch, x 1;

b, petals, x 2; c, androecium, x 2; d, gynoecium, x 4. (From

Stirton 9456, cultivated in Stellenbosch.)

10-18 x 4—8 mm; anterior three oblanceolate with

narrow claws, slightly reflexed, 7-12 x 2-4 mm.

2n=22. Fig. 40.

Diagnostic features

Erect to decumbent, much-branched, rose-cam-

phor-scented subshrub, young stems purplish and

villous. Lamina reniform, more or less crisped, his-

pid. Pseudo-umbels with 2-5 flowers each. Flowers

white to pinkish-purple, relatively small, pedicel

longer than hypanthium.

P. englerianum occurs from Lokenburg near Cal-

vinia south-eastwards to the Swartberg Pass near

Prince Albert (Fig. 41). It is extremely plentiful in

the Cedarberg Mountains where it grows on coarse

sandy soil, often in the shade of overhanging sand-

stone boulders. The distribution area falls entirely in

the winter rainfall region. The summers are very hot

and dry, and snowfalls occur regularly during the

winter months.

P. englerianum flowers from August to April with

a marked peak during the summer months Novem-

ber-January.

In the past, most herbarium specimens of P. engle-

rianum have been erroneously identified as P.

crispum (Berg.) L'Herit. These errors can be as-

cribed to the similarities which exist between the

leaves of the two species. The differences between

P. englerianum and P. crispum are tabulated under

the latter species.

Bothalia 15, 3 & 4 (1985)

375

FIG. 41. — Known geographical distribution of Pelargonium eng-

lerianum.

Floral and vegetative characteristics indicate a re-

lationship between P. englerianum , P. scabroide

Knuth and P. sublignosum Knuth.

CAPE PROVINCE. — 3119 (Calvinia): Lokenburg (-CA),

Acocks 17554 (PRE). 3218 (Clanwilliam): Boskloof, Cedarberg

Mountains (-BB), Esterhuysen 7594 (BOL; PRE); Clanwilliam

(-BB), Pocock 58 (PRE). 3219 (Wuppertal); Krakadouw Peak

(-AA), Esterhuysen 7499 (BOL), Stokoe s.n. (SAM); road to

Heuningvlei (-AA). Esterhuysen 21128 (BOL; NBG), Middle-

most 1907 (NBG). Stokoe s.n. (SAM); Scorpion’s Poort (-AA).

Esterhuysen 7527 (BOL); Pakhuis Pass (-AA), Esterhuysen 3381 ,

7408, 21153, 21926 (BOL), Galpin 11207 (PRE), Wisura 746

(NBG); Uitkyk Pass (-AC), Leipoldt s.n. (BOL), Van der Walt

1418 (STEU); Cedarberg Mountains (-AC), Stokoe 7315 (BOL),

Stokoe s.n. (SAM); Tafelberg (-AC), Barnard s.n. (SAM); Se-

derhoutkloof (-AC), Esterhuysen 20047 (BOL; NBG), Taylor

6183 (PRE); Duikerfontein (-AC), Schonken 82 (STEU); Alge-

ria (-AC), Esterhuysen 7167 (BOL); Middelburg (-AC); Acocks

19856 (PRE), Esterhuysen 2784, 7166, 7280 (BOL), Stokoe s.n.

(SAM); Maltese Cross (-AC), Esterhuysen 17994 (BOL); Drie-

hoek Valley (-AC), Esterhuysen 22421 (BOL; PRE); Hoogver-

toon (-AC), Haynes 1204 (PRE; STE); Wuppertal (-AC), Thode

A1969 (PRE); Wolfberg (-AD), Esterhuysen 22425 (BOL; PRE);

Elandskloof (-CA), Lewis 22058 (BOL); Grootberg (-CA). Es-

terhuysen 4150 (BOL); Duivelskop (-CA), Stokoe s.n. (SAM);

Donkerkloofkop (-CA), Stokoe s.n. (SAM); Sandfontein (-CB),

Rycroft 2642 (NBG), Van der Walt 1409 (STEU); Zuurvlakte

(-DC), Esterhuysen 12708 (BOL). 3319 (Worcester); Rosendal-

fontein (-AA), Pillans 9663 (BOL; PRE); nearGydo (-AB), Van

der Walt 644 (STEU); Agter-Witsenberg (-AB), Marais 46

(STEU); Koue Bokkeveld (-AB), Adamson 1566 (PRE);

Skurfteberg (-AB). Edwards 87 (BOL); Buffelshoek Pass (-AC),

Marsh 46 (PRE); near Prince Alfred Hamlet (-AD), Marais 39

(STEU); Ceres (-AD) Rogers 17575 (Z); Ertjieslandkloof

(-AD), Leighton 2270 (BOL); Valsgatkloof (-AD), Esterhuysen

1531 (BOL); Merino (-AD), Cillies.n. (STEU); Katbakkies Pass

(-BA), Taylor 6097 (PRE); Roodeberg (-BC), Compton 8379

(NBG), Esterhuysen 20917 (BOL); Matroosberg (-BC), Ester-

huysen 28628 (BOL), Phillips 1952 (SAM); Orchard (-BC), Es-

terhuysen 10907 (BOL); near Karoopoort (-BC), Stokoe s.n.

(SAM); near Osplaats (-BC), Rogers 16720 (BOL; Z); Hex River

Valley (-BD), Tyson 738 (BOL; MO; SAM); Bonteberg (-BD),

Compton 9943 (NBG), Esterhuysen 3663, 3723 (BOL; NBG);

Slanghoek Needle (-CA), Esterhuysen 17774 (BOL); between

McGregor and Stormsvlei (-DD), Esterhuysen 4302 (BOL). 3320

(Montagu): Tweedside (-AB), Marloth 10805 (PRE); Witteberg

(-AD), Esterhuysen 28877 (BOL); Fisantekraal (-BC), Compton

21117 (BOL; NBG). 3321 (Ladismith); Rooiberg (-CB), Oliver

5414 (STE). 3322 (Oudtshoorn): Swartberg Pass (-AC), Bolus

11452 (BOL), Moffett 701 (STEU).

18. Pelargonium greytonense J. J. A. v.d. Wall

in J1 S. Afr. Bot. 3:256-258 (1984) Type: Cape

Province, Happy Valley near Greyton, Esterhuysen

20755 (BOL, holo.!; K! ; PRE!).

An erect, much-branched, non-aromatic to aro-

matic shrub, up to 1 m high and 0,75 m in diameter.

Stems herbaceous when young but soon becoming

woody, hirtellous and with a few long soft hairs and

glandular hairs interspersed, green but becoming

greyish-brown with age. Leaves hirtellous and with

long hairs and many glandular hairs interspersed,

green; lamina cordiform in outline, shallowly

3-(5-8)-palmatilobate to palmatipartite, conspicu-

ously veined, base cordate, apices of lobes obtuse,

margin coarsely dentate, (15-) 35 (-60) x (20-) 40

(-70) mm; petiole (20—) 50 (-80) mm long; stipules

cordiform to triangular, apiculate to cuspidate, 4—8

x 4-6 mm. Inflorescence: flowering branches with

normal and smaller foliar leaves, peduncles 10-70

mm long, hirtellous to hirsute and densely inter-

spersed with glandular hairs; involucral bracts ovate

to narrowly ovate, cuspidate, indumentum as on pe-

duncles, 4-8 x 4-5 mm; pseudo-umbels with 2-9

flowers each. Pedicel 8-20 mm long, relatively thin,

indumentum as on peduncles. Hypanthium 3-8 mm

long, prominently thickened at the base. Sepals 5,

lanceolate, cuspidate, indumentum abaxially as on

peduncles, green, ca 12 x 2—4 mm. Petals 5, white to

pale pink; posterior two spathulate to obovate,

apices obtuse, with dark red markings, reflexed at

more than 90°, ca 20 x 8 mm; anterior three nar-

rowly spathulate with short claws, reflexed at less

than 90°, ca 18 x 2-3 mm. 2n=22. Fig. 42.

FIG. 42. — Pelargonium greytonense. a, flowering branch, x 1; b.

petals, x 2; c, androecium, x 2; d, gynoecium, x 3; e, flower

with petals removed, x 1. (From Van der Walt 520. culti-

vated in Stellenbosch.)

376

Bothalia 15, 3 & 4 (1985)

Diagnostic features:

Erect, much-branched, non-aromatic to aromatic

shrub. Lamina cordiform, shallowly 3-(5-8)-pal-

matilobate to palmatipartite, hirtellous and with

long hairs interspersed. Pseudo-umbels with 2-9

flowers each. Flowers white to pale pink, anterior

three petals much narrower than posterior two, ped-

icel relatively thin and longer than hypanthium.

P. greytonense flowers between September and

January with a peak in October and November.

P. greytonense is restricted in distribution to a

small area in the south-western Cape (Fig. 43). It is

common on the southern slopes of the Riviersonder-

end Mountains where it is often found in ravines.

The distribution area receives rain predominantly

during the winter months and is frost free because of

its close proximity to the coast. High temperatures

are experienced during the dry summer months. Al-

though restricted in distribution, it occurs locally in

very large numbers.

P. greytonense exhibits considerable morphologi-

cal variation especially of leaf characteristics, leaving

the impression of a relatively young species. It is,

however, already a well-established species produc-

ing large quantities of viable seed, and it is probably

just a matter of time before its distribution range will

be expanded.

P. greytonense is probably of hybrid origin with

the sympatric P. hermanniifolium (Berg.) Jacq. and

P. papilionaceum (L.)L’Herit. as possible parent

species. Many characters of P. greytonense, such as

the shape and indumentum of the leaves, are inter-

mediate between these two species. There is also

evidence of backcrossing between P. greytonense

and P. hermanniifolium. The habitat requirements

of P. greytonense are also between those of the two

putative parent species: P. papilionaceum growing in

a rather moist, semi-shaded habitat and P. herman-

niifolium in drier situations with direct sunlight.

It should be noted that a cytogenetic study did not

readily confirm the proposed hybrid origin of P.

greytonense. P. hermanniifolium is a diploid species

(2n = 22) and P. papilionaceum a tetraploid species

(2n=44). One would expect P. greytonense to be a

triploid species, but it is in fact also a diploid species

(2n = 22) with no meiotic aberrations. This could be

explained by the assumption that an original diploid

form of P. papilionaceum had been involved.

Some herbarium specimens of P. greytonense have

been determined as P. semitrilobum Jacq. No type

specimen of P. semitrilobum could be traced and the

drawing in Jacq., Hort. Schoenbr. 2: t . 130 (1797)

should be considered as the iconotype of this

species. The shape of the leaves, as depicted in the

drawing, shows some resemblance to those of P.

greytonense. However, the indumentum of the

leaves and the shape of the posterior petals in the

drawing differ completely from those of P. greyto-

nense. I am of the opinion that P. semitrilobum is a

hybrid, most probably an artificial one.

CAPE PROVINCE. —3419 (Caledon): Caledon (-AB), Prior

s.n. (PRE); Swartberg, Caledon (-AB), Zey her 2091 (MEL; W);

Steenbok River (-AD), Schlechler 9780 (Z); Greyton (-BA), Es-

FIG. 43. — Known geographical distribution of Pelargonium

greytonense.

terhuysen 20755 (BOL; K; PRE), Van der Walt 708, 801 (STEU),

Van der Walt & Vorster 1317, 1318, 1319, (STEU); Genadendal

(-BA). Van der Walt 1099 (STELI); Olifantsboskloof nearTyger-

hoek (-BB), Van der Walt 520 (STEU); Riviersonderend Moun-

tains (-BB). Ecklon & Zey her 635 (SAM), 2090 (SAM; Z), Ester-

huysen 18776 (BOL), Leighton s.n. (BOL), Stokoe s.n. (SAM);

near Oubos (-BD). Van der Walt 710. 711 (STEU). 3420 (Bredas-

dorp): Zandfontein (-AA), Galpin 3817 (PRE).

19. Pelargonium crispum (Berg.) L'Herit. in

Ait., Hort. Kew. edn 1,2: 430 (1789); L’Herit., Ge-

ran.: t. 32 ( 1792); Salisb., Prodr.: 316 ( 1796); Willd. ,

Sp. PI. 3: 682 (1800); Pers., Syn. PI. 2: 233 (1806);

Willd., Enum. 2: 709 (1809); Desf., Arb. 1: 464

(1809); Ait. f., Hort. Kew. edn 2,4; 180 (1812); DC.,

Prodr. 1; 677 (1824); Eckl. & Zeyh., Enum. 1: 81

(1835); Harv. in FI. Cap. 1: 304 (1860); Knuth in

Pflanzenr. 4,129: 469 (1912); Compton in Trans.

Roy. Soc. S. Afr. 19: 294 (1931). Type: Cape Prov-

ince, ‘e Cap. b. Spei’, Grubb s.n. (SBT 138, holo.!).

Geranium crispum Berg., Descr. PI. Cap.: 176 (1767); L..

Mant.: 257 (1771); Cav.^Diss. 4: 252, t. 109 (1787); Thunb.,

Prodr. 1: 115 (1794); Thunb.. FI. Cap. edn 2: 525 (1823).

Pelargonium crispum (Berg.) L’Herit. var. latifolium L'Herit.,

Geran.: t. 33 (1792) nonr. nud.; ex Harv. in FI. Cap. 1: 304

(1860). Iconotype: L'Heritier, Geran.: t. 33 (1792).

P. crispum (Berg.) L'Herit. var. majus DC., Prodr. 1: 677

(1824). Iconotype: L'Heritier, Geran.: t. 32 (1792).

Erect to decumbent, much-branched, strongly

lemon-scented subshrub or shrub, up to 0,75 m high

and 0,5 m in diameter. Stems herbaceous when

young but soon becoming woody, densely pubescent

to strigose with glandular hairs interspersed, green

but soon becoming brownish. Leaves strigose and

densely interspersed with glandular hairs, green;

lamina reniform, 3-palmatilobate to 3-palmatisect,

crisped, base cordate, apices of lobes obtuse to

acute, margins coarsely dentate-serrate, (2-) 5 (-10)

x (3-) 7 (-15) mm; petiole (0,5-) 4 (-15) mm long;

stipules cordiform, often apiculate, 2-4 x 2-5 mm.

Inflorescence: flowering branches with normal and

smaller foliar leaves; peduncles 5-10 mm long, pu-

bescent to strigose with glandular hairs interspersed;

involucral bracts ovate to triangular, indumentum as

on leaves, 3-4 x 1,5-2, 5 mm; pseudo-umbels with

1-2 (-3) flowers each. Pedicel 2-7 mm long, sparsely

Bothalia 15, 3 & 4 (1985)

377

strigose and densely interspersed with glandular

hairs. Hypanthium 5-8 mm long, indumentum as on

pedicel. Sepals lanceolate, indumentum abaxially as

on pedicel, green but sometimes with a reddish tint,

ca 8 x 2-4 mm. Petals white to dark pink or almost

purple; posterior two broadly spathulate, apices

sometimes emarginate to cleft, with dark red to dark

purple markings, reflexed at ca 90°, ca 18 x 10 mm;

anterior three spathulate with narrow claws, slightly

reflexed, ca 15 x 3 mm. 2n=22. Fig. 44.

P. crispum flowers from August to April with a

marked peak in spring (September-October).

The relationship between P. crispum and P. her-

manniifolium (Berg.) Jacq. is discussed under the

latter species. The leaves of P. crispum can also be

confused with those of P. englerianum Knuth. Dif-

ferences between these two species are tabulated:

P. crispum

1. Leaves lemon-

scented

2. Leaves strigose.

3. Lamina crisped and

relatively small.

4. Young stems green.

5. Flowers relatively

large.

P. englerianum

1. Leaves rose-

camphor-scented.

2. Leaves hispid.

3. Lamina less crisped

and larger.

4. Young stems purplish.

5. Flowers relatively

small.

FIG. 45. — Known geographical distribution of Pelargonium

crispum.

FIG. 44. — Pelargonium crispum. a, flowering branch, x 1; b,

leaf and stipules, x 3; c, petals, x 2; d, androecium, x 2; e,

gynoecium, x 4. (From Van der Walt 612, cultivated in Stel-

lenbosch.)

Diagnostic features

Erect to decumbent, much-branched, strongly

lemon-scented subshrub or shrub. Lamina reniform,

3-palmatilobate to 3-palmatisect, crisped, strigose,

base cordate. Pseudo-umbels with 1-2 (-3) flowers

each. Flowers white to dark pink or almost purple,

anterior petals much narrower than posterior ones,

pedicel usually shorter than hypanthium.

P. crispum occurs in the south-western part of the

Cape Province. It is confined to the one degree

squares which include the towns of Worcester, Mon-

tagu and Bredasdorp (Fig. 45). Its distribution area

falls entirely in the winter rainfall region, but is usu-

ally found on the lower slopes of mountains or on

hills where the rainfall is relatively low in compari-

son with habitats higher up on the mountains. It of-

ten grows on sandy soil in the shelter of sandstone

boulders. Temperatures are very high during the

summer and frost occurs during the mid-winter

months.

CAPE PROVINCE. — 3319 (Worcester): Molenaarsberg

(-CA). Esterhuysen 14071 (BOL); Brandwacht (-CB), Van Breda

258 (PRE); Walters 164 (NBG); Veld Reserve (-CB), Van Breda

4 (PRE); near Worcester (-CB), Cooper 1704 (E; NH; PRE; Z);

Brandvlei (-CB), Rehmann 2380 (Z); Du Toitskloof (-CC),

Compton 20105, 22835 (NBG). Esterhuysen 14352, 18919a

(BOL); Wemmershoek (-CC), Barker 275 (NBG), Esterhuysen

9119 (BOL), Van der Walt 635 (STEU); Stettynskloof (-CD),

Taylor 5395 (SRGH; STE); between Worcester and Villiersdorp

(-CD). Bolus 5135 (BOL; SAM); Louwshoek (-CD), Esterhuy-

sen 17648 (BOL; PRE); De Dooms (-CD), Lamb 1574 (SAM);

Kwadouw Mountains (-DA), Esterhuysen 10336 (BOL); Keer-

omsberg (-DA), Esterhuysen 27595 (BOL); Hexrivier (-DA),

Bolus 13050 (BOL), Marloth 2701 (PRE); Naudesberg (-DA),

Barker 9124 (NBG); Jonaskop (-DC), Boucher 3016 (STE), Van

der Walt 914, 972 (STEU); Klaasvoogds (-DD), Esterhuysen

22691 (BOL; PRE); Robertson (-DD), Anon s.n. (NBG), Comp-

ton 5737 (BOL; PRE). Melle s.n. (BOL), Nel s.n. (STEU). 3320

(Montagu): Voetpadsberg (-AA), Barker 7493 (NBG); Tweed-

side (-AB), Marloth 12069b (PRE), Van der Walt 820 (STEU);

Konstabel (-AD), Van der Walt 805 (STEU); Pieter Meintjies

(-AD), Rogers 21186 (PRE; Z); Witteberg (-BA), Adamson s.n.

(BOL), Compton 12229 (NBG), Pearson 1558 (NBG), Van der

Walt 817 (STEU); Kogmanskloof (-CC), Esterhuysen 23860

(BOL; PRE), Fredfold 417 (PRE); SW of Kogmanskloof (-CC),

Michell 52 (PRE); Bonnievale (-CC), Marloth 11830 (PRE);

Keurkloof. Montagu (-CC), Van Huyssteen s.n. (STEU); Mon-

tagu Baths (-CC), Compton 18345 (NBG), Lewis 1770 (SAM),

Page 126 (PRE), Page s.n. (BOL; Z); Montagu (-CC), Joubert

s.n. (STEU), Levyns 461 (STEU), Stephensen s.n. (STEU). Van

Breda 1240 (PRE). 3420 (Bredasdorp): near Stormsvlei (-AA),

Ecklon & Zeyher s.n. (E; PRE). Simon 6159 (PRE), Zeyher 2087

378

Bothalia 15, 3 & 4 (1985)

(MEL; PRE; SAM; Z); Hessequas Poort (-AA), Taylor 3966

(PRE; STE); near Swellendam (-AB), Galpin 3818 (GRA;

PRE), Liebenberg 6448 (PRE; STE), Van der Walt 1291 (STEU),

Wurts 395 (NBG); Bontebok Park (-AB). Barnard 702 (PRE),

Grobler 509 (PRE; STE) , J ordaan s.n. (STEU), Van der Walt 612

(STEU); Heidelberg (-BB), Leipoldt s.n. (BOL).

20. Pelargonium hermanniifolium (Berg.) Jacq.,

Icon. PI. Rar. 3: t. 545 (1794); Willd., Sp. PI. 3: 682

(1800), Pers., Syn. PI. 2; 233 (1806); DC., Prodr. 1:

677 ( 1824); Spreng. , Syst. Veg. 3: 58 (1826); Eckl. &

Zeyh., Enum. 1: 81 (1835); Knuth in Pflanzenr.

4,129: 464 (1912); J. J. A. van der Walt & Vorster,

Pelarg. S. Afr. 2: 69, fig. (1981). Type: Cape Prov-

ince, ‘e Cap. b. Spei’, Grubb s.n. (SBT 159, holo. !).

Geranium hermanniifolium Berg., Descr. PI. Cap.: 177 (1767);

L., Mant.: 569 (1771); Cav., Diss. 4: 240 (1787); L.f., Suppl.: 305

(1781); Thunb., Prodr. 2: 115 (1800); Thunb., FI. Cap. edn 2: 524

(1823). Pelargonium crispum var. hermanniifolium (Berg.) Harv.

in FI. Cap. 1: 304 (1860).

Erect, often many-stemmed, usually non-aromatic

shrub, up to 1 m high and 0,5 m in diameter. Stems

herbaceous when young but soon becoming woody,

densely hirsute with glandular hairs interspersed,

green but becoming brownish with age. Leaves disti-

chous, sparsely strigose to strigose with glandular

hairs interspersed, green; lamina narrowly to

broadly obovate, 3-palmatilobate to 3-palmatisect

with the segments variably incised, often somewhat

crisped, base cuneate, apex obtuse to acute, margins

crenate-serrate, (5-) 10-15 (-20) x (6-) 8-12 (-20)

mm; petiole (0-) 1-3 (-8) mm long; stipules cordi-

form, often cuspidate, ca 5 x 4 mm. Inflorescence:

flowering branches with normal and smaller foliar

leaves, peduncles 5-8 mm long, hirsute to strigose

with glandular hairs interspersed; involucral bracts

lanceolate to ovate, acuminate, indumentum as on

peduncles, 4-5 x 1-2 mm; pseudo-umbels with

1— 2(— 3 ) flowers each. Pedicel ca 5 mm long. Hypan-

thium 6-8 mm long, indumentum as on peduncles.

Sepals lanceolate, abaxially strigose, green except

posterior one which is partially reddish-brown, mar-

gins white, ca 9 x 2-4 mm. Petals white to pink;

posterior two broadly spathulate, apices sometimes

emarginate to cleft, with dark red markings, reflexed

at ca 90°, ca 19 x 13 mm; anterior three spathulate

with narrow claws, reflexed at less than 90°, ca 16 x 6

mm. 2n=22. Fig. 46.

Diagnostic features

Erect, often many-stemmed shrub. Leaves disti-

chous, strigose; lamina narrowly to broadly obovate,

base cuneate, petiole relatively short. Pseudo-um-

bels with 1-2 (-3) flowers each. Flowers relatively

large, white to pink, pedicel shorter than hypan-

thium.

P. hermanniifolium occurs in the Cape Province

from the district of Worcester southwards to Cale-

don and eastwards to the district of Swellendam

(Fig. 47). It is associated with mountainous habitats,

and is particularly common on the Riviersonderend

Mountains as a component of Fynbos, where it

grows in sandy soil derived from sandstone. The

closely related P. crispum (Berg.) L’Herit. grows in

a drier habitat and the distribution of the two species

on Jonaskop, Villiersdorp, is very interesting.

There, P. crispum is found on the lower and drier

FIG. 46. — Pelargonium hermanniifolium. a, flowering branch, x

1; b. petals, X 1; c, androecium. x 2; d, gynoecium, x 2; e,

inflorescence with petals of flowers removed, x 1. (From

Ward-Hilhorst 2 B. collected at McGregor.)

slopes and P. hermanniifolium on the higher slopes

where the precipitation is considerably higher.

P. hermanniifolium flowers from September to

April with a marked peak in spring.

The problem of distinguishing between P. her-

manniifolium and P. crispum goes back a long time

in the history of plant taxonomy. Bergius (1767) de-

scribed them as separate species, whereas L’Heritier

(1789) did not recognize P. hermanniifolium. Har-

vey (1860) considered it to be a variety of P.

crispum , but Knuth (1912) again maintained them as

distinct species.

The following features distinguish this species

from P. crispum:

P. hermanniifolium

1. Leaf base cuneate.

2. Lamina narrowly

obovate to

broadly obovate.

3. Leaves distichous.

4. Leaves usually

not aromatic.

P. crispum

1. Leaf base cordate.

2. Lamina reniform.

3. Leaves usually

not distichous.

4. Leaves usually

strongly lemon-

scented.

CAPE PROVINCE. — 3319 (Worcester): Elandskloof near

Villiersdorp (-CD), Galpin 12378 (PRE), Gillett 736 (STE); be-

tween Villiersdorp and Franschhoek (-CD), Bolus 5134 (BOL;

SAM); near Villiersdorp (-CD), Oliver 5487 (STE), Osrin 15

(STE); Jonaskop (-DC), Boucher 3026 (PRE; STE), Negin 2

Bothalia 15, 3 & 4 (1985)

379

FIG. 47. — Known geographical distribution of Pelargonium her-

manniifolium.

(NBG). Rycroft 2867 (NBG), Van der Walt 915, 969 (STEU);

Onklaarberg (-DC), Marloth 85 (PRE), Stokoe 1199 (PRE); 16

km E of McGregor (-DD), Marsh 996 (PRE; STE). 3419 (Cale-

don): Houhoek Pass (-AA), Werdermann & Oberdieck 688

(PRE); Caledon (-AB), Bolus s.n. (BOL), Ecklon & Zeyher 636

(SAM), Elbrecht 19013 (PRE). Fischer 263 (STEU), Guthrie 2226

(NBG). Marloth 1578, 7060 (PRE), Purcell 591 7 (SAM); between

Caledon and Villiersdorp (-AB). Rycroft 1724 (NBG); Swart-

berg, Caledon (-AB), Esterhuysen 18940 (BOL; PRE), Mac-

Owan 2950 (SAM), Schlechter 5565 (Z), Schlechter 9767 (E; MO;

PRE; Z), Zeyher 2085 (SAM; Z); Boesmans Pass near McGregor

(-BA), Esterhuysen 4297 (BOL; PRE), Hugo 417 (PRE; STE),

Rycroft 3216 (MO); Genadendal (-BA). Bolus 7377 (BOL), Gil-

len 846 (STE); Greyton (-BA), Taylor 9526 (MO; SRGH), Van

der Walt 803, 1094 (STEU), Van der Walt & Vorster 1316

(STEU); Riviersonderend (-BB), Esterhuysen 18773, 23794

(BOL); Riviersonderend Mountains (-BB), Stokoe s.n. (BOL;

SAM), Wilman 970 (BOL); Tygerhoek (-BB), Van der Walt 796

(STEU); Boesmansrivier (-DA), Barker 1147 (NBG). 3420 (Bre-

dasdorp): Stormsvlei (-AA), Compton 18522 (NBG).

21. Pelargonium cordifolium (Cav.) Curtis in

Curtis’s bot. Mag. 5: t. 165 (1792); Salisb., Prodr.:

314 (1796); H.E. Moore in Baileya 3: 10, fig. 4

(1955); Batten & Bokelmann, Wild. Flow. E. Cape

Prov.: 86 (1966); J. J. A. v.d. Walt, Pelarg. S. Afr.

1: 9, fig. (1977). Type: Cape Province, ‘Habitat ad

Caput Bona Spei. V.S. apud D.D. de Jussieu et

Thouin ex collectis a DD. Thunbergio et Sonnerat’,

(P-JU, lecto.!; MA!, both specimens with Cavanil-

les’s handwriting).

Geranium cordifolium Cav., Diss. 4: 240, t. 117, fig. 3 (1787);

Thunb.. Prodr. 2: 114 (1800); Andrews, Geran. 1: C, ic (1805);

Thunb., FI. Cap. edn 2: 521 (1823). Pelargonium cordatum

L'Herit. in Ait., Hort. Kew. edn 1,2: 427 (1789); Willd. , Sp. PI. 3:

670 (1800); Pers., Syn. PI. 2; 231 (1806); Willd., Enum. 2; 705

(1809); Desf., Arb. 1: 461 (1809); Ait. f., Hort. Kew. edn 2,4: 173

(1812); Sweet, Geran. 1: 67 (1821); DC., Prodr. 1: 671 (1824);

Spreng., Syst. Veg. 3: 58 (1826); G. Don, Gen. Syst. 1: 740

(1831); Eckl. & Zeyh., Enum. 1: 79 (1825); Harv. in FI. Cap. 1:

302 (1860); Knuth in Pflanzenr. 4, 129: 464 (1912); Courtenay-La-

timer & Smith, Flow. PI. Tsitsikamma: t. 40 (1967).

Geranium lanatum Thunb., Prodr. 2: 114 (1800); FI. Cap. edn

2: 518 (1823). Pelargonium lanatum (Thunb.) DC., Prodr. 1: 681

(1824); Eckl. & Zeyh., Enum. 1: 79 (1835). P. cordatum L’Herit.

var. lanatum (Thunb.) Harv. in FI. Cap. T. 302 (1860); Knuth in

Pflanzenr. 4, 129: 464 (1912). Type: Cape Province, locality and

collector unknown (UPS, holo.l, specimen with Thunberg’s hand-

writing).

P. rubrocinctum Link, Enum. Hort. Berol. 2: 191 (1822); DC.,

Prodr. 1: 671 (1824); Spreng., Syst. Veg. 3: 58 (1826); Steud.,

Nom. Bot. edn 2,2: 289 (1841); ex descr. P. cordatum L’Herit.

var. rubrocinctum (Link) Harv. in FI. Cap. 1: 302 (1860); Knuth

in Pflanzenr. 4,129: 464 (1912). Type: Cape Province, ‘Hab. in Pr.

b. Sp.' (Bt).

P. dregeanum Turcz. in Bull. Soc. Nat. Moscow 31: 423 (1858).

Type: Cape Province, George, Drege 7450a (El; Gl; MO!; OXF!;

P!; PRE!; TCD!).

Erect, branched, aromatic shrub, up to 1,75 m

high and 1 m in diameter. Stems herbaceous when

young but soon becoming woody, pubescent to vil-

lous and with long glandular hairs interspersed,

green but becoming brownish with age. Leaves: in-

dumentum extremely variable but always with long

glandular hairs; lamina cordiform, adaxially glabrate

to sparsely villous and green, abaxially pubescent to

villous or tomentose with a velvety texture, often

greyish, base cordate, apex usually acute, margin

dentate and sometimes reddish with the teeth often

reddish, (20-) 65 (-100) x (15-) 55 (-90) mm; pe-

tiole (15-) 70 (-120) mm long; stipules narrowly tri-

angular to triangular, cuspidate, indumentum as on

stems, 8-12 x 3-7 mm. Inflorescence: flowering

branches profusely branched, with smaller foliar

leaves; peduncles 10-50 mm long, pubescent to

densely villous and with long glandular hairs inter-

spersed; involucral bracts ovate to narrowly triangu-

lar, acuminate, pubescent and with long glandular

hairs and soft hairs in between, 6-10 x 2-4 mm;

pseudo-umbels with 3-12 flowers each. Pedicel 5-15

mm long, relatively thin, pubescent to densely vil-

FIG. 48. — Pelargonium cordifolium. a, flowering branch, x 1; b,

petals, X 1; c, androecium, x 2; d, gynoecium, x 2; e, inflo-

rescence with petals of one flower removed, x 1. (From

W ard-Hilhorst 54, collected at Stormsrivier.)

380

Bothalia 15, 3 & 4 (1985)

lous with many long glandular hairs in between. Hy-

panthium 2-12 mm long, prominently thickened at

base, indumentum as on pedicel. Sepals lanceolate,

apiculate, indumentum abaxially as on pedicel,

green with a reddish-brown tint, ca 12 x 2-3 mm.

Petals variable in shape and colour, white to pale

pink to bright purple; posterior two spathulate to

obovate, apices obtuse or emarginate, with dark

purple feather-like markings, reflexed at more than

90°, 20-30 X 5-15 mm; anterior three lorate to nar-

rowly lanceolate, slightly reflexed, 15-25 x 2-4 mm.

2n=22. Fig. 48.

Diagnostic features

Erect, branched and aromatic shrub. Lamina cor-

diform, abaxial side usually velvety and much more

hairy and lighter in colour than adaxial side. Flow-

ering branches profusely branched. Pseudo-umbels

with 3-12 flowers each. Flowers white to light pink

to bright purple, anterior petals much narrower than

posterior ones, pedicel relatively thin and usually

longer than hypanthium.

P. cordifolium occurs from Potberg near Bredas-

dorp in the south-western Cape, eastwards to the

district of Stutterheim in the eastern Cape (Fig. 49).

It is particularly common in the southern Cape

where it often grows in close proximity to the coast.

In the southern Cape it receives rain in winter as well

as in summer, and in the eastern Cape predomi-

nantly summer rains. It is usually found on sandy,

well-drained soil in semi-shaded situations near run-

ning water. Temperatures are high during the sum-

mer and frost can occur at some of the more inland

localities.

P. cordifolium flowers from August to January al-

though the odd flower may be found as early as

June. There is a definite flowering peak in spring

from September to October.

The cordiform leaves of P. cordifolium are so typ-

ical that it can hardly be confused with other species

of the section Pelargonium. It is apparently related

to P. hispidum (L.f.) Willd. and P. papilionaceum

(L.)L’Herit., which have a similar floral structure

with the posterior petals much wider than the ante-

rior ones.

The degree of hairiness of the leaves varies consid-

erably and this led to the recognition of three varia-

ties and even different species by previous workers.

The populations in the western part of the distribu-

tion area have less hairy leaves than those in the

east. There is, however, a gradual transition from

almost glabrous leaves to villous leaves over the dis-

tribution area as a whole, and it is therefore not ad-

visable to recognize infraspecific taxa.

CAPE PROVINCE. — 3226 (Fort Beaufort): Katberg Pass

(-BC), Dyer 759 (PRE), Sidey 3743 (PRE), Story 392 (PRE), Van

der Walt 987 (STEU), Werdermann & Oberdieck 1066 (PRE);

Kettlespoort Falls (-DB), Hilliard & Burtt 10943 (MO); Hogs-

back (-DB), Barker 936 (NBG), Bokelmann s.n. (NBG), Dahl-

strand 1549 (GRA), Giffen 308 (PRE), Rattray 410 (GRA), Sidey

3765 (PRE), Stayner 5 (GRA), Stirton 6236 (MO; PRE; SRGH).

3227 (Stutterheim): Evelyn Valley (-CB), Compton 19159

(NBG), Leighton 2683 (BOL), Taylor 4261 (NBG; PRE); Ko-

logha (-CB), Acocks 9004 (PRE), Flanagan 2181 (BOL; PRE);

near King William’s Town (-CD), Sim 1326 (BOL). 3321 (Ladis-

mith): Garcia’s Pass (-CC), Bolus 11226 (BOL; Z). Van der Walt

FIG. 49. — Known geographical distribution of Pelargonium cor-

difolium.

622, 831, 1298 (STEU); Cloetes Pass (-DD), Middlemost 2015

(NBG). 3322 (Oudtshoorn): Spitzkop (-AD), Rycroft 3108

(NBG); Langkloof (-CB), Ecklon & Zeyher 619 (S; SAM);

Ruitersbos (-CC), Hops 1 (BOL), Van Niekerk 22, 129 (BOL):

Robinson Pass (-CC), Maguire 796 (NBG), Van der Walt 1128

(STEU); near George (-CC), Drege 7450a (E; K; MO; OXF; P;

PRE; TCD), Gillett 2070 (BOL), Guthrie 4284 (NBG), Paterson

1226 (GRA), Schlechter 2307 (Z), Stephany s.n. (Z), Tennant 2

(NBG); Montagu Pass (-CD), Compton 7574 (NBG), Hiemstra

329 (NBG), Hutchinson 1198 (BOL; PRE), Neser s.n. (STEU),

Schonken 194 (STEU), Stokoe s.n. (SAM), Taylor 3679 (PRE),

Thorne s.n. (SAM), Van Breda 1121 (PRE), Van der Walt 424,

1306 (STEU); 13 km E of George (-CD), Hutchinson 1265

(BOL); Sandplaats (-CD), Thorne s.n. (SAM); Joubertsberg

(-CD), Taylor 1672 (PRE); Mannetjiesberg (-DB), Esterhuysen

6487 (BOL); Kaymansgat (-DC), Drege 7450 (MO); Outeniqua

Pass (-DC), Compton 24418 (NBG), Lewis 3911 (SAM); Saas-

veld (-DC), Schonken 184, 195 (STEU), Van der Walt 679

(STEU); Karatara Pass (-DD), Marsh 597 (PRE). 3323 (Willow-

more): Krantzkop near Avontuur (-CA), Stokoe s.n. (SAM); 10

km S of Avontuur (-CA), Marsh 630 (PRE), Thompson 570

(PRE; STE); Prince Alfred’s Pass (-CC), Britten s.n. (GRA),

Salter 6741, 6797 (BOL), Van der Walt 720 (STEU); Gouna Re-

serve (-CC), Taylor 1316 (SAM); Paardekop near Knysna (-CC),

Steyn 714 (BOL); near Keurboomsrivier (-CD), Gillett 4576

(PRE), Marsh 1329 (PRE), Story 3112 (MO; PRE), Theron 1778

(PRE; Z); Louterwater (-DC), Zinn s.n. (SAM); Helpmekaar

Peak (-DC), Esterhuysen 4575 (BOL; PRE); Grootrivier Pass

(-DC), Schlieben & Ellis 12327 (PRE); Stormsrivier Pass (-DD),

Liebenberg 6370 (PRE). 3324 (Steytlerville): Onder Kouga

(-CB), Bayliss 5967 (MO; NBG); Kareedouw Pass (-CC), Gillett

2030 (PRE); Kareedouw (-CC), Neser s.n. (STE), Thode A754

(PRE); Grootrivier (-CC), Sidey 1706 (PRE); 33 km from Karee-

douw on Knysna road (-CD), Story 3646 (PRE); Baviaanskloof

(-DA), Bayliss 4303 (MO; NBG; Z). 3325 (Port Elizabeth):

Springfield (-CB), Keet s.n. (PRE); Otterford Forest Reserve

(-CC), Dahlstrand 793 (GRA), Rodin 1136 (BOL; MO; PRE),

Schonken 127 (STEU), Thompson 1836 (PRE); Van Stadensri-

vier (-CC), Bolus 783 (BOL), Ecklon & Zeyher 620 (S; SAM),

Long 640 (PRE), Paterson 891 (GRA), Paterson 2525 (PRE);

Longmore Forest Reserve (-CC), Dahlstrand 718 (PRE); Loerie

Plantation (-CC), Dix 42 (BOL). 3420 (Bredasdorp): Potberg

(-BC), Esterhuysen 23243 (BOL), Pillans 9337 (BOL), Wall gate

911 (PRE). 3421 (Riversdale): Corente River (-AA), Muir 40

(PRE); Tysmanshoek (-AB), Muir 5427 (PRE). 3422 (Mossel

Bay): Victoria Bay (-BA), Compton 15781 (NBG); Belvidere

near Knysna (-BB), Van der Walt 841 (STEU). 3423 (Knysna):

Knysna (-AA), Wurts 2123A (NBG); Paardekop (-AA), Steyn

714 (BOL); Nanutzi forest (-AA), Hardy 867 (PRE; SRGH);

Noetzie (-AA), Middlemost s.n. (NBG); Kranskop West (-AA),

Horn s.n. (PRE); Goudveld Forest Reserve (-AA), Boucher 40

(STEU); Concordia (-AA), Kapp 70 (PRE), Phillips 98 (GRA);

Kruisfontein Mountains (-AA), Galpin 3821 (PRE); Portland

(-AA), Duthie 1164 (BOL); Plettenberg Bay (-AB), Ecklon &

Zeyher 618 (S; SAM), Rogers 26829 (PRE), Rogers 27881 (GRA;

Bothalia 15, 3 & 4 (1985)

381

PRE; SAM; Z), Rogers 27975 (Z); between Keurboomsrivier and

Stormsrivier (-AB), Gillett 4576 (BOL; PRE); Tsitsikamma

(-BB), Baltin s.n. (PRE), Zeyher s.n. (SAM); Stormsrivier

(-BB), Hall 212 (NBG), Schonken 131 (STEU), Taylor 3714,

5927 (NBG), Tyson 985, 3014 (SAM). 3424 (Humansdorp):

Witelsbos (-AA), Fourcade 1368 (BOL), Thompson 883 (PRE);

Clarkson (-AB), Thode A755 (PRE); Hofmansbos (-BB), Britten

1195 (GRA; PRE); Humansdorp (-BB), Britten 1012 (PRE),

Rogers 3018 (PRE).

22. Pelargonium hispidum (L.f.) Willd., Sp. PI.

3: 677 (1800); Pers., Syn. PI. 2; 232 (1806); Ait. f.,

Hort. Kew. edn 2,4: 177 (1812); Eckl. & Zeyh.,

Enum. 1: 79 (1835); DC., Prodr. 1: 679 (1824);

Harv. in FI. Cap. 1: 307 (1860); Knuth in Pflanzenr.

4,129; 474 (1912); J. J. A. v.d. Walt & Vorster, Pel-

arg. S. Afr. 2: 73, fig. (1981). Type: Cape Province,

‘Habitat in Cap. bonae Spei’, Back s.n. sub LINN

858:22 (LINN, holo.!).

Geranium hispidum L.f., Suppl.: 304 (1781); Murray, Syst.

Veg. 14: 614 (1784); Cav., Diss. 4: 248, t. 110, fig. 1 (1787);

Thunb. , Prodr. 2: 115 (1800) ;Thunb. , FI. Cap. edn 2: 521 ( 1823).

Erect, branched, faintly lemon-scented shrub, up

to 2,5 m high and 1 m in diameter. Stems herbaceous

when young but soon becoming woody, pubescent to

hirsute and with glandular hairs interspersed, green

but becoming greyish with age. Leaves pubescent to

hirsute and with glandular hairs interspersed, green

to dull green; lamina palmatilobate to palmatisect

with segments sometimes irregularly incised, con-

spicuously veined abaxially, base cordate, apices of

lobes acute, margins irregularly sinuous to dentate,

(50-) 90-120 (-250) x (60-) 100-130 (-270) mm; pe-

tiole (50-) 80 (-130) mm long; stipules lanceolate,

FIG. 50. — Pelargonium hispidum. a, flowering branch, x 1; b,

petals, x 2; c, androecium, X 2; d, gynoecium, X 4. (From

Van der Merwe s.n. (sub STEU 2743), cultivated in Stellen-

bosch.)

7-10 x 2-3 mm. Inflorescence: flowering branches

profusely branched, with smaller foliar leaves, pe-

duncles 10-40 mm long, hirsute and densely inter-

spersed with glandular hairs; involucral bracts lan-

ceolate, indumentum as on peduncles, 6-8 x 1-2

mm; pseudo-umbels with 6-12 flowers each. Pedicel

5-7 mm long; indumentum as on peduncles. Hypan-

thium 3^4 mm long, prominently thickened at the

base. Sepals lanceolate, indumentum as on pedun-

cles, green with reddish-brown bases and white mar-

gins, ca 8 x 2-3,5 mm. Petals pale to deep pink;

posterior two asymmetric-obovate, with wine-red

feather-like markings, reflexed at more than 90°, ca

12 x 7 mm; anterior three spathulate with long and

narrow claws, with darker pigmentation towards

bases, practically straight, ca 8 x 2 mm. 2n=44. Fig.

50.

Diagnostic features

Erect, branched, faintly lemon-scented shrub.

Leaves relatively large, pubescent to hirsute, green

to dull green, lamina palmatilobate to palmatisect.

Flowering branches profusely branched, pseudo-um-

bels with 6-12 flowers each. Flowers pale to deep

pink, two posterior petals much wider than the ante-

rior three, pedicel longer than hypanthium.

P. hispidum occurs commonly in the mountains of

the south-western Cape as well as in the Swartberg

Range in the southern Cape, from Piketberg east-

wards for about 450 km to Meiringspoort in the

Oudtshoorn District (Fig. 51). Although consistently

associated with mountains, it is confined to the lower

slopes where it occurs in shady ravines, usually near

streams between boulders or on scree. It has been

recorded at altitudes of between 300 and 1 000 m,

but there is one record of it occurring at 1 350 m

from the Swartberg Pass area. It is a component of

Fynbos or may even be associated with ravine forest

precursors. The soil in these situations is usually

sandy, often with a high percentage of organic mat-

ter. Temperatures are high during summer, and frost

may occur during winter. The sheltered niches occu-

pied by P. hispidum are probably less exposed to en-

vironmental extremes than the area in general. Usu-

ally growing close to streams, it probably has a con-

stant supply of moisture available throughout the

year, irrespective of rainfall.

P. hispidum flowers throughout the summer, from

September to the following April, but with a marked

peak in October and November.

Inflorescence and floral characters indicate a close

relationship between P. hispidum, P. cordifolium

(Cav.) Curtis, P. papilionaceum (L.) L’Herit. and P.

tomentosum Jacq. P. hispidum may be confused with

P. papilionaceum, but in the latter species the leaves

are only shallowly lobed and the flowers are larger

with a conspicuous white spot on the two posterior

petals.

CAPE PROVINCE. —3218 (Clanwilliam): Piquetberg (-DB),

Pillans 7187 (BOL). 3318 (Cape Town); Paarlberg (-DB), Drege

s.n. (PRE), Drege 9119 (PRE); Pic Blanche near Paarl (-DB),

Esterhuysen 1672 (BOL). 3319 (Worcester): Winterhoekberg

(-AA ),' Bolus 5324 (BOL), Ecklon & Zeyher 616 (SAM); Twenty

Four River Mountains (-AA), Esterhuysen 29954 (BOL), Van der

Walt 1016 (STEU); farm Roodesand near Tulbagh (-AC). Cillie

s.n. (STEU); Tulbagh Waterfall (-AC), Guthrie 3108 (NBG);

382

Bothalia 15, 3 & 4 (1985)

FIG. 51. — Known geographical distribution of Pelargonium his-

pidum.

Steendal near Tulbagh (-AC), MacOwan 1559 (SAM); Witzen-

berg (-AC), Esterhuysen 22505 (BOL), Zeyher 200 (BOL; E;

PRE; SAM); Michell’s Pass near Ceres (-AD), Bolus 2601 (BOL;

SAM), Van der Walt 1283 (STEU); Skilderberg near Ceres

(-AD), Stokoe 2653 (BOL); Du Toitskloof (-CA), Barker 4827

(NBG), Esterhuysen 22295 (BOL; PRE), Stokoe s.n. (SAM);

Bainskloof (-CA), Compton 16916 (NBG), Esterhuysen 25633

(BOL). Schlechter 9196 (PRE); near Worcester (-CB), Fine 916

(PRE), Rehrnann 2454 (PRE; Z); Audensberg near Worcester

(-CB). Compton 9792 (NBG), Esterhuysen 22906 (BOL; PRE);

Waaihoek near Worcester (-CB), Esterhuysen 22680 (BOL;

PRE); Malkopskloof (-CB). Esterhuysen 3810 (BOL); Brand-

wacht Mountains near Worcester (-CB), Acocks 15253 (PRE);

Wemmershoek (-CC), Compton 10141 (BOL), Esterhuysen 4032

(BOL; NBG); Du Toit’s Peak (-CC), Esterhuysen 16640 (BOL);

Haalhoek (-CC), Esterhuysen 13545 (BOL); Klein Drakenstein

Mountains (-CC), Esterhuysen 20855 (BOL); Molenaarsberg

(-CC), Esterhuysen 14099 (BOL; MO); Klaasvoogds (-DD), Es-

terhuysen 22705 (BOL; PRE). 3320 (Montagu); Montagu Baths

(-CC), Page 118 (PRE), Page s.n. (BOL); Donkerkloof, Mon-

tagu (-CC), Compton 18458 (NBG), Lewis 1778 (SAM). 3321

(Ladismith): Seven Weeks Poort (-AD), Van der Walt 629

(STEU), Wurts 1218 (NBG); Waterkloof near Ladismith (-BD),

Van der Walt 1121 (STEU); Bailey’s Peak (-CB), Esterhuysen

22331 (BOL). 3322 (Oudtshoorn): Swartberg Pass (-AC), Stokoe

9044 (BOL), Van der Wall 726 (STEU), Wall s.n. (NBG);

Meiringspoort (-BC), Esterhuysen 24879 ( BOL; MO; PRE). Van

der Walt 1145 (STEU). 3419 (Caledon); Genadendal (-BA), Eck-

lon & Zeyher 615 (SAM), Pappe s.n. (MEL), Schlechter 9848

(BOL; E; GRA; MO; PRE; Z), Van der Walt 1092 (STEU);Ty-

gerhoek (-BB), Van der Walt 519 (STEU). 3420 ( Bredasdorp):

Potberg (-CA), David s.n. (NBG), Esterhuysen 23227 (BOL;

PRE), Pillans 9309 (BOL; NBG), Lilians 9469 (BOL).

23. Pelargonium papilionaceum (L.) L' Merit, in

Ait., Hort. Kew. edn 1,2: 423 (1789); Salisb.,

Prodr.: 316 (1796); Willd., Sp. PI. 3: 671 (1800);

Pers., Syn. PI. 2: 231 (1806); Willd., Enum.: 706

(1809); Ait. f., Hort. Kew. edn 2,4: 174 (1812);

Sweet, Geran. 1: 27 (1820); DC., Prodr. 1: 671

(1824); Spreng., Syst. Veg. 3: 59 (1826); Eckl. &

Zeyh., Enum. 1: 79 (1835); Harv. in FI. Cap. 1: 305

(1860); Knuth in Pflanzenr. 4,129: 465 (1912); Phil-

lips in Rep. S. Afr. Ass. Advmt. Sci.: 456 (1918);

J. J. A. v.d. Walt, Pelarg. S. Afr. vol. 1: 32, fig.

(1977). Lectotype: ‘Habitat in Africa’, LINN 858.5!

(LINN).

Geranium papilionaceum L., Sp. PI. edn 1,2; 676 (1753); L..

Sp. PI. edn 2,2: 945 (1763); Burm. f.. Prodr. FI. Cap.; 18 (1768);

Mill., Gard. Diet, edn 8: 27 (1768); Murray, Syst. Veg.: 613

(1784); Cav.. Diss. 4: 244, t. 112, fig. 1 (1787); Thunb., Prodr. 2:

114 (1800): Thunb., FI. Cap. edn 2: 521 (1823).

Erect, much-branched, strongly aromatic shrub

with an unpleasant odour, up to 2,5 m high and 1,5

m in diameter. Stems herbaceous when young but

soon becoming woody, villous and with glandular

hairs interspersed, green but becoming greyish-

brown with age. Leaves sparsely villous and densely

interspersed with glandular hairs, green; lamina cor-

diform in outline, shallowly 3- (5-8)-lobed, con-

spicuously veined, base cordate, apices of lobes usu-

ally obtuse but sometimes acute, margin crenate-

dentate, (25-) 85 (-190) x (25-) 100 (-250) mm; pe-

tiole ( 10 ) 50-70 (-250) mm long; stipules cordiform

to triangular, 4-10 x 3-12 mm. Inflorescence: flow-

ering branches profusely branched, with smaller fo-

liar leaves, peduncles 20-100 mm long, villous and

with glandular hairs interspersed; involucral bracts

ovate, apiculate, indumentum as on peduncles, 5-8

x 4—6 mm; pseudo-umbels with 4—20 flowers each.

Pedicel 7-15 mm long, relatively thin, sparsely vil-

lous and densely interspersed with glandular hairs.

Hypanthium 2-5 mm long, prominently thickened at

the base. Sepals lanceolate, apiculate, indumentum

abaxially as on pedicel, green with a reddish-brown

tint, ca 8 x 3-4 mm. Petals pale pink to carmine;

posterior two spathulate to obovate, apices obtuse to

emarginate, with a dark red-purplish and white

blotch, reflexed at more than 90°, ca 20 x 7 mm; an-

terior three narrowly spathulate with short claws, re-

flexed at less than 90°, ca 7 x 2 mm. 2n=44. Fig. 52.

Diagnostic features

Erect, much-branched shrub with an unpleasant

odour. Lamina cordiform in outline, shallowly 3-8-

lobed, conspicuously veined, sparsely villous. Flow-

ering branches profusely branched. Pseudo-umbels

with 4—20 flowers each. Flowers pale pink to car-

mine, anterior three petals much shorter and nar-

rower than posterior two, pedicel relatively thin and

longer than hypanthium.

P. papilionaceum occurs in a relatively narrow

strip along the coast from Somerset West in the

south-western Cape, eastwards to Humansdorp in

the eastern Cape (Fig. 53). This area receives rain

chiefly during the winter months. The species is con-

sistently associated with mountains and is usually

found on the margins of ravine forests near streams

where it grows in sandy soil with organic matter.

Temperatures are high during the dry summer

months, but a mild microclimate is created in the

moist and semi-shaded habitats. The winters are

frost-free due to close proximity to the coast.

P. papilionaceum flowers from August to January

with a peak in September-October. Except for the

mid-winter months, the odd flower can be found

throughout the year.

P. papilionaceum could be confused with P. to-

mentosum (L.) L'Herit. and P. hispidum (L.f.)

Willd. The morphological differences between P.

papilionaceum and these two species, are discussed

under P. tomentosum and P. hispidum respectively.

The leaves of P. papilionaceum resemble those of P.

vitifolium (L.) L’Herit. to a certain extent. These

Bothalia 15, 3 & 4 (1985)

383

FIG. 52. — Pelargonium papilionaceum. a, flowering branch, x

1; b, petals, x 1,5; c, androecium, x 2; d, gynoecium, x 2.

(From Ward-Hilhorst 87, collected at Helderberg Nature Re-

serve.)

two species could, however, easily be distinguished

by inflorescence and floral characteristics. P. vitifo-

lium has a capitulum-like inflorescence and the flow-

ers are less zygomorphic than those of P. papiliona-

ceum.

FIG. 53. — Known geographical distribution of Pelargonium pa-

pilionaceum.

CAPE PROVINCE. — 3318 (Cape Town): Jonkershoek,

Guardian Peak (-DD), Esterhuysen 11949, 24127 (BOL);

Jonkershoek, Swartboskloof (-DD), Van der Merwe 23-59

(PRE); Jonkershoek, Langrivier (-DD), Van der Walt 507, 639

(STEU); Stellenbosch (-DD), Guthrie 2327 (NBG). 3319 (Wor-

cester): French Hoek Pass (-CC), Esterhuysen 24375 (BOL);

French Hoek (-CC), Marloth 5307 (PRE), Phillips 1054 (SAM);

Klaasvoogds (-DD), Esterhuysen 22704 (BOL). 3320 (Montagu):

Tradouw Pass (-DC), Esterhuysen 24610 (BOL; MO; PRE), Van

der Wall 827, 1292 (STEU); Grootvadersbos, Swellendam

(-DD), Esterhuysen 19261, 25031 (BOL), Marloth 3512 (PRE),

Zeyher 2093 (PRE; SAM); Ten O’Clock Mountain, Swellendam

(-DD), Wurts 354 (NBG); Voormansbos, Swellendam (-DD),

Zeyher s.n. (SAM); near Duiwelsbos (-DD), Van der Walt 1289

(STEU); Langeberge near Swellendam (-DD), Schlechter 5710

(BOL; Z); Hermitage Kloof near Swellendam (-DD), Esterhuy-

sen 24625 (BOL; PRE). 3321 (Ladismith): Garcia’s Pass (-CC),

Galpin 3816 (PRE), Marloth 12583 (PRE), Moffett 1074 (STEU),

Muir 2919 (PRE), Thornes. n. (SAM), Van der Walt 619 (STEU).

3322 (Oudtshoorn): Robinson Pass (-CC), Salter 6343 (BOL),

Van der Walt 1126 (STEU); 10 km E of George (-CD), Schonken

186 (STEU); Saasveld (-DC), Van der Walt 719 (STEU). 3323

(Willowmore): N slopes of Tsitsikamma Mountains near Jouber-

tina (-DC), Esterhuysen 22792 (BOL). 3418 (Simonstown): Hel-

derberg (-BB), Dimmer 564 (E), Stokoe s.n. (BOL), Van der

Walt 496 (STEU), Van der Walt & Vorster 1320 (STEU); Diepgat

near Somerset West (-BB), Esterhuysen 8233 (BOL). 3419 (Cale-

don): Kleinrivier near Caledon (-AD), Ecklon & Zeyher 617

(SAM); Vogelgat Kloof near Caledon (-AD), Williams 2606

(MO); Greyton (-BA), Van der Walt 707, 800, 1315 (STEU); Ri-

viersonderend Mountains (-BB), Esterhuysen 25089 (BOL; MO;

PRE); Oubos (-BD), Van der Walt 713 (STEU). 3420 (Bredas-

dorp); near Swellendam (-AB), Ecklon 613 (PRE; S); Buffels-

jagrivier (-BA), Drege s.n. (S); Zuurbraak (-BA), Schlechter

5710 (Z); Potberg (-BC), Pillans 9309 (PRE). 3421 (Riversdale):

Riversdale (-AB), Muir 2919 (BOL). 3424 (Humansdorp): Clark-

son (-AB), Thode A757 (PRE).

24. Pelargonium tomentosum Jaccp, Icon. PI.

Rar. 3: 10, t. 537 (1794); Willd., Sp. PI. 3: 677

(1800); Curtis in Curtis’s bot. Mag. 15: t. 518 (1801);

Pers., Syn. PI. 2: 232 (1806); Ait. f., Hort. Kew. edn

2,4: 177 (1812); Sweet, Geran. 2: 168 (1823); DC.,

Prodr. 1: 671 (1824); Spreng., Syst. Veg. 3: 61

(1826); Harv. in FI. Cap. 1: 305 (1860); Knuth in Pf-

lanzenr. 4,129: 458 (1912); J. J. A. v.d. Walt &

Vorster, Pelarg. S. Afr. 2: 145, fig. (1981). Type:

Locality and collector unknown (W, holo.!, speci-

men with Jacquin’s handwriting).

Geranium tomentosum (Jacq.) Poir., Encycl. Suppl. 2; 754

(1811). Geraniospermum tomentosum (Jacq.) Kuntze, Rev. Gen.

1: 95 (1891).

Pelargonium micranthum Eckl. & Zeyh., Enum. 1: 79 (1835).

Type: Cape Province, ‘In montium lateralibus prope Zwellen-

dam’, Ecklon & Zeyher 614 (S! ; SAM!).

P. corymbosum Turcz. in Bull. Soc. Nat. Moscow 31,1: 422

(1858). Type: Cape Province, ‘C.b. spei’, Zeyher 2095 (MEL!;

PRE!; SAM!; Z!).

Decumbent, much-branched, peppermint-scented

subshrub, up to 0,5 m high and 1,5 m in diameter.

Stems herbaceous and brittle, becoming somewhat

woody with age, villous and densely interspersed

with glandular hairs, greyish-green but becoming

brownish with age. Leaves with soft hairs and nu-

merous glandular hairs interspersed, greyish-green;

lamina 3-(5)-palmatilobate to 3-(5)-palmatipartite,

base cordate, apices of lobes mostly obtuse (rarely

acute), margins irregularly crenate-serrate, adaxially

villous, abaxially tomentose, (25-) 40-60 (-110) x

(35-) 50-70 (-120) mm; petiole (30-) 80-130 (-180)

mm long; stipules triangular to ovate, acute to acu-

minate, 6-20 x 4—12 mm. Inflorescence: flowering

branches profusely branched, with or without

smaller foliar leaves; peduncles 30-150 mm long, vil-

lous and densely interspersed with glandular hairs;

involucral bracts narrowly ovate to lanceolate, acu-

minate, indumentum as on peduncles, 4—5 x 1-2

mm; pseudo-umbels with 4—15 flowers each. Pedicel

384

Bothalia 15, 3 & 4 (1985)

18-20 mm long, indumentum as on peduncles. Hy-

panthium ca 2 mm long. Sepals elliptic to lanceolate,

indumentum abaxially as on peduncles, green with

white margins, ca 6 x 2-3,5 mm. Petals white with

purple markings; posterior two elliptic to obovate

with eared bases, reflexed at ca 90°, ca 9 x 5 mm;

anterior three linear-spathulate with very narrow

claws, practically straight, ca 11 x 1,5 mm. 2n=44.

Fig. 54.

FIG. 54. — Pelargonium tomentosum. a, flowering branch, x 1;

b, petals, x 2; c, androecium, x 4; d, gynoecium, x 4. (From

Van der Walt 1349, cultivated in Stellenbosch.)

Diagnostic features

Decumbent, much-branched, peppermint-scented

subshrub. Lamina 3-(5)-palmatilobate to 3-(5)-pal-

matipartite, adaxially villous, abaxially tomentose.

Flowering branches profusely branched. Pseudo-

umbels with 4—15 flowers each. Flowers white, post-

erior petals auriculate, anterior petals longer and

much narrower than posterior ones, pedicel much

longer than hypanthium.

So far P. tomentosum has been collected on the

Hottentots Holland Mountains near Somerset West,

on the Riviersonderend Mountains near Greyton,

and on the Langeberg range from Swellendam to

Riversdale (Fig. 55). It is rare on the Hottentots

Holland Mountains but abundant on the Langeberg

range, especially near Swellendam. There is also a

specimen in the herbarium of the University of

Zurich (Z) which was collected by Bolus in April

1884 at Wynberg (Cape Peninsula), but the species is

probably now extinct in this urbanized area.

P. tomentosum is confined to mountains where it

occurs in semi-shaded, moist habitats. It is usually

found on the margins of ravine forests near streams,

where it grows in sandy soil derived from sandstone.

The flowering time extends from October to Janu-

ary.

The species may be confused with P. papiliona-

ceum (L.) L'Herit. when they are not in flower. The

main differences between these two related species

are tabulated below:

P. tomentosum

1. Leaves villous and

tomentose.

2. Leaves peppermint-

scented.

3. Flowers white.

4. Posterior petals

shorter than

anterior petals.

P. papilionaceum

1. Leaves sparsely

villous.

2. Leaves with an

unpleasant odour.

3. Flowers light pink to

carmine.

4. Posterior petals

longer than

anterior petals.

FIG. 55. — Known geographical distribution of Pelargonium to-

mentosum.

CAPE PROVINCE. — 3320 (Montagu): Tradouw Pass (-DC),

Hafstrom & Acocks 1982 (BOL; PRE), Van der Walt 1293

(STEU); above Grootvadersbosch near Swellendam (-DD), Es-

terhuysen 18275 (BOL; PRE), Esterhuysen 25029 (BOL; MO;

NBG); Voormansbosch near Swellendam (-DD), Ecklon &

Zeyher 613 (K; MEL; OXF; W), Zeyher 2095 (MEL; PRE; SAM;

Z); Ten O’Clock Mountain near Swellendam (-DD) Wurts 447,

510 (NBG); Strawberry Hill near Swellendam (-DD), Stokoe s.n.

(NBG); mountains near Swellendam (-DD), Ecklon & Zeyher

614 (S; SAM). 3418 (Simonstown): Wynberg (-AB), Bolus s.n.

(Z); Helderberg Nature Reserve (-BB), Van der Walt 1349

(STEU); Helderberg ravine (-BB), Esterhuysen 7656 (BOL;

PRE); Valleiberg, Hottentotsholland (-BB), Esterhuysen 29865

(BOL). 3419 (Caledon): near Genadendal (-BA), Ecklon &

Zeyher 615 (P). 3421 (Riversdale): Corente River (-AA), Muir

5062 (PRE); Riversdale (-AB), Peers s.n. (BOL); De Hoek near

Riversdale (-AB), Muir 2964 (PRE).

NATURAL HYBRIDS

The following natural hybrids of the section Pelar-

gonium have been identified:

1 Hybrids between members of the section Pelargo-

nium

Putative parent species:

1.1 P. citronellum x P. hispidum

3321 (Ladismith): near Waterkloof, Ladi-

smith (-AD), Van der Walt 1124 (STEU).

Bothalia 15, 3 & 4 (1985)

385

1.2 P. cucullatum subsp. cucullatum x P. betuli-

num

3418 (Simonstown): near Betty’s Bay (-BD),

Volschenk 24 (STEU).

1.3 P. cucullatum subsp. tabulare x P. betulinum

3418 (Simonstown): Klein Leeukoppie

(-AB), Volschenk 17 (STEU).

1.4 P. panduriforme x P. quercifolium

3322 (Oudtshoorn): Duiwerivier, Baviaans-

kloof (-DC), Taylor 412 (BOL).

1.5 P. scabrum x P. glutinosum

3322 (Oudtshoorn): Meiringspoort (-BC),

Van der Walt 1445 (STEU).

1.6 P. scabrum x P. hispidum

3219 (Wuppertal): near Keerom (-CC), Pil-

lans 8730 (BOL).

1.7 P. scabrum X P. scabroide

3319 (Worcester): Groot Winterhoek (-AA),

Van der Walt 1075 (STEU).

1.8 P. scabrum x P. sublignosum

3218 (Clanwilliam): Farm Grootfontein near

Porterville (-CC), Van der Walt 910 (STEU).

2 Hybrids between members of the section Pelargo-

nium and the section Eumorpha (Eckl. & Zeyh.)

Har\’.

Putative parent species:

2.1 P. cucullatum subsp. strigifolium x P. patu-

lum Jacq.

3418 (Simonstown): Swartboskloof ,

Jonkershoek (-BB), Van der Walt 1013

(STEU).

2.2 P. cucullatum subsp. tabulare x P. patulum =

P. dodii Schlt. ex Knuth

3418 (Simonstown): Orange Kloof (-AB),

Wolley Dod 2160. (K, isotype).

2.3 P. scabrum x P. patulum

3418 (Simonstown): Stellenboschberg (-BB),

Van der Walt 1326 (STEU).

2.4 P. scabrum x P. grandiflorum (Andr.) Willd.

3118 (Vanrhynsdorp): Gifberg (-DC), Ester-

huysen 21960 (BOL).

2.5 P. sublignosum x P. grandiflorum

3218 (Clanwilliam): Farm Berghof near Por-

terville (-CC) Van der Walt 1089 (STEU).

2.6 P. sublignosum x P. patulum

3219 (Wuppertal): Hexberg (-CA), Esterhuy-

sen 18430 (PRE).

2.7 P. tomentosum x P. patulum

3418 (Simonstown): Helderberg Nature Re-

serve (-BB), Van der Walt 1350 (STEU).

3 Hybrids between members of the section Pelargo-

nium and the section Glaucophyllum Harv.

Putative parent species:

3.1 P. cordifolium x P ternatum (L.f.) Jacq. = P.

riversdalense Knuth

3321 (Ladismith): Garcia’s Pass (-CC), Bolus

11233 (BOL, holotype).

3.2 P. scabrum x P. fruticosum (Cav.) Willd.

3322 (Oudtshoorn): Meiringspoort (-BC),

Van der Walt 704 (STEU).

3.3 P. scabrum x P. lanceolatum (Cav.) Kern. =

P. tricuspidatum L’Herit.

3319 (Worcester): Sandhills (-DA), Van der

Walt 632 (STEU).

ACKNOWLEDGEMENTS

I thank all collaborators of the Pelargonium pro-

ject who contributed in one or other way to this

study. A special word of thanks to Dr P. J. Vorster

for reading the manuscript, Prof. F. Albers for the

valuable cytogenetic information, the botanical art-

ist, Ellaphie Ward-Hilhorst, for the illustrations, and

Mrs E. M. J. Prinsloo for proof reading the manu-

script.

A sincere word of thanks to all the curators of her-

baria who kindly lent material to us. I am grateful

for the financial support received from the CSIR and

the University of Stellenbosch.

UITTREKSEL

Vier-en-twintig spesies van die seksie Pelargonium,

wat laas in 1912 deur Knuth hersien is, word in hier-

die taksonomiese studie onderskei. Meeste spesies

kom in suidwes-, suid- en oos-Kaap voor, waarhulle

meesal in halfskadu in betreklike vogtige habitatte

groei. 'n Sleutel vir die identifikasie van die spesies is

opgestel, en vir elke spesie is daar minstens een illus-

trasie asook ’n verspreidingskaart. Die seksie word as

die primitiefste in die genus beskou met 'n basiese

chromosoomgetal van x = 11.

REFERENCES

ALBERS, F. & VAN DER WALT, J. J A., 1984. Untersu-

chungen zur Karyologie und Mikro-sporogenese von Pel-

argonium sect. Pelargonium (Geraniaceae). PI. syst. Evol.

147: 177-188.

DE CANDOLLE, A. P., 1824. Prodromus systematis naturalis

regni vegetabilis Vol. 1. Paris: Treuttel & Wiirtz.

ECKLON, C. F. & ZEYHER, K.L., 1835. Enumeratio planta-

rum africae australis extratropicae Vol. 1. Hamburg.

HARVEY, W. H., 1860. Flora capensis Vol. 1. Dublin: Hedges

& Smith.

JESSOP, J. P., 1964. Itinerary of Rudolf Schlechter's collecting

trip in southern Africa. Jl S. Afr. Bot. 30 : 129.

KNUTH, R., 1912. Geraniaceae. Pflanzenreich 4, 129 : 1-640.

L'HERITIER. C.-L.. 1789. In W. Aiton. Hortus Kewensis , edn

1, vol. 2. London: George Nicol.

SWEET, R., 1822. Geraniaceae Vol. 1. London: James Ridgway.

VAN DER WALT, J. J. A., 1979. Notes on the nomenclature of

Pelargonium (Geraniaceae). Jl S. Afr. Bot. 45 : 377-380.

VAN DER WALT, J. J. A. & VORSTER, P. J., 1981. Typifica-

tion of the genus Pelargonium L’Herit. (fam. Gerania-

ceae). Taxon 30 : 307.

VAN DER WALT, J. J. A. & VORSTER, P. J., 1981. Pelargo-

niums of southern Africa Vol. 2. Cape Town: Juta.

VAN DER WALT, J. J. A. & VORSTER. P. J., 1983. Phytogeo-

graphy of Pelargonium. Bothalia 14 : 517-323.

Bothalia 15, 3 & 4: 387-503 (1985)

Conspectus of the African species of Restionaceae

H. P. LINDER*

Keywords: African, conspectus, keys, new combinations, new taxa, nomenclature, Restionaceae, typification

ABSTRACT

A conspectus of the African species of the Restionaceae is given. 318 species, arranged in the generic classifica-

tion of Linder (1984) are recognized. Detailed notes on the nomenclature and the typification and brief notes on

the taxonomy and distribution, are given for each species. Keys to the genera and the species are provided. Fifty-

five new specific epithets are published, representing 51 new species (Anthochortus capensis Esterhuysen, Aski-

diosperma rugosum Esterhuysen, Calopsis adpressa Esterhuysen, C. clandestina Esterhuysen, C. dura Esterhuy-

sen, C. pulchra Esterhuysen, Chondropetalum decipiens Esterhuysen, Elegia atratiflora Esterhuysen, E. caespi-

tosa Esterhuysen, E. fucata Esterhuysen, E. grandispicata Linder, Hypodiscus montanus Esterhuysen, H. procur-

rens Esterhuysen, H. squamosus Esterhuysen, Ischyrolepis affinis Esterhuysen, I. caespitosa Esterhuysen, I. curvi-

bracteata Esterhuysen, /. feminea Esterhuysen, I. karooica Esterhuysen, /. longiaristata Pillans ex Linder, I. nana

Esterhuysen, I. nubigena Esterhuysen, /. papillosa Esterhuysen, I. pratensis Esterhuysen, /. rivula Esterhuysen, /.

sporadica Esterhuysen, /. unispicata Linder, /. wittebergensis Esterhuysen, Nevillea singulars Esterhuysen, Platy-

caulos acutus Esterhuysen, Restio colliculospermus Linder, R. corneolus Esterhuysen, R. fragilis Esterhuysen, R.

implicatus Esterhuysen, R. inconspicuus Esterhuysen, R. ingens Esterhuysen, R. inveteratus Esterhuysen, R. mon-

tanus Esterhuysen, R. nuwebergensis Esterhuysen, R. peculiaris Esterhuysen, R. perseverans Esterhuysen, R.

pillansii Linder, R. pulvinatus Esterhuysen, R. pumilus Esterhuysen, R. rarus Esterhuysen, R. rupicola Esterhuy-

sen, R. singularis Esterhuysen, R. vallis-simius Linder, R. verrucosus Esterhuysen, R. versatilis Linder, R. zuluen-

sis Linder, Staberoha ornata Esterhuysen, Thamnochortus cinereus Linder, T. arenarius Esterhuysen, T. rigidus

Esterhuysen, Willdenowia rugosa Esterhuysen). One new subspecies is described, Askidiosperma chartaceum (Pil-

lans) Linder subsp. alticolum Esterhuysen, and 84 new combinations are made. Fourteen of the new species are

illustrated. Extensive notes on the taxonomic history of the African Restionaceae, especially in so far as it affects

typification of the names, are given.

CONTENTS

Introduction 387

Historical background 388

Key to the genera 394

1. Staberoha 395

2. Ischyrolepis 397

3. Elegia 418

4. Chondropetalum 427

5. Dovea 431

6. Askidiosperma 431

7. Platycaulos 434

8. Restio 437

9. Calopsis 464

10. Thamnochortus 471

11. Rhodocoma 478

12. Ceratocaryum 479

13. Cannomois 480

14. Nevillea 482

15. Hydrophilus 484

16. Anthochortus 484

17. Mastersiella 487

18. Hypodiscus 488

19. Willdenowia 493

Excluded names 497

Nomina non satis cognitae 497

Acknowledgements 497

Uittreksel 497

References 497

Alphabetical index to the names applied to

the African Restionaceae 498

4 Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

INTRODUCTION

The Restionaceae is a remarkable family of South-

ern Hemisphere, evergreen, rush-like plants. Some

400 species in about 40 genera are known, the ma-

jority of which occur on the poor sandy soils of the

winter-rainfall regions of the SW and S portions of

the Cape Province, South Africa, and of S and SW

Australia. There are outliers further north in Africa,

Madagascar, Indo-China and Chile.

Of the 318 species of Restionaceae recognized in

this paper in Africa, all but four occur within the

Cape Floristic Region (sensu Goldblatt, 1978). This

is 4-5 % of the total flora of the region. Within the

Region the family is of great ecological importance

(Taylor, 1978). A taxonomic knowledge of the Res-

tionaceae is therefore of great importance to ecolog-

ists and land-managers. The last revision and keys to

the species were published in 1928. Since then, 70

species have been described, and in this paper a fur-

ther 51 new species are described.

The recent completion of a generic revision of the

African Restionaceae (Linder, 1984), in which sev-

eral genera were redelimited, necessitated some 74

new combinations. I have also had the opportunity

to study the historically important collections at K,

H, BM, LINN, P, Z, B, MEL, S, UPS, E, BOL, C

and OXF, and consequently have been able to typify

most of the names published for the African Restio-

naceae. In addition, Miss Esterhuysen made avail-

able her information on 51 new species and one new

subspecies. For 48 of the species and for the subspe-

cies, she also provided diagnoses and consequently,

the species names to which her name is attached

should be attributed to her. As it will still be some

388

Bothalia 15, 3 & 4 (1985)

years before the treatment of the family will be

ready for publication in the Flora of Southern

Africa, the Flora Zambesiaca and the Flora of Trop-

ical East Africa, it was decided to publish this Con-

spectus.

It must be emphasized that this Conspectus is not

a critical revision. Although some species complexes

have been resolved, there are still numerous in-

stances where the specific delimitations are highly

dubious. Several species complexes are very diffi-

cult, and may only be resolved after extensive field-

work — this applies particularly to genera like Can-

nomois and sections of Restio and Anthochortus. I

am also aware of the existence of several more

species, which are not described here due either to

the absence of adequate material at Kew (indeed

any herbaria), or because they are part of a complex

which has not yet been adequately resolved. Several

species are also listed in the Conspectus which, on

further study, may be found to be part of other

species. Where possible, I have made notes in the

Conspectus indicating possible taxonomic problems.

I have not yet produced any hypotheses on the

phylogeny of the species within the genera, and con-

sequently have not been able to produce infra-gen-

eric classifications. Consequently, I have listed the

species alphabetically within the genera and have

not given them numbers. The genera are arranged in

the sequence suggested by Linder (1984).

As the keys in this Conspectus are essentially draft

keys, often based to some extent on manuscript keys

which N. S. Pillans had produced, I would much ap-

preciate any comments on them, so that they can be

improved for the Flora accounts.

HISTORICAL BACKGROUND

There are not many pre-Linnaean records of Res-

tionaceae. Breynius illustrated a sterile twig in his

Centuriae Primae , tab. 91 (1678). The specimen de-

picted is in the BM, and is Restio quadratus. He

called it ‘Equisetum junceum nigrinodum capitis bo-

nae spei\

Plukenet in his Almcigesti Botanici Mantissa (1700:

109), refers to ‘Juncus Africanus lignoso calamo, ad

nodos involucris nigris circu voluto, panicula arundi-

nacea, ex Prom. Bon. Spei’; which is clearly also a

Restionaceous plant, although the identity of it is

obscure.

Scheuchzer refers to a ‘Canna Capitis bonae spei,

spicis iuliformibus; Juncus e Capitie bonae spei, pan-

iculis sulcis, juliformibus’ in his Agrostographia

(1709: 352). He illustrated the flower and the spike-

let. It may be Ischyrolepis ocreata or I. capensis.

During the lifetime of Linnaeus several botanists

worked on Restionaceae, but only a few species

were described. Linnaeus himself published Schoe-

nus capensis (= Ischyrolepis capensis ), but he con-

fused the i&ue by first changing the name to Restio

dichotomus, and then changing the diagnosis, so that

it fitted Thamnochortus lucens. Bergius (1767) de-

scribed Thamnochortus fruticosus, and Burman fil-

ius (1768) Equisetum capense (= Elegia capensis )

(see Schelpe, 1967). The major work on the Restio-

naceae during this period was by Rottboell (1772,

1773). He described and published 10 new species,

based on Konig collections. The holotypes of most

of the species are in C, with occasional isotypes in

other herbaria.

Linnaeus filius is the last person of this era, but he

already had access to Thunberg collections, on

which he based his three new species. The types of

two of them are at LINN, the type of the third is at

UPS.

C. P. Thunberg collected at the Cape between

1772 and 1774. His herbarium, which has been pre-

served intact at UPS, includes 61 sheets of Restiona-

ceae. On this rich collection Thunberg described 21

new species in several papers (1788, 1790, 1803).

The work was also repeated in his treatments of the

Flora Capensis (1794, 1811, 1823). Unfortunately his

descriptions are often inadequate (Nees, 1830) and

his treatments of the Flora were not updated to take

into account new material or descriptions {fide Steu-

del, 1929: 131). Neither Nees nor Kunth studied

Thunberg’s herbarium, and this resulted in many

FIG. 1. — Holotype of Restio filiformis Poir. in P.

Bothalia 15, 3 & 4 (1985)

FIG. 2. — Holotype of Hypodiscus tristachyus Mast, in B. The

two labels at the bottom (a) are in the hand of C.G.D. Nees

ab Esenbeck. Similar labels also occur in the Sonder material

at MEL. The pencilled determination label above (b) is by

Masters, whereas the determination label (c) below ‘TYPUS’

is by Pillans.

Thunberg names being misapplied or neglected.

This was only partially corrected when Masters

(1874) and Pillans (1928) studied his herbarium.

In 1804 Poiret published the botanical sections of

Lamarck’s Encyclopedic, in which he listed 35

species of Restio and 3 of Willdenowia, including 6

new species. Several species were illustrated in 1823.

The work is based on collections in Lamarck’s her-

barium and a few specimens in herb. Poiret. in the

general herbarium at P. (Fig. 1). There is no indica-

tion of who collected the material, or how it got to

Paris. This means that only holotypes can be recog-

nized.

Robert Brown (1810) laid the foundation of the

classification of the Australian Restionaceae. In this

work he suggested that several of the South African

species of Restio should be transferred to Thamno-

chortus, but he did not actually make any combina-

tions.

Sprengel’s account of the Systema Vegetabilium

(1824) is generally regarded as being poor (fide

Hoppe, 1825). He described several new species, but

his diagnoses are totally inadequate and his herbar-

ium has been scattered, so that it is difficult to find

out what his names refer to.

389

In 1828 Desvaux published a study on the Restio-

naceae, based on the manuscripts of Baron Palisot

de Beauvois. There is no indication of the material

on which this work was based and consequently no

types can be located. Fortunately, only one new

species is described in this paper, and this species is

adequately illustrated.

a

FIG. 3. — Holotype of Ceratocaryum fimbriatum (Kunth)

Linder. The writing on the ticket in the bottom lefthand cor-

ner (a) is probably that of Kunth, while the number is on a

typical blue Drege ticket (b). Note also the ‘vidi N.S.P.’ (c)

indicating that Pillans had seen the sheet.

By the end of the second decade of the nineteenth

century the first collections of Ecklon reached

Europe. These were being sold through the ‘Natur-

historischer Reiseverein’ of Wiirttemberg. These

Ecklon collections are all numbered and have local-

ity data. Steudel (1829) produced a listing of the

Restionaceae in this collection and described three

new species. There is some material of this collection

in herb. Sonder in MEL, in the Benhardi collection

at MO, and the occasional collection in some Euro-

pean herbaria, but I have not found any material of

many of the numbers.

The lengthy and somewhat critical review of the

Restionaceae which Nees ab Esenbeck published in

Linnaea 5 in 1830 was probably based on the same

Ecklon collection. Unfortunately Nees did not give

the collection numbers, only the locality data. This

390

makes it difficult to identify the types. As Nees con-

tinued working on the group after 1830, annotations

on specimens could have been made after the publi-

cation of the names. Nees described 11 new species

and recognized numerous varieties, several of which

eventually formed the basis of further new species.

These varieties are not validly published, as Nees did

not consistently comply with Articles 24—27 of the

Code, but labelled some of them with Greek letters

and others with polynomials.

Nees continued his work on the Restionaceae

after 1830 and in 1836 he described several new gen-

era. He also appears to have worked on the later

Ecklon & Zeyher collections, as on much of the ma-

terial in European herbaria there are printed labels

giving a Nees manuscript name, as well as the Eck-

lon & Zeyher locality code number (see Drege,

1847). In herb. Sonder in MEL this information is on

handwritten labels, but they are written in Sonder’s

hand. Masters (1865, 1868) took up many of these

Nees manuscript names. I have found very few

specimens annotated in Nees’s hand (see Fig. 2).

Kunth made a detailed study of the Restionaceae

for his Enumeratio Plantarum (1841). This was

FIG. 4. — Isotype of Restio festuciformis Mast, in MEL. Both

handwritten labels (a) are in Sonder's hand and give a Nees

manuscript name, as well as an Ecklon-Zeyher locality code

number. The label on the left also gives a number ‘37’, which

appears to be part of a numbering series for the species of

Restionaceae in herb. Sonder (see text). There is often a sec-

ond sheet giving the same information in a much neater

hand.

Bothalia 15, 3 & 4 (1985)

Si

It

b

a

FIG. 5. — Isotype of Cannomois nitida (Mast.) Pillans. The label

on the right (a) is in Zeyher's hand. The pencilled note (b) is

probably Sonder’s writing. There are several such pencilled

notes giving Nees manuscript names on sheets in MEL.

based largely on the very rich collections of Drege

(see Gunn & Codd, 1981), a few Garnot sheets,

some other material (largely from herb. Willdenow),

but strangely, no Ecklon and Zeyher collections.

Drege’s numerous duplicates can generally be recog-

nized by a small blue tag with a number (see Fig. 3).

The numbering is not chronological and was prob-

ably done after Drege had returned from his collect-

ing trip (1826-1832). Pillans prepared a list of the

Drege Restionaceous numbers and their determina-

tions (manuscript in Bolus Herbarium). A similar

list, based on the Drege specimens in Meyer’s her-

barium at Liibeck, was prepared by N. E. Brown

and is at Kew. Drege collections can either be local-

ized from his Zwei Pflanzengeographische Docu-

mente (1841) or from the original labels. Williams

(1972) reported finding a complete set of Drege col-

lections with the original labels of Leucadendron at

P. However, although there were some Drege orig-

inal labels of Restionaceae at P, these were rather

scrappy specimens and no types were represented.

In 1897 Kew received the Liibeck Drege collection

on loan and N. E. Brown copied all the locality data

onto the Kew duplicate set. The Liibeck collection

was acquired by B in 1916 and was largely destroyed

in the last war. Some of the surviving material has

original Drege labels.

Bothalia 15, 3 & 4 (1985)

391

As Kunth worked at Berlin, he presumably based

his work on the herb. Berol. set of Drege specimens

(see Fig. 3). The other two sets at B are from Liibeck

(acquired 1916) and Altona Museum (acquired

1924). I consider the herb. Berol. material as holo-

types of Kunth names, and material from Altona

and Liibeck as lectotypes if no herb. Berol. material

was extant.

Kunth produced detailed descriptions of 63 new

species. This number is high because he frequently

gave the male and female specimens of the same

species different names, although they were col-

lected at the same locality and were given conse-

cutive numbers (i.e. Restio setiger, based on Drege

2503 (female) and R. fuirenoides, based on Drege

2504 (male). Some of the confusion was at generic

level. All male specimens of Cannomois were de-

scribed as Thamnochortus and only female material

was included in Staberoha. Kunth did not study the

Thunberg herbarium and misapplied several Thun-

berg names (i.e. Restio parviflora Thunb. applied to

Elegia). Kunth names were widely accepted and

used, presumably because they were based on Drege

collections, which had duplicates in all major her-

baria, and were easily related to Kunth’s descrip-

tions by Drege’s excellent numbering system. Al-

though Kunth indicated several varieties, he did not

FIG. 6. — Lectotype of Restio pannosus Mast, in MEL. Note the

Zeyher ticket (a) with a number (1742) as well as a locality

code number. There are several species with such numbers

(between 1 700 and 500) in the Restionaceae.

FIG. 7. — Holotype of Ischyrolepis subverticellata Steud. in P.

Note the label reading ‘Herbarium Steudel’.

give them names and these varieties cannot be

upheld.

In 1844—1845 Ferdinand Krauss published his ac-

count of the plants which he collected during his tra-

vels in the Cape and Natal in 1838-1839. The four

new species of Restionaceae in this account were de-

scribed by Hochstetter and are attributed to him, but

the one new combination is best attributed to

Krauss. I have not been able to trace the Krauss col-

lection, but there are some duplicates in OXF. Al-

though there are original Krauss labels at NY, I

found no Krauss types there. Hochstetter also sug-

gested (as a footnote in the same publication) that

several of Kunth's species of Thamnochortus should

be transferred to Cannomois, but he did not make

the actual combinations.

Steudel described several new species in his Syn-

opsis Plantarum Glumacearum (1855). Two of these

species were already mentioned in a report on the

status of the Synopsis in Flora 33 (Steudel, 1850) but

they were only validly published in 1855. Steudel's

herbarium is in P, and many of the types of his

names may be found in it, (Fig. 7) but of several

species no material was found.

During the second half of the 19th Century Max-

well Masters was established as the specialist on the

Restionaceae. Between 1865 and 1901 he published

392

Bothalia 15, 3 & 4 (1985)

FIG. 8. — Lectotype of Restio quadratus Mast, in K. Note the

Burchell label (a). The locality is given in N.E. Brown's hand

(b), and the determination (c) is in Masters’s hand.

144 new species and created the general framework

of the generic classification.

In his first two papers on the Restionaceae (1865,

1868), Masters produced a review of the African

species of the family with descriptions of new species

and copious notes on the other species. This review

was based on the herbaria of Hooker, Harvey, Dau-

beny and Sonder. Sonder’s herbarium was the most

important, as it was more complete and contained

many of Nees von Esenbeck’s typical species, la-

belled by the author himself, and frequently also ac-

companied by manuscript notes (Masters, 1868:

211). Hooker’s herbarium is now at K, that of Har-

vey at TCD, but the whereabouts of Sonder’s her-

barium is somewhat puzzling. According to

Nordenstam (1980), Sonder’s African herbarium

was sold to Stockholm, and he presents evidence

that for the groups investigated it is indeed as S.

Gandoger (1913) claimed that he had bought the

herbarium and was intending to leave it to Z. There

is no evidence that it is there, and there does not

seem to be much Sonder herbarium material at Lyon

(although I have not studied the herbarium at LY

myself). However, I have only found a few frag-

ments of Sonder’s Restionaceae at S. At MEL there

are some 330 sheets of Restionaceae from the Son-

der herbarium. This includes 78 species numbered

consecutively (with some gaps) from 1 to 98 (see Fig.

4). For most of the numbers there are two sheets.

One is labelled with a Nees manuscript name or a

previously published name, an Ecklon & Zeyher

locality code number and the herbarium number in

Sonders’ hand, whereas the other has the same in-

FIG. 9. — Type material of Restio praeacutus Mast, in K. The

writing marked ‘a’ is of N.E. Brown, done when the Liibeck

material was on loan to Kew in 1897. The writing marked ‘b’

is by Masters, indicating that before 1897 he included this

material in R. furcatus. Note that Masters often wrote in pen-

cil. I have not been able to identify the writing below that.

The collection number ‘1608’ is a typical Drege number on a

blue tag (c).

formation in a much neater hand. I have not found

the sheets labelled in Nees’s hand, to which Masters

referred, nor are any of the sheets annotated by

Masters. Several types which Masters claimed to

have seen in herb. Sonder are not in the MEL collec-

tion. So, although it cannot be doubted that there is

much Sonder herbarium material in MEL, there is

still some critical material missing.

In addition to the Ecklon & Zeyher material, the

MEL herb. Sonder material also comprised some

Drege collections, with Drege numbers on blue or

pink tags, some early Ecklon collections, a few late

Zeyher collections (Fig. 6) and a few Bergius,

Schlechter, Burchell and MacOwan collections.

Bothalia 15, 3 & 4 (1985)

393

dot.

ETL:"'

FIG. 10. — Lectotype of Ischyrolepis coactilis Mast. Note the typ-

ical Schlechter labels and the pencilled determination label of

Masters.

Pianiae Scla-cMenaniiK Ausiro-Africanae.

Typus!

In his first paper. Masters dealt only with the

genus Restio and in his second paper he dealt with

additional material of Restio, as well as the remain-

ing genera. Prominent among the additional ma-

terial which he studied for the 1868 paper are the

Burchell collections (Fig. 8). Although Burchell

made his African collections in 1811-1815, Kew only

acquired the material in 1865. Burchell kept the ma-

terial until his death in 1863, when he bequeathed it

to his sister, who donated the material to Kew in

1865 (Hooker, 1865). The duplicates of the collec-

tion were distributed in 1867 (Hooker, 1867). It is

very unfortunate that this material was not available

for study when it was first collected, as the material

is in superb condition and is chronologically num-

bered. It would have made much better type ma-

terial than the Ecklon & Zeyher collections.

In 1874 Masters studied the Thunberg herbarium.

This resulted in some nomenclatural corrections.

However, as Masters followed the ‘Kew Rule’, many

necessary name changes were left for Pillans, as

Thunberg only recognized three genera, and most of

his species were described in Restio.

In 1878 Masters summarized his treatment of the

family (extended to include a few new species and

varieties, as well as the Australian taxa) in A. De

Candolle’s Monographiae Phanerogamarum, giving

full descriptions for all the species. In the last quar-

ter of the century the collections of Rehmann, Bo-

lus, MacOwan and others started arriving in Europe

and in 1886 Masters described some new species

based on these collections. The material from Mey-

er’s herbarium in Liibeck was loaned to Kew in 1897

and was studied by both Masters and N. E. Brown.

Of special importance was that the locality data of

the Drege collections were transcribed by N. E.

Brown onto the Kew sheets (Fig. 9). On the basis of

his earlier work, as well as the Liibeck material,

Masters wrote up the Restionaceae for the Flora Ca-

pertsis (1897).

The material from Berlin was loaned to Kew in

1900. This included the Schlechter collections (Fig.

10), made between 1891 and 1898, as well as various

Ecklon and Zeyher collections. It is not clear

whether only unnamed material was sent. Although

Schlechter collected numerous duplicate sets, it is

obvious that the entire collection was sent to Mas-

ters, as numerous duplicates in various European

herbaria carry Masters’s det. slips. The material was

probably distributed after Masters had named it.

Consequently there are no holotypes. As the top set

was probably kept in Berlin, where this material is

extant, I have selected lectotypes from it. In the 1900

paper Masters also described several species based

on Ecklon and Zeyher material which is only at B.

FIG. 11. — Holotype of Restio strictus N.E. Br. in K, with a Gai-

pin label (a). Note the Pillans determination label (b).

394

Bothalia 15, 3 & 4 (1985)

In 1900 N. E. Brown published an appendix to

Flora Capensis, in which he described 15 new species

based on collections made by Galpin in 1897 along

the coastal mountains from Humansdorp to Cape

Town (see Fig. 11). Unfortunately N. E. Brown’s

work was not critical and most of his species were

reduced to synonymy by Pillans (1928).

Dod published a paper with new species from the

Cape Peninsula in 1901. In this paper two new

species of Restionaceae are described. These are

correctly attributed to Masters, as Dod noted in the

introduction to his paper ‘I am indebted to Dr Mas-

ters and to Mr N. E. Brown for the descriptions of

the species to which their names are appended’.

Towards the end of the last century tropical Africa

became better known botanically and in 1897 N. E.

Brown described Hypolaena mahonii (= Restio ma-

honii ) from Malawi. In 1922 Chermozen described

Restio madagascariensis from Madagascar; the tropi-

cal Southern African Restionaceae are still being lo-

cated and their ranges extended (i.e. Restio zuluen-

sis, in this work).

From 1920 until his death in 1964 Neville Pillans

was the specialist on the African Restionaceae. His

first papers were descriptions of new species and

new combinations (1921, 1922). In 1922 he spent

some months at Kew, studying the collections of

Masters, as well as receiving on loan the Restiona-

ceae of all the major European herbaria. With this

rich and important material available he was able to

place the nomenclature on a very sound footing, so

that in the present study less than 20 nomenclatural

name changes have to be made. At the Bolus Her-

barium in Cape Town, where he was based, he had

available the rich local collections made by Marloth,

MacOwan, Bolus, Stokoe and many others. Based

on this material he resolved most of the specific deli-

mitation problems and reduced many of Masters’s

species to synonymy. His monograph (1928) can

only be flawed on the generic concepts: his nomen-

clature and species concepts are generally very good.

In the 1930’s Miss Esterhuysen started collecting

Restionaceae. These collections, with those of

Levyns, Stokoe and a few others, resulted in a fur-

ther 70 species being described in three publications

by Pillans (1942, 1945, 1952). Since 1952 material of

a further 55 species has been collected by Miss Ester-

huysen. These species are described in this paper.

KEY TO THE GENERA AND SOME SPECIES OF THE AFRICAN RESTIONACEAE

(Numbers preceding names indicate genus numbers used in the text. Species are arranged alphabetically)

la Style solitary:

2a Fruit dehiscent 2. Ischyrolepis feminea

2b Fruit indehiscent:

3a Perianth chartaceous; a rare plant of the Great Swartberg 1. Staberoha stokoei

3b Perianth cartilaginous:

4a Male spikelets erect 9. Calopsis monostylis

4b Male spikelets pendent 10. Thamnochortus

lb Styles 2-3:

5a Styles 2:

6a Ovary dehiscent:

7a Styles fused at the base, if free, bracts longer than the flowers 2. Ischyrolepis

7b Styles free; bracts shorter than the flowers 4. Chondropetalum microcarpum

6b Ovary indehiscent:

8a Culms simple:

9a Male flowers in spikelets:

10a Male spikelets pendent; nut enclosed by the perianth 1. Staberoha

10b Male spikelets erect; perianth shorter than the nut, if as large, then hyaline:

11a Nut flattened on one side; female spikelets usually with several flowers 13. Cannomois

lib Nut round in cross-section:

12a Male bracts acute; female flowers always with a single flower 18. Hypodiscus

12b Male bracts rounded; female spikelets with several flowers 14. Nevillea

9b Male flowers racemose:

13a Tepals of the male flowers valvate, lanceolate 12. Ceratocaryum

13b Tepals of the male flowers linear 19. Willdenovvia

8b Culms branching:

14a Female spikelets with several flowers:

15a Bracts of female flowers hyaline-chartaceous 15. Hydrophilus

15b Bracts of the female flowers coriaceous to osseous:

16a Bracts shorter than the perianth 4. Chondropetalum microcarpum

16b Bracts longer than the perianth:

17a Nut less than 5 mm long

9. Calopsis esterhuyseniae

Bothalia 15, 3 & 4 (1985)

395

17b Nut more than 10 mm long 13. Cannomois

14b Female spikelets with a solitary flower:

18a Bracts shorter than the perianth and nut 3. Elegia

18b Bracts overtopping the perianth and nut:

19a Bracts of the male flowers like the perianth lobes, linear, hyaline; male flowers in racemes

19. Willdenowia

19b Bracts of the male flowers larger than the perianth, not linear; male flowers in spikelets:

20a Bracts of the female flowers osseous; male flowers in cone-like spikelets

17. Mastersiella

20b Bracts of the female flowers chartaceous; male flowers in few-flowered spikelets

16. Anthochortus

5b Styles 3:

21a Sheaths usually deciduous; bracts shorter than the flowers, or if longer, then hyaline and lacerated:

22a Ovary unilocular, indehiscent 3. Elegia

22b Ovary 3-locular, usually at least some locules dehiscent:

23a Culms branching 5. Dovea

23b Culms simple:

24a Bracts longer than the flowers 6. Askidiosperma

24b Bracts shorter than the flowers, or if longer, then not hyaline 4. Chondropetalum

21b Sheaths persistent; bracts usually cartilaginous, often overtopping the flowers;

25a Culms flattened; sheaths green with a stout mucro; seeds white with a fragile surface

7. Platycaulos

25b Culms terete, if flattened, then not with the above characters:

26a Culms with a simple central axis, and with sterile and/or fertile branches clustered at the

nodes; male spikelets pendent 11. Rhodocoma

26b Culms usually branching, if simple, then without branches clustered at the nodes; male spike-

lets erect:

27a Fruit a capsule, if a nut, then the bracts narrowly lanceolate, concolorous:

28a Culms branching 8. Restio

28b Culms simple:

29a Bracts shorter than the flowers, spikelets numerous at several nodes

4. Chondropetalum

29b Bracts at least as long as the flowers; spikelets few to several 8. Restio

27b Fruit a unilocular nut:

30a Culms branching 9. Calopsis

30b Culms simple 1. Staberoha

KEY TO THE SPECIES OF STABEROHA

la Lateral sepals keeled:

2a Bracts 0,6-1 cm long; perianth 2-2,5 mm long

2b Bracts 1, 2-1,5 cm long; perianth 3-3,5 mm long:

3a Spikelets solitary, about 2 cm long, elliptic; ovary elliptic -

3b Spikelets 1-3, 2,5-5 cm long, oblong or lanceolate; ovary obovate

lb Lateral sepals winged:

4a Wings cartilaginous, deeply lacerated with slender irregular tough spidery lobes

4b Wings membranous to cartilaginous, entire or finely lacerated:

5a Tepals acute:

6a Tepals linear-lanceolate; styles 2; spikelets more than 2 cm long

6b Tepals oblanceolate; style solitary; spikelets up to 2 cm long

5b Tepals obtuse:

7a Tepals 2,5-3 mm long; wings finely lacerated

7b Tepals 3-5 mm long; wings entire or crenate;

8a Ovary acute; styles 2; spikelets 1(2)

8b Ovary rotund; styles 3; spikelets several

5. vaginata

5. remota

5. banksii

5. ornata

. . S. distachyos

5. stokoei

S. cernua

S. aemula

S. multispicula

396

Bothalia 15, 3 & 4 (1985)

STABEROHA

1. Staberoha Kunth, Enum. PI. 3: 442 (1841);

Linder in Bothalia 15:63 (1984).

Staberoha aemula (Kunth) Pillans in Trans. R.

Soc. S. Afr. 16 : 386 (1928).

Thamnochortus aemulus Kunth, Enum. PI. 3 : 439 (1841).

Type: Cape, 3319 (Worcester): Du Toits Kloof, 900-1 200 m

(-CA), Drege 1652 cf (B, holo.!; BM!; BOL!; K!;NY!; OXF!; P!;

S!).

Staberoha stenoptera Kunth, Enum. PI. 3 : 443 (1841). Thamno-

chortus imbricatus (Thunb.) Mast. var. stenopterus (Kunth) Mast,

in J. Linn. Soc., Bot. 10 : 231 (1868); in A. DC., Monogr. Phan. 1

: 324 (1878); in FI. Cap. 7 : 127 (1897). Staberoha imbricata

(Thunb.) Kunth var. stenoptera (Kunth) Dur. & Schinz, Consp.

FI. Afr. 5 : 521 (1894). Tvpe: Cape, 3319 (Worcester): Du Toits

Kloof, 900-1 200 m (-CA), Drege 1636 9 (B, holo.!; BM!; BOL!;

K!; MO!; P!).

Restio tetrasepalus Steud., Syn. PI. Glum. 2 : 251 (1855). Type:

Cape, 3319 (Worcester) : Du Toits Kloof, 900-1 200 m (-CA),

Drege 27 9 (P, holo.!; BM!; K!; NY!).

Notes

1. The specimen of Drege 27 in P is in herb. Steu-

del, so it can be regarded as the holotype of R. tetra-

sepalus.

2. The types of Staberoha stenoptera and Thamno-

chortus aemulus are probably the female and male of

the same collection.

3. S. aemula is generally a high-altitude species,

occurring between 1 000 and 1 800 m in the moun-

tains from the Cedarberg to the Klein Swartberg, Ri-

viersonderend Mountains and the Caledon Paarde-

berg. A few collections reach as low as 600 m. It is a

fairly common species.

Staberoha banksii Pillans in Trans. R. Soc. S.

Afr. 16 : 385 (1928); in Adamson & Salter, FI. Cape

Penins. 153 (1950). Syntypes: Cape, 3418 (Simons-

town): Mountains at Simonstown (-AB), Pillans

3405 $ (BOL, lecto.!); 3405 C? (BOL!); between

Wynberg and Constantia (-AB), Burchell 789 cf

(BOL!; K!); Princess Vlei (-AB), MacOwan 1789 cf

(BOL!; K!); Smitswinkel Bay (-AD), Dod 2548 cf,

2549 $ (BOL!; K!); Muizenberg Mountain (-AB),

Moss 845 9 (BM!; BOL!); 3318 (Cape Town): Cape

Flats (-DC), Burke s.n. cf (BM!; BOL!; BR!; K!);

Zeyher s.n. cf (BOL!; K!); Mamre (-CB), Baur 1180

Cf (BOL!; K!); 3419 (Caledon): Zoetmelk Vlei

(-DB), Wallich 9 (BM!; BOL!; K!); without precise

locality, Banks s.n. C f (BM!;BOL!); Verreaux s.n. cf

& $ (BOL!).

Notes

1 . This species is very close to S. vaginata, with the

only difference being in size. This may be biologi-

cally trivial.

2. S. banksii ranges from Bredasdorp to the Hex

River Mountains, and to the Cape Peninsula, from

sea level to about 1 600 m.

Staberoha cernua (L.f.) Dur. & Schinz, Consp.

FI. Afr. 5 : 520 (1894); Pillans in Trans. R. Soc. S.

Afr. 16 : 387 (1928); in Adamson & Salter, FI. Cape

Penins. 154 (1950). Type: Cape, 3318 (Cape Town):

Hills at E base of Table Mountain (-CD), in herb.

Thunb. 23226 (UPS, lect.!; H!; S!).

Restio cernuus L.f., Suppl. 425 (1781); Thunb., FI. Cap. edn 1,

319 (1811); edn Schultes 83 (1823). Thamnochortus cernuus (L.f.)

Kunth, Enum. PI. 3 : 439 (1841); Mast, in A. DC., Monogr. Phan.

1 : 325 (1878); in FI. Cap. 7 : 128 (1897).

leones: A. DC., Monogr. Phan. 1 : t.2 f.28-37 (1878). Rice &

Compton, Wild Flow. Cape G.H. t.250 f.4, 5 (1950).

Notes

1. Linnaeus filius described R. cernuus from a

Thunberg collection. There is no material of this

species in LINN, but there is excellent material at

UPS, S and H. These may all be isotypes (unless the

species was actually described by Thunberg!), so I

lectotypify the material at UPS.

2. S. cernua is wide-ranging, reaching from the

Kouga and Swartberg Mountains to Ceres, Piket-

berg and the Cape Peninsula. There is only one col-

lection from the Cedarberg.

Staberoha distachyos (Rottb.) Kunth , Enum. PI.

3: 444 (1841); Pillans in Trans. R. Soc. S. Afr. 16:

389 (1928); in Adamson & Salter, FI. Cape Penins.

154 (1950). Type: Descriptionum et Iconum Ra-

riores, t.3, fig. 5 (Iconotype).

Restio distachyos Rottb., Descriptionum et Iconum Rariores 8

(1773).

Restio imbricatus Thunb., Diss. Restio 9 (1788); FI. Cap. edn 1,

316 (1811); edn Schultes, 83 (1823). Staberoha imbricata (Thunb.)

Kunth, Enum. PI. 3: 442 (1841). Thamnochortus imbricatus

(Thunb.) Mast, in J. Linn. Soc., Bot. 10: 231 (1868); in A. DC.,

Monogr. Phan. 1: 323 (1878); in FI. Cap. 7: 127 (1897). Type:

Cape, without precise locality, in herb. Thunberg. 23237 9 (UPS,

holo!; BM!).

Restio umbellatus Thunb., Diss. Restio 11 (1788); Thunb., FI.

Cap. edn 1, 320 (1811); edn Schultes, 84 (1823). Thamnochortus

umbellatus (Thunb.) Kunth, Enum PI. 3: 440 (1841); Mast, in

A. DC., Monogr. Phan. 1: 324 (1878); in FI. Cap. 7: 126 (1897).

Syntypes: Cape, without precise locality, in herb. Thunberg.

23261 Cf (UPS, lecto.!; S!); 23260 (UPS!).

Icon: Rottb., Descriptionum et Iconum Rariores t.3 f. 5 (1773).

Notes

1. There is no specimen of R. distachyos in the

Rottboell herbarium in Copenhagen, nor have I

seen any Konig collections which might be isotypes.

As the figure which Rottboell published is quite ad-

equate, it may be regarded as an iconotype.

2. There are two collections annotated as 'Restio

umbellatus’ in herb. Thunberg. The one is identical

to the specimen annotated as ‘Restio cernuus' (Thun-

berg herb. 23226), and the other agrees with Restio

distachyos. I am following the implicit lectotypifica-

tion of Pillans (1928) in regarding R. umbellatus as a

synonym of R. distachyos.

3. S. distachyos is a wide-ranging species, re-

corded from clayey, sandy or gravelly soils, from sea

level to 1 000 m. It is distributed from Bredasdorp to

Van Rhynsdorp, and is quite common on the lower

slopes and sandy coastal forelands.

Staberoha multispicula Pillans in J1 S. Afr. Bot.

18: 117 (1952). Type: Cape, 3419 (Caledon): Frans-

kraal (-CD), Leighton 1897 $ (BOL, lecto.!; K!);

1897 Cf (BOL!; K!).

Note

1. S. multispicula occurs in the Caledon and Bre-

dasdorp areas, generally below 300 m, in sandy or

peaty places.

Staberoha ornata Esterhuysen, sp. nov. a S. dis-

tachyote (Rottb.) Kunth marginibus tepalorum

laceratis, carinis sepalorum lateralium lobatis dif-

fert.

Bothalia 15, 3 & 4 (1985)

397

TYPE. — Cape, 3319 (Worcester): between Bertsberg

and Buffelshoek Peak, above Ceres Plain (-AD), Esterhuysen

30823, $ (BOL, holo.!; B; C; E; F; GRA; K; L; LD; M; MO;

NBG; NY; PRE; RSA; S; STE; TCD; UC; US; W; WAG).

Plants caespitose, tussocks 15-40 cm tall, culm

bases aggregated. Culms solid, terete, simple, to 1

mm in diameter, surface smooth to obscurely

wrinkled. Sheaths closely convoluted, 20-35 mm

long, body coriaceous, green to brown, very acute,

somewhat chartaceous towards the margins, awn

very slender, aciculate, 2-3 mm long, older sheaths

decaying and lacerated, culms with only 1-2 sheaths.

Male inflorescence a single cluster of 3-6, 8-25 x

5-10 mm pendent spikelets. Basal spathe like the

sheaths, 1-2 cm long, branches flattened, about as

long as the spikelets. Spathes and bracts similar, im-

bricate, cartilaginous, acute, 4—6 x 1,5-2 mm,

shortly mucronate to emucronate. Flowers subses-

sile, perianth 4 mm long. Tepals chartaceous, mar-

gins finely serrate to lacerated, acute; lateral sepals

conduplicate, 4 mm long, subcarinate; petals and

odd sepal flat, 3,5 x 1 mm, similar. Anthers not ex-

serted at anthesis, 1,9-2, 3 mm long. Stylodium min-

ute. Female inflorescence a solitary, 15—35 x 5 mm

spikelet. Spathe and lower bracts like the sheaths,

about half as long as the spikelets. Upper bracts fer-

tile, like the lower bracts but 8-12 mm long, imbri-

cate, scarcely mucronate. Flowers sessile, perianth

2,3-4 mm long. Sepals cartilaginous; laterals 2, 5-3, 5

mm long, piliferous, conduplicate, margins finely

lacerated, carina irregularly and deeply lobed, lobes

0,5-1 mm long, margins sparsely finely lacerated;

odd sepal 2,5-3 x 0,5 mm, acute, margins finely

lacerated. Petals 2-3 x 0,8 mm, lanceolate, acute,

margins lacerated, with one to numerous tooth-like

processes towards the base of the midrib. Stami-

nodes minute, 0,3 mm long. Styles 2, swollen and

fused at the base, about 3 mm long, strap-shaped,

pilose on the inner margin. Fruit an elliptical nutlet,

shed with the perianth. Fig. 12.

S. ornata occurs widespread between the Cedar-

berg and the Hex River Valley, between 600 and

1 200 m. The populations occur on the sandy flats or

plateaux, often along the margins of damp spots.

Flowering occurs in early spring, in July and August.

This species is unique in Staberoha in having cu-

riously lobed and lacerated tepals. It is probably

most closely related to S. distachyos by the winged

sepals, and the general size of the bracts and flowers.

Note that the flower size in this species is very vari-

able.

CAPE. — 3219 (Wuppertal); South Cedarberg, W of Apollo

Peak (-CA), Esterhuysen 34156 (BOL; K; M; MO; S); South Ce-

darberg, Suurviakte (-CA), Esterhuysen 34252 (BOL; K; MO;

S); between Donkerhoek Kop and the Apollo Plateau (-CA), Es-

terhuysen 34153 (BOL; C; E; F; K; L; LD; M; MO; NBG; NY; S;

STE; TCD; UC; US; W; WAG); Northern Cold Bokkeveld, Vre-

delus (-CC), Esterhuysen 34680 (BOL; C; E; K; L; M; MO; S);

Schurweberg between Bokkeveld Tafelberg and Sneeuberg

(-CD), Esterhuysen 32620 (BOL; K; S). 3319 (Worcester): 24

River Mountains, Suurviakte (-AA), Esterhuysen 34848 (BOL;

K); Schurfteberg above Kweperfontein-Merwede (-AD), Ester-

huysen 34088 (BOL; K; S); Milnervlakte, Hex River Mountains

(-AD), Esterhuysen 35529 (BOL; C; E; L; M; MO; S); between

Bertsberg and Buffelshoek Peak (-AD), Esterhuysen 30823 (B;

BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY; PRE; RSA;

S; STE; TCD; UC; US; W; WAG).

Staberoha remota Pillans in Trans. R. Soc. S.

Afr. 29: 351 (1942). Type: Cape, 3319 (Worcester):

Roodeberg, Ceres, 1 800 m (-BC), Esterhuysen

1493 $ (BOL, lecto.!; K!); 1492 cf (BOL!; K!).

Note

1. This is a high-altitude species, occurring be-

tween ( 1 200) 1 500 - 2 100 m in the mountains from

Matroosberg (Ceres) to Paarl and Fransch Hoek.

Most of the collections are from marshy places.

Staberoha stokoei Pillans in J1 S. Afr. Bot.

18:118 (1952). Type: Cape, 3322 (Oudtshoorn):

Swartberg Pass area (-AC), Stokoe 9011 f (BOL,

lecto.!); 9011 cf (BOL!; K!).

Note

1. S. stokoei is only known from the Swartberg be-

tween the Swartberg Pass and Meiringspoort. It oc-

curs above 1 800 m in moss on sloping rock and

cliffs.

Staberoha vaginata (Thunb.) Pillans in Trans.

R. Soc. S. Afr. 16:384 (1928); in Adamson & Salter,

FI. Cape Penins. 153 (1950). Type: Cape, without

precise locality, in herb. Thunberg. 23262 $ (UPS,

holo.!).

Restio vaginatus Thunb., Diss. Restio 10 (1788); Thunb.; FI.

Cap. edn 1, 317 (1811); edn Schultes, 83 (1823); Kunth, Enum.

PI. 3:408 (1841).

Restio spicifer Poir. in Lam. Tabl. Encycl. 3:399 (1823). Type:

Lam., Tabl. Encycl. t. 804 f. 2 (1799) (Iconotype).

Icon: Lam., Tabl. Encycl. t. 804 f. 2 (1799).

Notes

1. The plate of Restio spicifer is not very clear, but

this name would appear to be best placed here.

2. This species is very similar to S. banksii and

may, in fact, be conspecific.

3. S. vaginata ranges from the Kogelberg to the

Cape Peninsula and the Cedarberg. It is known from

rocky areas between 450 and 1 200 m, but is not

common.

ISCHYROLEPIS

2. Ischyrolepis Steud., Syn. PI. Glum. 2:249

(1855); Linder in Bothalia 15:63 (1984).

Notes

1. In couplet 1 of the key, the number of flowers

per spikelet overlaps. If a species has 1-3 flowers per

spikelet, both la and lb should be tried.

2. The shape of the apex of the coriaceous portion

of the sheath (couplet 3) can be difficult to deter-

mine, and it is often not obvious on older sheaths

where the hyaline shoulders have decayed.

3. Woolly scales in the axils of the sheaths (couplet

35) do not usually occur in all the sheaths.

4. Rhizomes are quite distinct from the culms,

whereas stolons are spreading and rooting culms. As

a rule of thumb rhizomes have- small, overlapping

scales, whereas stolons have sheaths similar to those

of the aerial culms.

398

Bothalia 15, 3 & 4 (1985)

FIG. 12. — Staberoha ornata Esterhuysen. a, habit; b, female plant, with very short rhizomes, simple culms and solitary spikelets, x

0,8; c, male plant, with clusters of 2—4 pendulous spikelets, x 0,8; d, female bract, x 4; e, female flower, showing deeply-lobed

lateral sepal wings, x 9; f, odd sepal of the female flower, x 9; g, petal of the female flower, x 9; h, ovary with slender strap-

shaped styles, x 9; i, male flower, x 9; j, male lateral sepal, x 9; k, male odd sepal, x 9; 1, male petal, x 9; m, anthers, x 9, n,

male bract, x 6. (From Esterhuysen 30823.)

Bothalia 15, 3 & 4 (1985)

399

KEY TO THE SPECIES OF ISCHYROLEPIS

la Spikelets 1-2-flowered:

2a Branches verticellate I. subverticeliata

2b Branching dichotomous:

3a Leaf-sheaths with the coriaceous portion truncate or subtruncate:

4a Rhizome well-developed, creeping /. sabulosa

4b Rhizome absent:

5a Male flowers 5-6 mm long; male spikelets 10-15 mm long /. pratensis

5b Male flowers 2-4 mm long; male spikelets 4-10 mm long:

6a Terminal sections of the culms very flexuose:

7a Perianth glabrous, 4-lobed; seed verrucate I. caespitosa

7b Perianth villous, 6-lobed; seed smooth I. cincinnata

6b Terminal sections of the culms straight or slightly flexuose:

8a Spikelets more than three:

9a Spikelets 5 mm long, slender, closely adpressed to the culm I. rottboellioides

9b Spikelets about 3 mm long, not closely adpressed to the culm I. duthieae

8b Spikelets 1-3:

10a Seed warty; plants straggly I. curviramis

10b Seed reticulate-foveate; plants caespitose:

11a Female spikelets usually with more than 1 flower; culms rarely tuberculate I. nana

lib Female spikelets single-flowered; culms usually tuberculate I. sporadica

3b Sheaths with the coriaceous portion acute to obtuse:

12a Culms roughly tuberculate:

13a Culms much-branched; sheaths with hyaline apices:

14a Terminal branches very flexuose I. cincinnata

14b Terminal branches straight:

15a Stylopodium pilose; seed reticulate-foveate I. karooica

15b Stylopodium glabrous; seed smooth I. esterhuyseniae

13b Culms sparsely branched; sheaths without hyaline apices I. unispicata

12b Culms finely tuberculate, rugulose or smooth:

16a Spikelets 1(2-3):

17a Plants with deep-seated shiny black rhizomes; growing in coastal sand /. eleocharis

17b Plants without rhizomes; not from coastal sand:

18a Spikelets 7-15 mm long; sheaths with woolly scales:

19a Sheaths and bracts muticous I. unispicata

19b Sheaths and bracts with slender awns I. coactilis

18b Spikelets 2-5 mm long; sheaths without woolly scales:

20a Ultimate branches flexuose; culms about 1 mm in diameter:

21a Perianth villous, 6-lobed /. cincinnata

21b Perianth glabrous, 4-lobed I. tenuissima

20b Ultimate branches straight; culms about 0,5 mm in diameter:

22a Plants stoloniferous, matted /. rivula

22b Plants not stoloniferous; caespitose /. caespitosa

16b Spikelets several to many:

23a Style solitary,' simple /. feminea

23b Styles 2 or deeply bilobed:

24a Culm bases swollen, reddish:

25a Male spikelets with several flowers; seed colliculate /. paludosa

25b Male spikelets single -flowered; seed papillose L papillosa

24b Culm bases not swollen, brown:

26a Plants 15-20 cm tall /. duthieae

26b Plants 30-50 cm tall /. monanthos

lb Spikelets several- to many-flowered:

27a Sheaths loosely convoluted to flat:

28a Culms branching:

29a Branches verficellate /. subverticeliata

400

Bothalia 15, 3 & 4 (1985)

29b Branches dichotomous:

30a Bracts without awns /. gaudichaudiana

30b Bracts with awns:

31a Culms tuberculate:

32a Coriaceous portion of the sheaths acute; culms tuberculate; tepals acute /. capensis

32b Coriaceous portion of sheaths truncate; tepals acuminate; culms roughly tuberculate

I. longiaristata

31b Culms rugulose or smooth:

33a Sheaths acute to acuminate; perianth 6-7 mm long I. ocreata

33b Sheaths obtuse to rounded; perianth 5-6 mm long I. fraterna

28b Culms simple:

34a Bract apices erect; styles free to the base I. fuscidula

34b Bract apices reflexed; styles united at the base I. aff. ocreata

27b Sheaths tightly convoluted:

35a Sheaths generally with an exserted woolly scale from the axil:

36a Style column conspicuously villous near the base I. gossypina

36b Style column glabrous:

37a Culms obsoletely tuberculate:

38a Flowers 6 mm long; spikelets 1 cm long I. distracta

38b Flowers 2-3 mm long; spikelets 5 mm long I. arida

37b Culms smooth or obscurely rugulose:

39a Sheaths nervose-striate, grey or yellow-speckled over the entire outer surface /. sieberi

39b Sheaths smooth and glossy, not speckled as above:

40a Perianth 4 mm long I. coactilis

40b Perianth 6-7 mm long I. laniger

35b Sheaths without a partially exserted woolly scale:

41a Lateral sepals glabrous I. pygmaea

41b Lateral sepals villous-carinate:

42a Awn on bracts at least as long as the body:

43a Plants 10-30 cm tall /. nana

43b Plants 30-100 cm tall:

44a Spikelets numerous; perianth 7-9 mm long; culms moderately branched I. hystrix

44b Spikelets 1-5; perianth 5-6 mm long; culms simple or sparingly branched:

45a Awns as long as the body of the bract /. ocreata

45b Awns longer than the body of the bract:

46a Male bracts 10 mm long, obscuring the flowers; from Swellendam I. affirtis

46b Male bracts 3 mm long, male flowers exposed; from Cedarberg I. setiger

42b Awn on bracts shorter than the body of the bract or absent:

47a Spikelets subulate or cylindrical, very acute at the apex (character obscured in fruiting material):

48a Culms filiform, rugose or finely tuberculate; coriaceous portion of the sheaths acute:

49a Young sheaths with hyaline shoulders; male spikelets curved; from Paarl to Port Eliza-

beth /. triflora

49b Young sheaths without hyaline shoulders; male spikelets straight; from the Kamies-

berg I. vilis

48b Culms slender, rugulose or smooth; coriaceous portion of the sheaths obtuse or rounded:

50a Spikelets 1-2, 1-2 cm long; perianth 3,5-4 mm long /. helenae

50b Spikelets several to many, 0,6-1 cm long; perianth 2,5-3 mm long /. wallichii

47b Spikelets oblong or ovate-lanceolate, oblong-elliptic or ovate:

51a Culms tuberculate:

52a Coriaceous portion of sheaths truncate I. macer

52b Coriaceous portion of the sheaths acute:

53a Culms very prominently tuberculate:

54a Stylopodium glabrous; seed smooth /. esterhuyseniae

54b Stylopodium pilose; seed reticulate-foveate /. karooica

53b Culms obsoletely or finely tuberculate:

55a Tepals 4; bracts with conspicuous shoulders; spikelets solitary /. marlothii

Bothalia 15, 3 & 4 (1985)

401

55b Tepals 6; bracts without conspicuous shoulders; spikelets 1-many:

56a Spikelets 6-8 mm long; bracts obtuse; perianth 2-2,5 mm long /. arida

56b Spikelets 10-20 mm long; bracts acute to acuminate; perianth 4-8 mm long:

57a Plants stoloniferous; male and female inflorescences identical, of 1-2 spikelets

/. schoenoides

57b Plants caespitose; male and female inflorescences differentiated:

58a Perianth 4-5 mm long; spikelets racemose /. vilis

58b Perianth 6,4-7 ,5 mm long; spikelets solitary or capitate /. wittebergensis

51b Culms obscurely wrinkled, smooth or almost so:

59a Sheaths distinctly grey or yellow-speckled over the whole outer surface:

60a Plants rhizomatous; perianth 2,5-3 mm long I. wallichii

60b Plants caespitose; perianth 4-6 mm long:

61a Culms sparingly or much branched, 60-100 cm long; spikelets several to many, 7-9

mm long /, gaudichaudiana var, luxurians

61b Culms sparingly branched or simple, mostly 40-60 cm long; spikelets 1-3(4), 8-15

mm long:

62a Plants forming compact cushions; seed smooth, shiny I. nubigena

62b Plants straggly; seed ridged /, sieberi

59b Sheaths not or indistinctly speckled:

63a Perianth 2,5-3 mm long; spikelets 5-6 mm long /. sabulosa

63b Perianth more than 3 mm long; spikelets more than 6 mm long:

64a Spikelets with 1—4 flowers, up to 7 mm long /. monanthos

64b Spikelets with more than 4 flowers, 8-15 mm long:

65a Plants rhizomatous; male and female inflorescences similar /, schoenoides

65b Plants caespitose; male and female inflorescences differentiated:

66a Female spikelets several /. vilis

66b Female spikelets 1(2):

67a Culms simple:

68a Bracts recurved /, curvibracteata

68b Bracts erect I. wittebergensis

67b Culms sparingly branched /. virgea

Ischyrolepis affinis Esterhuysen, sp. nov., 1. seti-

gero (Kunth) Linder affinis, a qua imprimis testa

retifoveolata, floribus et bracteis masculis minori-

bus, floribus masculis expositis differt.

TYPE. — Cape, 3320 (Montagu): Langeberg,

Kruispad Ridge between Goedgeloof Peak and Pro-

tea Valley, 1 050 m (-CD), Esterhuysen 35612 9

(BOL, holo.!; B; C; E; F; K; L; LD; M; MO; NBG;

PRE; S; STE; TCD; UC; US; W; WAG).

Plants caespitose, tussocks 50-70 cm tall. Fertile

culms sparsely branched, solid, terete, to 1,5 mm in

diam., sometimes compressed by drying, densely

and finely tuberculate; sterile culms shorter, flexu-

ose and much-branched. Sheaths closely convoluted,

2-4 cm long, body coriaceous, brown, acuminate-pi-

liferous, with the upper margins flanking the hair

membranous; on the sterile culms the sheaths are

about 5 mm long, with a stout green aciculate awn as

long as, to twice as long as the body. Male inflores-

cence of 2-10, 7 x 2,5 mm spikelets, aggregated into

a single head. Spathes similar to the sheaths,

overtopping the spikelets, 1-3 cm long. Bracts in

total about 1 cm long, coriaceous body concave, 5

mm long, with a narrow hyaline margin, awn 5 mm

long, aciculate, erect or reflexed, bracts all fertile,

obscuring the flowers. Flowers subsessile, perianth

4,5 mm long. Sepals cartilaginous, acute, densely vil-

lous along the midrib, laterals conduplicate, cari-

nate, odd sepal flat. Petals membranous to subchar-

taceous, acute, 3,8 x 1 mm. Anthers exserted at an-

thesis, 2,5 mm long, mucro about 0,5 mm long. Fe-

male inflorescence of 1-3, 10-15 x 3-7 mm spikelets,

more or less aggregated. Spathes and bracts as in the

male, but the bracts are 10-15 mm long, all fertile.

Flowers subsessile, perianth 6 mm long. Sepals carti-

laginous; laterals conduplicate, carina densely vil-

lous, slender, acute to piliferous, 6 mm long; odd

sepal concave, glabrous, very acute, 5 x 1,5 mm.

Petals chartaceous to membranous, acute, 4x2

mm. Styles 2, about 10 mm long, lower 1/3 fused,

branches villous on the inner surface. Ovary bilocu-

lar; fruit a unilocular capsule; seed 2, 3-2, 8 x 1,5

mm, triangular in cross-section, ridge present, sur-

face reticulate-foveolate.

At present /. affinis is known from a small area in

the Langeberg near Swellendam, on Leeurivierberg

and west of Goedgeloof Peak. It occurs in moist

areasatabout 1 050m. Flowering occurs in January.

I. affinis is related to I. setiger (Kunth) Linder by

having long-awned floral bracts, the structure of the

sheaths, etc. The female spikelets and flowers can

only be separated on the seed characters (see

Linder, 1984: 25). The males differ significantly by

the flowers, which are twice as large, and the bracts,

which are about three times as large, and which com-

pletely obscure the flowers. Geographically the two

402

Bothalia 15, 3 & 4 (1985)

species are widely separated, with I. setiger only oc-

curring in the Cedarberg. There is also a relationship

to I. hystrix and I. capensis, but I. hystrix does not

have aggregated spikelets, whereas I. capensis has

loosely convoluted sheaths. From the /. virgeus com-

plex I. affinis is readily separated by the longer awns

on the floral bracts and by the tuberculate culms.

CAPE. — 3320 (Montagu): Leeurivierberg, rocky south slope

(-CD), Esterhuysen 27758 (BOL, K); along path from Goedge-

loof Peak to Protea Valley (-CD), Esterhuysen 35008 (BOL; C;

E; F; K; L; M; MO; UC); Esterhuysen 35612 (B; BOL; C; E; F;

K; L; LD; M; MO; NBG; PRE; S; STE; TCD; UC; US; W;

WAG).

Ischyrolepis arida (Pillans) Linder, comb. nov.

Restio aridus Pillans in Ann. Bolus Herb. 3; 84 (1921); in Trans.

R. Soc. S. Afr. 16 : 283 (1928). Syntypes: Cape, 3320 (Montagu):

hills at Touws River (-AC), Bolus 7456 $ (BOL!); 9210 9 (BOL,

lecto.!; K\),9210 cf (BOL\); Marloth 3157 cf & 9 (B!; BOL!; K!).

Notes

1. Some collections of this species are single-flow-

ered, but this appears to be a rare situation. There

are also some collections that have woolly scales in

the sheath axils.

2. I. arida is known from the Witteberg at Laings-

burg, the mountains at Touwsrivier, the Hex River

Mountains and the Lokenberg near Calvinia. This is

an arid country species, occurring in the very arid

northern and north-eastern margins of the fynbos re-

gion.

Ischyrolepis caespitosa Esterhuysen, sp. nov., I.

cincinnatae (Mast.) Linder similis, sed testa verru-

cata, perianthiis femineis glabris, 4-lobatis differt.

TYPE. — Cape, 3419 (Caledon): Pheasantshoek

near Viljoenshof (-DA), Esterhuysen 34358 $

(BOL, holo.!; B; C; E; F; GRA; K; L; LD; M; MO;

NBG; NY; PRE; RSA; S; STE; TCD; UC; US; W;

WAG).

Plants caespitose, occasionally densely tangled,

tussocks 10-20 cm tall. Culms solid, terete,

branched, usually flexuose in upper half and tan-

gled, to 0,2 mm in diam., surface wrinkled to tu-

bercled. Sheaths 3-8 mm long, convoluted, coria-

ceous portion green to brown, subacute to subtrun-

cate, awn stout, acute, up to 1/3 of the total length of

the sheath, often somewhat reflexed; hyaline shoul-

ders shorter than the awn. Male inflorescence race-

mose, one to several spikelets on every side branch,

the central axis often continuing beyond the spike-

lets, spikelets 2-3 mm long, 1-2-flowered. Spathe

shorter than the spikelet, awn stout, as long as the

body, body with a hyaline margin, about 2 mm long.

Bract 2-3 mm long, chartaceous with a membranous

margin, acute, lower bracts often with a double sca-

brid midrib, awns short. Flowers sessile, perianth

2-2,5 mm long. Tepals 4-5, subequal, chartaceous,

acute, 2 x 0,7 mm, sepals slightly bigger than the

petals, sometimes very sparsely pilose on the midrib.

Anthers exserted at anthesis, 1,5 mm long, emucro-

nate, 2-3. Filaments flat. Female inflorescence of

solitary, single-flowered, 2-3 mm long spikelets.

Spathe as in the male, about 3 mm long. Bracts

2-2, 5-3 mm long, slender-acute, cartilaginous, up-

per margins widely membranous, outer surface mu-

ricate. Flowers subsessile, perianth 2 mm long. Te-

pals 4, subequal, acute, chartaceous to cartilaginous,

2 x 0,7 mm, petals somewhat smaller than the se-

pals. Staminodes absent. Styles 2, about 4 mm long,

lower 1 mm fused, inner surface of the style

branches villous. Ovary 1 (2)-locular; fruit a single

locular capsule; seed 1 mm long, round, irregularly

warty.

I. caespitosa occurs widespread in the Caledon

and Bredasdorp divisions, below 600 m. Most of the

collections are from the low mountains between the

Riviersonderend Mountains and the coast, and oc-

cur in moist places or swamps, often in shallow sand

over rock. There are a few collections from gravelly

and shaly conditions from the base of mountains or

from the lowlands.

This species is allied to I. cincinnata on account of

the caespitose habit, the structure of the sheaths, the

tuberculate culms and the simple spikelets. There

are three segregates of the I. cincinnata complex: I.

cincinnata, /. saxatilis and I. caespitosa, which can be

separated both geographically and by the seed sur-

face. /. caespitosa further differs from I. cincinnata

by the reduction of the perianth and the ovary, the

smaller, fewer-flowered male spikelets and the

smaller plants.

CAPE. — 3418 (Simonstown): Kogelberg Reserve, NW of Wy-

nand Louwsbos (-BD), Esterhuysen 35446 (BOL; K); banks of

Palmiet River at Oudebos (-BD), Esterhuysen 31001 (BOL; K).

3419 (Caledon): Houhoek Pass, above the top of the pass (-AA),

Esterhuysen 35021 (BOL; K; S); Caledon Swartberg, valley above

the hospital (-AB), Esterhuysen 34726 (BOL; K; MO; S); near

Caledon along the national road to Bot River at base of the Swart-

berg (-AB), Esterhuysen 31262 (BOL; K); ridge between Gal-

pin’s Kop and Aasvogelkop, mountains above Hermanus (-AC),

Esterhuysen 35302 (BOL; K; S); Klein River Mountains above

Stanford (-AD), Esterhuysen 33666 (BOL; C; E; K; L; LD; M;

MO; NBG); north slopes of the Klein River Mountains (-AD),

Esterhuysen 32876 (BOL; C; E; K; L; M; MO; NBG); Klein

River Mountains above Stanford (-AD), Esterhuysen 32873

(BOL, K); Riviersonderend Mountains, Galgeberg (-BA), Ester-

huysen 34539 (BOL; K; L; M; MO; S); Pheasantshoek near Vil-

joenshof (-DA), Esterhuysen 34358 (B; BOL; C; E; F; GRA; K;

L; LD; M; MO; NBG; NY; PRE; RSA; S; STE; TCD; UC; US;

W; WAG); four miles from Elim to Stanford (-DA), Esterhuysen

31250 (BOL, K). 3420 (Bredasdorp): Rietfontein (-CA), Ester-

huysen 34912 (BOL, K).

Ischyrolepis capensis (L.) Linder, comb. nov.

Schoenus capensis L., Amoen. 4 : 264 (1755). Restio dichoto-

my L., Syst. Nat. 12: 735 (1767), nom. illeg. , superfluous name

of Schoenus capensis. Thamnochortus dichotomus (L.) Spreng, ,

Syst. Veg. 1 : 187 (1824); Kunth, Enum. PI. 3: 433 (1841); nom.

illeg. Type: Cape, without precise locality, s.l. in herb. Linn.

1164.3 (LINN, holo.!).

Restio cuspidatus Thunb. in Phytogr. Bl. 1: 8 (1803); Thunb.,

FI. Cap. edn 1, 331 (1811); edn Shultes, 87 (1823); Kunth, Enum.

PI. 3 : 413 (1841); Mast, in A. DC., Monogr. Phan. 1 : 233 (1878);

in FI. Cap. 7: 66 (1897); Pillans in Trans. R. Soc. S. Afr. 16 : 267

(1928); in Adamson & Salter, FI. Cape Penins. 140 (1950). Type:

Cape, without precise locality, Thunberg in herb. Thunberg 23227

Cf (UPS, holo.!).

Restio squarrosus Poir. in Lam., Encycl. 6: 174 (1804); Mast, in

A. DC., Monogr. Phan. 1 : 236 (1878); in FI. Cap. 7 : 71 (1897).

Type: Cape, without precise locality, s.l. in herb. Lam. (P,

holo.!).

Restio sprengelii Mast, in J. Linn. Soc., Bot. 8 : 224 (1865).

Syntypes: Cape, 3418 (Simonstown): False Bay (-AB), Robertson

s.n. Cf (BM!); Simons Bay (-AB), Milne 223 cf (K!). 3318 (Cape

Town): at the base of Table Mountain (-CD), Ecklon 84 cf (?);on

Lions Rump (-CD), Ecklon 848 9 (?); Table Mountain (-CD),

Ecklon 848 cf (K! ) ; without precise locality, Sieber s.n. cf (K!);

Harvey 388b cf (K!); Thom s.n. (?); Bergius s.n. cf (K, lecto.!);

Ecklon & Zeyher 77.9; Drege 44 9 (?); Wallich s.n. 9 (?)•

Icon: Lam., Tabl. Encycl. t. 804 f.l (1799).

Bothalia 15, 3 & 4 (1985)

403

Notes

1. In the Linnaean Herbarium in London, sheet

1164.3 is marked ‘ Schoenus capensis' by Linnaeus,

and 1164.4 is marked 1 Schoenus capensis funale' .

The former sheet is Ischyrolepis cuspidatus Thunb.,

whereas the latter is Restio bifidus Thunb. The diag-

nosis in the protologue of Schoenus capensis is in full

agreement with LINN 1164.3. Previous students of

the group were confused by Linnaeus altering his

later diagnosis of the species to indicate that the spi-

kelets are pendulous. This suggests that Schoenus

capensis is in fact a Thamnochortus. Further confu-

sion was caused by Linnaeus’s illegitimate change of

the name to Restio dichotomus. The consequence of

all this is that the name Schoenus capensis was re-

jected as a nomen confusion (Pillans, 1928), and

Restio dichotomus was applied to one of the most

common species of Thamnochortus (T. lucens).

2. Masters (1865) was under the impression that

Sprengel (1824) used the name Restio squarrosus in

a sense different from that of Poiret (1804), the

author, so he suggested R. sprengelii as a new name

for R. squarrosus Sprengel non Poiret.

3. I. capensis is very common at most altitudes

from Port Elizabeth to the Cape Peninsula and to

Ceres. There are a few collections from further

north, from the Cedarberg and the Piketberg. This

species can also tolerate a wide range of soil types:

sandy gravelly or clayey.

Ischyrolepis cincinnata (Mast.) Linder, comb.

nov.

Restio cincinnatus Mast, in J. Linn. Soc., Bot. 8: 240 (1865); in

A. DC., Monogr. Phan. 1: 267 (1878); in FI. Cap. 7 : 79 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 287 (1928); in Adamson &

Salter, FI. Cape Penins. 141 (1950). Type: Cape, 3418 (Simons-

town): mountains near Simonstown, May (-AB), Ecklon &

Zeyher s.n. 9 (K, lecto.l; B!; BM!; BR!; MEL!; MO!; P!; S!);

s.n. cf (B!; BM! K!; MEL!; MO!; P!; S!; Z!).

Notes

1. There are several sheets of the type in some

herbaria (B; MEL; MO; P; S) labelled ‘ R . ludwigii

(R. nutans Steud.)’ by Nees and Sonder.

2. This species is close to I. caespitosa but can be

distinguished on details of floral morphology. The

taxa are also allopatric.

3. I. cincinnata occurs on the Cape Peninsula, with

a few collections from the Riversonderend Moun-

tains and the mountains at Hermanus. The usual

habitat is amongst rocks and boulders.

Ischyrolepis coactilis (Mast.) Linder, comb.

nov.

Restio coactilis Mast, in Bot. Jb. 29 Beibl. 66 : 3 (1900); Pillans

in Trans. R. Soc. S. Afr. 16 : 274 (1928). Type: Cape, 3319 (Wor-

cester): near Tulbagh Waterfall (-AC), Schlechter 7515 $ (B,

lect. ! ; BOL!), see Fig. 10.

Note

1. This species is only known from the area be-

tween Ontong’s Kop and the Tulbagh Waterfall,

near Tulbagh.

Ischyrolepis curvibracteata Esterhuysen, sp.

nov., ab I. virgea (Mast.) Linder culmis simplicibus,

testa retifoveolata differt.

TYPE. — Cape, 3319 (Worcester): Du Toits

Peak, c. 1 800m (-CA), Esterhuysen 29894 9 (BOL,

holo.!; B; C; E; F; K; L; M; MO; S; STE; UC).

Plants caespitose, tussocks 25-45 cm tall. Culms

solid, terete, simple, 1,2-1, 5 mm in diam., smooth.

Sheaths closely convoluted, 2-3,5 cm long, coria-

ceous portion acuminate with the apex continued as

a setaceous awn, the upper margins of the sheath

membranous and soon decaying; sheaths at the base

of the culms shorter, reddish, imbricate with the

hyaline margins and awns reduced. Male inflores-

cence of 1-5, 8-15 x 2-^1 mm many-flowered spike-

lets more or less aggregated into a single head.

Spathes similar to the sheaths, as tall as or taller than

the spikelets. Basal bracts obtuse, about 7 mm long,

with two villous Carinas; the remaining bracts are all

fertile, 8-12 x 3,5 mm, coriaceous to cartilaginous,

very acute with the apex continued as a 2-4 mm long

awn, the apical part of the bracts and the awn usually

spreading to reflexed. Flowers subpedicellate, per-

ianth 6,5 mm long, obscured by the bracts. Sepals

chartaceous, acute; laterals conduplicate, carina pi-

lose, 6,5 mm long; odd sepal flat, glabrous, 6x1

mm. Petals chartaceous to membranous, subacute, 6

x 1 mm. Anthers exserted at anthesis, 3 mm long,

mucronate, filaments strap-shaped. Female inflores-

cence a solitary 10-20 x 5-7 mm, many-flowered

spikelet. Spathes and bracts as in the male, 10-20

mm long, basal 1-2 bracts sterile. Flowers subpedi-

cellate, perianth 5-7 mm long. Sepals cartilaginous,

acute; laterals, conduplicate, villous-carinate, 5-7

mm long; odd sepal glabrous, flat, 4—6,5 x 1,5-2

mm. Petals membranous-chartaceous, 3,5-5 x

1,5-2, 5 mm, rounded to acute. Styles 2, 8-10 mm

long, pilose on the inner surface of the branches,

fused in the basal 1 mm, outer edges of the lower

portion of the styles usually pilose. Ovary bilocular.

Fruit a 1-2-locular capsule; seed 3 x 1,3 mm, round

in cross-section, ridge scarcely present, end obtuse

to acute, surface silvery, reticulate-foveolate with

the fovea circular.

1. curvibracteata is a high-altitude species, occur-

ring above 1 200 m in the mountains between the

Breede River Valley and the Cape Flats. Many col-

lections are from rocky summits and ridges, whereas

others are from the margins of swampy areas. This

species appears to be locally abundant.

This new species is very close to I. virgea and its

allies. It has the same type of sheath as I. virgea and

I. wittebergensis, but is separated from the latter by

its larger, less aggregated male spikelets and by the

larger, more recurved floral bracts. The floral bracts

are similar to those in I. ocreata , but in the latter

species the sheaths are loosely convoluted and the

female spikelets are usually much longer. It is more

difficult to separate it from I. virgea. It can be distin-

guished by the reticulate-foveolate rather than

striate seed and by the simple culms.

CAPE. — 3318 (Cape Town): Jonkershoek Forestry Reserve,

Disa Vlei (-DD), Esterhuysen 33115 (BOL; K; MO; S). 3319

(Worcester): Du Toits Peak (-CA), Esterhuysen 30566 (BOL; K);

Esterhuysen 29894 (B; BOL; C; E; F; K; L; M; MO; S; STE;

UC); Slanghoek Mountains, Grassy Dome (-CA), Esterhuysen

32815a (BOL; K; MO; S); Upper Wellington Sneeukop, N aspect

(-CA), Esterhuysen 30775 (BOL; K); Wemmershoek Mountains,

Winterberg (-CC), Esterhuysen 35401 (BOL; K; MO; S). 3418

404

Bothalia 15, 3 & 4 (1985)

(Simonstown): Somerset Sneeukop, N - NE slopes (-BB), Ester-

huysen 31796 (BOL; K; S); Esterhuysen 27066 (BOL; K; NBG).

3419 (Caledon): Jonkershoek, Victoria Peak (-AA), Esterhuysen

33089a (BOL; K; MO; S).

Ischyrolepis curviramis (Kunth) Linder, comb,

nov.

Restio curviramis Kunth, Enum. PI. 3 : 395 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 263 (1878); in FI. Cap. 7 : 78 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 282 (1928). Syntypes: Cape,

3319 (Worcester): Du Toits Kloof, 900-1 200m (-CA), Drege

1626 $ (B,lecto.!;BM!;K!;MO!;NY;OXF!;P!);57$ &cf(B!;

BM!; K!; MO!; OXF!; P! ; S!) without precise locality, Drege 2490

(K!).

Notes

1. The sheets of Drege 1626 and 57 in Berlin are

determined by Kunth. The sheet of Drege 2490 at K

is 1. sieberi and is the type of Restio divaricatus.

2. I. curviramis is very closely related to I. nana ,

but can be distinguished by the more lanky growth

and the warty seed coat. I. nana occurs at higher alti-

tudes than I. curviramis and the plants form tighter

tussocks.

3. I. curviramis occurs in the mountains from

Clanwilliam to Swellendam and the Cape Peninsula.

However, it is rather rare east and south of Paarl. It

generally occurs below 1 200 m, usually in shallow

sand over rock, in slight seepages.

Ischyrolepis distracta (Mast.) Linder, comb.

nov.

Restio distractus Mast, in FI. Cap. 7 : 70 (1897). Restio laniger

Kunth var. distractus (Mast.) Pillans in Trans. R. Soc. S. Afr. 16 :

274 (1928). Type: Cape, 3224 (Graaff-Reinet): Gnadouw, 2 000m

(-BB), Bolus 2633 9 (K, lecto.!; BOL!; Z!) ; 2633 cf (BOL!; K!;

Z!).

Notes

1. The distinction between this species and I. lan-

iger (presence/absence of tubercles) may be, as Pil-

lans suspected, trivial.

2. I. distracta is a high-altitude species that gener-

ally occurs above 1 500 m on the rocky summits of

peaks from King William’s Town (Gaikas Kop) to

the Cedarberg at Clanwilliam. It is absent from the

wetter coastal mountains.

3. Presumably ‘Gnadouw’ is the same as ‘Na-

douw’. I have not been able to locate a Mt Gna-

douw, but Nadouw is a 2 100 m peak to the east of

Graaff-Reinet.

Ischyrolepis duthieae ( Pillans ) Linder, comb.

nov.

Restio duthieae Pillans in Trans. R. Soc. S. Afr. 16 : 287 (1928).

Type: Cape, 3318 (Cape Town): slopes of the Eerste Rivier ter-

race, Stellenbosch Flats (-DD), Duthie 1604 9 (BOL, holo.!; K!).

Note

1. I. duthieae occurs below 150m on the coastal

forelands, low mountains and foothills of the Stel-

lenbosch-Paarl-Malmesbury area, with one collec-

tion from near Wolseley at the entrance to Bains-

kloof. This species is reported to grow in sandy,

gravelly or clayey soils.

Ischyrolepis eleocharis (Mast.) Linder, comb.

nov.

Restio eleocharis Nees ex Mast, in J. Linn. Soc., Bot. 8: 238

(1865); in A. DC., Monogr. Phan. 1: : 266 (1878); in FI. Cap. 7 :

81 (1897); Pillans in Trans. R. Soc. S. Afr. 16 : 288 (1928); in

Adamson & Salter, FI. Cape Penins. 141 (1950). Syntypes: Cape,

3318 (Cape Town): Cape Flats June (-DC), Ecklon 1 cS (MEL);

560 c f (?); Drege 2489 cf (?). 3418 (Simonstown): Zeekoevlei

(-BA), Ecklon 78 9 (K, lecto.!; MEL!); 894 9 (?).

Notes

1. There are two sheets in the Sonder herbarium

in MEL labelled T2, Restio eleocharis , Nees, 60.6’

by Sonder. They could possibly be Ecklon 1 of Mas-

ters.

2. Although the female spikelets are usually 1-2-

flowered, there are some that are up to 3-flowered.

The size of the spikelets may also be bigger than in-

dicated by Pillans (1928), namely up to 1cm long.

3. I. eleocharis is a coastal species that occurs on

stabilized sea-dunes from Cape Town to East Lon-

don. Some populations are found some kilometres

inland on sandy river banks or blown sand.

Ischyrolepis esterhuyseniae (Pillans) Linder,

comb. nov.

Restio esterhuyseniae Pillans in Trans. R. Soc. S. Afr. 30: 248

(1945). Syntypes: Cape, 3319 (Worcester): Baviaansberg (-BA),

Esterhuysen 2574 9 (BOL, lecto.!; K!); 2573 cf (BOL!; K!).

Notes

1. This species is rather close to I. karooica, with

which it is sympatric.

2. I. esterhuyseniae occurs in the arid mountains

around Karoo Poort (between Ceres and Touwsri-

vier) and the mountains around Touwsrivier, gene-

rally on rocky slopes.

Ischyrolepis feminea Esterhuysen, sp. nov., a

speciebus quas Linder ad Ischyrolepidem ascripsit

stylo singulari, a I. papillosa Esterhuysen et I. palu-

dosa (Pillans) Linder floribus parvioribus, testis col-

liculatis, inflorescentiis ramificantibus diversa.

TYPE. — Cape, 3418 (Simonstown): near the

Four Seasons, W of Betty’s Bay (-BD), Esterhuysen

31662 $ (BOL, holo.!; C; E; K; L; M; MO; NBG;

PRE; S; STE).

Plants caespitose, tussocks 15-25 cm tall, spread-

ing to form extensive patches. Culms solid, terete,

branching, to 1 mm in diam., swollen at the base and

aggregated. Sheaths closely convoluted, 5-15 mm

long, coriaceous portion yellowish, acute, with a

narrow membranous margin, the apex extended into

an acute, erect, 1-4 mm long awn; at the base of the

culms the sheaths are imbricate, reddish, with the

awn lost. No male plants are known. Female inflo-

rescence compound-racemose, culms often branch-

ing directly below the basal spikelet, spikelets sub-

imbricate, 3-4 mm long, single-flowered, numerous.

Spathes as tall as the spikelets, largely obscuring

them, similar to the sheaths. Bracts 2, enclosing the

flowers, 3-3,5 mm long, acute, cartilaginous, the

outer pilose on the midrib. Flowers subsessile, per-

ianth 3 mm long. Tepals 4, acute, cartilaginous, gla-

brous, concave, the sepals 3 mm long, the petals 2,5

mm long. Style solitary, 2-3 mm long. Ovary pilose

on the apex, unilocular. Fruit a unilocular dehiscent

capsule; seed 1,2 x 0,8 mm, round in cross-section,

ends obtuse to round, surface colliculate.

I. feminea occurs on the coastal plateau at Betty’s

Bay, in marshy localities. The plants form extensive

Bothalia 15, 3 & 4 (1985)

405

pure stands, flowering in December. Despite de-

tailed searching, no males were found, and seed-set

is prolific — presumably the species is apomictic.

Although this species is closely related to the /.

paludosa-I. papillosa complex, it can easily be dis-

tinguished by the single style, the pilose ovary apex,

the colliculate seed surface and the culms branching

at the base of the inflorescence so that the whole in-

florescence is compound. Note that in Linder, 1984:

p. 25, ‘Restio feminens’ is placed in Group 1, with

reticulate-foveolate seeds. The collection on which

this is based may well not be conspecific with the

Betty’s Bay populations that constitute I. feminea

s.s.

CAPE. — 3418 (Simonstown): Betty’s Bay (-BD), Esterhuy-

sen 34482a (BOL; K; MO); Esterhuysen 32154 (BOL; C; E; K; L;

M; MO; S; STE); near the Four Seasons, W of Betty’s Bay

(-BD). Esterhuysen 33254 (BOL; K; S); Esterhuysen 31662

(BOL; C; E; K; L; M; MO; NBG; PRE; S; STE); Esterhuysen

31686 (BOL; E; K; L; M; MO; S).

Ischyrolepis fraterna (Kunth) Linder, comb,

nov.

Restio fraternus Kunth, Enum. PI. 3 : 386 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 235 (1878); in FI. Cap. 7 : 67 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 270 (1928). Syntypes; Cape,

Ceres Division, Bokkeveld Mountains at Uien Valley, 600-750 m,

Drege 1623 (B. lecto.!; B!; BM!; BOL!; K!; MO!; OXF!; P!).

3318 (Cape Town): mountains near Stellenbosch, 900-1 200 m

(-DD), Drege 45 (BM!; K!; MO!; OXF!).

Notes

1. There are several sheets of Drege 1623 at B, one

of which is named in Kunth’s hand and this is clearly

the best lectotype. Kunth (1841) also cites Drege

1623 as a syntype of Calopsis peronata. Superficially

the two species are similar and could easily be con-

fused.

2. This species is very close to I. ocreata, with

which it may be conspecific. The character on which

they are separated (acute vs obtuse sheaths) is vari-

able even among the syntypes of I. fraterna.

3. I. fraterna occurs in the mountains from the Ce-

darberg to the mountains in the Stellenbosch - Cale-

don area, with most of the collections from the the

mountain block between Worcester, Stellenbosch

and Caledon.

Ischyrolepis fuscidula (Pillans) Linder, comb.

nov.

Restio fuscidulus Pillans in Trans. R. Soc. S. Afr. 30 : 249

(1945). Syntypes: Cape, 3319 (Worcester): Audensberg (-CB),

Esterhuysen 3232 $ (BOL, lecto!; K!); 3231 cf (BOL!; K!).

Note

1. This very curious species is only known from

the Audensberg near Worcester. It occurs between

900 and 1 700 m in hot, dry and rocky areas.

Ischyrolepis gaudichaudiana (Kunth) Linder,

comb. nov.

Restio gaudichaudianus Kunth, Enum. PI. 3 : 387 (1841); Mast,

in A. DC., Monogr. Phan. 1 : 246 (1878); in FI. Cap. 7 : 68 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 279 (1928); in Adamson &

Salter, FI. Cape Penins. 140 (1950). Syntypes: Cape, 3319 (Wor-

cester): Du Toits Kloof, 300-600 m (-CA), Drege 52 cf (B, lecto!;

K!; MEL!; MO!; OXF!; P!; S!); without precise locality. Gaudi-

chaud s.n. (?).

Restio ferruginosus Link ex Kunth, Enum. PI. 3 : 393 (1841);

Mast, in A.DC., Monogr. Phan. 1 : 245 (1878); in FI. Cap. 7 : 67

(1897). Syntypes: Cape, 3319 (Worcester): Du Toits Kloof

(-CA), Drege 1619b cf (K, lecto.!; BM!; MEL!; OXF!; P!; S!);

Drege 1816 $ (?); Drege 1 9 (B!; BOL!; K!; OXF!; P!; S!).

Restio luceanus Kunth, Enum. PI. 3 : 385 (1841), Syntypes:

Cape, without precise locality, in herb. Willdenow 18279.2 (B,

lecto.!); in herb. Luceano ( K ! ) ; Ecklon 839 p.p. (?); Ecklon 320

(?)■

Restio elatus Mast, in J. Linn. Soc., Bot. 8 : 226 (1865); in

A.DC., Monogr. Phan. 1 : 246 (1878); in FI. Cap. 7 : 68 (1897).

Type: Cape, 3318 (Cape Town): Paarl Mountains, 300-600 m

(-DB), Drege 93 9 (K, lecto.!; MO!; P!); 93 cf (BOL!; K!; MO!;

P!).

Restio cirratus Mast, in Bot. Jb. 29 Beibl. 66 : 5 (1900). Type:

Cape, 3218 (Clanwilliam): hills near Brakfontein (-DB),

Schlechter 10778 cf (B, holo.!; BR!; K!; MO!; P!; S!; Z!).

Restio gaudichaudianus Kunth var. microstachyus Nees ex

Mast, in A.DC., Monogr. Phan. 1 : 247 (1878); in FI. Cap. 7 : 68

(1897). Syntypes: Cape, 3319 (Worcester): around Tulbagh

Waterfall, Nov. (-AC), Ecklon & Zeyher s.n. cf (MEL!; K!); Du

Toits Kloof (-CA), Drege 52 cf (K, lecto.!; B!; OXF!; P!; S!);

without precise locality, Pappe 78 C f (?). Name illegitimate, based

on the same specimen as R. gaudichaudianus.

Restio gaudichaudianus Kunth var. luxurious Pillans in Trans.

R. Soc. S. Afr. 16 : 280 (1928). Syntypes: Cape, 3319 (Worces-

ter): Hex River Mountains (-AD/BC), Bolus 4234 9 (BOL!; K!);

4235 cf (BOL!; K!); 15949 cf (BOL!); 15950 $ (BOL!); Wilde

Paarde Berg (?), Stokoe in BOL 17665 cf (BOL!). 3320 (Mon-

tagu); Langeberge at Montagu (-CC), Marloth 3118 cf (BOL!);

3119 $ (BOL!); Page in BOL 16540 c f & 9 (BOL!); Mitchell in

BOL 16626 9 (BOL!). 3321 (Ladismith): Zwarteberge (-AC/

AD), Marloth 3170 cf & 9 (BOL!); Glen Leith, foot of the

Lange Berge (-?), Muir 3078 9 (BOL, lecto.!; K!); 3078 cf

(BOL!; K!). 3323 (Willowmore): Aasvogelberg (-AC), Andreae

940 Cf & 9 (BOL!); locality uncertain, ‘Zwartwaterpoort’, Bur-

chett 3409 9 (BOL!; K!).

Notes

1 . The identity of Restio luceanus is somewhat un-

certain. The herb. Willd. specimen (which Kunth

certainly saw) is I. gaudichaudiana, whereas the

herb. Luceano specimen at Kew (which he may not

have seen) is I. capensis. Consequently, I have lecto-

typified the former sheet.

2. R. ferruginosus is largely based on Drege 1619b

but Kunth also referred to Drege 1816 and based the

description of the fruit on Drege 1. Drege 1 is the

type of I. subverticellata.

3. R. gaudichaudianus var . microstachyus is lecto-

typified by the same specimen as R. gaudichaudia-

nus. This collection, Drege 52, is annotated by Mas-

ters in K as ‘var. microstachyus’ , whereas the other

syntype is not annotated. The result of this lectotypi-

fication is that the varietal name is superfluous and

therefore illegitimate.

4. R. gaudichaudianus var. luxurians is very dis-

tinct from typical R. gaudichaudianus, especially in

habit. It also appears to grow on shaly soils, whereas

the typical form grows on soils derived from sand-

stone. Further research is needed to show whether it

is a distinct species.

5. I. gaudichaudiana is ubiquitous in the Cape

Flora, distributed from Humansdorp to Clanwilliam.

It generally occurs in dry habitats, often in rocky

areas.

Ischyrolepis gossypina (Mast.) Linder, comb.

nov.

Restio gossypinus Mast, in Bot. Jb. 29 Beibl. 66 : 3 (1900); Pil-

lans in Trans. R. Soc. S. Afr. 16 : 276 (1928). Syntypes: Cape,

406

Bothalia 15, 3 & 4 (1985)

3219 (Wuppertal): Koude Bokkeveld, Tafelberg, 1 650 m (-CD),

Schlechter 10083 $ (B, lecto.!; BM!: BOL!; BRL K!; MO!; P!; S!;

Z!); 10082 cf (BM!; BOL!; BR!; K!; MO!; P!; S!; Z!); Kers Kop,

1 050 m (-A), Schlechter 8793 cf (B!; BOL!; BR!; K!; MO!; P! ; S! ;

Z!); 8794 $ (BM!; BOL!; BR!; K!; MO!; P!; S!; Z!). 3118 (Van

Rhynsdorp): Koude Berg, 1 050 m (-DC), Schlechter 8751 cf

(BOL!: BR!; K!; MO!; P!; S!; Z!).

Notes

1. I. gossypina occurs in dry to very arid fynbos,

both along the northern ridges of the Cape moun-

tains from Meiringspoort in the Swartberg to Clan-

william, with outliers in the Kamiesberg and at Lo-

kenberg in the Roggeveld Mountains, as well as in

the Cedarberg, the Piketberg and the mountains at

Villiersdorp.

Ischyrolepis helenae (Mast.) Linder, comb. nov.

Restio helenae Mast, in J. Linn. Soc., Bot. 8: 233 (1865); in

A. DC., Monogr. Phan. 1 : 249 (1878); in FI. Cap. 7 : 74 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 293 (1928). Type: Cape, 3319

(Worcester): Du Toits Kloof (-CA), Drege 1616 9 (=1618) (K,

lecto.!; B!; MO!); 1616 cf (B!; K! MO!).

Restio procurrens Mast, in Bot. Jb. 29 Beibl. 66: 3 (1900). Syn-

types: Cape, 3419 (Caledon): Riversonderend, 200 m (-BB),

Schlechter 9891 cf (B, lecto.! BM!: BR!; K!; MO!; P!; S!; Z!).

3226 (Fort Beaufort): Winterberge, 1 500-1 800 m (-AC/AD),

Drege 77 $ (?).

Restio productus Mast, in Bot. Jb. 29 Beibl. 66 : 5 (1900). Type:

Cape 3319 (Worcester): hills near Vogelvlei, 60 m (-AC),

Schlechter 10496 9 (B, lecto.!; B!; BM!; BR!; K!; MO!; P!; S!;

Z!).

Notes

1. The type of R. helenae may be found under two

different numbers, either Drege 1616 or 1618.

2. The locality of Drege 77 is quite unlikely. I have

not found any female material of that collection, but

at P there is a male specimen of Drege 77, with an

original Drege label, giving the locality as ‘Tarka’s

Winterberg, 5000-6000 ft’. This must be an error.

3. I. helenae has a very curious distribution. Most

of the collections are from the area from the Koue

Bokkeveld to the Riviersonderend, but there are re-

cords from the Zuurberg (near Grahamstown) and

Alexandria (Natal). There is also the record of

Drege 77 from the Winterberg at Fort Beaufort.

More collections are required to substantiate this cu-

rious distribution.

Ischyrolepis hystrix (Mast. ) Linder, comb. nov.

Restio hystrix Mast., in J. Linn. Soc., Bot. 10 : 276 (1868); in

A. DC., Monogr. Phan. 1 : 252 (1878); in FI. Cap. 7 : 76 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 271 (1928). Type: Cape, 3321

(Ladismith): Garcias Pass (-CC), BurcheU 6991 9 (K, lecto.!;

BOL!); 6991 cf (BOL!; K!).

Notes

1. In the protologue Masters gives the types as

Burchell s.n. and does not give a locality. The sheets

at K are clearly numbered and named by Masters. It

is most certainly the type, and the lack of informa-

tion in the protologue is a mystery.

2. I. hystrix has a wide distribution, ranging from

Ceres to Uniondale. It occurs in the arid Swartberg

and the mountains at Willowmore, as well as in the

much wetter coastal mountains at Stormsrivier and

Riversdale.

Ischyrolepis karooica Esterhuysen, sp. nov., ab

I. esterhuyseniae (Pillans) Linder testa retifoveolata.

stylopodiis pilosis, inflorescentiis masculis laxis, spi-

culis masculis paucifloris differt.

TYPE. — Cape 3319 (Worcester): Karoopoort,

on mountain slopes above the poort (-AA), Ester-

huysen 30458 $ (BOL, holo.!; BM; C; E; F; GRA;

K; L; LD; M; MO; NBG; NY; PRE; RSA; S; STE;

TCD; UC; US; W; WAG).

Plants tangled, up to 30 cm tall. Culms solid, te-

rete, branching, to 1 mm in diam., sometimes flexu-

ose, roughly tubercled. Sheaths tightly convoluted,

5-15 mm long, body acute, grading from a coria-

ceous lower 2/3 to a membranous upper 1/3, awn

slender, 2-3 mm long, on older culms the sheaths are

largely decayed. Male inflorescence of 3-8 widely

spaced, 3-5 mm long spikelets, each with about 3

flowers. Spathe similar to the sheaths, about as tall

as the spikelets, soon decayed. Bracts acute, shortly

mucronate, cartilaginous, about 3 mm long, as tall as

or shorter than the flowers. Flowers subsessile, per-

ianth 3-3,5 mm long. Tepals 4-6, chartaceous, sepals

conduplicate to subcarinate, sparsely pilose, petals

marginally shorter or longer than the sepals, more

membranous. Anthers exserted at anthesis, 1,3-2

mm long, mucronate. Female inflorescence of 1-3

lax, 6 x 1,5 mm spikelets. Spathes acute, aristate,

coriaceous with the upper margin membranous, as

long as the spikelets. Bracts acute with an acute, 1

mm long awn, body coriaceous with a narrow mem-

branous margin, 3,5—5 mm long, the lower two

bracts often sterile. Flowers subsessile, perianth

3,6-4 mm long. Tepals 4—6, cartilaginous; lateral se-

pals conduplicate to subcarinate, villous along the

midrib. Petals shorter than the sepals, acute to ob-

tuse. Styles 2, fused into a basal, 2,5 mm long, pilose

stylopodium which widens towards the ovary. Ovary

bilocular; fruit a unilocular capsule; seed 1,4 x 0,9

mm, round in cross-section, surface reticulate-foveo-

late.

I. karooica is known from the area between Ceres

and Laingsburg, on the arid mountains which form

the border between fynbos and the Karoo. The alti-

tude at which this species occurs is about 1 000 m.

Flowering probably occurs in October.

I. karooica is very similar to I. esterhuyseniae, with

which it is sympatric. It can be distinguished by the

more numerous, lax, and fewer-flowered male spi-

kelets, by the pilose stylopodium and by the reticu-

late-foveolate testa. However, very few collections

of these two species are available, and it may be

found that the variation in these characters will not

correlate in future collections.

CAPE. — 3319 (Worcester): Karoopoort (-AA), Esterhuysen

30458 (B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY;

PRE; RSA; S; STE; TCD; UC; US; W; WAG); Bonteberg, near

Touws River, on upper south slopes (-BD), Esterhuysen 32643

(B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY; PRE;

RSA; S; STE; TCD; UC; US; W; WAG).

Ischyrolepis laniger (Kunth) Linder, comb. nov.

Restio laniger Kunth, Enum. PI. 3 : 386 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 238 (1878); in FI. Cap. 7 : 69 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 273 (1928). Type: Cape, 3118

(Van Rhynsdorp): Giftberge, 450-750 m (-DC), Drege 51 cf (B,

holo.!; B!; BOL!; K!; MO!; OXF!; P!;).

Note

1. /. laniger ranges from Van Rhynsdorp to

Oudtshoorn. It occurs in mountains, generally above

Bothalia 15, 3 & 4 (1985)

407

1 000 m. Generally the populations occur in damp

localities and along seepage lines.

Ischyrolepis leptoclados (Mast.) Linder, comb.

nov.

Restio leptoclados Mast, in J. Linn. Soc., Bot. 8 : 241 (1865); in

A. DC., Monogr. Phan. 1 : 265 (1878); in FI. Cap. 7 : 80 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 289 (1928). Type: Cape,

without precise locality, Drege 3 9 (K, holo.l; Bl; BOLL P!).

Restio penicillatus Mast, in Bot. Jb. 29 Beibl. 66 : 5 (1900). Syn-

types: Cape, 3419 (Caledon): hills near Ratel River, 10 m

(-DA), Schlechter 9720 cf (B. lecto.l; BM!; BR!; K!;MO!;P!;S!;

Z! ) ; Bolus 8698 cf (BOL!; K!). 3421 ( Riversdale): Melkhoutfon-

tein (-AD), Galpin 4788 cf ( K!).

Notes

1 . Although there is both male and female ma-

terial under Schlechter 9720, Masters only described

the male material, and noted 'femina latet.’

2. I. leptoclados occurs on old sea-dunes or low

hills near the sea, from Gansbaai near Caledon to

Knysna. It is often associated with limestone out-

crops, especially in the Bredasdorp area.

Ischyrolepis longiaristata Pillans ex Linder, sp.,

nov., a speciebus quas Linder ad Ischyrolepis ascrip-

sit combinatione vaginis truncatis, laxe convolutis,

spiculis plurifloris et testa striata et reticulata di-

versa.

TYPE. — 3118 (Vanrhynsdorp); Driekoppen,

Gifberg (-DC), Esterhuysen 30749 $ (BOL, holo.!;

B; C; E; F; GRA; K; L; LD; M; MO; NBG; NY;

PRE; RSA; S; STE; TCD; UC; US; W; WAG).

Plants tangled, 30-60 cm tall. Culms solid, terete,

branching, to 1 mm in diameter, very roughly tuber-

culate. Sheaths loosely convoluted, 8-25 mm long,

coriaceous portion reddish-brown, obscurely tuber-

culate. more or less truncate, awn very slender, aci-

culate, 1/4— 1/3 of the total sheath length, flanked by

two large membranous shoulders, as tall as the awns.

Male inflorescence of 1-several very lax, 6-10 x 3-5

mm spikelets. Spathes and bracts similar to the

sheaths, shorter than the spikelet but obscuring the

flowers, the upper bracts shorter than the lower

bracts, more acute, all fertile. Flowers subsessile,

perianth 5 mm long. Tepals chartaceous, lateral se-

pals conduplicate, carina pilose, 5 mm long, pilifer-

ous; remaining tepals equal, acute, 4 x 0,6-1 mm.

Anthers exserted at anthesis, 3 mm long, filaments

flat. Female inflorescence of solitary, several-flow-

ered, 8-15 x 3 mm spikelet. Spathe and bracts as in

the male, bracts all fertile or the basal bract with a

rudimentary flower. Flowers pedicellate, perianth

4,5-6 mm long. Sepals subcartilaginous; lateral se-

pals 5-6 mm long, piliferous, conduplicate, carina

pilose; odd sepal glabrous, flat, acute, 4—5 x 0,8-1

mm. Petals chartaceous to membranous, acute, 3,5

x 1 mm. Styles 2, 7-9 mm long, basal 0,8 mm fused,

inner surface of the branches pilose. Ovary bilocu-

lar. Fruit a unilocular capsule; seed 1,6 x 1 mm,

round in cross-section, surface complex, with fine re-

ticulate pattern, a rougher striate pattern and a few

tubercles at one end. Fig. 13.

I. longiaristata is known from the Gifberg and the

Nieuwoudtville escarpment, where it occurs be-

tween 450 and 1 000 m, in sand, often in damp

places. Flowering occurs in October.

This is a peculiar species, not closely related to

any presently known species. The truncate sheaths

are also found in the I. curviramis group, but the

sheaths in that group are always closely convoluted

and the spikelets few to one-flowered. The seed sur-

face is quite unique. More research is needed to es-

tablish the affinities of this species.

CAPE. — 3118 (Vanrhynsdorp): Gifberg (-DC), Esterhuysen

30749 (B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY;

PRE; RSA; S; STE; UC; US; W; WAG); Esterhuysen 22081

(BOL; K; S). 3119 (Calvinia): Keysersfontein (-AC), Goldblatt

669 (BOL; E; K; L; M; MO; S).

Ischyrolepis macer (Kunth) Linder, comb. nov.

Restio macer Kunth, Enum. PI. 3 : 390 (1841); Mast, in A. DC.,

Monogr. Phan. 1 : 251 (1878); in FI. Cap. 7 : 77 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 291 (1928). Type: Cape, 3218 (Clanwil-

liam): near Groen and Waterfalls Rivers, Onder Bokkeveld,

750-900 m (-BB). Drege 2487 cf (B, holo.l; K!; MO!; OXF!; PI).

Note

1. I. macer has a wide but patchy distribution,

from Calvinia to Bredasdorp. It occurs in the moun-

tains between Clanwilliam and Calvinia, in very arid

conditions on sandstone soils; in much wetter condi-

tions in the Koue Bokkeveld and on the Piketberg;

on the sandy flats between Malmesbury and Paarl,

as well as around Wolseley and Bredasdorp. It

seems to prefer arid conditions with seasonal water-

logging. The species is morphologically quite vari-

able and needs more detailed research.

Ischyrolepis marlothii ( Pillans ) Linder, comb.

nov.

Restio marlothii Pillans in Ann. Bolus Herb. 3 : 83 (1921); in

Trans. R. Soc. S. Afr. 16 : 273 (1928). Type: Cape, 3320 (Mon-

tagu): Witteberge. 1 250 m (-BC), Marloth 3146 9 (BOL, holo.l;

BL Kl).

Restio marlothii Pillans var. parviflorus Pillans in Trans. R. Soc.

S. Afr. 16 : 273 (1928). Type: Cape, 3320 (Montagu): Witteberge,

west of Matjiesfontein (-BC), Pillans in BOL 16475 9 (BOL,

holo.l; Kl).

Note

1. I. marlothii occurs on the arid northern ridges

and slopes of the Cape mountains, from Laingsburg

to the Cedarberg and Calvinia. Some collections

from the Cedarberg may be from better watered

habitats on ‘suurvlaktes’. It probably grows both on

sandstone-derived soils as well as on gravels and

shales.

Ischyrolepis monanthos (Mast.) Linder , comb.

nov.

Restio monanthos Mast, in J. Linn. Soc., Bot. 8 : 238 (1865); in

A. DC.. Monogr. Phan. 1 : 264 (1878); in FI. Cap. 7 : 79 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 283 (1928). Type: Cape, 3018

(Kamiesberg): between Buffels River and Pedros Kloof, 600-900

m (-AC), Drege 2486 9 (K, holo.l; BOLL PI).

Notes

1. I. monanthos ranges from Bredasdorp to Cal-

vinia and the Kamiesberg. In the southern portion of

the range (from Bredasdorp to Piketberg) the

species occurs on the sandy coastal forelands and in

the foothills of the coastal mountains, but from Mal-

mesbury futher northwards it occurs up to 1 000 m

up in the mountains.

408

Bothalia 15, 3 & 4 (1985)

FIG. 13. — Ischyrolepis longiaristata Pillans ex Linder, a, base of plant, showing aggregated culm bases and loosely convoluted

sheaths, x 1; b, detail of sheaths, with truncate coriaceous portion and large hyaline shoulders, x 4,5; c, male culms, x 1; d,

female spikelet, note tuberculate culms, x 6,5; e, female flower, somewhat opened, lateral sepals piliferous, and with pilose

Carinas, x 10; f, ovary, x 10; g, malespikelets, x 6,5 ; h, male flower with exserted anthers, x 10. (From Esterhuysen 22081.)

Bothalia 15, 3 & 4 (1985)

409

2. In the Cold Bokkeveld the female spikelets of /.

monanthos are often more than 2-flowered.

Ischyrolepis nana Esterhuysen , sp. nov., ab I.

curvirami (Kunth) Linder testa retifoveolata, culmis

plerumque laevibus, spiculis femineis plerumque

plurifloris, spiculis masculis 1-3 differt.

TYPE. — Cape, 3319 (Worcester): Lower

Wellington Sneeukop, at W base of the shale band

(-BA), Esterhuysen 32658 $ (BOL, holo.!; C; E; K;

L; M; MO; S; STE).

Plants caespitose, 10-20 cm tall, tufts spreading by

stolons to form small colonies. Culms solid, terete,

branching, to 1 mm in diam., sometimes spreading

as stolons, surface smooth to rarely very finely and

densely tuberculate. Sheaths convoluted, 5—15 mm

long, coriaceous portion green when young, brown

on older culms, of the total length of the sheaths,

truncate to very obtuse, extended as a cylindrical,

acute awn; hyaline shoulders variable, taller or

shorter than the awn, acute to rounded, soon decay-

ing; basal portion of the culms with nitid, reddish-

brown, imbricate, coriaceous, obtuse-mucronate

sheaths. Male inflorescence of 1-3 generally lax,

5-10 mm long spikelets. Spathes similar to the

sheaths, awn about \ of the total length, shorter than

the spikelet. Bracts obscuring the flowers, acute,

stoutly and acutely awned, coriaceous, 3-5 mm long

with hyaline margins, the lowest bracts sterile. Flow-

ers subsessile, perianth about 3 mm long. Sepals

chartaceous to subcartilaginous, acute; laterals con-

duplicate, carina pilose, apex piliferous; odd sepal

flat, 3 x 0,6 mm. Petals like the odd sepal, but more

membranous. Anthers exserted at anthesis, 2 mm

long, mucronate. Pistillode absent. Female inflores-

cence with 1 (3) lax, 4-10 mm long spikelets, some-

what smaller than the male spikelets, (1 ^-flow-

ered. Bracts obscuring the flowers, coriaceous,

acute, margins membranous, 5-7 mm long, acute

with a tapering awn, at the base a few smaller sterile

bracts. Spathe similar to the bracts but with a longer

awn. Flowers subsessile, perianth 2,5^4 mm long.

Sepals chartaceous; laterals conduplicate, piliferous,

carina villous to sparsely pilose, 2,5^4 mm long; odd

sepal flat, acute, 2x1 mm. Petals membranous, 1,5

x 1 mm. Staminodes absent. Styles 2, about 5 mm

long, fused at the base, the upper | on the inner side

of the branches pilose, the base of the styles murica-

te. Ovary bilocular. Fruit bilocular, rarely a unilo-

cular capsule; seed 1 mm long, round, ends obtuse,

ridge weakly developed, surface reticulate-foveo-

late.

Ischyrolepis nana is a montane species that rarely

occurs below 1 000 m. It appears to be quite common

on the higher mountains between Worcester and

Clanwilliam, generally occurring in shallow sandy

soil over rock, in conditions which are frequently

waterlogged in winter. Where the habitat is suitable,

the species becomes locally dominant, often forming

extensive tussocks or colonies, which grow by stolo-

niferous extension.

Superficially, I. nana is very similar to I. curvira-

mis and has, until now, been included in it. The two

species can always be distinguished by their seed sur-

faces (see Linder, 1984: p.25). In addition, I. nana

has a very different habit, forming spreading tus-

socks less than 20 cm tall. The female spikelet often

has up to 4 florets, the culms are usually smooth and

there are fewer male spikelets. I. nana also grows at

higher altitudes than /. curviramis. It is likely that

the two species will be more easily distinguished in

the field than from herbarium specimens.

CAPE. — 3218 (Clanwilliam): Piketberg Mountain, summit of

Zebrakop (-DB), Esterhuysen 30386 (BOL; K; M; MO; S); Ester-

huysen 35337 (BOL; K). 3219 (Wuppertal): Central Cedarberg,

Cedarhout Kop (-AC), Esterhuysen 30010a (BOL; K; S); Central

Cedarberg, Uitkyk Peak (-AC), Esterhuysen 30102 (BOL; E; K;

L; M; MO; S); Esterhuysen 30103 (BOL; C; E; F; K; L; LD; M;

MO; NBG; S; STE; UC); Central Cedarberg, on plateau above

the Wolfberg Cracks (-AD), Esterhuysen 29977 (BOL; E; K; L;

M; MO; STE) South Cedarberg, below Sneeuberg Hut (-CA),

Esterhuysen 30312 (BOL; K); South Cedarberg, Sneeuberg sum-

mit (-CA), Esterhuysen 30093 (BOL; K); Esterhuysen 30092 (B;

BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY; RSA; PRE;

S; STE; TCD; UC; US; W; WAG); Esterhuysen 30304 (BOL; K;

M; MO; S); South Cedarberg, Sneeuberg Pup (-CA), Esterhuy-

sen 34395 (BOL; K); Olifants River Mountains, Turret Peak

(-CC), Esterhuysen 29661 (BOL; K; L; M; MO; S); Cold Bokke-

veld, Schoongezicht Peak, (-CC), Esterhuysen 29656 (BOL; K);

Bokkeveld Sneeukop, below shale band (-CD), Esterhuysen

30104 (BOL; K; MO; S). 3319 (Worcester): Tulbagh, at the head

of Sneeugat Kloof (-AA), Esterhuysen 29944 (BOL; E; K; L; M;

MO; S; STE); Klein Winterhoek (-AA), Esterhuysen 28534

(BOL; K); Paarl Distr., Lower Wellington Sneeukop (-BA), Es-

terhuysen 32658 (BOL; C; E; K; L; M; MO; S); Hex River Moun-

tains, Sanddrift Peaks (-BC), Esterhuysen 29747a (BOL; K);

Bainskloof, above the waterfall in Baviaanskloof (-CA), Ester-

huysen 31083 (BOL, K); Hex River Mountains, on plateau of

Brandwacht Peak (-CB), Esterhuysen 29928 (BOL; E; K; L; M;

MO; S); Esterhuysen 30136 (BOL; K); Fonteintjiesberg (-CB),

Esterhuysen 30580 (BOL; K; L; M; MO; S); between Du Toit’s

Peak and Goudini Sneeukop (-CC), Esterhuysen 30988 (BOL;

K); Goudini Sneeukop (-CD), Esterhuysen 28543 (BOL; K; S);

Esterhuysen 29074 (BOL; K; M; MO; S).

Ischyrolepis nubigena Esterhuysen sp. nov., ab.

I. sieberi (Kunth) Linder testa nitida, alveolata,

plantis pulvinatis, culmis implexis differt.

TYPE. — Cape, 3319 (Worcester): Matroos-

berg, 2 100 m (-BC), Esterhuysen 27780 $ (BOL,

holo.!; E; K; L; M; MO; S; STE).

Plants forming tangled masses or large cushions,

up to 1 m deep. Culms solid, terete, branching,

somewhat flexuose, to 2 mm in diameter, surface

smooth, basal portion of the culms stoloniferous.

Sheaths tightly convoluted, 8-15 mm long, coria-

ceous portion reddish-brown with very fine golden

mottling, acute with the apex extended into an

acute, slender, 1,3 mm long awn, upper margin of

the body chartaceous and eventually decaying; on

the stoloniferous portion the sheaths are more imbri-

cate, acute to obtuse, awn lost. Male inflorescence of

3-8 somewhat aggregated, curved, acute, 5-8 x 1,5

mm spikelets. Spathes slender acuminate, with the

apical awn about \ of the total length, body coria-

ceous with a wide membranous margin, spathes as

long as or shorter than the spikelets. Bracts charta-

ceous, obtuse to subacuminate, 3 x 2,5 mm, upper

margins membranous and soon decaying. Flowers

subsessile, perianth 2,5-3 mm long. Sepals subcarti-

laginous, subacute; lateral sepals 2,7 mm long, con-

duplicate, villous-carinate; odd sepal 2,5 x 0,7 mm,

glabrous, flat. Petals subacute, chartaceous, 2,2 x

0,8 mm. Anthers about 2 mm long, mucronate. Fe-

male inflorescence of 1 (2), 10 x 2,5 mm, acute, 1-4-

flowered spikelets. Spathes and bracts similar, coria-

410

ceous with a narrow membranous margin, acute,

body 4-6 mm long, awn slender acuminate, erect,

1-2,5 mm long. Flower pedicellate, perianth 4,5-6

mm long. Sepals cartilaginous, acute; lateral sepals

conduplicate, villous-carinate, 5,5-6 mm long; odd

sepal flat, glabrous, 5-5,5 x 1,3-2 mm. Petals carti-

laginous, subacute, 4—5 x 1, 5-2,5 mm. Styles 2,

fused in the basal 3 mm, arising from a flattened cap;

ovary bilocular. Fruit a 1-locular capsule; seed 2 x

1,5 mm, elliptical in cross-section, ends obtuse, sur-

face smooth, shiny, white with an irregular brown

reticulate pattern.

I. nubigena occurs above 1 650 m on the summits

of peaks from the Hex River Mountains to the Ri-

viersonderend Mountains. Many collections are

from rocky localities, although some are from shale

bands. The plants are reported to form springy large

cushions — a typical alpine growth-form.

This species is a high-altitude segregate of I. sie-

beri, to which it is very similar. However, the habit is

quite different. I. sieberi forms tussocks, usually

somewhat disorganized, but not cushions. The cush-

ioned habit is reflected in the somewhat flexuose

culms of I. nubigena. The seed-coat is very distinct

and may be the most reliable way of distinguishing

the new taxon. Linder (1984: p.25) included both /.

sieberi and I. nubigena (as Restio nubigena) in his

Group 1, but there are clear detailed and visual dif-

ferences between them.

CAPE. — 3319 (Worcester): Waaihoek, on SE side near the

summit (-AD), Esterhuysen 31062 (BOL; K; M; MO; S; STE);

Groothoek Peak, on the shale band (-BC), Esterhuysen 27682

(BOL; K; S); Matroosberg (-BC), Esterhuysen 27780 (BOL; E;

K; L; M; MO; S); Sonklip, N of Matroosberg (-BC), Esterhuysen

32852 (BOL; C; E; K; L; M; MO; S); Hex River Mountains,

Windsor Peak (-CB), Esterhuysen 30082a (BOL; K; L; M; MO;

S); Stettynsberg, S side of the summit (-CD), Esterhuysen 34940

(BOL; C; K; L; M; MO; S); Klein Winterhoek Pk, at summit

(-AA), Esterhuysen 35892 (B; BOL; C; E; F; K; L; LD; M; MO;

NBG; NY; RSA; PRE; S; STE; TCD; UC; US; W; WAG). 3419

(Caledon): Pilaarkop, Riviersonderend Mts (-BB), Esterhuysen

31812 (BOL; K; L; M; MO; S).

Ischyrolepis ocreata (Kunth) Linder, comb.

nov.

Restio ocreatus Kunth, Enum. PI. 3 : 385 (1841); Mast, in

A. DC. , Monogr. Phan. 1: 234 (1878); in FI. Cap. 7 : 67 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 269 (1928); in Adamson &

Salter, FI. Cape Penins. 140 (1950). Type: Cape, Ceres Division,

Bokkeveld Mountains at Uien Valley, 600-750 m, Drige 2505 $

(B, holo.l; B!; K!; MEL!; MO!; OXF!; P!; S!)

Notes

1. Dr 'ege 1623, part of the type of Calopsis pero-

nata, is I. ocreata. However, C. peronata is lectotypi-

fied by the other element in its protologue and is

placed in the synonymy of C. viminea.

2. The distinction between I. fraterna and I.

ocreata are very small and the two taxa may be syno-

nymous.

3. I. ocreata is widespread in the western portion

of the Cape, being found from Ladismith to Stellen-

bosch and northwards, as far as Steinkopf near

Springbok. It generally occurs in dry places and is

most common in the Cedarberg area. It is related to

I. capensis , with which it is parapatric. In the overlap

Bothalia 15, 3 & 4 (1985)

zone some curious forms occur, which deserve more

detailed study.

4. In the Hex River Mountains around Waaihoek

and Fonteintjiesberg, at about 1 800 m, a curious

taxon occurs. It is related either to I. ocreata or to I.

curvibracteata and is distinct in having simple culms

and small recurved bracts. Material of this taxon has

been distributed as ‘ Restio aff. ocreatus' .

Ischyrolepis paludosa ( Pillans ) Linder, comb.

nov.

Restio paludosus Pillans in Ann. Bolus Herb. 3: 142 (1922); in

Trans. R. Soc. S. Afr. 16 : 292 (1928); in Adamson & Salter, FI.

Cape Penins. 141 (1950). Type: Cape, 3418 (Simonstown):

Smitswinkel Vlei (-AD), Pillans 4145 $ (BOL, holo.!; K!).

Notes

1. This species is part of a complex which includes

/. feminea, I. papillosa and I. paludosa. The last two

species can only be distinguished by the male spike-

lets and seed coats.

2. I. paludosa ranges from the Koue Bokkeveld

and Piketberg Mountain to the Bredasdorp flats.

The habitat is generally swampy or marshy. Towards

the southern portion of the distribution range the

species is confined to the coastal flats, but at the

northern end it occurs at up to 1 200 m.

Ischyrolepis papillosa Esterhuysen, sp. nov., ab

I. paludosa (Pillans) Linder spiculis masculis uniflor-

is, floribus masculis in dimidio superiore expositis, a

I. feminea Esterhuysen inflorescentiis simplicibus,

ab ambabus testis tuberculatis recedit.

TYPE. — Cape, 3418 (Simonstown): Kenil-

worth Race Course (-AB), Esterhuysen 30849 9

(BOL, holo.!; B; C; E; F; GRA; K; L; LD; M; MO;

NBG; NY; PRE; RSA; S; STE; TCD; UC; US; W;

WAG).

Plants caespitose, non-rhizomatous, often forming

mats, 15-30 cm tall. Culms solid, terete, branching,

up to 1 mm in diameter, swollen at the base, aggre-

gated, surface smooth to tuberculate. Sheaths closely

convoluted, swollen at the base, 5-20 mm long, co-

riaceous body green to pale yellow-brown, truncate

to rounded with a small membranous margin, awn

somewhat flattened, up to half of the total length of

the sheath, occasionally twice as long as the body,

erect to reflexed; sheaths at the base of the culms

reddish, imbricate, awns obsolete. Male inflores-

cence with 3-10 lax to subimbricate, single-flowered,

about 5 mm long spikelets. Spathe as tall as to some-

what shorter than the spikelet, usually more or less

obscuring the spikelet, obtuse to acute with a mem-

branous margin and a re flexed or straight awn. Bract

chartaceous, acute, as tall as the flower, 5 x 1,5 mm,

with two pilose ribs. Perianth 4 mm long, flowers

subsessile. Tepals subequal, chartaceous, glabrous,

acute to rounded, 3,5 x 0,6 mm, sepals marginally

taller than the petals. Anthers exserted at anthesis,

2,5-3 mm long, emucronate, filaments slender. Fe-

male inflorescence similar to the male, bract similar

but is rounded or acute, as long as or shorter than

the flower, and with 1 (2) more acute, glabrous or

pilose, taller bracts present. Flower subpedicellate,

perianth 3-^1 mm long. Tepals 5, cartilaginous.

Bothalia 15, 3 & 4 (1985)

411

acute, 2-3,5 mm long; sepals larger than the petals,

minutely pilose on the midrib. Styles 2-4,5 mm long,

fused in the basal 2 mm, inner surfaces of the

branches pilose. Ovary bilocular. Fruit a unilocular

capsule; seed 0,8 mm long, round in cross-section,

ends obtuse, surface densely tuberculate, either

white or black.

I. papillosa occurs widespread on the coastal fore-

lands from Malmesbury to Humansdorp. All collec-

tions are from below 300 m. The substrate varies

from recent sand to clay. Some collections also ap-

pear to be from gravelly Table Mountain Sandstone

derived conglomerates. Many populations also occur

in seasonally wet or marshy places, but some are

from well-drained sites. Flowering occurs in spring,

from September to December.

This species is closely related to I. paludosa and I.

feminea because of the swollen culm bases, the more

or less truncate, long-awned sheaths, the single-

flowered spikelets and the large spathes. All occupy

the coastal forelands and generally occur in moist to

wet habitats. The differences between I. paludosa

and /. papillosa are small. I. papillosa can be separ-

ated by the single-flowered male spikelets, in which

the flowers are obscured by the bracts, by the flexu-

ose culms and by the papillose or tuberculate seed

coat, compared to the more or less colliculate sur-

face in I. paludosa. From I. feminea our species can

be distinguished by the simple inflorescences, the

seed coat and the two styles.

CAPE. — 3318 (Cape Town): Paardeberg, above Lemoenkloof

(-BB), Esterhuysen 33628 (BOL; K); Mamre Road, Riverlands

(-BC), Esterhuysen 34658 (BOL; C; E; K; L; M; MO; S; STE);

Camps Bay, above Fiskaal Road (-CD), Esterhuysen 33991

(BOL; K; S); Rondebosch Common (-CD), Esterhuysen 30858

(BOL; C; E; K; L; M; MO; NBG; PRE; S; STE); along the Cape

Town-Paarl Road, before the Malmesbury turn-off (-DC), Ester-

huysen 31590 (BOL; K; L; M; MO; S); 2-3 miles along the Lyne-

doch Road from Faure to Stellenbosch (-DD), Esterhuysen 31174

(BOL; K; S); Stellenbosch Golf Course (-DD), Duthie 1604a

(BOL; K). 3418 (Simonstown): Kenilworth Race Course (-AB),

Esterhuysen 34059a (BOL; K); Esterhuysen 31749a (BOL; K);

Esterhuysen 33323 (BOL; C; E; K; L; LD; M; MO; S; STE; UC);

Esterhuysen 31049 (BOL; K; STE); Esterhuysen 30849 (B; BOL;

C; E; F; GRA; K; L; LD; M; MO; NBG;' NY; PRE; RSA; S;

STE; TCD; UC; US; W; WAG); between Hangberg and Karbon-

kelberg (-AB), Esterhuysen 31969 (BOL; K); between Fish Hoek

and Kommetjie (-AB), Esterhuysen 31679 (BOL; K); flats be-

tween Strand and Gordon’s Bay (-BB), Esterhuysen 34766 (BOL;

K; MO; S); Esterhuysen 34030 (BOL; K; MO; S); Vergelegen,

near base of Landdrost Kloof (-BB), Esterhuysen 34021 (BOL;

K); Esterhuysen 33924 (BOL; K; L; M; MO; S); Oudebos, Pal-

miet River Valley (-BD), Esterhuysen 31000 (B; BOL; C; E; F;

GRA; K; L; LD; M; MO; NBG; NY; RSA; PRE; S; STE; TCD;

UC; US; W; WAG). 3419 (Caledon); at the turn-off to Boskloof

on the Villiersdorp-Caledon road (-AA), Esterhuysen 33715

(BOL; K; MO; S); base of Babylon’s Tower in Flemel-en-Aarde

(-AD), Esterhuysen 32322a (BOL; K); Pheasantshoek near Vil-

joenshof (-DA), Esterhuysen 34370a (BOL; K; MO; S); 15 miles

from Elim to Strays Bay (-DB), Esterhuysen 29690a (BOL; K; S);

c. 1 mile south of Viljoenshof (-DC), Esterhuysen 32012 (BOL,

K). 3424 (Humansdorp): Kromme River, ‘Osbosch’ (-BB), Cowl-

ing 349 (BOL; K).

Ischyrolepis pratensis Esterhuysen, sp. nov., I.

curvirami (Kunth) Linder proxima, bracteis et semi-

nibus similibus, vero culmis partibus basalibus stolo-

niformibus, spiculis masculis grandioribus differt.

TYPE. — Cape, 3318 (Cape Town): Paarl, Re-

serve on Paarl Mountain (-DB), Esterhuysen 31206

$ (BOL, holo.!; C; E; F; K; L; M; MO; S; STE;

UC).

Plants caespitose, 15-20 cm tall, culms spreading

at the base, stoloniferous. Culms solid, terete,

branching, to 30 cm long, 1 mm in diam., tubercled.

Sheaths on the aerial portions closely convoluted,

5-15 mm long, coriaceous portion green to brown,

truncate to truncate-rounded, §-| of the total length

of the sheath, with a truncate, cylindrical to flat-

tened, recurved, stout awn and hyaline shoulders

shorter than the awn and soon decaying; on the sub-

terranean portions golden-brown, subimbricate, ob-

tuse, with the awn scarcely developed. Male inflores-

cence of 1-2, 10 x 2,5 mm, many-flowered spikelets,

older spikelets soon decaying. Spathes cartilaginous,

about 7 mm long, acute, apex extended into a short

awn flanked by hyaline shoulders. Bracts similar to

the spathes, awn somewhat longer than the hyaline

apex of the bracts and somewhat reflexed. Flowers

sessile, perianth 5-5,5 mm long, very compressed.

Sepals acute, papyraceous, lateral sepals condupli-

cate, sparsely pilose on the carina; petals like the

odd sepal. No pistillode present. Anthers exserted at

anthesis, 3 mm long, mucronate. Female inflores-

cence similar to the male, but the spikelets are

smaller and with 1-4 flowers. Spathe similar to the

sheaths. Bracts as in the male, obscuring the flowers.

Flowers almost sessile perianth 4-5 mm long. Sepals

acute, 4-5 x 1 mm, papyraceous, brown to red; lat-

erals subcarinate, sparsely pilose on the carina; odd

sepal flat, somewhat shorter than the laterals. Petals

acute, 3x1 mm, papyraceous, reddish to membra-

nous-hyaline. Staminodes absent. Ovary bilocular.

Styles two, about 6 mm long, fused in the lower 1-2

mm, pilose at the base. Fruit a capsule, 1-2 locular;

seed about 1 mm long, triangular, surface irregular

colliculate. Fig. 14.

Ischyrolepis pratensis occurs in the intermontane

valleys and coastal forelands between Worcester and

the Cape Peninsula, generally below 300 m. The

plants have been collected from sand or gravel, often

in seasonally wet areas. Due to the stoloniferous ha-

bit, they often form tussocks up to 2 m in diameter.

Flowering occurs in October and November.

This rather distinct species is related to I. curvira-

mis because of the truncate sheaths and the curious

seed walls (see Linder, 1984: 25). However, I. pra-

tensis can be readily distinguished by the fewer

larger male spikelets, the stoloniferous habit, and

the more numerous flowers in each female spikelet.

CAPE. — 3318 (Cape Town): Rondebosch Common (-CD),

Esterhuysen 30859 (BOL; C; E; K; L; M; MO; NBG); Esterhuy-

sen 29903 (BOL; C; E; K; L; M; MO; S; STE); Kenilworth Race

Course (-CD), Esterhuysen 31828 (BOL; K; S); Esterhuysen

33993 (BOL; E; K; L; M; MO; S; STE); Esterhuysen 32345

(BOL; E; K; L; M; MO; NBG; S); Paarl Mountain Reserve

(-DB), Esterhuysen 31206 (BOL; C; E; K; L; M; MO; S; STE);

sandy flats along the old road to Paarl, NE of Kraaifontein

(-DC), Esterhuysen 31560 (BOL; K; M; MO; S); on municipal

land at Durbanville (-DC), Esterhuysen 31316 (B; BOL; C; E; F;

GRA; K; L; LD; M; MO; NBG; NY; PRE; RSA; S; STE; TCD;

UC; US; W; WAG). 3319 (Worcester): Romansrivier (-AC), Es-

terhuysen 31355 (BOL; C; E; K; L; MO; S; STE). 3418 (Simons-

town): Somerset West, Vergelegen, (-BB), Esterhuysen 32149

BOL; K; M; MO; S); Esterhuysen 34022 (BOL; K; M; MO; S);

flats between Gordons Bay and the Strand (-BB), Esterhuysen

32288 (BOL; K; M; MO; S).

412

Bothalia 15, 3 & 4 (1985)

FIG. 14. — Ischyrolepis pratensis Esterhuysen. a, female plant, showing culms rooting at the base and the stoloniferous habit, x

0,8; b, detail of the culm and sheaths, showing tubercles, long, reflexed awns on the sheaths and large hyaline shoulders, x 4; c,

male plant, x 0,8; d, male spikelet, x 3; e, male flower with exserted anthers and papyraceous perianth, x 9; f, dissection of

the male flower, showing subequal tepals, x 9; g, female inflorescence, x 3; h, female flower, x 9; i, dissection of female

flower, x 9. (Male material from Esterhuysen 33995, female material from Esterhuysen 31316.)

Bothalia 15, 3 & 4 (1985)

413

Ischyrolepis pygmaea (Pillans) Linder , comb.

nov.

Restio pygmaeus Pillans in Trans, R. Soc. S. Afr. 30 : 253

(1945). Type: Cape, 3319 (Worcester): Milner Peak (-AD), Es-

terhuysen 8710 $ (BOL, lecto.l; Kl); 8710 cT (BOL!; Kl).

Note

1. This is a high-altitude species, which occurs in

the Hex River Mountains and the Keeromsberg be-

tween 1 500 and 2 200 m, with most collections from

above 1 800 m. It has been most frequently recorded

from rock ledges and poorly drained areas.

Ischyrolepis rivula Esterhuysen , sp. nov., ab 7.

tenuissima (Kunth) Linder culmis partibus basalibus

stoloniformibus, plantis implexis differt.

TYPE. — Cape, 3219 (Wuppertal): Boboskloof

(-CD), Esterhuysen 31912 $ (BOL; holo.!; C; E; F;

K; L; LD; M; MO; NBG; PRE; S; STE; TCD; UC;

US; W; WAG).

Plants forming densely tangled, spreading mats.

Culms solid, terete, branching, slender, to 1 mm in

diam. at the base, more slender near the terminal

branches, surface smooth to very obscurely rugu-

lose, faintly mottled, at the base often stoloniferous.

Sheaths convoluted, with a swelling at the base, 3-8

mm long, coriaceous portion acute, brown, extend-

ing into a slender awn 3-? of the total sheath length,

flanked and overtopped by obtuse membranous

lobes, the margins between the membranous and co-

riaceous portions are not always distinct. Male inflo-

rescence with 1-3, 3-6 x 1 mm, lax, usually 3-flow-

ered spikelets. Spathes like the bracts, shorter than

the spikelets. Bracts obtuse, obscuring the flowers,

3-5 mm long, cartilaginous with a wide membranous

margin. Flowers subsessile, perianth 2,5-3, 5 mm

long. Tepals chartaceous to membranous; lateral se-

pals conduplicate, 3-3,5 mm long, very acute to pili-

ferous, carina very sparsely pilose; odd sepal acute,

flat, 3 x 0,8 mm; petals obtuse, more membranous

and somewhat smaller than the odd sepal. Anthers

exserted at anthesis, 1,5 mm long. Female inflores-

cence of 1-2 lax, 2, 5-3, 5 mm long, 1 (2)-flowered

spikelets. Spathe like the sheaths, about § of the

length of the spikelet. Bracts similar to the male,

chartaceous-cartilaginous, subacute to obtuse, as tall

as the flowers. Flowers subsessile, perianth 2,3-3

mm long. Sepals chartaceous, acute, margins sub-

membranous, glabrous, 2,3-3 mm long; odd sepal

marginally shorter than the laterals. Petals membra-

nous, acute to subacute, 1,5-2 x 0,8-1 ,2 mm. Styles

2, 3-4 mm long, fused in the lower 1 mm, inner sur-

faces of the branches pilose. Staminodes absent.

Ovary bilocular; fruit a 1-2-locular dehiscent cap-

sule; seed 1 mm long, round, ends obtuse, surface

finely reticulate-foveolate.

I. rivula occurs in the Cold Bokkeveld from the

South Cedarberg to the Witzenberg, at altitudes of

900-1 300 m. The collections are all from stream-

banks and marshy places, where it forms large tan-

gled colonies. Flowering occurs in February and

March.

This species is very closely related to 7. tenuissima,

but the habit is quite different. 7. tenuissima is tus-

socked, whereas 7. rivula forms tangled mats. In ad-

dition, the male spikelets are larger and more promi-

nent and the seed is smoother. Both taxa occur in

Boboskloof and maintain their distinction.

CAPE. — 3219 (Wuppertal): South Cedarberg, Suurvlakte, S

of Apex Peak (-CA), Esterhuysen 32435 (BOL; C; E; F; K; L;

LD; M; MO; NBG; NY; PRE; S; STE; TCD; UC; WAG);

Rosendal (-CD), Esterhuysen 33942 (BOL; K; MO; S); Bobos-

kloof (-CD), Esterhuysen 31954 (BOL; K; L; M; MO; S); Ester-

huysen 31616 (B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG;

NY; PRE; RSA; S; STE; TCD; UC; US; W; WAG); Esterhuysen

31912 (BOL; C; E; F; K; L; LD; M; MO; NBG; NY; PRE; S;

STE; TCD; UC; US; W).

Ischyrolepis rottboellioides (Kunth) Linder,

comb. nov.

Restio rottboellioides Kunth, Enum, PI. 3: 394 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 261 (1878); in FI. Cap. 7: 77 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 284(1928). Type: Cape, 3018

(Kamiesberg): Kamiesberg near Leliefontein, 1 200-1 500 m

(-AB), Drege 2494 5 (B, lecto.l; BM!; BOL!; Kl; NY!; PI); 2494

Cf (B!; BM!; BOL!; K!; MO!; NY!; P!).

Note

1. 7. rottboellioides generally occurs in damp to

wet habitats, such as in shallow sand over rock, at

between 800 and 1 200 m, from the Kamiesberg to

Clanwilliam and the Piketberg.

Ischyrolepis sabulosa ( Pillans ) Linder, comb.

nov.

Restio sabulosus in Trans. R. Soc. S. Afr. 16: 285 (1928). Type:

Cape, 3318 (Cape town): four miles east of Mowbray (-DC), Pil-

lans 4396 $ (BOL, holo.!; K!).

Notes

1. This species is unique in the genus by its 5 mm

thick, well developed, creeping rhizomes.

2. 7. sabulosa occurs in marshy hollows on the

Cape Flats (Monte Vista to Phillippi) and the Bre-

dasdorp coastal flats around De Dam.

Ischyrolepis schoenoides (Kunth) Linder, comb.

nov.

Restio schoenoides Kunth, Enum. PL 3 : 391 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 237 (1878); in FI. Cap. 7 : 71 (1897).

Restio sieberi Kunth var. schoenoides (Kunth) Pillans in Trans. R.

Soc. S. Afr. 16 : 278 (1928). Type: Cape, 3027 (Lady Grey): Witte

Bergen, 1 500-1 800 m (-C/D), Drege 50 $ (B, holo.!; BM!;

BOL!; K!; MO!; OXF!; P!).

Notes

1. Within the western Cape, in the Ladismith-

Montagu area, this species grades into a complex in-

volving 7. virgea, I. sieberi and 7. wittebergensis. It is

possibly because of these intermediate forms that

Pillans (1928) included 7. schoenoides in 7. sieberi.

Resolution of the complex still awaits more detailed

work.

2. 7. schoenoides is an ‘Afro-montane’ species,

that grows in montane grassland from King Wil-

liam’s Town to the Blaauwberg near Pietersberg,

generally at altitudes over 1 800 m. To the south,

this species enters the Cape region along the Swart-

berg.

As defined here, 7. schoenoides differs from its

relatives in its chartaceous bracts, stoloniferous ha-

bit and very similar male and female inflorescences.

Ischyrolepis setiger (Kunth) Linder, comb. nov.

Restio setiger Kunth, Enum. PI. 3 : 385 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 236 (1878); in FI. Cap. 7 : 72 (1897);

414

Bothalia 15, 3 & 4 (1985)

Pillans in Trans. R. Soc. S. Afr. 16 : 270 (1928). Type: Cape, 3219

(Wuppertal): on the Koude Berg near Heuning Vlei (-AA),

Drege 2503 9 (B, holo.!; BM!; K!; MO!; NY!; OXF!; P!; S!).

Restio fuirenoides Kunth, Enum. PI. 3 : 386 (1841). Type:

Cape. 3219 (Wuppertal): on the Koude Berg near Heuning Vlei

(-AA), Drege 2504 cf (K, lecto.!; BM!; BOL!; MO!; NY!; OXF!;

P!).

Notes

1. Drege 2503 and 2504 are clearly the female and

male of the same collection.

2. /. setiger is closely related to I. affinis from

Swellendam. The females can only be distinguished

by the seed coats, but the males are very different.

The two species also occupy very different habitats.

3. /. setiger is known from the rather arid northern

Cedarberg, from altitudes between 900 and 1 200 m,

on well-drained sandy soils, or among rocks.

Ischyrolepissieberi (Kunth) Linder, comb. nov.

Restio sieberi Kunth, Enum. PI. 3 : 387 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 244 (1878); in FI. Cap. 7 : 73 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 276 (1928); in Adamson &

Salter, FI. Cape Penins. 140 (1950). Type: Cape, without precise

locality, Sieber 228 cf (B, holo.!; K!; MEL!; MO!; P!; S!;).

Restio scoparius Kunth, Enum. PI. 3: 390 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 240 (1878). Type: Cape, 3320 (Mon-

tagu): Konstable (-AB), Drege 9450 9 (B, holo.!; K!; MO!;

OXF!; P!).

Restio venustulus Kunth, Enum. PI. 3 : 388 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 240 (1878); in FI. Cap. 7 : 70 (1897).

Restio sieberi Kunth var. venustulus (Kunth) Pillans in Trans. R.

Soc. S. Afr. 16 : 278 (1928). Type: Cape, 3320 (Montagu):

Konstable (-AB ) , Drege 9450 cf ( B , holo. ! ; K! ; MO ! ; OXF! ; P! ).

Restio intermedius Kunth, Enum. PI. 3 : 388 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 242 (1878); in FI. Cap. 7 : 73 (1897).

Syntypes; Cape, 3218 (Clanwilliam): between Lange Vlei and

Heerenlogement (-BA), Drege 2475 9 (B, lecto.!; K!; MO!;

OXF!; P!); 2475 cf (B!; K!; OXF!; P!); between Grasberg River

and Waterfalsrivier (-BB), Drege 2500 p.p. cf (K!), nom. illeg. ,

later homonym for Restio intermedius Steud.

Restio divaricatus Mast, in J. Linn. Soc., Bot. 8 : 236 (1865).

Type: Cape, without precise locality, Drege 2490 cf (K, holo.!).

Restio subfalcatus Nees ex Mast, in J. Linn. Soc., Bot. 8 : 231

(1865); in A. DC., Monogr. Phan. 1 : 243 (1878); in FI. Cap. 7 : 75

(1897). Syntypes: Cape, 3218 (Clanwilliam): at Olifantsrivier and

at Brakfontein (-BD), Ecklon s.n. 9 (K, lecto.!; BR!; MEL!;

MO!; S!); s.n. cf (BR!; K!; MEL!; MO!; S!); 3319 (Worcester):

Du Toits Kloof (-CA), Drege 75 cf (B ! ; K!; P!); without precise

locality, Ecklon & Zeyher 50 (?).

Restio neesii Mast, in J. Linn. Soc., Bot. 8 : 232 (1865); in

A. DC., Monogr. Phan. 1 : 244 (1878); in FI. Cap. 7 : 75 (1897).

Syntypes: Cape, 3319 (Worcester): Winterhoeksberg, 300-1 500

m, Nov. (-AA), Ecklon & Zeyher s.n. cf (?); Tulbagh Waterfall,

300-600 m, Dec., (-AC), Ecklon & Zeyher s.n. 9 (?).

Restio pycnostachyus Mast, in J. Linn. Soc., Bot. 8 : 232 (1865);

in A. DC., Monogr. Phan. 1 : 243 (1878); in FI. Cap. 7 : 74 (1897).

Type: Cape, 3319 (Worcester): Du Toits Kloof, 900-1 200 m

(-CA), Drege 79 cf (K, holo.!; B!; MO!; OXF!; P!).

Restio consimilis Mast, in Bot. Jb. 29 Beibl. 66 : 4 (1900). Type:

Cape, 3218 (Clanwilliam): Piquenierskloof, 450 m (-DB),

Schlechter 7946 9 (B, lecto.!; BM!; BOL!; BR!; K!; MO!; S!; Z!>;

Cf (?)■

Restio spiculatus Mast, in Bot. Jb. 29 Beibl. 66 : 4 (1900). Syn-

types: Cape, 3319 (Worcester): Fransch Hoek, 900 m (-CC),

Schlechter 9267 c f (B, lecto.!; BM!; BOL!; BR!; K!; MO!; P!; S!;

Z !); 9268 9 (BM!; BOL!; BR!; K!; MO!; P!; S!; Z!).

Notes

1 . Restio scoparius and R. venustulus are based on

male and female plants of the same collection. They

represent a morph with woolly scales exserted from

the sheath axils. This morph generally occurs in the

higher peaks from the northern Cedarberg to the

Hex River Valley. Pillans (1928) regarded this

morph as a subspecies of I. sieberi. It may well be

distinct at specific level.

2. Restio intermedius is an illegitimate name, as it

is a later homonym of R. intermedius Steud. ( = Ele -

gia intermedia).

3. The type of R. divaricatus (Drege 2490) is also

cited as a type specimen of R. curviramis. R. curvira-

mis is lectotypified by another collection.

4. I have not been able to trace either of the syn-

types of R. neesii, but from the description it is prob-

ably I. sieberi.

5. I. sieberi is a very variable species, but the only

variation that might be significant is the presence of

woolly scales in the sheath axes of some high-alti-

tude populations.

6. I. sieberi is ubiquitous in the Cape Flora, ex-

tending from Grahamstown to Namaqualand. It also

occurs in the Richtersveld. Generally it occurs in

soils derived from sandstone or quartz.

Ischyrolepis sporadica Esterhuysen, sp. nov., ab

I. curvirami (Kunth) Linder culmis partibus basali-

bus stoloniformibus, inflorescentiis masculis spiculis

paucioribus, a I. pratensi Esterhuysen spiculis et

floribus minoribus, spiculis femineis semper unifloris

recedit.

TYPE. — 3318 (Cape Town): Riverlands near

Mamre Road (-BC), Esterhuysen 34656 $ (BOL,

holo.!; C; E; K; L; M; MO; S; STE).

Plants caespitose, tussocks 15-25 cm tall, spread-

ing by stolons. Culms solid, terete, branching, ter-

minal sections either straight or flexuose, to 1 mm in

diam., surface densely tuberculate, basal portions of

the culms often functioning as stolons. Sheaths con-

voluted, 3-8 mm long, the coriaceous portion trun-

cate, awn 5-| of the total length of the sheath, sub-

clavate to subaciculate, hyaline shoulders smaller

than the awn; base of the sheath swollen. Male inflo-

rescences of 1-2, 4—6 mm long spikelets. Spathes sim-

ilar to the sheaths, somewhat shorter than the spike-

lets. Bracts cartilaginous with membranous margins,

about 4 mm long, obscuring the flowers, acute with a

slender acute awn. Flowers subsessile, perianth

3- 3,5 mm long. Sepals cartilaginous, acute; laterals

conduplicate, villous to sparsely pilose in the upper

third, 3 mm long; odd sepal concave, glabrous, 2,5 x

0,6 mm. Petals acute, 2,5-3 x 0,6 mm. Pistillode ab-

sent. Anthers exserted at anthesis, 1,8-2 mm long,

scarcely mucronate. Female inflorescence of 1 (2),

4— 6 mm long, single-flowered spikelets. Spathe simi-

lar to sheath, about as tall as the spikelet. Bracts 2,

acute with the apex continued as an aciculate, 1 mm

long awn, body 4-5 mm long, cartilaginous with

membranous margins. Flowers subpedicellate, per-

ianth 3-3.5 mm long. Sepals cartilaginous, reddish

above in young flowers, concave; laterals sparsely

pillose to villous on the midrib, piliferous-acute; odd

sepal very acute, 3 x 0,7 mm. Petals chartaceous to

membranous, acute to sometimes piliferous, 2,5 x

0,6-1 mm. Staminodes lost. Styles 2, about 5 mm

long, lower 1,5 mm fused, minutely muricate. Ovary

bilocular; fruit a unilocular capsule.

Bothalia 15, 3 & 4 (1985)

415

/. sporadica is not well understood. It occurs

mostly in the coastal forelands between Malmesbury

and Bredasdorp, often in seasonally wet sand. Some

collections are from gravelly or more mountainous

regions.

This new species is related to I. pratensis and /.

curviramis on account of the truncate sheaths, the

flower structure and the tuberculate culms. It is sep-

arated from I. curviramis by the stoloniferous habit

and the fewer male flowers, and from I. pratensis by

the smaller flowers and spikelets and by the consist-

ently single-flowered female spikelets.

CAPE. — 3318 (Cape Town): Souterivier near Philadelphia

(-DA), Esterhuysen s.n. (BOL; K); sandveld in the Mamre Road

area (-BC), Esterhuysen 34276 (BOL; K; MO; S); Riverlands

(-BC), Esterhuysen 34656 (BOL; C; E; K; L; M; MO; S; STE);

field near Malmesbury (-BC), Esterhuysen 32532a (BOL; K;

MO; S); Rondebosch Common (-CD), Esterhuysen 29905 (BOL;

C; E; K; L; M; MO; S). 3418 (Simonstown): along the Swartkops

Range between Smitswinkel and Millers Point (-AB), Esterhuy-

sen 32525 (BOL; K); flats between Gordon’s Bay and the Strand

(-BB), Esterhuysen 32286 (BOL; K; S); Esterhuysen 34034

(BOL; K; S). 3419 (Caledon): Pheasantshoek near Viljoenshof

(-DA), Esterhuysen 34359a (BOL; K).

Ischyrolepis subverticellata Steud., Syn. PI.

Glum. 2 : 249 (1855). Type: Cape, 3319 (Worces-

ter): Tulbagh Waterfall (-AC), Dr'ege 1 $ (P, holo. ! ;

B!; BOL!; K!; MO!; NY!; OXF!; P!; S!), Fig. 7.

Restio subverticellatus (Steud.) Mast, in J. Linn. Soc., Bot. 8 :

227 (1865); in A. DC., Monogr. Phan. 1 : 248 (1878): in FI. Cap.

7 : 69 (1897): Pillans in Trans. R. Soc. S. Afr. 16 : 281 (1928).

Restio ameles Steud., Syn. PI. Glum. 2 : 252 (1855). Type:

Cape, 3319 (Worcester): Du Toits Kloof (-CA), Drege 1620 cf (P,

holo.!; MO!; NY!; OXF!; P!).

Icon: A. DC., Monogr. Phan. 1 : t. 5 f. 2 (1878).

Notes

1. It is not clear whether Drege 1 is from Tulbagh

Waterfall or from Genadendal. The latter locality is

given on the sheets in some herbaria.

2. I. subverticellata occurs along streams, often in

half-shade, in the mountains from Caledon to Paarl.

There is one record from the Hex River Mountains.

It is absent from the Cape Peninsula.

Ischyrolepis tenuissima (Kunth) Linder , comb.

nov.

Restio tenuissimus Kunth, Enum. PI. 3 : 394 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 268 (1878); in FI. Cap. 7 : 81 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 285 (1928); in Adamson &

Salter, FI. Cape Penins. 141 (1950). Syntypes: Cape, 3319 (Wor-

cester): Tulbagh Waterfall (-AC), Drege 1970 9 (B, lecto.!; BM!;

BOL!; K!; MEL!; MO!; NY!; OXF!; P!; S!); Du Toits Kloof

(-CA), Drege 1625 cf (B!; BM!; K!; MO!; OXF!; P!).

Restio nutans Steud. in Flora, Jena 12 : 134 (1829), non Thunb.

(1803). Type: Cape, without exact locality, Ludwig s.n. (OXF,

lecto.!).

Restio ludwigii Steud., Syn. PI. Glum. 2 : 254 (1855); Mast, in

A. DC., Monogr. Phan. 1 : 268 (1878); in FI. Cap. 7 : 80 (1897).

Nom. nov. for R. nutans Steud. non Thunb.

Hypolaena tenuissima Pillans in J1 S. Afr. Bot. 18 : 119 (1952).

Type: Cape, 3219 (Wuppertal): Olifants River (-CC), Esterhuy-

sen 14286 $ (BOL, holo.!).

Notes

1. Although it seems unlikely that Steudel saw the

Ludwig specimen at OXF, it may well be an isotype .

2. I. tenuissima is very close to 1. rivula, from

which it may be distinguished by its habit and seed

coat morphology.

3. I. tenuissima ranges from Swellendam to the

Koue Bokkeveld near Ceres, the Piketberg and the

Cape Peninsula. Curiously, it is absent from the Ce-

darberg. Most of the collections are from mountain-

ous areas, where it occurs along small streams, on

rock flushes and in marshes. A few collections from

the Caledon-Bredasdorp area occur on hard, dry, of-

ten gravelly soils.

Ischyrolepis triflora (Rottb.) Linder, comb. nov.

Restio triflorus Rottb., Descriptiones Plantarum Rariorum 10

(1772); Thunb., FI. Cap. ed 1, 33 (1811); ed Schultes, 86 (1823);

Kunth. Enum. PI. 3 : 391 (1841); Mast, in A. DC., Monogr. Phan.

1 : 249 (1878); in FI. Cap. 7 : 73 (1897); Pillans in Trans. R. Soc.

S. Afr. 16 : 290 (1928); in Adamson & Salter, FI. Cape Penins.

141 (1950). Type: Cape, without precise locality, Konig cf s.n. (C,

holo.!; C!; K!).

Restio kunthii Steud., Syn. PI. Glum. 2 : 251 (1855). Type:

Cape, 3318 (Cape Town): Paarl Mountains, 300-600 m (-DB),

Drege 69 9 (K, lecto.!; B!; BM!; K!; MO!; NY!; OXF!; P!; S!).

Restio araneosus Mast, in Bot. Jb. 29 Beibl. 66 : 3 (1900). Type:

Cape, 3419 (Caledon): Houw Hoek (-AA), Schlechter 7444b 9

(B. lecto.!; BOL!)

Icon: Rottb., Descriptionumeticonumrariorest. 2f. 2(1773).

Notes

1. Although there are two sheets of Drege 69 at P,

neither are from herb Steud. and so neither are holo-

types.

2. There is a specimen of Schlechter 7 444 at K, but

not Schlecjtter 7444b. They are probably the same,

but I cannot prove that.

3. I. triflora ranges from Paarl and the Cape Pen-

insula to Grahamstown, generally along the foothills

of the mountains. It appears to prefer clayey soils,

derived from shale or granite and is often associated

with grassy fynbos or renosterveld.

Ischyrolepis unispicata Linder , sp. nov., ab /.

sieberi (Kunth) Linder culmis brevioribus glaucis,

testa retifoveolata differt.

TYPE. — Cape, 3219 (Wuppertal): Northern

Cold Bokkeveld, Vredelus at E base of Schoonge-

zicht Peak (-CC), Esterhuysen 29670 $ (BOL,

holo.!; C; E; K; L; M; MO; S).

Plants caespitose, stoloniferous, tussocks rather

disorganized, 20-40 cm tall. Culms solid, terete,

branching, glaucous, to 1 mm in diam., rather

densely and obscurely tuberculate, rarely smooth,

often somewhat flexuose. Sheaths tightly convo-

luted, 5-10 mm long, acute, mucronate, body coria-

ceous, reddish with yellow mottling, turning grey,

apex scarcely chartaceous, occasionally a woolly

scale present in the axil of the sheath. Male inflores-

cence of 1-3 imbricate, 5-15 x 1-1,5 mm, acute, of-

ten curved spikelets. Spathes and bracts similar, ob-

tuse to acute, awn acute, about 1 mm long, body car-

tilaginous with a membranous upper margin, 3-5

mm long, as long as the flowers, almost all fertile.

Flowers subpedicellate, perianth 3 mm long. Sepals

cartilaginous, acute; lateral sepals conduplicate, car-

ina densely villous in the upper half; odd sepal 3,2 x

0,8 mm, flat. Petals more membranous, subacute,

2,9 x 0,8 mm. Anthers 2,5 mm long, exserted at an-

416

Bothalia 15, 3 & 4 (1985)

ch're. smrft\

r ^ er J i; ^-^owe; with

exserted anthers, x 8. (From Esterhuysen 30205; 30502.)

Bothalia 15, 3 & 4 (1985)

417

thesis. Female inflorescence a solitary, 7-15 x 1,2

mm, 1-3-flowered spikelet. Spathes similar to the

bracts, coriaceous with a narrow subchartaceous

margin, acute, 4,5-7 mm long, awn stout, 0,5-1 ,5

mm long, acute, usually obscuring the flowers. Flow-

ers subsessile, perianth 3,5-6 mm long, occasionally

visible above the bracts. Sepals cartilaginous, acute;

lateral sepals conduplicate, villous-carinate, 3,5-6

mm long; odd sepal flat, 3-5,5 x 1 mm, somewhat

villous on the midrib. Petals subcartilaginous, obtuse

to acute, 2-5 x 1-2 mm. Styles 2, fused in the lower

1,5 mm. Ovary bilocular. Fruit a unilocular capsule;

seed 1, 8-2,1 x 1, 3-1,5 mm, more or less triangular

in cross-section, ends obtuse, surface reticulate-fo-

veolate. Fig. 15.

I. unispicata occurs along the arid inland fringe of

the Cape sandstone mountains, from Clanwilliam to

Uniondale. The altitude range of the species is be-

tween 900 and 1 800 m. The substrate varies from

sandy flats to gravelly and stony slopes, from pla-

teaux to shale bands to exposed ridges. This is quite

a common species, and occurs widespread and com-

monly in well-drained conditions.

Generally the culms of this species are tubercu-

late, but at the summit of the Swartberg Pass be-

tween Oudtshoorn and Prince Alfred is a population

with smooth culms and the bracts somewhat shorter

than the perianth. I am including this population

under I. unispicata , although further research may

indicate that it is distinct.

I. unispicata tends to be somewhat intermediate

between I. sieberi and I. distracta. It is distinct in its

bluish, often flexuose culms, the solitary female and

only 1-3 male spikelets, and the small, somewhat

loosely organized tussocks. It is clearly a segregate

from I. sieberi , a widespread and variable species.

CAPE. — 3219 (Wuppertal): east base of Schoongezicht Peak,

Cold Bokkeveld (-CC), Esterhuy sen 29670 (BOL; C; E; K; L; M;

MO; S); near Sneeuberg hut (- ), Esterhuysen 30098 (BOL; C; E;

F; K; L; LD; M; MO; NBG; S; STE; UC); Bokkeveld Sneeukop,

upper slopes of the shale band (-CD), Esterhuysen 33939a (BOL;

K; S). 3319 (Worcester): Tulbagh district, neck at head of Sneeu-

gat Kloof (-AA), Esterhuysen 29947 (BOL; K); Ceres district,

Baviaansberg (-BA), Esterhuysen 29788 (BOL; K; M; MO; S);

Hex River Mountains, Buffelshoek Kloof (-CC), Esterhuysen

33339 (BOL; K; M; MO; S); Soutrivier on the Koo Road, c. 14

miles S of Matroosberg Station (-DB), Jessop s.n. (BOL; K; S);

Jona's Kop (Langberg) (-DC), Esterhuysen 32699a (BOL; K; S).

3320 (Montagu): between Touws River and Constable (-AA),

Esterhuysen 32635 (BOL; K; M; MO; S). 3322 (Oudtshoorn): N

of Tierberg, upper slopes (-AA), Esterhuysen 29553 (BOL; K).

Ischyrolepis vilis (Kunth) Linder , comb. nov.

Restio vilis Kunth, Enum. PI. 3 : 389 (1841); Mast, in A. DC.,

Monogr. Phan. 1 : 241 (1878); in FI. Cap. 7 : 72 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 275 (1928). Syntypes: Cape, 3018 (Ka-

miesberg): Kamiesberge. 900-1 200 m (-AC), Drege 2476 $ (B,

lecto.l; BOL!; K! MO!; OXL!; P!); 2476 o" (B!; K!; OXF!; P!);

without precise locality, Drege 75 c f ( B ! ; K! ; MO!; NY!; P!).

Notes

1. Drege 75 is I. sieberi , the type of Restio subfal-

catus Mast. The specimen in B is labelled ‘ R . vilis' by

Kunth.

2. The material of Drege 2476 at B is ex herb.

Meyer, so it cannot be holotype material, but is iso-

type material.

3. I. vilis is known only from the Kamiesberg,

where it occurs on the high granitic peaks.

Ischyrolepis virgea (Mast.) Linder , comb. nov.

Restio virgeus Mast, in Bot. Jb. 29 Beibl. 66: 4 (1900); Pillans in

Trans. R. Soc. S. Afr. 16 : 272 (1928). Syntypes: Cape, 3319

(Worcester): Gydouwsberg, 1 850 m (-AB), Schlechter 10231 cf

(B, lecto.!; BM!; BOL!; BR!; K!; MO!; P!; S!; Z1)- 10232 $

(BM!; BOL!; BR!; K!; MO!; P!; S!; Z!)

Notes

1 . I. virgea is very closely related to I. wittebergen-

sis and I. schoenoides and collections from the Mon-

tagu-Ladismith area are difficult to determine.

2. I. virgea occurs in dry or well-drained habitats

in the mountains from Caledon to Clanwilliam, but

it is not known from either the Cape Peninsula or the

Piketberg. It is a high-altitude species, with most col-

lections from above 1 500 m.

Ischyrolepis wallichii (Mast.) Linder , comb.

nov.

Restio wallichii Mast, in J. Linn. Soc., Bot. 8 : 234 (1865); in

A. DC., Monogr. Phan. 1 : 251 (1878); in FI. Cap. 7 : 76 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 291 (1928). Type: Cape,

without precise locality, Wallich s.n. 9 (K, lecto.!; BM!; BOL!

K!).

Restio humilis Pillans in Ann. Bolus Herb. 3 : 84 (1921). Syn-

types: Cape, 3319 (Wuppertal): Koue Bokkeveld, Kleinvlei,

1 500 m (-CB), Schlechter 10218 9 (BOL, lecto.!; K! ; MO!; P!;

S!; Z!). 3419 (Caledon): Houw Hoek, 600 m (-AA), Schlechter

7416 9 (B!; BM!; BR!; S!; Z!).

Notes

1. There are two sheets of Wallich s.n. at K, both

annotated by Masters. One is selected as lectotype.

2. I. wallichii ranges from Keysersfontein in Na-

maqualand to Riviersonderend. Most of the collec-

tions are from sandbanks in riverbeds, a few are

from seepages. Although this species occurs over

such a wide area, it is not often collected.

Ischyrolepis wittebergensis Esterhuysen, sp.

nov., ab. I. virgea (Mast.) Linder culmis plerumque

simplicibus, tuberculatis, mucronibus antherum 0,5

mm longis differt.

TYPE. — Cape, 3320 (Montagu); upper rocky

S slopes of the Witteberge (-BC), Esterhuysen 30479

$ (BOL, holo.! C; E; F; K; L; LD; M; MO; NBG;

S; STE; UC).

Plants caespitose, to 50 cm tall. Culms simple or

very rarely branched, about 1,3 mm in diam.,

smooth to very finely and densely tuberculate,

aggregated at the base to form short ascending root-

stocks, young culms often drying sulcate. Sheaths

convoluted, to 4 cm long, piliferous, coriaceous,

grading into a very wide membranous margin, older

sheaths decaying to short, very lacerated structures.

Male inflorescence of numerous spikelets aggregated

into a single, 1-2 cm in diam. globose capitulum.

Spathes to 2,5 cm long, narrowly ovate with long

awns. Spikelets about 8x3 mm. Bracts acuminate,

5-6 x 2-2,5 mm, somewhat awned, coriaceous with

submembranous margins. Flowers subsessile, per-

ianth 4-5,5 mm long. Tepals chartaceous to subcarti-

laginous; lateral sepals conduplicate, piliferous, Car-

ina pilose, 4,5-5 mm long; odd sepals and petals

418

Bothalia 15, 3 & 4 (1985)

acute, 3,8 x 0,9 mm. Anthers exserted at anthesis, 2

mm long with a 0,5 mm long mucro. No trace of a

pistillode. Female inflorescence of 1-6, 10-20 mm

long several-flowered spikelets aggregated into a sin-

gle globose capitulum. Spathes overtopping the spi-

kelets, acuminate-piliferous, coriaceous with broad

membranous margins. Bracts cartilaginous, acumi-

nate with the apex awn-like, about 10 mm long,

obscuring the flowers, the lower bracts sterile. Flow-

ers subsessile, perianth 6, 5-7, 5 mm long. Sepals car-

tilaginous; laterals conduplicate, carina densely vil-

lous, especially in the lower §, apex slender acute,

6, 5-7, 5 mm long; odd sepal acute, flat, 5 x 1,8 mm.

Petals chartaceous-cartilaginous, acute, 4,5 x 1,8

mm. No staminodes visible. Styles 2, about 8 mm

long, fused in the lower 3, inner surface of the

branches pilose. Fruit a bilocular capsule; seed 2 x

1,3 mm, obscurely triangular in cross-section, ends

obtuse, ridge faintly developed, surface very finely

reticulate-foveate.

Ischyrolepis wittebergensis appears to be restricted

to the summit ridge of the Witteberg, near Matjies-

fontein, at 1 350-1 500 m, where it grows among

rock outcrops.

This new species is closely related to I. virgea

(Mast.) Linder. This species, in turn, is difficult to

delimit from the south-western extension of I. schoe-

noides (Kunth) Linder. I. wittebergensis can be sep-

arated from the whole complex by the long mucro on

the anther and the generally simple culms. It is pos-

sible that a second segregate, distributed as ‘Restio

sp. aff. wittebergensis', based on the material from

the adjacent Little Swart Berg, can also be recog-

nized, but that would first require the resolution of

the I. schoenoides - I. virgea complex.

CAPE. — 3320 (Montagu): Witteberg, along the rocky summit

(-BC), Esterhuysen 31500 (BOL; E; K; L; M; MO; S); Esterhuy-

sen 28857 (BOL; K); Esterhuysen 30479 (BOL; C; E; F; K; L;

LD; M; MO; NBG; S; STE; UC).

ELEGIA

3. Elegia L. , Mant. Alt. 162 (1771); Linder in

Bothalia 15 : 63 (1984).

Notes

1. ‘Deciduous’ or ‘caducous’ sheaths or spathes

have distinct abscission zones. They are often per-

sistent until immediately before anthesis, but rarely

or never persist during anthesis.

2. If the culms are ‘distinctly fistular’ (couplet 11),

then they are easily compressed.

3. Couplet 18 is difficult to use, as robust or de-

pauperate individuals are easily placed into the

wrong group. However, by not being too rigid, the

majority of specimens key out correctly.

4. Couplet 31 may also cause some difficulty, as it

is difficult to describe the inflorescence shape in 31a

accurately.

KEY TO THE SPECIES OF ELEGIA

la Culms branching:

2a Branches verticellate or subverticellate:

3a Inflorescence 15-20 cm long; spathes deciduous E. capensis

3b Inflorescence 6-12 cm long; spathes persistent E. equisetacea

2b Branches dichotomous:

4a Styles 3:

5a Bracts 7-8 mm long, oblong, obtuse; inner perianth segments oblong E. muirii

5b Bracts about 2 mm long, ovate-lanceolate, acute; inner perianth segments rotundate or orbicular

E. hutchinsonii

4b Styles 2:

6a Styles slender, curled in the upper half, with the free parts longer than the stylopodium:

7a Culms rarely branched, spathellae prominent E. stokoei

7b Culms much branched, spathellae not visible E. stipularis

6b Styles stout, free parts shorter than the stylopodium:

8a Ovary much exceeding the perianth in length:

9a Culms much branched, branchlets almost filiform; sheaths 10-15 mm long; inflorescence

1 ,5—2,5 cm long; perianth about 1 mm long E. prominens

9b Culms sparingly branched; sheaths 3-6 cm long; inflorescence 4—6 cm long; perianth 2-2,5

mm long E. neesii

8b Ovary not exceeding the perianth in length:

10a Perianth 2-3,5 mm long E. neesii

10b Perianth more than 5 mm long E. grandis

lb Culms simple, very rarely branched:

11a Culms distinctly fistular E. fistulosa

lib Culms solid or rarely with a narrow, usually interrupted, internal passage:

12a Leaf-sheaths persistent, tightly convoluted:

13a Bracts twice as long as the perianth, slender acuminate; sheaths with purple apices E. fucata

13b Bracts at most as long as the perianth; sheaths without purple apices:

Bothalia 15, 3 & 4 (1985)

419

14a Plants rhizomatous; culms terete to somewhat compressed E. asperiflora

14b Plants caespitose; culms very compressed E. caespitosa

12b Sheaths deciduous:

15a Culms compressed:

16a Bracts and tepals lacerated E. asperiflora

16b Bracts entire; tepals somewhat lacerated:

17a Plants rhizomatous E. coleura

17b Plants caespitose E. caespitosa

15b Culms terete:

18a Culms at most 50 cm long, or at most 1,25 mm in diam.:

19a Bracts twice as long as the perianth, slender acuminate; sheaths with purple apices E. fucata

19b Bracts at most as long as the perianth; sheaths without purple apices:

20a Rhizomes creeping:

21a Perianth 5-7 mm long E. esterhuyseniae

21b Perianth less than 4 mm long:

22a Culms tufted on the rhizome E. verreauxii

22b Culms evenly spread on the rhizome:

23a Culms closely arranged, 10-20 cm long E. squamosa

23b Culms sporadic, 20-40 cm long E. vaginulata

20b Rhizomes absent or very short:

24a Spathellae conspicuous, aristate, 10-12 mm long E. fastigiata

24b Spathellae inconspicuous or absent, less than 5 mm long:

25a Spathes 10-15 mm long, erect, dorsally convex, not obscuring the flowers E. filacea

25b Spathes 10-30 mm long, flat, recurved at the apex, obscuring the flowers:

26a Spathes 10-20 mm long E. rigida

26b Spathes 20-30 mm long E. spathacea

18b Culms at least 50 cm long, or at least 1,25 mm in diam.:

27a Bracts aristate:

28a Awns 1^1 mm long; sheaths pale brown E. grandispicata

28b Awns longer than 5 mm; sheaths chestnut-brown E. cuspidata

27b Bracts not aristate, at most subulate-mucronate:

29a Rhizomes long, creeping; culms tufted:

30a Spikelets dark chestnut-brown or blackish; flowers almost rotund E. galpinii

30b Spikelets pale brown to straw-coloured; flowers oblong E. extensa

29b Rhizomes absent, or if present, culms evenly spaced:

31a Inflorescence linear or tapering; spathes erect to erect-spreading, imbricate, much tal-

ler than the flowers; flowers rarely more than 10 per node:

32a Flowers 5-7 mm long E. esterhuyseniae

32b Flowers less than 5 mm long:

33a Spathes concolorous, in the male early deciduous E. thyrsoidea

33b Spathes variegated, the margins much paler than the body, in the male usually persistent:

34a Perianth lobes acute E. racemosa

34b Perianth lobes rounded E. persistens

31b Inflorescence oblong; spathes less imbricate, spikelets often exposed; flowers usually

more than 10 per node:

35a Spathes papery E. intermedia

35b Spathes coriaceous to cartilaginous:

36a Spathes, or rarely the spathellae, obscuring the spikelets, spathes persistent...E. juncea

36b Spathes only partially obscuring the spikelets:

37a Tepals papillose, reddish E. atratiflora

37b Tepals smooth, pale brown:

38a Bracts taller than the flowers; perianth about 3 mm long E. grandispicata

38b Bracts shorter than the flowers; perianth 2-3 mm long:

39a Plants rhizomatous; spikelets 5-8-flowered E. fenestrata

39b Plants caespitose; spikelets with 3-4 flowers E. thyrsifera

420

Bothalia 15, 3 & 4 (1985)

Elegia altigena Pillans in Trans. R. Soc. S. Afr.

30: 257 (1945). Type: Cape, 3322 (Oudtshoorn):

Mannetjiesberg, Kamannassie, 1 800 m (-DB), Es-

terhuysen 6399 $ (BOL, lecto.!; K!); 6399 cf (BOL!;

K!).

Note

1. This species is distinct from E. juncea by having

two, instead of three, styles. However, this could

well be a trivial character. It is only known from the

type collection.

Elegia asperiflora (Nees) Kunth , Enum. PI. 3 :

474 (1841); Mast, in A. DC., Monogr. Phan. 1 : 355

(1878); in FI. Cap. 7 : 110 (1897); Pillans in Trans.

R. Soc. S. Afr. 16 : 328 (1928); in Adamson &

Salter, FI. Cape Penins. 147 (1950). Syntypes: ‘In

planitie Capensi, Decembri’, Ecklon s.n. (SAM);

‘Femina in horto Schwetzingensicolitur' (B, lecto. !).

Restio asperi floras Nees in Linnaea 5 : 656 (1830).

Elegia dregeana Kunth, Enum. PI. 3: 469 (1841). Syntypes:

Cape. 3319 (Worcester): Du Toits Kloof, 600-900m (-CA),

Drege 1640 9 (B. lecto.!; BM!; MO!; P!). 3323 (Willowmore):

between Welgelegen and Onzer, 450-600 m (-CD), Drege 102 C ?

(B!: K!; MOP P!).

Elegia glauca Mast, in J. Linn. Soc., Bot. 21 : 579 (1885); in FI.

Cap. 7 : 107 (1897); Pillans in Trans. R. Soc. S. Afr. 16 : 331

(1928), but see below. Syntypes: Cape, 3419 (Caledon): Hotten-

tots Holland Mountains near Grietjies Gat (-AA), Bolus 4221 9

(K, lecto.!). 3319 (Worcester): mountains above Worcester

(-CB/DA). Rehmann 2564 9 (K!; BM!). 3321 (Ladismith): sum-

mit of Kampsche Berg, Riversdale (-CD), Burchell 7105 cf

(BOL!; K!; OXF!).

Elegia ciliata Mast, in Bot. Jb. 29 Beibl. 66 : 8 (1900). Type:

cultivated in Schwetzingen Gardens, 9 (B, holo.!), nom. illeg.,

superfluous name for Restio asperiflora Nees.

Elegia lacerata Pillans in Ann. Bolus Herb. 3 : 144 (1922).

Elegia asperiflora (Nees) Kunth var. lacerata (Pillans) in

Trans. R. Soc. S. Afr. 16 : 330 (1928). Syntypes: Cape, 3326 (Gra-

hamstown): Stony hills near Grahamstown (-BC), MacOwan

1359 9 (BOL, lecto.!; MEL!); 1360 cf (BOL!). 3227 (Stutter-

heim): summit of Mt Pirie (-CC), Galpin 5944 cf (BOL!); Dohne

Hill (-CB), Sim 2843 cf (BOL!). 3322 (Oudtshoorn): Zwartberg

Pass (-AC), Bolus 11661 cf (BOL!). 3319 (Worcester); Du Toits

Kloof (-CA), Bolus 16997 cf (BOL!).

Icon: A. DC., Monogr. Phan. 1 : t.3 f. 29-32; t.5 f . 14 (1878).

Notes

1. The only Zeyher collection from the Cape Flats

which I have been able to trace is in Cape Town

(SAM) and is most likely an isosyntype. At B and Z

there are Ecklon and Zeyher collections from T. 12’,

i.e. Tulbagh Waterfall, collected in December.

Maybe Nees made a mistake with his localities.

However, at B is a female specimen, cultivated in

the Schwetzingen Gardens. It is annotated as being

the type of Elegia ciliata Mast., but is also probably

the second syntype cited by Nees. It is obviously the

best lectotype for E. asperiflora, as Nees probably

saw it. This changes E. ciliata into a superfluous

name.

2. Pillans (1928) was clearly mistaken as to the

identity of E. glauca Mast. Bolus 4221 is selected as

lectotype, as Masters annotated this sheet at K. The

Burchell collection at K has an original label with a

Burchell ms name ‘Restio glaucus’.

3. E. asperiflora is quite variable, but I doubt that

var. lacerata (Pillans) Pillans can be recognized. The

variation patterns in E. asperiflora need to be stu-

died in detail.

4. E. asperiflora is ubiquitous in the Cape Floral

Region, ranging from Port Alfred to Clanwilliam,

from the Cape Penninsula to Laingsburg, from sea

level to 1 500 m. It occurs in marshy areas and

among streams and seepage lines. It is usually asso-

ciated with sandy soils.

Elegia atratiflora Esterhuysen, sp. nov., ab E.

fenestrata Pillans rhizomatibus absentibus, tepalis

acutis, a E. thyrisifera Persoon vaginis et floribus

atro-castaneis, ab ambabus tepalis et bracteis papil-

losis recedit.

TYPE. — Cape, 3418 (Simonstown): Oudebos,

Palmiet River (-BD), Esterhuysen 35251 $ (BOL,

holo.!; C; E; F; GRA; K; L; LD; M; MO; NBG;

PRE; RSA; S; STE; TCD; UC; US; W; WAG).

Plants caespitose, tussocks 40-70 cm tall. Culms

solid, compressed, simple, to 2 mm in diameter, sur-

face smooth. Sheaths deciduous, 1-2 per culm,

rather loosely convoluted, to 9 cm long, shiny, dark

brown, cartilaginous with the margins chartaceous,

acute, often subpersistent and largely decayed. Male

inflorescence up to 10 x 1,5 cm. Spathes subpersis-

tent, similar in texture to the sheaths, 1-5 cm long,

not obscuring the inflorescence. Inflorescence

branches numerous, branching, up to 3 cm long,

with very numerous, 1-4-flowered, subglobose, 3^4

mm in diameter spikelets. Spathellae narrowly lan-

ceolate, acute, 0,5-1, 5 cm long. Bracts rounded to

acute, 1-1,5 mm long, lanceolate to orbicular, often

subcarinate, upper margins finely lacerate, backs fi-

nely papillose. Flowers 2-2,5 mm long, petals

overtopping the sepals. Tepals acute, cartilaginous,

somewhat papillose on the outside in the upper half,

concave; sepals 1,5 x 1 mm, petals 2x1 mm. Anth-

ers dehisce inside the perianth, 1mm long, mucro-

nate. Pistil lode minute. Female inflorescence similar

to the male, but somewhat smaller, and with the

spathes largely obscuring the spikelets, inflorescence

branches fewer and less branched than in the males

and shorter than the spathes. Spikelets 2-4-flowered,

3-5 mm in diameter. Bracts acuminate to obtuse,

cartilaginous, half as long as to twice as long as the

flowers, papillose on the backs, especially along the

carina. Flowers 2,5-3 mm long, perianth lobes about

the same length. Tepals equal, 2-2,4 x 1 mm, con-

cave, mucronate, acute, papillose, especially along

the midrib. Staminodes minute. Styles three, free.

Ovary unilocular, fruit a round, black, 2 mm long

unilocular nutlet.

E. atratiflora is only known from the mountains

between the Cape Flats and the Klein River at Her-

manns. The species occurs roughly between 300 and

900 m, generally in marshy places, often in shallow

soils. Several collections record single plants. Flow-

ering occurs in September.

This new species is allied to E. thyrsifera and E.

fenestrata. It has the same type of inflorescence as

these two taxa, but lacks the rhizome and the obtuse

to rounded tepals of E. fenestrata, and the sheaths

and the flowers are much darker in colour than in E.

thyrsifera and the tepals are more papillose and less

acute. E. fenestrata occupies a seasonally much

Bothalia 15, 3 & 4 (1985)

421

drier, sandier habitat in the coastal forelands,

whereas the habitat of E. thyrsifera is rather similar

to that of E. atratiflora.

CAPE. — 3418 (Simonstown): Oudebos, Palmiet River (-BD),

Esterhuysen 33235 (BOL;K;MO;S); Ester hay sen 35277a (BOL;

K; S); Palmiet River Valley (-DB), Esterhuysen 35251 (B; BOL;

C; E; F; GRA; K; L; LD; M; MO; NBG; NY; PRE; RSA; S;

STE; TCD; UC; US; W; WAG); Kogelberg Reserve, between

Blousteenberge and Wynand Louwsbos (-BD), Esterhuysen

35445 (BOL; C; K; L; M; MO: S). 3419 (Caledon); mountains

above Hermanus, ridge below Aasvogelkop (-AD), Esterhuysen

35301 (BOL; K; MO; S).

Elegia caespitosa Esterhuysen, sp. nov., ab E.

asperiflora (Nees) Kunth plantis caespitosis, culmis

complanatis differt.

TYPE. — Cape, 3419 (Caledon): Palmiet River

Mountains, on the lower marshy gradual slopes

(-AC), Esterhuysen 34836 ? (BOL, holo.!; C; K; L;

M; MO; S).

Plants caespitose, non-rhizomatous, 30-60 cm tall.

Culms solid, compressed, simple, to 2 mm wide, sur-

face smooth. Sheaths subpersistent, chartaceous to

sub-cartilaginous, 3-8 x 0,8 cm, acute, awn slender

aciculate, to 6 mm long, yellowish with fine and

dense red mottling. Male inflorescence 20-60 x

15-25 mm, dense. Spathes persistent, chartaceous,

to 25 x 6 mm, acute, not obscuring the flowers, pale

yellow with sparse mottling. Inflorescence branches

about as long as the spathes, with very numerous,

globose, 2 mm long, 3-6-flowered spikelets. Spathel-

lae up to 12 x 3 mm, acute, pale yellow, much

overtopping the spikelets, subcarinate. Spathellae

and lower bracts chartaceous, slender-acuminate,

2-3 times as tall as the flowers, dominating the ap-

pearance of the spikelets, upper bracts glandular-pi-

lose on the midrib. Flowers subpedicellate, perianth

1,5 mm long. Tepals acute, sparsely glandular-pi-

lose; sepals 1,2 x 0,6 mm; petals 1,5 x 0,6 mm.

Anthers dehisce inside the perianth, 1 mm long. Pis-

tillode minute. Female inflorescence subglobose to

somewhat elongated, 2-3 x 2 cm. Spathes up to 35 x

10 mm, similar to the males. Inflorescence branches

shorter than the spathes, rather sparsely branched.

Spathellae as in the males, but obscuring the spike-

lets. Spikelets about 3 mm long, with 1-4 flowers.

Bracts much longer than the flowers, conduplicate,

acute, with a long acute awn; carina and awn

sparsely pilose; lower bracts much longer than the

upper bracts; generally all bracts fertile. Flowers

subsessile, perianth l,5-2mm long, lower flowers

maturing long before the upper flowers. Tepals char-

taceous, obtuse, the upper margins finely lacerated,

mucronate, subcarinate, sparsely glandular-pilose,

sub-equal with the sepals marginally shorter than the

petals, petals fused at the base. Styles 3, style bases

adjacent, styles about 1 mm long, slender, pilose.

Ovary unilocular. Fruit a unilocular, subtriangular,

1,5-2 mm long black nutlet.

E. caespitosa occurs in the Palmiet River Valley

and the surrounding mountains, and is known from

single collections from Swellendam and Bainskloof.

All collections are from very marshy areas, on sand

or almost clayey soil. The flowering period is ex-

tended by the sequential flowering of the flowers in

the female spikelets and occurs over December.

This species is clearly a segregate of E. asperiflora.

It is distinct in lacking a rhizome and very com-

pressed culms. There is also a difference in the ap-

pearance of the male inflorescence. However, it is

difficult to assign some herbarium specimens to their

species, as the rhizomes are frequently not collected

and the degree of culm compression character over-

laps. Esterhuysen 32724, from Bainskloof, has a

somewhat anomalous female spikelet, which closely

resembles the male.

CAPE. — 3319 (Worcester): Bainskloof, marshy places on

slopes between the road and Observation Point (-CA), Esterhuy-

sen 32724 (BOL; C; E; K; L; M; MO; NBG; S). 3320 (Montagu):

Swellendam Hiking Trail, Protea Valley (-CD), Esterhuysen

35613 (BOL; C; E; K; L; M; MO; NBG; S; UC). 3418 (Simons-

town): Palmiet River Valley, 1-3 miles above the bridge above

the lagoon (-BD), Esterhuysen 34770 ( BOL; C; K; L; M; MO; S);

Palmiet River Mountains near Platteberg (-BD), Stokoe 9527

(BOL); slopes above the left bank of the Palmiet River (-BD),

Esterhuysen 35439 (BOL; C; E; K; L; M; MO; S); plateau above

the Palmiet River near the bridge (-BD), Esterhuysen 32542

(BOL; K; STE); near Somersfontein (-BD), Boucher 718 (K);

Esterhuysen 32745 (BOL; K). 3419 (Caledon): Houw Hoek

Mountain (-AA), Esterhuysen 33687 (BOL; C; K; L; M; MO; S);

Esterhuysen 32782 (BOL; K): Palmiet River Mountains (-AC),

Esterhuysen 34836 (BOL; C; K; L; M; MO; S).

Elegia capensis (Burm.f) Schelpe in J1 S. Afr.

Bot. 33 : 156 (1967). Type: Cape, without precise

locality, in herb. Burman s.n. (G, holo.).

Equisetum capense Burm. f. , Prodr. FI. Cap. 31 (1768).

Restio verticillaris L.f. , Suppl. 425 (1781); Thunb., FI. Cap.,

edn 1, 336 (1811); edn Schultes, 88 (1823). Elegia verticillaris

(L.f.) Kunth, Enum. PI. 3 ; 471 (1841); Mast, in A. DC., Monogr.

Phan. 1 : 351 (1878); in FI. Cap. 7 : 105 (1897); Pillans in Trans.

R. Soc. S. Afr. 16 : 315 (1928); in Adamson & Salter, FI. Cape

Penins. 145 (1950). Type: Cape, 3319 (Worcester): ‘In Roode

Sand’ (-AA), Thunberg s.n. (LINN, holo.!; Cl; SI; SBT!).

Notes

1 . I have not seen the type of Equisetum capense.

2. The type specimen of Restio verticillaris in

LINN (LINN 1164.2) is named in Linnaeus the

Younger’s hand. There are numerous duplicates of

the Thunberg collection in Scandinavian herbaria.

3. The locality given by Thunberg (1788) is ‘Infra

Winterhoek in Roode sand, in Attaquas Kloof et

alibi.’ I have chosen the first as the ‘type locality,’

but this note does suggest that Thunberg collected

this species several times and pooled the material. If

this explains the numerous duplicates, can these du-

plicates be regarded as isotypes?

4. E. capensis is ubiquitous in the Cape Lloral Re-

gion, ranging from Uitenhage to Clanwilliam. It of-

ten forms large monospecific stands in wet seepage

zones and along streams.

Elegia coleura Nees ex Mast, in A. DC.,

Monogr. Phan. 1 : 358 (1878); in FI. Cap. 7 : 106

(1897); Pillans in Trans. R. Soc. S. Afr. 16 : 330

(1928); in Adamson & Salter, FI. Cape Penins. 147

(1950). Syntypes: Cape, 3318 (Cape Town): Cape

Flats near Rondebosch (-CD), Burchell 168 cf (K,

lecto.!; BOL!); Doornhoogte, August (-DC), Eck-

lon s.n. (B!). 3319 (Worcester): in the Tulbagh basin

and at Vogelvlei, November (-AC), Ecklon s.n. (B!;

P!; MEL!; S!).

Elegia exilis Mast, in Bot. Jb. 29 Beibl. 66 : 7 (1900). Syntypes:

Cape, 3419 (Caledon): Riviersonderend, 220 m (-BB). Schlechter

422

Bothalia 15, 3 & 4 (1985)

9893 9 (B, lecto.!; BR!; K!; P!; S!; Z!); 9894 cf (B!; BR!; K!; P!;

S!; Z!).

Notes

1. The specimen of Burchell 168 at K is named by

Masters as ‘E. coleura Nees var.’. However, this

specimen is cited without further comment in the

protologue.

2. The type of E. obtusiflora Mast, is also an Eck-

lon collection from locality 77, but it was collected in

September.

3. Masters (1900) published the type numbers of

E. exilis as Schlechter 9883 and 9884. However, it is

clear from the specimens that it was a printing error

and that they ought to be 9893 and 9894.

4. There is rather little material of this species

available. It ranges from Humansdorp to Paarl and

from sea level to about 1 000 m. It occurs in season-

ally marshy conditions in sandy soils.

Elegia cuspidata Mast, in J. Linn. Soc., Bot. 10 :

240 (1868); in A. DC., Monogr. Phan. 1 : 354 (1878);

in FI. Cap. 7 : 109 (1897); Pillans in Trans. R. Soc. S.

Afr. 16 : 341 (1928); in Adamson & Salter, FI. Cape

Penins. 149 (1950). Syntypes: Cape, 3418 (Simons-

town): Simons Bay (-AB), MacGillivray 437 9 (K,

lecto.!; BM!); Milne cf & 9 (K!); Robertson cf

(BM!); Wright 484 cf & ? (P!; NY!).

Icon: Rice & Compton, Wild Flow. Cape G.H. t. 250 f. 3

(1950).

Note

1. This strange species occurs on the coastal hills

and mountains between Kleinmond and Gordons

Bay, on the southern Cape Peninsula, and in the

‘sandveld’ near Mamre Road. The majority of habi-

tats are recorded as marshy, or along streams, but

some collections are from dry sandy areas. All col-

lections appear to be from below 300 m.

Elegia equisetacea (Mast.) Mast, in J. Linn.

Soc., Bot. 21 : 583 (1885); in FI. Cap. 7 : 106 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 317 (1928). Syn-

types: Cape, 3320 (Montagu): foot of the Lange-

berge near Swellendam (-CD), Burchell 7429 $ (K,

lecto.!; P!). 3419 (Caledon): Genadendal (-BA),

Dr'ege 1644 cf (B!; P!); Distr. Swellendam, ‘Puspas-

vallei,’ Ecklon & Zeyher s.n. C f & $ (B!).

Elegia propinqua (Nees) Kunth var. equisetacea Mast, in J.

Linn. Soc., Bot. 10 : 242 (1868).

Notes

1. The Ecklon & Zeyher collection in B is anno-

tated by Nees as ‘Restio parviflorus var. equisetacae

N. ab E.\ The Drege collection in P approaches E.

juncea, which is typical of the more western popula-

tions of E. equisetacea. This may make it very diffi-

cult to separate this species from the E. juncea com-

plex.

2. This species is very close to E. juncea and ma-

terial without sterile branches may be very difficult

to separate. This is often the case at the southern or

western ends of the distribution range of E. equiseta-

cea. There are several collections from the Kogel-

berg which are intermediate between the two

species.

3. The distribution range of E. equisetacea is from

Humansdorp along the Tsitsikamma-Outeniqua-

Langeberg Mountains to Swellendam, the Potberg

and possibly the Kogelberg. Populations generally

occur in damp to wet conditions.

Elegia esterhuyseniae Pillans in Trans. R. Soc.

S. Afr. 30 : 259 (1945). Type: Cape, 3219 (Wupper-

tal): Cedarberg, Langberg, 1 650 m (-AC), Ester-

huysen 7353 $ (BOL, lecto.!; K!); 7353 cf (BOL!;

K!).

Elegia esterhuyseniae Pillans var. dispar Pillans in Trans. R.

Soc. S. Afr. 30 : 260 (1945). Type: Cape, 3319 (Worcester): Hex

River Mountains, Slab Peak, 1 800 m (-AD), Esterhuysen 8451 $

(BOL, lecto,!; K!); 8451 cf (BOL!; K!).

Notes

1. The var. dispar of Pillans can probably not be

upheld, as it most likely falls within the variation

range of the species.

2. E. esterhuyseniae is a high-altitude species,

known from the mountains from the Cedarberg to

the Riviersonderend Mountains from altitudes usu-

ally between 1 500 and 2 100 m, with a few collec-

tions from as low as 1 000 m. The collections are

from rocky or somewhat peaty places, sometimes

from ledges, generally somewhat sheltered locali-

ties.

Elegia extensa Pillans in Trans. R. Soc. S. Afr.

29 : 343 (1942). Type: Cape, 3319 (Worcester): Wol-

seley (-AC), Esterhuysen 8654 9 (BOL, lecto.!; K!);

8654 Cf (BOL!; K!).

Note

1. E. extensa occurs in shaly soils, usually asso-

ciated with renosterveld, between Tulbagh and Bre-

dasdorp. It appears to prefer localities which be-

come wet or sodden in winter. It is quite rare.

Elegia fastigiata Mast, in Bot. Jb. 29 Beibl. 66 :

7 (1900); Pillans in Trans. R. Soc. S. Afr. 16 : 328

(1928). Syntypes: Cape, 3319 (Worcester): moun-

tains above Mitchell’s Pass 450 m (-AD), Schlechter

9955 $ (K, lecto.!; BM!; BOL!; BR!; K!; MO!; P!;

S!; Z!); 9954 cf (B!; BM!; BOL!; BR!; K!; MO!; P!;

S!; Z!).

Note

1. This species is only known from the type collec-

tion. It may be an aberrant form of E. asperiflora or

of E. rigida. It differs from these taxa largely in hav-

ing well developed spathellae, a variation common

in some variable taxa such as E. juncea.

Elegia fenestrata Pillans in Trans. R. Soc. S.

Afr. 16 : 338 (1928); in Adamson & Salter, FI. Cape

Penins. 148 (1950). Type: Cape, 3418 (Simonstown):

hills west of Buffels Bay (-AD), Pillans 4144 $

(BOL, lecto.!; K!); 4144 cf (BOL!; K!).

Note

1. E. fenestrata is known from the southern Cape

Peninsula and from the Bredasdorp coastal flats.

The populations occur below 100 m, often in marine

sand, in seepages or along streams.

Elegia filacea Mast, in J. Linn. Soc., Bot. 21 :

589 (1885); in FI. Cap. 7 : 112 (1897). Syntypes:

Bothalia 15, 3 & 4 (1985)

423

Cape, 3418 (Simonstown); Nieuw Kloof, Houw

Hoek Mountains (-BB), Burchell 8121 9 (K, lecto.!;

BOL!); Lowry’s Pass, 300 - 600 m (-BB), Drege 126

(B!; K!); without precise locality, Drege 110 cf (B!;

BM!; K!; OXF!).

Elegia parviflora Pillans var. filacea (Mast.) Pillans in Trans.

R. Soc. S. Afr. 16 : 326 (1928).

Elegia gracilis N.E.Br. in FI. Cap. 7 : 754 (1900). Type: Cape,

3419 (Caledon): mountains at Houw Hoek, 450 m (-AA), Galpin

4800 cf (K, holo.!; B!; BOL!).

Elegia rehmanni Mast, in Bot. Jb. 29 Beibl. 66 : 7 (1900). Type:

Cape Town, Cape Flats, Rehmann 1806 (B, holo.!; BR!; Z!).

Elegia parviflora Pillans in Trans. R. Soc. S. Afr. 16 : 325

(1928); in Adamson & Salter, FI. Cape Penins. 147 (1950), nom.

illeg., later homonym for E. parviflora (Thunb.) Kunth, Enum.

PI. 3 : 467 (1841).

Notes

1. Kunth (1841) misapplied Restio parviflorus

Thunb. to a species of Elegia based on Drege 120,

110, 9435, 1646, and 121. The type of Restio parvi-

florus Thunb. is a male specimen of a Cannomois.

Since 1841 ‘Elegia patviflora’ has been used in the

sense of Kunth. Pillans (1928) attempted to legiti-

mize the situation by citing the name as 'E. parvi-

flora, Kunth, Enum. 3 : 467 (1841), excl. syn. omni.’

But this excludes the type from E. parviflora

(Thunb.) Kunth and so effectively describes a new

species, Elegia parviflora Pillans, which is a later

homonym of E. parviflora (Thunb.) Kunth and is il-

legitimate.

The next name available is Elegia rigida Mast.,

which Pillans (1928) recognized as a variety of his E.

parviflora. I remove it here as a distinct species.

That leaves us with E. filacea Mast., the type of

which is also somewhat different from the common

form of the species, but which I think can be safely

included in the species.

3. Elegia filacea is a very variable species and even

displays anatomical variation. The variation patterns

are complex and need further study.

4. E. filacea is ubiquitous in the Cape Floral Re-

gion, ranging from Port Elizabert to Clanwilliam,

from sea level to 1 800 m. It may be very common,

forming monospecific stands on ‘suurvlaktes’ in the

mountains and is generally a component in the vege-

tation. It appears to occur only in sandy soils.

Elegia fistulosa Kunth, Enum. PI. 3 : 467 (1841);

Mast, in A. DC., Monogr. Phan. 1 : 356 (1878); in

FI. Cap. 7 : 110 (1897); Pillans in Trans. R. Soc. S.

Afr. 16 : 333 (1928); in Adamson & Salter, FI. Cape

Penins. 147 (1950). Type: Cape, 3318 (Cape Town):

Paarl Mountains, 300 - 600 m (-DB), Drege 117 9

(B, holo.!; BM!; K!; MO!; P!).

Elegia fistulosa Kunth var. parviflora Pillans in Trans. R. Soc.

S. Afr. 16 : 334 (1928). Type: Cape 3325 (Port Elizabeth): Witte-

klip, near Van Stadens (-CC), MacOwan 2153 $ (K, lecto.!;

MEL!; SAM!); 2153 cf (SAM).

Notes

1. The locality on the Drege sheet in BM is given

as ‘Cape Flats, 1000 - 2000 ft’. Since the Cape Flats

rarely exceed 20 m, this is obviously false and could

refer to the Paarl Mountains, which flank the flats.

2. The var. parviflora of Pillans probably falls

within the variation range of the species.

3. This species is common between Port Elizabeth

and Malmesbury, occurring between sea level and

1 000 m, between the coast and the first range of

mountains. The habitat is marshy, often quite wet.

Elegia fucata Esterhuysen , sp. nov., E. caespi-

tosa Esterhuysen et E. asperiflora (Nees) Kunth affi-

nis, culmis teretibus, vaginis et bracteis purpureis

differt.

TYPE. — 3319 (Worcester): Jona’s Kop, on

slopes below the summit (-DC), Esterhuysen 35631

(BOL, holo.!; B; C; E; F; K; L; LD; M; MO; NBG;

NY; PRE; RSA; S; STE; TCD; UC; US; W;

WAG).

Plants probably caespitose, 15-30 cm tall. Rhi-

zomes about 2 mm in diam., covered by nitid, imbri-

cate scales. Culms solid, terete, or subcompressed as

a result of drying, simple, to 1,5 mm in diameter, the

surface smooth. Sheaths deciduous, 1-2 per culm,

30-35 mm long, loosely convoluted, acute with a

slender aciculate 3 mm long awn, basal | green, up-

per 3 purplish, margins membranous; basal sheaths

shorter, imbricate, persistent, grading into the rhi-

zome scales. Male inflorescence globose to subglo-

bose, 2-3 cm long. Spathe papyraceous, 10-25 x 5-8

mm, acute with an aciculate, up to 2 mm long awn,

purplish in the middle with the margins paler and

often undulate, imbricate, partially obscuring the

flowers. Inflorescence branches several from each

node, shorter than the spathes, much branched. Spa-

thellae and bracts slender acuminate, to 10 x 1 mm,

bracts about twice as long as the flowers, subcarinate

and papillose on the carina, occasionally bracts with-

out acuminate apices, lacerated, about 1 mm long.

Spikelets weakly defined, with 2-6 flowers and no

sterile bracts. Flowers subsessile, perianth 1,5-2 mm

long. Tepals chartaceous, acute, glandular-papillose

on the backs, especially in the upper half; sepals 10

x 0,6 mm; petals 1,5 x 0,6 mm. Anthers about 1 mm

long. Female inflorescence with the flowers obscured

by the spathes, inflorescence branches shorter and

less branched. Bracts and spikelet organization as in

the males. Flowers subsessile, 1,5 mm long. Tepals

acute, chartaceous, subequal .with the sepals margi-

nally shorter than the petals, glandular-pappillose,

about 1,5 x 1 mm. Styles 3, fruit a three-angled, 1,5

mm long, unilocular nutlet.

E. fucata is known from the Wildepaardeberg and

from Jona’s Kop, at an altitude of about 1 500 m. It

occurs in areas which are peaty and probably wet in

winter. Flowering occurs in November.

This curious endemic is probably related to the

Elegia asperiflora group, with which it shares the pa-

pillose perianth, the general inflorescence organiza-

tion, the subpersistent sheaths and the long-acumi-

nate bracts. It is readily distinguished from the

species in this complex by the purple sheaths and

spathes, the small stature and the terete culms.

CAPE. — 3319 (Worcester): Jona’s Kop, slopes below the sum-

mit c. 1 500 m (-DC), Esterhuysen 35631 (B; BOL; C; E; F; K; L;

LD; M; MO; NBG; NY; PRE; RSA; S; STE; TCD; UC; US; W;

WAG); Esterhuysen 32684 (BOL; C; E; K; L; M; MO; S; STE;

UC); Wilde Paarde Berg (-DC) Stokoe 2784, 2785 (BOL; K).

Elegia galpinii N.E. Br. in FI. Cap. 7: 754

(1900); Pillans in Trans. R. Soc. S. Afr. 16: 340

424

Bothalia 15, 3 & 4 (1985)

(1928). Type: Cape, 3321 (Ladismith): mountains at

Garcias Pass, 350 m (-CC), Galpin 4803 9 (K,

lecto.!; B!; BOL!); 4803 cf (B!; BOL!; K!).

Note

1. This very curious species occurs at the northern,

arid base of the Langeberg between Garcias Pass

and Montagu, on the Rooiberg and in the arid

mountains around Calitzdorp.

Elegia grandis (Nees) Kunth , Enum. PI. 3: 475

(1841); Pillans in Trans R. Soc. S. Afr. 16: 322

(1928). Syntypes: Cape, 3419 (Caledon): Genaden-

dal (-BA), Ecklon in herb. Zeyher (B, lecto.!);

Zeyher s.n. (K!; S!); forma B: ‘in planitie tabulari,

Ecklon s.n. (B!; P!; S!).

Restio grandis Spreng. ex Nees in Linnaea 5: 660 (1830). Lam-

procaulis grandis (Nees) Mast, in A. DC., Monogr. Phan. 1: 349

(1878); in FI. Cap. 7: 115 (1897).

Icon: A. DC., Monogr. Phan. 1: t. 3 f. 22-28; t. 5 f. 13 (1878).

Notes

1. Nees did not clearly cite the material on which

he based his descriptions, making the typification of

his names very difficult. However, the identity of

this species is beyond doubt. Forma B is clearly Ele-

gia neesii (see below) and has been excluded from

this species by Masters (1868).

2. This species is very distinct in its blue-green

colouration. Even in dried specimens the sheaths

still appear green.

3. E. grandis occurs at about 1 200 m in the moun-

tains from Bainskloof to Kanonkop at Genadendal.

Populations generally occur on better-drained

slopes, on shale bands, amongst rocks or on sandy

flats.

Elegia grandispicata Linder , sp. nov., ab E. in-

termedia (Steud . ) Pillans spathis cartilagineis differt .

TYPE. — Cape, 3319 (Wuppertal); Koue

Bokkeveld, Skoongesig (-CC), Hanekom 1227 $

(K, holo.!; PRE).

Plants caespitose, tussocks to about 70 cm tall.

Rhizomes short, spreading, about 4 mm in diameter,

scales dark brown, obtuse, imbricate. Culms solid,

simple, terete, 60-70 cm long, 2,5-4 mm in diam.

Sheaths deciduous, 3-6 cm long, acute with an up to

2 mm long awn, cartilaginous, loosely convoluted to

flat. Male inflorescence 5-15 x 2-3,5 cm, compound.

Spathes similar to the sheaths, 3-5 cm long, persist-

ent. Inflorescence branches several per node,

branching freely, forming spikes longer than the

bracts. Spathellae cartilaginous, narrowly lanceolate,

acute, up to 15 mm long, not prominent. Spikelets

with 5-8 flowers, globose, to 5 mm in diam. Bracts

shorter than the flowers, acute, shortly awned, co-

riaceous with a narrow membranous margin, or

much longer than the flowers, long-awned. Perianth

3 mm long. Sepals about 2 mm long, acute, the lat-

erals shortly keeled, obscurely rough along the keel.

Petals 3 mm long, acute to subacuminate, subspathu-

late. Anthers dehiscing inside the petals, mucronate,

2 mm long. Female inflorescence 7-9 x 2-3 cm,

spathes as in male, largely obscuring the female

spikes. Spikes similar to male, but less branched and

smaller. Spathellae narrowly lanceolate, up to 15 mm

long, very acute, fairly prominent. Spikelets with 1-4

flowers. Bracts taller than the flowers, coriaceous,

acute, the lower bracts with awns as long as or longer

than the body of the bract. Flowers 3 mm long; te-

pals acute, coriaceous, the margins narrowly mem-

branous, lacerate; sepals slightly shorter than the pe-

tals, keel somewhat better developed. Ovary unilo-

cular, styles 3, slender, villous, free to the base. Fruit

a shiny trigonous black nutlet, 2,5 mm long.

E. grandispicata occurs in the mountains between

the Hex River Valley and the Cedarberg, between

1 200 and 1 800 m, both on streamsides and on grav-

elly mountain slopes.

This species agrees with E. glauca in Pillans

(1928). But the types of E. glauca all belong to Ele-

gia asperi flora, which is quite distinct.

E. grandispicata is rather close to E. intermedia.

The latter species is probably best construed as being

endemic to Table Mountain on the Cape Peninsula,

but this leaves several collections from the inland

mountains between Worcester, Stellenbosch and

Riversdale, unaccounted for.

Elegia hutchinsonii Pillans in Trans. R. Soc. S.

Afr. 29: 344 (1942). Type: Cape, 3319 (Worcester):

gorge west of Ceres (-AD) Hutchinson 617 $ (BOL,

lecto.!; BM!; K!); 617 $ (BM!; BOL!; K!).

Note

1. E. hutchinsonii occurs on rocky slopes or sandy

flats, or along streams, between 450 and 1 200 m in

the mountains from Bainskloof to Porterville and to

Ceres.

Elegia intermedia (Steud.) Pillans in Trans. S.

Soc. R. Afr. 16: 332 (1928); in Adamson & Salter,

FI. Cape Penins. 147 (1950). Type: Cape, 3318

(Cape Town): ‘In humides summitatis montis Tabu-

laris, Nov. (-CD), Ecklon 836 cf (B, lecto.!; MEL!;

MO!; P!).

Restio intermedius Steud. in Flora, Jena 12: 132 (1829).

Restio membranaceus Nees in Linnaea 5: 657 (1830). Elegia

membranacea (Nees) Kunth, Enum. PI. 3: 474 (1841); Mast, in FI.

Cap. 7: 108 (1897). Type: Cape, 3318 (Cape Town): ‘In humides

verticis montis Tabularis, inter rupis, Nov.’ (-CD), Zeyher s.n. cf

(B, holo.!; MEL!; P!), nom. illeg.

Notes

1. Both basionyms are based on the same collec-

tion from the summit of Table Mountain. Conse-

quently, R. membranaceus is illegitimate, as it is a

superfluous name.

2. This species is close to E. grandispicata , of

which it may be a geographic vicariant. It can be

readily distinguished by the more delicate flowers

and plants and the paler, chartaceous, spathes.

3. E. intermedia is endemic to Table Mountain,

where it occurs in seepages and swampy areas.

Elegia juncea L., Mant. Alt. 297 (1771);

Thunb., FI. Cap. edn 1, 311 (1811); edn Schultes, 81

(1823); Kunth, Enum. PI. 3 : 464 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 357 (1878); in FI. Cap. 7 :

109 (1897); Pillans in Trans. R. Soc. S. Afr. 16 : 334

(1928); in Adamson & Salter, FI. Cape Penins. 148

(1950). Type: Cape, without precise locality, Konig

s.n. (?); LINN 1164a. 4 (LINN, neot.!).

Bothalia 15, 3 & 4 (1985)

425

Restio elegia Murray, Syst. Veg. 738 (1784), nom. illeg. , super-

fluous name for E. juncea L. Restio junceus (L.) Nees in Linnaea

5 : 658 (1830).

Restio propinquus Nees in Linnaea 5 : 653 (1830). Elegia pro-

pinqua (Nees) Kunth, Enum. PI. 3 : 473 (1841); Mast, in A. DC.,

Monogr. Phan. 1 ; 356 (1878); in FI. Cap. 7 : 108 (1897). Syn-

types; 3318 (Cape Town): ‘In fissura rupis ad planitem montis

Tabularis’ (-CD), Ecklon s.n. (?); ‘Montis Leonis’ (-CD), Ecklon

s.n. (?). 3418 (Simonstown): ‘In latere orientali ejusdem montis

prope Wittebome’ (-AB), Ecklon s.n. (?).

Elegia propinqua (Nees) Kunth var. minor Mast, in FI. Cap. 7 :

109 (1897). Type: Cape, 3318 (Cape Town): flats near Ronde-

bosch (-CD), Burchell 822 cf (?).

Elegia juncea L. var. geniculata Pillans in Trans. R. Soc. S. Afr.

16 : 336 (1928). Type: Cape, 3318 (Cape Town): summit of Koe-

berg (-DA), Pappe s.n. $ (BOL, lecto. !); s.n. cf (BOL!; K!).

Notes

1. Linnaeus (1771) stated that his Elegia juncea is

based on a Konig collection. There is no evidence to

suggest that any material at LINN or at S is of Konig

origin. There is a sheet at LINN annotated ‘Elegia

juncea’, but this is in Dahl’s writing. Dahl only

started working with Linnaeus after the publication

of the Mantissa Altera. As Linnaeus did probably see

the specimen in LINN, but not at the time he wrote

the Mantissa Altera, I have designated it as a neo-

type.

2. I have not succeeded in tracing any of the type

specimens of Restio propinquus and place this name

here on the basis of its description.

3. This species is part of a complex which includes

E. equisetacea, E. vaginulata and E. spathacea. Al-

though a good 90 per cent of the material can be sep-

arated with ease by the characters indicated in the

key, there are some collections which are difficult,

such as material of E. equisetacea from the Kogel-

berg and some large specimens of E. vaginulata. Ele-

gia altigena, from the Kamannassie Mountains, is

probably not distinct from E. juncea.

4. E. juncea ranges from Uitenhage to Ceres, from

sea level to 1 500 m. It is probably restricted to

sandy soils and usually occurs in somewhat moist

habitats.

Elegia muirii Pillans in Trans. R. Soc. S. Afr. 16

: 319 (1928). Type: Cape, 3421 (Riversdale): Alber-

tinia (-BA), Muir 1354 $ (BOL, lecto.!); 1354 cf

(BOL!).

Note

1. E. muirii occurs on the coastal forelands be-

tween Albertinia and Bredasdorp. The plants grow

either in sand or on limestone outcrops, in dry, well-

drained habitats.

Elegia neesii Mast, in J. Linn. Soc., Bot. 10 : 246

(1868); Pillans in Trans. R. Soc. S. Afr. 16 : 321

(1928); in Adamson & Salter, FI. Cape Penins. 146

(1950). Syntypes: Cape, 3318 (Cape Town): Table

Mountain (-CD), Burchell 569 J (K, lecto.!; P!).

3418 (Simonstown): eastern slopes of Table Moun-

tain at Constantia (-AB ) , Ecklon s.n., cf (B ! ; MEL! ;

P!; S!); ‘In planitie Capensi, 184T (-AB), Ecklon

560b $ (BOL!; S!).

Lamprocaulis neesii (Mast.) Mast, in A. DC., Monogr. Phan.

1 : 350 (1878); in FI. Cap. 7 : 115 (1897).

Lamprocaulis schlechteri Gilg-Ben. in Pflanzenfam. edn 2, 15a :

23 (1930). Type: Cape, 3424 (Humansdorp): Humansdorp, 70 m

(-BB), Schlechter 6024 cf (Z, lecto.!; BOL!; K!).

Notes

1. Masters (1868) cites in his synonymy ‘ Restio

grandis, var. B, N. ab E. Linnaea 5 p. 66T. As Nees

did not apply Latin binomials to his infra-specific

taxa, this citation is meaningless in terms of priority.

The Ecklon collection from the eastern slopes of

Table Mountain in B is annotated ‘ Restio (Elegia)

grandis, Constantia, Zeyher’ by Nees. This is prob-

ably the type of Restio grandis forma B of Nees. The

Ecklon collection in S is annotated as ‘ Restio ( Ele-

gia) grandis v. gracilis N. ab E.’.

2. There is a form of E. neesii in which the ovary is

much longer than the perianth. However, in all

other characters it agrees with E. neesii. It is only

known from the high mountains at the head of the

Jonkershoek Valley, where it occurs in marshy

areas. It could be a high-altitude variant of E. neesii.

The rank at which this variant should be recognized

has not been resolved. Pillans (ms) suggested speci-

fic status and suggested the name ‘E. ligulata’.

3. E. neesii s.s. ranges from the North Cedarberg

to the Cape Peninsula, with a few collections reach-

ing to Humansdorp. It generally occurs in moist to

marshy areas. From Malmesbury to Humansdorp

most of the collections are from near sea level, or

below 1 000 m, but in the Cedarberg it grows be-

tween 1 200 and 1 500 m.

Elegia persistens Mast, in Bot. Jb. 29 Beibl. 66 :

8 (1900). Syntypes: Cape, 3419 (Caledon): Houw

Hoek, 350 m (-AA), Schlechter 7772 $ (B, lecto.!;

BM!; BOL!; BR!; K!; MO!; P!; S!; Z!); 7771 cf

(BM!; BOL!; BR!; K!; MO!; P!; S!; Z!).

Notes

1. Pillans (1928) included this species in the syno-

nymy of E. racemosa. Although it is superficially

similar, the shape of the female flowers is quite dif-

ferent, being almost globose rather than oblong, the

perianth segments are rounded and the ovary has

two styles.

2. E. persistens occurs in the Riviersonderend

Mountains, the mountains between Villiersdorp and

Cape Agulhas and in the mountains around Cale-

don. The majority of the collections are from below

400 m, but a few reach up to 900 m. It usually occurs

on rocky slopes, with a few from sandy slopes.

Elegia prominens Pillans in Trans. R. Soc. S.

Afr. 16 : 320 (1928); in Adamson & Salter, FI. Cape

Penins. 145 (1950). Type: Cape, 3418 (Simonstown):

Near Zeekoevlei (-BA), Pillans 4139 9 (BOL,

holo.!; K!).

Note

1. E. prominens occurs on the coastal flats be-

tween Bredasdorp and Malmesbury, generally in

sand, at low altitudes. Many of the collections are

from somewhat marshy, peaty areas.

Elegia racemosa (Poir.) Pers., Syn. PI. 2 : 607

(1807); Kunth, Enum. PI. 3 : 463 (1841); Pillans in

Trans. R. Soc. S. Afr. 16 : 337 (1928); in Adamson

& Salter, FI. Cape Penins. 148 (1950). Type: Cape,

426

Bothalia 15, 3 & 4 (1985)

without precise locality, in herb. Lam. cf (P, lecto.!)

(for 9 see below).

Restio racemosus Poir. in Lam., Encycl. 6 : 177 (1804); in Lam.,

Tabl. Encycl. 3 : 400 (1823). Dovea racemosa (Poir.) Mast, in J.

Linn. Soc., Bot. 21 : 578 (1885); in FI. Cap. 7 : 103 (1897).

Elegia fusca N.E.Br. in FI. Cap. 7 : 754 (1900). Type: Cape,

3322 (Oudtshoorn): Cradock Mountain, near George, 750 m

(-CD), Galpin 4802 9 (K, holo.!; BOL!).

Elegia arnoena Pillans in Trans. R. Soc. S. Afr. 30 : 258 (1945).

Type: Cape, 3319 (Worcester): Witteberg, 1 800 m (-CA), Ester-

huysen 8660 9 (BOL, lecto.!; K!); 8660^C? (BOL!; K!).

Elegia bella Pillans in Trans. R. Soc. S. Afr. 30 : 258 (1945).

Type: Cape, 3323 (Willowmore): Helpmekaar Peak (-DC), Es-

terhuysen 4583 9 (BOL, lecto.!; K!); 4583 cf (BOL!; K!).

Icon: Lam., Tabl. Encycl. t.804 f.4f (1799).

Notes

1. There are two elements in Restio racemosus

Poir. The female part is clearly Chondropetalum mu-

cronatum, whereas the male is an Elegia. Since both

match the description and Pillans (1928) explicitly

excluded the female from his concept of Elegia race-

mosa (‘ R . racemosus Poir. ... quoad plant, masc.’),

this is followed here.

2. I am treating the species here in a wide sense,

including the variants which Pillans (1945) segre-

gated. I suspect that the number of styles, on which

Pillans based this grouping, may be a variable

character in Elegia.

3. E. racemosa occurs in mountains, generally be-

tween 600 and 1 800 m, from Uniondale to the Cape

Peninsula and to Worcester. Populations generally

occur on steep rocky slopes.

Elegia rigida Mast, in J. Linn. Soc., Bot. 21: 587

(1885); in FI. Cap. 7 : 113 (1897). Type: Cape, 3319

(Worcester): Drakenstein Mountains near Bains-

kloof(-CA), Bolus 4100 $ (K, holo. ! ; BM! ; BOL!).

Elegia parviflora Pillans var. rigida (Mast.) Pillans in Trans. R.

Soc. S. Afr. 16 : 327 (1928).

Elegia obtusiflora Mast, in FI. Cap. 7 : 112 (1897). Type: Cape,

3319 (Worcester): nearTulbagh, September (-A), Zeyher s.n. 9

(K, lecto.!; B!; MEL!); s.n. cf (B!, K!; MEL!).

Elegia spathacea Mast. var. attenuata Pillans in Trans. R. Soc.

S. Afr. 16 : 328 (1928). Syntypes: Cape, 3319 (Worcester): moun-

tains at Tulbagh Waterfall, 600 m (-AC), Schlechter 7458 9

(BOL, lecto.!; B!; BR!; K!; MO!; S!; Z!); 7457 cf (B!; BOL!;

BR!; K!; S!; Z!).

Notes

1. This segregate can be distinguished from E. fila-

cea by the closely imbricate spathes with recurved

apices and from E. spathacea by smaller spathes. It is

very close to the latter species and further studies

may show that it is best regarded as a subspecies of

it, or as the end of a cline in spathe size.

2. E. rigida is known from the mountains between

Tulbagh and Bainskloof. It grows between 300 and

900 m, in seasonally marshy places.

Elegia spathacea Mast, in J. Linn. Soc., Bot.

21 : 588 (1885); in FI Cap. 7 : 113 (1897); Pillans in

Trans. S. Soc. S. Afr. 16 : 327 (1928). Syntypes:

Cape, 3419 (Caledon): Donkerhoek Mountains

(-AB), Burchell 7959 9 (K, lecto.!; BOL!); summit

of Baviaanskloof mountains (-BA), Burchell 7710 cf

(BOL!; K!; P!).

Notes

1. Pillans described a var. attenuata, but this is

here included in Elegia rigida. These two taxa are

very similar, and may well be subspecies or be con-

specific.

2. E. spathacea occurs in the mountains between

Jonkershoek, Genadendal, Hermanus and Kogel-

berg. Generally, collections are from between 300

and 1 000 m. The habitat usually reported is coarse

sand, somewhat marshy.

Elegia squamosa Mast, in J. Linn. Soc., Bot. 10;

244 (1868); in A. DC., Monogr. Phan. 1: 359 (1878);

in FI. Cap. 7: 111 (1897); Pillans in Trans. R. Soc. S.

Afr. 16: 322 (1928); in Adamson & Salter, FI. Cape

Penins. 146 (1950). Syntypes: Cape, without precise

locality, Pappe 105 cf (K, lecto.!). 3321 (Ladismith):

near Garcias Pass (-CC), Burchell 7141 cf (BOL!;

K!). 3318 (Cape Town): Doornhoogte, August

(-DC), Ecklon & Zeyher s.n. cf (?)• 3219 (Wupper-

tal): near Ezelbank, 900 m (-AC), Drege 2519 9

(K!). 3319 (Worcester): bank of the Berg River in

the Klein Drakenstein Mountains below 150 m

(-CC), Drege 1648 cf (K!; P!); Cape, without precise

locality, Drege 111 9 (K!; P!). 3325 (Port Eliza-

beth): flats at Van Stadens River Mountains (-CC),

Ecklon 795 cf (MEL!).

Elegia pectinata Pillans in Ann. Bolus Flerb. 3: 145 (1922). Syn-

types: Cape, 3318 (Cape Town): Lions Head (-CD), Pappe in

SAM 12850 Cf & 9 (SAM); in SAM 12975 (SAM); Devils Peak

(-CD), Pillans 3401 9 (BOL, lecto.!); 3401 cf (BOL!); above

Oude Kraal (-CD), Pillans 3403 (BOL!). 3418 (Simonstown): Sir

Lowry’s Pass (-BB), Guthrie in BOL 16622 (BOL!). 3419 (Cale-

don): Zwartberg near Caledon (-AB/BA), Guthrie in BOL 16016

(BOL!)

Notes

1. Masters (1868) based his E. squamosa on a very

mixed set of specimens. I have not been able to trace

material of Ecklon & Zeyher (from Doornhoogte).

Drege 2519 and 111 are E. filacea. This differs from

E. squamosa in the absence of the rhizome. Drege

1648 is E. coleura which is distinguished by the tal-

ler, compressed culms. Burchell 7141 is E. vaginu-

lata, which differs by its taller culms, which are

placed further apart on the rhizome. Pappe 105 is E.

squamosa, as construed by Pillans (1928). Of all the

syntypes, this is the only one which Pillans (1928)

included in his cited specimens. The sheet at K is

also annotated by Masters as E. squamosa — the

only syntype to be so annotated. Pappe 105 agrees

with the phrase ‘rhizoma repens’ in the protologue,

but not with ‘culmi . . . 8-12 pollicares’. Burchell

7141 fits the protologue better than Pappe 105, but

as Pillans had already set a precedent and as Masters

only annotated the Pappe sheet, and as I do not

think that Pappe 105 can be regarded as ‘being in

serious conflict with the protologue’ (ICBN, rule 8),

I uphold Pillans’s implicit lectotypification.

2. E. squamosa occurs between Malmesbury and

Bredasdorp, below 200 m. The usual habitat is in

hard clayey or rarely gravelly soils at the base of

mountains.

Elegia stipularis Mast, in J. Linn. Soc., Bot. 21:

587 (1885); in FI. Cap 7: 112(1897); Pillans in Trans.

R. Soc. S. Afr. 16: 318 (1928); in Adamson & Salter,

Bothalia 15, 3 & 4 (1985)

427

FI. Cape Penins. 145 (1950). Syntypes: Cape, 3422

(Mosselbay): hills near the landing place at Mossel-

bay (-AA), Burchell 6290 9 (K, lecto.!; BOL!);

without precise locality, R. Brown s.n. C f (BM!);

Masson cf (BM!; BOL!); Verreaux (?); ‘Olifants

River,’ Gill s.n. cf (K!).

Notes

1. This species can be distinguished by its much-

branched habit and golden bracts.

2. E. stipularis ranges from Mosselbay to Malmes-

bury. Generally collections are from near the coast,

all localities are below 200 m. The populations occur

in well-drained sandy soils, even in the foothills of

the mountains.

Elegia stokoei Pillans in Trans. R. Soc. S. Afr.

29: 345 (1942). Syntypes: Cape, 3319 (Worcester):

Tulbagh waterfall (-AC), Stokoe 1433 9 (BOL,

lecto.!; K!); 1431 cf (BOL!; K!).

Notes

1. The culms are very sparsely branched and on

some specimens they are simple.

2. E. stokoei is known from rather few collections

from the mountains between Villiersdorp and Tul-

bagh and around Worcester. It occurs on dry, rocky

slopes, up to 1 200 m.

Elegia thyrsifera (Rottb.) Pers., Syn. PI. 2: 607

(1807); Kunth, Enum. PI. 3: 465 (1841); Mast, in FI.

Cap. 7: 107 (1897); Pillans in Trans. R. Soc. S. Afr.

16: 336 (1928); in Adamson & Salter, FI. Cape

Penins. 148 (1950). Type: Cape, without precise

locality, Konig s.n. 9 (C, holo.!).

Restio thyrsifer Rottb., Descriptiones Plantarum Rariorum 11

(1772).

Elegia acuminata Mast, in J. Linn. Soc., Bot. 21: 580 (1885); in

FI. Cap. 7: 108 (1897). Syntypes: Cape, 3321 (Ladismith): lower

part of Kampsche Berg (-CD), Burchell 6993 9 (K, lecto.!;

BOL!). 3320 (Montagu): near Swellendam at the foot of the

Langeberg (-CD), Burchell 7435 cf (K!). 3318 (Cape Town):

Table Mountain above Kirstenbosch, 750 m (-CD), Bolus 4640 cf

& $ (BM!; BOL!; K!); without precise locality, Drege 104 cf ( B ! ;

K!); Thom 2824 cf (K!); Verreaux s.n. (?).

leones: Rottb., Descriptionum et iconum rariores t.3 f.4

(1773). Lam., Tabl. Encycl. t.804 f.3 (1799).

Notes

1. This species ranges from Riversdale to the Cape

Peninsula. It grows between sea level and 900 m in

swamps and along river banks, in wet habitats. It is

only known from sandstone derived soils.

2. E. thyrsifera is sometimes erroneously called

Elegia thyrsiflora.

Elegia thyrsoidea (Mast.) Pillans in Ann. Bolus

Herb. 3: 146 (1922); in Trans. R. Soc. S. Afr. 16: 339

(1928). Type: Cape, 3322 (Oudtshoorn): summit of

Postberg near George (-CD), Burchell 5895 9 (K,

lecto.!; BM!); 5895 cf (BM!; K!).

Dovea thyrsoidea Mast, in J. Linn. Soc., Bot. 10: 251 (1868); in

A. DC., Monogr. Phan. 1: 309 (1878); in FI. Cap. 7: 104 (1897).

Notes

1. The type collection is a freak, having a three-

locular capsule and larger than average bracts. This

is probably the reason why Masters placed it in Do-

vea.

2. Superficially, this species is similar to Chondro-

petalum aggregatum, but it can be distinguished by

its persistent female spathes and the pale brown

sheaths and spathes.

3. As in E. racemosa , the size of the spathes and

the plants is very variable.

4. E. thyrsoidea is known from the Tsitsikamma

and Outeniqua Mountains, from Joubertina to

George. It grows between 300 and 1 200 m on moist

mountain slopes.

Elegia vaginulata Mast, in J. Linn. Soc., Bot.

21: 586 (1885); in FI. Cap. 7: 111 (1897); Pillans in

Trans. R. Soc. S. Afr. 16: 324 (1928); in Adamson &

Salter, FI. Cape Penins. 146 (1950). Syntypes: Cape,

3321 (Ladismith): near Garcias Pass (-CC), Burchell

7141 cf (K, lecto.!; BOL!). 3319 (Worcester):

Drakenstein Mountains near Bainskloof (-CA), Bo-

lus 4100 ? (BM!; BOL!; K!).

Notes

1. Bolus 4100 is also the type of Elegia rigida. The

basic differences between E. vaginulata and E.

rigida are (1) that the former has spreading rhizomes

whereas the latter is caespitose, and (2) the inflores-

cences of the former are about 1 cm long, whereas in

the latter they are much longer. Masters does not

mention rhizomes in the protologue, but he does

state ‘inflorescentia spiciformi brevi (1 cm).’ This

agrees better with Burchell 7141 than with Bolus

4100, so I follow the precedent set by Pillans (1928),

and lectotypify Burchell 7141.

2. Masters (1868) cited Burchell 7141 as a syntype

of E. squamosa, but this species is lectotypified by

Pappe 105 (see above).

3. E. vaginulata is ubiquitous in the Cape Floral

Region, ranging from Port Elizabeth to the Cedar-

berg and from sea level to 1 500 m. Populations oc-

cur on sandy soils, often in marshy or peaty places.

4. There is a Gueinzius collection of E. vaginulata

at K, reportedly collected from Port Natal. This

locality is probably wrong.

Elegia verreauxii Mast, in J. Linn. Soc., Bot. 21:

589 (1885); in FI. Cap. 7: 114 (1897); Pillans in

Trans. R. Soc. S. Afr. 16: 323 (1928); in Adamson &

Salter, FI. Cape Penins. 146 (1950). Type: Cape,

without precise locality, Verreaux s.n. 9 (BM,

holo.!).

Note

1. E. verreauxii occurs on the coastal forelands,

below 150 m, between Bredasdorp and Malmesbury.

The species is fairly common in shallow seasonal

pans, usually in sand, but sometimes in sand over

clay.

CHONDROPETALUM

4. Chondropetalum Rottb., Descriptiones Plant-

arum Rariorum 11 (1772); Linder in Bothalia 15: 64

(1984).

428

Bothalia 15, 3 & 4 (1985)

KEY TO THE SPECIES OF CHONDROPETALUM

la Culms branching; styles 2 C. microcarpum

lb Culms simple; styles 3:

2a Sheaths persistent:

3a Sheath solitary; petals longer than the sepals C. deustum

3b Sheaths several; petals as long as the sepals C. acockii

2b Sheaths deciduous:

4a Flowers less than 3 mm long:

5a Petals as long as the sepals; plants with long slender rhizomes C. rectum

5b Petals longer than the sepals; plants sometimes rhizomatous:

6a Petals scabrid over the entire outer surface; median nerve prominent C. nudum

6b Petals smooth or scabrid in the apical half; median nerve not prominent C. tectorum

4b Flowers more than 3,5 mm long:

7a Spathes 5-10 cm long C. mucronatum

7b Spathes less than 5 cm long:

8a Spikelets with 3—4 flowers C. aggregatum

8b Spikelets with 1-2 flowers:

9a Inflorescence subcapitate; bracts as long as the tepals; culms with 1-2 nodes C. decipiens

9b Inflorescence slender, lax; bracts shorter than the tepals; culms with more than 2 nodes

C. ebracteatum

Chondropetalum acockii Pillans in Trans. R.

Soc. S. Afr. 29 : 342 (1942). Type: Cape, 3318 (Cape

Town): flats near Kraaifontein (-DC), Acocks 4157

$ (BOL, lecto.!; K!; S!); 4157 cf (BOL!; K!).

Note

1. This species occurs on the sandy flats from

Brackenfell to Malmesbury, in typical ‘Sandveld fyn-

bos’. This vegetation type is strangely undercol-

lected and is severely threatened by urban and agri-

cultural extension.

Chondropetalum aggregatum (Mast.) Pillans in

Trans. R. Soc. S. Afr. 16 : 307 (1928).

Dovea aggregate! Mast, in FI. Cap. 7 : 101 (1897). Syntypes:

Cape, 3419 (Caledon): Genadendal, 900-1 200 m (-BA). Drege

1622 2 (K, lecto.!; PI). 3318 (Cape Town): mountains near Cape

Town (-CD), Zeyher s.n. cf ( K! ) .

Notes

1. At K there are two specimens of Drege 1622,

one in Herb. Hook., the other in Herb. Benth. The

former is labelled in Master’s hand and is designated

as the lectotype. There is also a capsule with a large

scrap of the Zeyher collection, from herb. Liibeck.

This was received in the beginning of 1897 or the end

of 1896 and was probably seen by Masters when he

described the species.

2. This species cannot readily be distinguished

from C. marlothii and the two taxa are probably con-

specific.

3. There is a superficial resemblance between C.

aggregatum and Elegia thyrsoidea, with which it is

sympatric. It can be separated from E. thyrsoidea by

the deciduous female spathes and the very dark

brown sheaths.

4. C. aggregatum has a wide distribution, from

Worcester and Cape Town to George, but seems to

be rather rare over the whole range, except the Ri-

viersonderend Mountains, from which Miss Ester-

huysen has made many collections.

Chondropetalum decipiens Esterhuysen, sp.

nov., C. ebracteato (Kunth) Pillans affinis, a qua im-

primis differt vaginis 1-2, inflorescentiis masculis

subcapitatis, bracteis longioribus quam floribus.

TYPE. — Cape, 3419 (Caledon): Vogelgat Re-

serve (-AD), Esterhuysen 35535 f (BOL, holo.!; K;

M; MO; S).

Plants caespitose, tussocks 30-70 cm tall. Culms

solid, terete to obscurely sulcate, simple, to 1,8 mm

in diameter, surface smooth. Sheaths deciduous, car-

tilaginous to coriaceous, 15-30 mm long, very

loosely convoluted, nitid, dark reddish-brown,

acute, awn aciculate, slender, 5-6 mm long; sheaths

at the base persistent, imbricate, shorter. Male inflo-

rescence compound, 15-30 x 10-15 mm, forming a

dense head. Spathes soon caducous, to 25 x 5 mm,

similar to the sheaths but the apical | rolled. Spike-

lets very densely clustered, generally 5x3 mm, with

3- 4 flowers. Bracts 4-5 x 2,5 mm, coriaceous, the

apical half conduplicate, shortly and stoutly mucro-

nate-acuminate, as tall as the flowers but not obscur-

ing them, sparsely villous on the midrib, lowermost

bracts often sterile. Flowers subsessile, perianth 3-4

mm long. Tepals subcoriaceous, all about the same

length; sepals subacuminate, 2, 5-3, 5 mm long, lat-

eral sepals conduplicate, sparsely villous on the Car-

ina, petals acute, 3-3,5 x 1 mm, concave. Anthers

dehisce inside the perianth, 1,3-2 mm long, acute,

mucronate. Pistillode 0,5 mm long. Female inflores-

cence 15-40 x 10 mm, more interrupted than the

male. Spathes more or less persistent, to 3 cm long,

similar to the males. Spikelets 1-4 per node, 7-10 x

4- 10 mm, lM-flowered. Bracts similar to the spa-

thellae, about as tall as or slightly taller than the

flowers, basal bract sterile. Flowers subsessile, per-

ianth 5-6 mm long, tepals coriaceous, acuminate,

subequal, red-tipped, lateral sepals sparsely villous

on the carina, petals slightly longer or shorter than

the sepals. Stand nodes almost 1 mm long. Styles 3,

about 3 mm long, villous, ovary 3-locular; fruit a 3-

locular, triangular capsule.

Bothalia 15, 3 & 4 (1985)

429

C. decipiens is restricted to the Hermanus and

Klein River Mountains, where it occurs between 300

and 900 m. The collections are all from peaty to

somewhat marshy habitats. Flowering occurs in Au-

gust and September.

This species is related to the C. hookerianum-C.

ebracteatum complex. It is easily distinguished by the

densely aggregated male spikelets, by the floral

bracts as long as the flowers, and by the culms with

only 1-2 sheaths. C. ebracteatum is sympatric with

C. decipiens, but occupies a drier habitat, along well-

drained ridges, whereas C. decipiens occurs in the

seepages and marshes. C. decipiens appears to be a

local, wetter habitat segregate of the C. ebracteatum

complex.

CAPE. — 3419 (Caledon): Klein River Mountains above Stan-

ford (-AD), Esterhuysen 33664 (BOL; K); Aasvogelkop above

Hermanus (-AD), Esterhuysen 35305 (BOL; K; L; M; MO; S);

Maanschynkop (-AD), Esterhuysen 31927 (BOL; K; MO; S); Es-

terhuysen 31734 (BOL; C; E; F; K; L; LD; M; MO; NBG; S;

STE; UC); Hermanus Mountain, on the middle slopes. S aspect

(-AD), Esterhuysen 33220 (BOL; C; K; L; M; MO; S); Vogelgat

Reserve (-AD), Esterhuysen 35535 (BOL; K; M; MO; S).

Chondropetalum deustum Rottb., Descriptiones

Plantarum Rariorum 12 (1772); Pillans in Trans. R.

Soc. S. Afr. 16 : 299 (1928); in Adamson & Salter,

FI. Cape Penins. 142 (1950). Type: Cape, without

precise locality, Konig s.n. cf (C, holo.!).

Elegia deusta (Rottb.) Kunth, Enum. PI. 3 : 460 (1841); Mast,

in A. DC., Monogr. Phan. 1 : 352 (1878); in FI. Cap. 7 ; 111

(1897).

Restio chondropetalum Nees in Linnaea 5 : 652 (1830), nom.

illeg. , superfluous name for Chondropetalum deustum Rottb.

Icon: Rottb.. Descriptionum et iconum rariores t.3 f.2 (1773).

Note

1. This species occurs on the Cape Peninsula and

along the mountains between Jonkershoek and Bet-

ty’s Bay and reaches as far as Bredasdorp. Many of

the collections are from seasonally wet localities —

peaty soils, or shallow sand over rock.

Chondropetalum ebracteatum (Kunth) Pillans

in Trans. R. Soc. S. Afr. 16 : 305 (1928); in Adam-

son & Salter, FI. Cape Penins. 144 (1950). Syntypes:

Cape, ‘Waagemmas Kersberg’, Drege 125 9 (B,

lect . ! ; B!; MO!; NY!; P!; S!); without precise local-

ity, Sieber 232 $ (BR!; K!; MEL!; MO!; P!; S!).

Dovea ebracteata Kunth. Enum. PI. 3 : 458 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 308 (1878); in FI. Cap. 7 : 102(1897).

Notes

1. There are two sheets of Drege 125 at B. The

sheet selected as lectotype has an original Drege la-

ble, giving a locality which I could not trace.

2. Sieber 232 is frequently labelled ‘Restio equise-

tum’.

3. I cannot clearly separate this species from C.

hookerianum. Pillans (1928, 1950) separates them

on flower size, but a complete range of intermediate

forms exist.

4. C. ebracteatum occurs on well-drained moun-

tain slopes and plateaux from Riversdale to the

Slanghoek Mountains and the Peninsula. Plants oc-

curing at higher altitudes or on shallow soils over

rock are frequently smaller and stouter than plants

from deeper soils or lower altitudes.

Chondropetalum hookerianum (Mast.) Pillans

in Trans. R. Soc. S. Afr. 16 : 304 (1928); in Adam-

son & Salter, FI. Cape Penins. 143 (1950). Syntypes;

Cape, 3418 (Simonstown): Nieuw Kloof, Houw

Hoek Mts (-BB), Burchell 8062 9 (K, lecto.!;

BOL!; OXF!; P!); 8062 cf (BOL!; K!; OXF!; P!).

3419 (Caledon): between Donkerhoek and Houw

Hoek Mts (-AA), Burchell 8007 cf (P!); Zoete-

melksvlei (-BA), Burchell 7568 cf (P!); mountains at

Grietjiesgat, June (-AA), Ecklon & Zeyhers.n. cf &

9 (MEL!); without precise locality, Thom 632 9

(K!); 901 Cf (K!); 1021 cf (K!); Ecklon960 cf &9 (?).

3419 (Caledon): Hemel en Aarde, Aug. (-AD),

Zeyher 4335 (MEL!).

Dovea hookeriana Mast, in J. Linn. Soc., Bot. 10 : 249 (1868);

in A. DC., Monogr. Phan. 1 : 306 (1878); in FI. Cap. 7 : 100

(1897).

Dovea bolusii Mast, in J. Linn. Soc., Bot. 21 : 576 (1885); in FI.

Cap. 7 : 102 (1897). Syntypes: Cape, 3418 (Simonstown): Muizen-

berg Mountain, near Kalk Bay (-AB), Bolus 3909 $ (K, lec-

to. !;BM!; BOL!); 3910 cf (BOL!; K!).

Notes

1. Burchell 8007 is a mixed collection containing

elements of both C. hookerianum and C. rectum.

2. C. hookerianum is probably conspecific with C.

ebracteatum. According to Pillans (1928) it is distin-

guished by having a female perianth 4-5 mm long,

instead of less than 4 mm in C. hookerianum. There

are numerous intermediate forms making this dis-

tinction trivial.

3. C. hookerianum is a montane species that

ranges from Riversdale to the Cape Peninsula. It

generally occurs on well-drained slopes and is quite

common.

Chondropetalum marlothii (Pillans) Pillans in

Trans. R. Soc. S. Afr. 16 : 306 (1928). Type: Cape,

Malmesbury division, Marloth 10158 9 (BOL,

lecto!; K!); 10158 cf (BOL!; K!).

Dovea marlothii Pillans in Ann. Bolus Herb. 3 : 143 (1922).

Note

1. This species should probably be included under

C. aggregation. Pillans (1928) separated them on the

size of the spathes and the spikelets, which is prob-

ably a trivial character. They key out together in my

key.

Chondropetalum microcarpum (Kunth) Pillans

in Trans. R. Soc. S. Afr. 16 : 303 (1928); in Adam-

son & Salter, FI. Cape Penins. 143 (1950). Type:

Cape, 3420 (Bredasdorp): St Sebastiansbay, Rhe-

nosterfontein (-B), Garnot s.n. 9 (B, holo.!; B!).

Dovea microcarpa Kunth, Enum. PI. 3 : 459 (1841).

Dovea rigens Mast, in Bot. Jb. 29 Beibl. 66 : 7 (1900). Type:

Cape, 3418 (Simonstown): Zeekoeivlei (-BA), Zeyhers.n. 9 (?)•

Icon: Mason, W. Cape Sandveld Flow, t.2 f.7 (1972).

Notes

1. Although I have not seen the type of Dovea

rigens, the description closely fits C. microcarpum. I

presume that the type was destroyed in Berlin.

430

Bothalia 15, 3 & 4 (1985)

2. This is a ‘coastal’ species, occurring in dunes

and on coastal limestone plateaux and shelves from

Melkbosstrand to Port Elizabeth. It is a very distinct

species, spreading by long, deep-seated rhizomes. It

has the same distribution and is superficially similar

to Ischyrolepis leptoclada, from which it is easily sep-

arated in the field by its spreading sheaths.

Chondropetalum mucronatum (Nees) Pillans in

Trans. R. Soc. S. Afr. 16: 308 (1928); in Adamson &

Salter, FI. Cape. Penins. 144 (1950). Type: Cape,

3418 (Cape Town): ‘Inter saxa et locis humidis sub-

paludosis verticis montis tabularis’, Nov. (-CD),

Zeyher s.n. 9 (S, lecto.!); s.n. cf (K!; S!).

Restio mucronatus Nees in Linnaea 5: 660 (1830). Elegia mu-

cronata (Nees) Kunth, Enum. PI. 3: 475 (1841). Dovea mucronata

(Nees) Mast, in J. Linn. Soc., Bot. 10: 251 (1868); in A. DC.,

Monogr. Phan. 1: 308 (1878); in FI. Cap. 7: 103 (1897).

Elegia panicoides Kunth, Enum. PI. 3: 470 (1841). Syntypes:

Cape, 3319 (Worcester): Du Toits Kloof. 900-1 200 m (-CA),

Drege 34 cf (K. lecto.!; MO!; P!); without precise locality, Garnot

s.n. Cf (P!).

leones: Lam., Tabl. Encycl. t. 804 f. 4, a-e (1799). J. Linn.

Soc., Bot. 10: t. 7b (1868). A. DC., Monogr. Phan. 1: t. 2f.l-6; t.

5 f. 6 (1878).

Notes

1. Nees (1830) refers to both the male and the fe-

male material in his description. The only female

material of the Zeyher collection that I could find is

in Stockholm. This material I designate as the lecto-

type.

2. Restio racemosus Poir. is based on both male

and female material. Pillans (1928) accepted the

male as the lectotype of Elegia racemosa (Poir.)

Pers. The female material is Chondropetalum mu-

cronatum. This ‘implicit’ lectotypification of Pillans

is formally accepted here.

3. C. mucronatum occurs in swampy places in the

mountains from Riversdale to Bainskloof and the

Cape Peninsula.

Chondropetalum nudum Rottb., Descriptiones

Plantarum Rariorum 12 (1772); Pillans in Trans. R.

Soc. S. Afr. 16: 299 (1928); in Adamson & Salter,

FI. Cape Penins. 142 (1950). Type: Cape, without

precise locality, Konig s.n. $ (C, holo.!).

Restio nudus (Rottb.) Nees in Linnaea 5: 651 (1830). Elegia

nuda (Rottb.) Kunth, Enum. PI. 3: 462 (1841); Mast, in A. DC.,

Monogr. Phan. 1: 353 (1878); in FI. Cap. 7: 114 (1897). Dovea

nuda (Rottb.) Pillans in Ann. Bolus Herb. 3: 146 (1922).

Restio acuminatus Thunb., Diss. Restio 13 (1788); Thunb., FI.

Cap. edn 1, 323 (1811); edn Schultes, 84 (1823), nom. illeg. , su-

perfluous name for Chondropetalum nudum Rottb.

Cuculifera dura Nees in Lindl., Introd. Nat. Syst. edn 2, 451

(1836); Kunth, Enum. PI. 3: 484 (1841), nom. illeg., superfluous

name for Restio acuminatus Thunb.

Elegia elongata Mast, in FI. Cap. 7: 114 (1897). Syntypes: Cape,

3318 (Cape Town): between Wynberg and Devils Peak (-CD),

Drege 147a cf (K, lecto.!; B !) ; between Paarl and the Lady Grey

Railway Bridge (-DB), Drege 9454 cf, 9 (MO!; P!).

leones: Rottb., Descriptionum et Iconum Rariores t. 3. f. 3

(1773). Mason, W. Cape Sandveld Flow. t. 3. f. 4 (1972).

Notes

1 . This species is close to the usually much taller

C. tectorum , but they can generally be separated on

the appearance of the petals. However, some popu-

lations show strange character combinations and are

in need of detailed study, as they could be of hybrid

origin.

2. C. nudum occurs on the sandy coastal forelands

from Darling to Albertina. A few collections are

from shaley soils. The habitats are generally marshy.

Chondropetalum rectum (Mast.) Pillans in

Trans. R. Soc. S. Afr. 16: 302 (1928); in Adamson &

Salter, FI. Cape Penins. 143 (1950). Type: Cape,

3419 (Caledon): between Donkerhoek and Houw

Hoek Mountains (-AA), Burchell 8007 , pp cf (K,

holo.!; BOL!; P!).

Dovea recta Mast, in FI. Cap. 7: 101 (1897).

Notes

1. Burchell 8007 is a mixed collection, also includ-

ing material of C. hookerianum (see above). The

sheet in K is named by Masters.

2. C. rectum is very distinct with its long rhizomes.

The fruit often appears to be indehiscent and is

dropped with the flower.

3. This species occurs on the coastal forelands

from Bredasdorp to Mamre Road near Malmesbury,

as well as near Wolseley in the Worcester Valley. It

is found on clayey, gravelly or sandy soils, usually in

wet or marshy habitats.

Chondropetalum tectorum (L.f.) Rafin., FI.

Tellur. 4: 33 (1838); Pillans in Trans. R. Soc. S. Afr.

16: 300 (1928); in Adamson & Salter, FI. Cape

Penins. 143 (1950). Type: Cape, without precise

locality from sandy fields beyond the Cape, Thun-

berg in herb. Linn. 1164.12 (LINN, holo.!).

Restio tectorum L.f., Suppl. 425 (1781); Thunb., FI. Cap. edn

1, 323 (1811); edn. Schultes, 85 (1823). Dovea tectorum (L.f.)

Mast, in J. Linn. Soc., Bot. 10: 249 (1868); in A. DC., Monogr.

Phan. 1: 306 (1878); in FI. Cap. 7: 99 (1897).

Dovea cylindrostachya Mast, in FI. Cap. 7: 100 (1897). Syn-

types: Cape, without precise locality, Thom 908 9 (K, lecto.!;

BOL!); Drege 96059 (BOL!).

Icon: Mason, W. Cape Sandveld Flow, t.2 f.6 (1972).

Notes

1 . In LINN there are two sheets of Chondropeta-

lum tectorum , but 1164.12 is labelled 'Restio tecto-

rum' in the hand of Linnaeus //7ms, whereas 1164.11

is annotated by J. E. Smith.

2. The sheet of Thom 908 at Kew has a fertile culm

(annotated by N. E. Brown as being identical to

Drege 9605 at Liibeck) and a sterile bit which clearly

belongs to another species. The sheet is labelled

'Dovea cylindrostachya' in the hand of Masters.

3. C. tectorum varies greatly in size, from 30 cm to

over 2 m tall. Some plants have well developed rhi-

zomes, others are caespitose and the petals may vary

from being half-scabrid to entirely smooth. This

range of variation still needs to be studied, as the

species as presently delimited may contain several

taxa. The limits to C. nudum also need to be studied.

4. C. tectorum occurs on the coastal forelands

from Grahamstown to Clanwilliam and favours mar-

shy habitats.

Bothalia 15, 3 & 4 (1985)

431

DOVEA

5. Dovea Kunth , Enum. PI. 3: 457 (1841);

Linder in Bothalia 15: 64 (1984).

Dovea macrocarpa Kunth , Enum. PI. 3: 458

(1841); Mast, in A. DC., Monogr. Phan. 1: 307

(1878); in FI. Cap. 7: 101 (1897). Type: Cape, 3218

(Clanwilliam): sandhills between Pretors and Piqui-

niers Kloof, 300-450 m (-DB), Drege 2523 § (B,

holo. !; BM!; BOL!; K!; MO!; NY!; P!).

Chondropetalum macrocarpum (Kunth) Pillans in Trans. R.

Soc. S. Afr. 16: 305 (1928).

Icon: Mason, W. Cape Sandveld Flow, t.3 f.l (1972).

Note

1. This very distinct species occurs at altitudes be-

tween 300 and 1 000 m in the rather dry fynbos on

sandy flats in the Piketberg and Citrusdal moun-

tains.

ASKIDIOSPERMA

6. Askidiosperma Steud., Syn. PI. Glum. 2: 257

(1855); Linder in Bothalia 15: 64 (1984).

Askidiosperma albo-aristatum ( Pillans ) Linder

comb. nov.

Chondropetalum albo-aristatum Pillans in J1 S. Afr. Bot. 18:

107 (1952). Type: Cape, 3219 (Wuppertal): North Cedarberg,

peak at Koupoort (-AC), Esterhuysen 12158 9 (BOL. lecto.l);

12158 cf (BOL!).

Note

1. This species is only known from the Cedarberg,

where it occurs above 1 300 m, usually in wet habi-

tats along streams and in smaller marshes.

Askidiosperma andreaeanum ( Pillans ) Linder,

comb. nov.

Chondropetalum andreaeanum Pillans in Trans. R. Soc. S. Afr.

16: 310 (1928). Type: Cape, 3319 (Worcester): mountains south

of Wemmershoek (-CC), Andreae 813 9 (BOL, lecto.l); 813 cf

(BOL!; K!).

Notes

1. This species is closely related to A. panicula-

tum. It does not seem possible to separate them on

the female flowers, but male plants are easily separ-

ated on flower size. The characters which Pillans

(1928) used in his key do not hold.

2. This species occurs amongst rocks on ridges or

on steep slopes above 1 200 m in the mountains from

Bainskloof to Jonkershoek. From this it would ap-

pear to occupy a different habitat to A. paniculatum

as well.

Askidiosperma capitatum Steud., Syn. PI.

Glum. 2: 257 (1855); Mast, in J. Linn. Soc., Bot. 10:

247 (1868); in A. DC., Monogr. Phan. 1: 304 (1878);

in FI. Cap. 7: 98 (1897). Type: Cape, 3219 (Wupper-

tal): Cedarberg, Ezelsbank (-AC), Drege 2510 (P,

holo.!; B!; K!; MO!; P!; OXF!).

Chondropetalum capitatum (Steud.) Pillans in Trans. R. Soc. S.

Afr. 16: 312 (1928).

Icon: A. DC., Monogr. Phan. 1: t.l f.32-36; t.5 f.5 (1878).

Notes

1. There are two sheets of the type collection in

Paris. One sheet is ‘Herb. Mus. Paris’, whereas the

other is ‘Herb. Steudel’. The latter is clearly the

holotype.

2. A. capitatum occurs above 1 000 m, usually in

somewhat marshy places, in the mountains from the

South Cedarberg to the Hex River Valley.

Askidiosperma chartaceum (Pillans) Linder,

comb. nov.

Dovea chartacea Pillans in Ann. Bolus Herb. 3: 144 (1922).

Chondropetalum chartaceum (Pillans) Pillans in Trans. R. Soc. S.

Afr. 16: 311 (1928). Type: Cape, 3418 (Simonstown): top of

Sneeuwkop (-BB), Marloth 3635 9 (BOL, holo.!).

Notes

1. The important key character (sepals as long as

the petals) may be difficult to observe in some speci-

KEY TO THE SPECIES OF ASKIDIOSPERMA

la Petals as long as the sepals:

2a Spikelets with 1 flower:

3a Male flowers 2-3 mm long A. paniculatum

3b Male flowers 3-4 mm long A. andreaeanum

2b Spikelets with several flowers A. chartaceum

lb Petals longer than the sepals:

4a Flowers 2-3 mm long A. esterhuyseniae

4b Flowers 4-7 mm long:

5a Flowers 4-5 mm long:

6a Spikelets narrowly oblong:

7a Spathes membranous, 2-3 cm long; from the Cedarberg A. albo-aristatum

7b Spathes chartaceous, 1-2 cm long; from the Caledon Mountains A. rugosum

6b Spikelets globose to obtriangular:

8a Spathes 5-6 cm long; culms more than 2 mm in diam A. insigne

8b Spathes 1-3 cm long; culms less than 2 mm in diam A. nitidum

5b Flowers 6-7 mm long; spikelets capitate:

9a Bracts coriaceous A. longiflorum

9b Bracts membranous A. capitatum

432

Bothalia 15, 3 & 4 (1985)

mens, as the sepals are very slender acuminate and

the apical sections break off rather easily.

2. A. chartacewn occurs above 1 000 m in marshy

places in the mountains from Ceres to Bainskloof

and southwards to Genadendal and the Kogelberg.

3. The specimens from the mountains between

Ceres and Worcester are distinct from those from

the mountains south of the Breede River Valley in

having firmer and darker spathes and ridged rather

than smooth seed. This suggests that they are best

treated as a subspecies of the southern populations:

Askidiospermum chartaceum (Pillans) Linder

subsp. alticolum Esterhuysen, subsp. nov., a subsp.

chartaceo spathis subcartilagineis, ferrugineis, testis

striatis differt.

TYPE — Cape, 3319 (Worcester): Waaihoek

Peak, on steep SE slope below the crest of the ridge,

1 900 m (-CB), Esterhuysen 31064 $ (BOL, holo.!;

Kl).

Flowering occurs in May, and seed-set in June.

This subspecies occurs in seepages above 1 800 m.

CAPE. — 3319 (Worcester): Waaihoek Peak (-CB), Esterhuy-

sen 32604 (BOL; K); 31654 (BOL; K).

Askidiosperma esterhuyseniae ( Pillans ) Linder,

comb. nov.

Chondropetalum esterhuyseniae Pillans in Trans. R. Soc. S.

Afr. 30: 255 (1945). Type: Cape, 3418 (Simonstown): Somerset

Sneeukop (-BB), Esterhuysen 8219 $ (BOL, holo.!; K!).

Notes

1. The fruit formation in this species appears to be

peculiar, as it often appears as though two cells are

dehiscent, and one cell is indehiscent. Similar irregu-

larities occur in several smaller flowered species of

Askidiosperma and Chondropetalum and deserve

more detailed study.

2. A. esterhuyseniae occurs in marshy or wet

places, usually above 1 000 m, in the mountains be-

tween the Slanghoek Mountains, Villiersdorp and

Sir Lowry’s Pass.

Askidiosperma insigne (Pillans) Linder , comb.

nov.

Chondropetalum insigne Pillans in J1 S. Afr. Bot. 18: 106

(1952). Type: Cape, 3319 (Worcester): Mt Superior, Waaihoek

(-CA), Esterhuvsen 18198 $ (BOL, lecto.!; K!); 18198 cf (BOL!;

K!).

Note

1 . This species occurs in swampy and wet places in

the Bokkeveld Mountains, the Hex River Mountains

and the Waaihoek Mountains. Most collections are

from about 1 800 m, although a few are from be-

tween 1 300 and 1 800 m.

Askidiosperma longiflorum (Pillans) Linder,

comb. nov.

Chondropetalum longiflorum Pillans in Trans. R. Soc. S. Afr.

30: 256 (1945). Type: Cape, 3319 (Worcester): Slab Peak (Mit-

chell's Pass) (-AD), Esterhuysen 6215 9 (BOL, lecto.!; K!); 6215

Cf (BOL!; K!).

Note

1. A. longiflorum occurs in dry habitats above

L 000 m in the Waaihoek Mountains between Ceres

and Worcester, with some collections from Turrett

Peak in the Bokkeveld Mountains and Slab Peak

above Mitchell’s Pass. One collection is from a

slightly marshy habitat.

Askidiosperma nitidum (Mast.) Linder, comb.

nov.

Dovea nitida Mast, in Bot. Jb. 29 Beibl. 66: 6 (1900). Chondro-

petalum nitidum (Mast.) Pillans in Trans. R. Soc. S. Afr. 16:312

(1928). Syntypes: Cape, 3219 (Wuppertal): Cold Bokkeveld,

Skurfteberge near Wagebooms Rivier. 1 700 m (-CA), Schlechter

10161 $ (B, lecto.!; BM!; BOL!; BR!; K!; MO!; P!; S!; Z!);

10160 Cf (B!; BM!; BOL!; BR!; K!; MO!; P!; S!; Z!).

Notes

1. There is some variation in the fruit formation.

In most populations all three locules are fertile and

dehiscent, but in one population the fruit appears to

be functionally unilocular and indehiscent. This

needs detailed investigation.

2. A. nitidum occurs in the mountains from

Jonkershoek at Stellenbosch to the northern Cedar-

berg at Clanwilliam, but has not been recorded from

the Cape Peninsula or the Piketberg. Most of the

collections are from marshy places, between 900 and

1 800 m.

Askidiosperma paniculatum (Mast.) Linder,

comb. nov.

Dovea paniculata Mast, in J. Linn. Soc., Bot. 21: 577 (1885); in

FI. Cap. 7: 102 (1897). Chondropetalum paniculatum (Mast.) Pil-

lans in Trans. R. Soc. S. Afr. 16: 309 (1928). Syntypes: Cape,

3319 (Worcester): Drakenstein Mountains near Bainskloof,

500-600 m (-CA), Bolus 4081 9 (BOL!; K!), 4082 cf (BOL!; K!);

4097 9 (K, lecto.!; BOL!); 4098 cf (BOL!; K!); Witzenberg

(-AA), Burchell 8719 cf (K!).

Note

1. This species is widespread, occurring in the

mountains from Swellendam to Paarl, Ceres and the

Cedarberg. It has not been recorded from the Penin-

sula and appears to be rare in the Cedarberg.

Askidiosperma rugosum Esterhuysen, sp. nov.,

ab A. alboaristato (Pillans) Linder floribus femineis

quam floribus masculis grandioribus, loculis laterali-

bus fertilibus differt.

TYPE. — Cape, 3419 (Caledon): Houwhoek

Mt. (-AA), Esterhuysen 34346 (BOL, holo.!; B; C;

E; F; K; L; LD; MO; NY; PRE; RSA; S; STE;

TCD; UC; US; W; WAG).

Plants caespitose, tussocks 60-100 cm tall. Rhi-

zomes about 3 mm in diameter, compact. Culms

solid, terete, simple, to 1,6 mm in diam., surface

smooth. Sheaths deciduous, loosely convoluted,

15-20 mm long, obtuse, finely mucronate, greenish.

Male inflorescence 9-15 cm long, compound. Spathes

15-30 x 4—8 mm, acute, cartilaginous with charta-

ceous margins, golden-green. Inflorescence

branches slightly longer than the spathes, with sev-

eral semi-globose to somewhat elongated, 4-7 mm in

diameter, many-flowered spikelets. Spathellae like

the bracts, but marginally longer and coriaceous at

the base. Bracts up to 10 x 0,5 mm, membranous,

long-acuminate, somewhat spreading. Flowers

shortly pedicellate, perianth 3 mm long. Sepals carti-

laginous, about 1,5 mm long; lateral sepals acute,

conduplicate, carinate; odd sepal flat, obtuse, 1,5 x

1 mm. Petals cartilaginous, concave, acute, 2,5-3 x

Bothalia 15, 3 & 4 (1985)

433

FIGinflor^ntf xTH ^^/sterhuysen. a habit; b, rhtzome and culm bases, note deciduous sheaths, X 0,8; c, female

sepals Xlabrous’ x Q-T nfl0rescence’ , * °>* e’male sP*elet, x 4; f, female flower, with the petals twice as long as the

largely included in the Deri a nth 6 xT^’ Snf ifr tha"the female flower, x 9; h, female flower dissected, note that the styles are

perianth, J ! Sh°Wing the piStiUode and the anthers “eluded in the

434

Bothalia 15, 3 & 4 (1985)

1 mm. Anthers 1,8 mm long, mucronate. Pistillode 1

mm tall. Female inflorescence 3-7 cm long, com-

pound. Spathes as in the male, 10-20 x 3-7 mm. In-

florescence branches about as long as the spathes.

Spikelets 1-2 per node, oblong, 6-15 x 6-8 mm.

Spathellae like the bracts, long-acuminate, to 10 mm

long with the apical § hair-like. Flowers shortly ped-

icellate, perianth 4,5-5 mm long. Sepals cartilagi-

nous, acute, red-tipped, 3 mm long; lateral sepals

conduplicate, carinate; odd sepal concave, 3 x 1,5

mm. Petals cartilaginous, concave, acute, 4 X 1,5-2

mm. Staminodes minute, to 1 mm long. Styles 3, free

to the base, filiform, densely villous. Ovary 3-locu-

lar. Fruit a 3-locular capsule with the lateral cells fer-

tile; seeds 1,5 x 1 mm, round in cross-section, ridge

faintly developed, ends obtuse, surface reddish, with

faint longitudinal ridging. Fig. 16.

A. rugosum is restricted to the mountains from

Houw Hoek to Grabouw, at altitudes of 200-700 m,

the collections are all from somewhat marshy areas

or on steep shaley slopes. Many of the collections

are from shale bands. At least some of the popula-

tions are extensive.

This species is clearly very closely related to A.

alboaristatum which is restricted to the Cedarberg.

Morphologically it can be distinguished by the fe-

male flowers being larger than the males, and by the

chartaceous spathes, which are only 1-2 cm long.

Geographically, the taxa are widely allopatric. Eco-

logically, A. alboaristatum occurs above 1 500 m, in

very poor sandstone-derived soils, whereas this

species occurs at much lower altitudes in shaley soils.

CAPE. — 3418 (Simonstown): Somersfontein (-BD), Ester-

huysen 32748a (BOL; K; S). 3419 (Caledon): Houw Hoek Moun-

tain (-AA), Esterhuysen 32779 (BOL; C; E; K; L; M; MO; S);

Esterhuysen 31123 (BOL; K; L; M; MO; S); Esterhuysen 34346

(B; BOL; C; E; F; K; L; LD; M; MO; NY; PRE; RSA; S; STE;

TCD; UC; US; W; WAG); Arieskraal Dam on the Palmiet River

(-AA), Powrie s.n. (BOL; K; S); Esterhuysen 31569 (BOL; C; E;

F; K; L; LD; M; MO; NBG; PRE; S; STE; TCD; UC; US; W;

WAG).

PLATYCAULOS

7. Platycaulos Linder in Bothalia 15: 64 (1984).

Note

1. The number of locules (couplet 2) is best seen

on fruit, where the capsules are dehisced.

Platycaulos acutus Esterhuysen , sp. nov., quoad

ovaris uniloculis ad P. ancepitem (Mast.) Linder ac-

cedit, sed ab ea differt marginibus culmis non incras-

satis et bracteis quam spathis longioribus.

TYPE. — Cape, 3320 (Montagu): Goedgeloof

Peak (-CD), Esterhuysen 34524 ^ (BOL, holo.!; C;

E; K; L; M; MO; S; STE).

Plants caespitose, to 50 cm tall. Culms solid, com-

pressed, branching, to 1 mm in diam., very finely

tuberculate. Sheaths closely convoluted, 1-2 cm

long, acute with an aciculate awn more than 1/3 of

the total sheath length, body coriaceous, pale

brown, with a very narrow hyaline margin. Male in-

florescence a solitary, 1-2 cm long spikelet. Spathes

like the sheaths, 1-1,5 cm long. Bracts all fertile, car-

tilaginous without a hyaline margin, acute, the lower

bracts often extended into an awn, to 15 x 2 mm,

subimbricate, about 5 per spikelet. Flowers on a very

short to 0,8 mm long pedicel, perianth 3,5-4 mm

long. Sepals cartilaginous, acute; lateral sepals con-

duplicate, villous along the carina; odd sepal 4 x 0,5

mm, shallowly concave, glabrous. Petals charta-

ceous, margins hyaline, acute, 3,5 x 0,7 mm. Anth-

ers exserted at anthesis, 2,6 mm long. Pistillode min-

ute, three-lobed. Female inflorescence of 1 (-2),

1-1,5 cm long, 1 (-2)-flowered spikelets. Spathes

like the sheaths, about 1 cm long. Bracts as in the

males, 2-A, lower bracts usually fertile, upper bracts

sterile. Flowers on a 1 mm long pedicel, perianth

5-5,5 mm long. Sepals cartilaginous, acute; lateral

sepals conduplicate to deeply concave, carinate in

the upper half, upper part of the carina pilose; odd

sepal 4,5 x 1 mm, shallowly concave, glabrous. Pe-

tals subcartilaginous, chartaceous along the margins,

acute, 4,5 x 1,5 mm. Staminodes 1 mm long. Ovary

with one fertile locule. Styles three, free. Fruit a cap-

sule, seed ellipsoid, 2,2 x 1,8 mm, surface smooth,

brittle, white. Fig. 17.

KEY TO THE SPECIES OF PLATYCAULOS

la Spikelets exceeding 2,5 cm in length P. major

lb Spikelets less than 2,5 cm in length:

2a Ovary unilocular:

3a Flowers 4-5 mm long; perianth half the length of the bract:

4a Tepals subequal; lateral sepals pilose P- acutus

4b Lateral sepals much longer than the other tepals; lateral sepals glabrous P. anceps

3b Flowers 2-3 mm long; perianth about equal to the length of the bract P. depauperatus

2b Ovary with two locules:

5a Plants erect shrubs, 1-3 m tall; only terminal branchlets compressed P. callistachyus

5b Plants sprawling, cushions or mats; all culms more or less compressed:

6a Spikelets with at least 3 flowers; perianth less than half the length of the bract P. compressus

6b Spikelets with 1-2 flowers; perianth at least half as long as the bracts:

7a Spikelets less than 5 mm long; culms without distinct margins

7b Spikelets 5-10 mm long; culms with distinct margins

P. subcompressus

.... P. cascadensis

Bothalia 15, 3 & 4 (1985)

435

FIG. 17. — Platycaulos acutus Esterhuysen. a, culm bases, aggregated and bearing the roots, x 0,8; b, female plant with com-

pressed and branching culms, x 0,8; c, detail of sheath, without a membranous margin and with a stout awn, x 3; d, female

spikelets with the flowers partially exposed, x 3; e, female flower with the lateral sepals villous carinate, x 5; f, dissection of

the female flower, x 5; g, male spikelet, x 3; h, male flower with exserted anthers, x 5; i, dissection of male flower showing

pistillode, x 5. (From Esterhuysen 34524.)

436

Bothalia 15, 3 & 4 (1985)

Platycaulos acutus occurs in marshes on the slopes

of Goedgeloof Peak in the Langeberg, between

1 350 and 1 550 m. It may be more widespread, but

the rest of the Langeberg is not botanically well

known.

This species is rather typical of this new genus, in

its morphology, seed-type and also its marshy habi-

tat. It is probably closest to P. anceps, from which it

differs in the lack of thickened margins on the culms

and by the spathes being shorter than the bracts.

CAPE. — 3320 (Montagu): Goedgeloof Peak (-CD), Esterhuy-

sen 34524 (BOL; C; E; K; L; M; MO; S; STE); Esterhuysen 35899

(B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY; PRE;

RSA; S; STE; UC; US; W; WAG).

Platycaulos anceps (Mast.) Linder , comb. nov.

Hypolaena anceps Mast, in J. Linn. Soc., Bot. 10 : 267 (1868);

in A. DC., Monogr. Phan. 1 ; 373 (1878); in FI. Cap. 7 : 131

(1897); Pillans in Trans. R. Soc. S. Afr. 16 : 397 (1928). Caloro-

phus anceps (Mast.) Kuntze, Rev. Gen. 747 (1894). Restio anceps

(Mast.) Pillans in J1 S. Afr. Bot. 18 ; 106 (1952). Syntypes: Cape,

3322 (Oudtshoorn): Post Berg near George (-CD), Burchett 5896

Cf (K, lecto.!; BOL!). 3321 (Ladismith): Kampsche Berg, Rivers-

dale (-CD), Burchett 7080 cf (K!).

Restio aspericaulis Pillans in Trans. R. Soc. S. Afr. 30 : 245

(1945). Type: Cape, 3323 (Willowmore): Joubertina, Tsitsi-

kamma Mountains (-DD), Esterhuysen 6879 9 (BOL, lecto.!;

K!); 6879 cf (BOL!; K!).

Notes

1. The original Burchell ticket on Burchell 7080 at

K is labelled in Burchell’s hand as 'Restio anceps

Neither sheet, however, is annotated in Masters’s

hand.

2. The material of R. aspericaulis is stouter and

more branched than the type of P. anceps, but this

could be due to the variation in the habitat.

3. P. anceps usually occurs along streams in the

wet sea-facing mountains from Hermanus and Swel-

lendam to Uniondale.

Platycaulos callistachyus (Kunth) Linder, comb.

nov.

Restio callistachyus Kunth, Enum. PI. 3 : 400 (1841); Mast, in

FI. Cap. 7 : 94 (1897); Pillans in Trans. R. Soc. S. Afr. 16 : 259

(1928). Type: Cape, Humansdorp Distr., mountains near the

Kromme River, Drege 30 9 (B, holo.!; BM!; K!; MO!; NY!;

OXF!; P!).

Restio polystachyus Kunth, Enum. PI. 3 : 402 (1841). Type:

Cape, without precise locality, Drege 32 cf (B , holo.!; BM!; K!;

MO!; OXF!; P!).

Restio spinulosus Kunth, Enum. PI. 3 : 402 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 292 (1878); in FI. Cap. 7 : 94 (1897).

Type: Cape, 3323 (Willowmore): between Welgelegen and Onzer

(-CD), Drege 37 cf (B, holo.!; BM!; K!; MO!; OXF!; P!; S!).

Restio concolor Steud., Syn. PI. Glum. 2 : 251 (1855); Mast, in

A. DC., Monogr. Phan. 1 : 241 (1878). Type: Cape, without pre-

cise locality, Drege 31 $ (P, holo.!; MEL!; P!).

Restio fastigiatus Nees ex Mast, in J. Linn. Soc., Bot. 8 : 250

(1865); nom. illeg. non R. Br. (1810), superfluous name for R.

callistachyus.

Restio mastersii F. Muell., Fragm. 8 : 68 (1872); Mast, in

A. DC., Monogr. Phan. 1 : 292 (1878); nom. nov. for R. fastigiatus'

Mast., nom. illeg., superfluous.

Notes

1. Kunth gave the male and the female collections

of Drege from the lower Langkloof separate names.

2. Restio fastigiatus Masters is both a later homo-

nym of R. fastigiatus R. Br. (corrected by Mueller in

1872), and also a superfluous name for R. callista-

chyus, as they are based on the same types.

3. Superficially, P. callistachyus appears to be a

Restio as only the ultimate branchlets are com-

pressed. However, the sheaths, seed coat and anat-

omy is typical of that of Platycaulos.

4. P. callistachyus occurs between Ceres and Port

Elizabeth. It is a tall shrub, to 3 m tall, growing by

streamsides and often in rocky riverbeds.

Platycaulos cascadensis (Pillans) Linder, comb.

nov.

Restio cascadensis Pillans in J1 S. Afr. Bot. 18: 102 (1952).

Type: Cape, 3418 (Simonstown): Cascades, Betty's Bay (-BC),

Parker 4516 $ (BOL, lecto.!; K!); 4516 cf (BOL!).

Note

1. P. cascadensis occurs along streams and water-

falls, often in shade, in the coastal areas of the Hot-

tentots Holland Mountains, i.e. Kogelberg and the

mountains at Betty’s Bay.

Platycaulos compressus (Rottb.) Linder in Bo-

thalia 15: 64 (1984). Type: Descriptionum et Iconum

Rariores T. 2, fig. 4 (1773) (Iconotype).

Restio compressus Rottb., Descriptiones Plantarum Rariorum

11 (1772); Kunth, Enum. PI. 3: 403 (1841); Mast, in A.DC., Mon-

ogr. Phan. 1: 290 (1878); in FI. Cap. 7: 95 (1897); Pillans in Trans.

R. Soc. S. Afr. 16: 255 .(1928); in Adamson & Salter, FI. Cape

Penins. 139 (1950).

Restio praefixus Mast, in Bull. Herb. Boissier, ser. 2, 1: 778

(1901). Type: Cape, 3321 (Ladismith): Langeberg near Rivers-

dale (-CC), Schlechter 1914 9 (K, holo.!; Z!).

Icon: Rottb., Descriptionum et Iconum Rariores t.2 f.4 (1773).

Notes

1 . I could not locate a specimen of P. compressus

in the Rottboell herbarium at C, nor anywhere else,

so the plate that Rottboell published has to serve as

an iconotype.

2. P. compressus occurs along streams and in

damp places in the mountains from the Cape Penin-

sula to Humansdorp.

Platycaulos depauperatus (Kunth) Linder,

comb. nov.

Restio depauperatus Kunth, Enum. PI. 3: 405 (1841); Mast, in

A.DC., Monogr. Phan. 1: 262 (1878); in FI. Cap. 7: 78 (1897);

Pillans in Trans. R. Soc. S. Afr. 16: 238 (1928). Syntypes: Cape,

3319 (Worcester): Fransch Hoek Mountains (-CC), Drege 2021 9

p.p. (B. lecto.!; BOL!; MEL!; NY!); 2021 p.p. C f (B!; BOL!;

MEL!; NY!).

Notes

1. The type collection is mixed, the other part of

Drege 2021 cf is the type of R. graminifolius Kunth.

The material in B is annotated by Kunth.

2. Kunth (1841) noted ‘ovario bilocularis; stylis 2,’

while the type in B is unilocular, with three styles.

3. P. depauperatus occurs in wet localities in the

mountains between Malmesbury, Tulbagh and Stel-

lenbosch.

Platycaulos major (Mast.) Linder, comb. nov.

Restio compressus Rottb. var. major Mast, in .1. Linn. Soc.,

Bot. 8: 249 (1865); in A.DC., Monogr. Phan. 1: 291 (1878); in FI.

Cap. 7: 95 (1897). Restio major (Mast.) Pillans in Trans. R. Soc.

S. Afr. 16: 256 (1928); in Adamson & Salter, FI. Cape Penins. 139

(1950). Syntypes: Cape, 3318 (Cape Town): near Cape Town

Bothalia 15, 3 & 4 (1985)

437

(-CD), Drege 48 cf (P, lecto.!; K!) ; Table Mountain (-CD), Eck-

lon 849 cf (P!). 3419 (Caledon): mountains near Grietjies Gat

(-AA), Zeyher4350 c f ( K! ; P!); Swartberg, near the baths (-AB),

Ecklon & Zeyher s.n. (?); without precise locality, Sieber 224 2

(BR! ; HI; Kl; MO!; PI); Ecklon 842 (?).

Notes

1. Although there are several syntypes at K, none

is annotated by Masters. Possibly the annotated ma-

terial is in herb. Sonder.

2. P. major has the same distribution range as P.

compressus, along the coastal mountains from the

Cape Peninsula to Humansdorp. The relationship

between the two taxa needs detailed investigation.

Platycaulos subcompressus (Pillans) Linder,

comb. nov.

Restio subcompressus Pillans in Trans. R. Soc. S. Afr. 30: 253

(1945). Type: Cape, 3418 (Simonstown): Landdrost Kop (-BB),

Esterhuysen 3601 2 (BOL, lecto.!; K!); 3601 cf (BOL!; K!).

Notes

1. This species is rather close to P. cascadensis,

and it may be difficult to maintain them as separate

species.

2. P. subcompressus occurs in marshes in the

mountains from Bainskloof to Sir Lowry’s Pass.

RESTIO

8. Restio Rottb., Descriptiones Plantarum Ra-

riorum 9 (1772); Linder in Bothalia 15: 64 (1984).

Craspedolepis Steud., Syn. PI. Glum. 2: 264

(1855).

Notes

1. The number of fertile locules (couplets 6, 13) is

most easily determined from old flowers, in which

the capsules have dehisced. Here dissection is usu-

ally not necessary. If only young material is avail-

able, the position of the styles is helpful: if central on

a trigonous ovary, it is usually 3-locular, if central on

a flattened ovary, it is 2-locular, if eccentric, the

ovary is usually unilocular.

2. The hairs on the lateral or anterior sepals are

occasionally lost from old flowers. Hairs can often

only be observed with a x 10 lens.

3. Ciliation of the bracts (couplet 20) may be very

fine and can often not be seen by naked eye.

KEY TO THE SPECIES OF RESTIO

la Culms 4-angled or compressed:

2a Culms 4-angled:

3a Plants 60-100 cm tall; about 3 mm in diameter at the middle; spikelets 6-8 mm long; lateral sepals

villous R. tetragonus

3b Plants 150-200 cm tall; about 5 mm in diameter in the middle; spikelets 4-6 mm long; lateral sepals

glabrous R. quadratus

2b Culms more or less compressed:

4a Spikelets longer than 1cm R. acockii

4b Spikelets shorter than 1 cm:

5a Perianth 1,5 mm long R. subtilis

5b Perianth 3 mm long R. stereocaulis

lb Culms terete:

6a Ovary with three fertile locules:

7a Spikelets 1 (2)-flowered:

8a Culms simple or rarely branched R. miser

8b Culms moderately to much branched:

9a Spikelets 6-7 mm long; flowers 3-4 mm long R. similis

9b Spikelets 10 mm long; flowers 7 mm long R. rupicola

7b Spikelets few to many-flowered:

10a Sheaths with linear foliaceous awns R. quinquefarius

10b Sheaths without such awns:

11a Flowers about half the length of the bracts; bracts acute R. pedicellatus

lib Flowers about the same length as the bracts; bracts obtuse:

12a Culms 0,5-1 mm in diameter; spikelets usually 2-4 R. confusus

12b Culms 0,3-0, 5 mm in diameter; spikelets solitary R. filicaulis

6b Ovary with 1-2 fertile locules:

13a Ovary with 2 fertile locules, occasionally one aborted:

14a Spikelets l-(2)-flowered:

15a Culms almost smooth:

16a Spikelets obovate; bract apices recurved R. obscurus

16b Spikelets elliptic; bract apices erect:

17a Spikelets with no sterile bracts R. perseverans

17b Spikelets with sterile bracts:

18a Bracts acute; seed colliculate R. aureolus

438

Bothalia 15, 3 & 4 (1985)

18b Bracts obtuse-mucronate; seed smooth R. pulvinatus

15b Culms more or less tubercled:

19a Perianth (5)7-8 mm long, bracts dark brown R. brachiatus

19b Perianth 2-5 (6) mm long, if 5-6 mm long then the bracts are straw-coloured:

20a Bracts lacerate-ciliate:

21a Awn on sheath at least as long as the body; perianth 2 mm long R. scaberulus

21b Awn on sheath shorter than the body; perianth 3-3,5 mm long R. stokoei

20b Bracts not ciliate:

22a Sheaths 1,5-3 cm long, the upper half hyaline; culms with dense flat-topped tubercles

R. peculiaris

22b Sheaths 0,5-1 ,5 cm long; culms smooth or finely tuberculate;

23a Spikelets 4-6 mm long; perianth 3-4 mm long:

24a Lateral sepals slightly villous; culms smooth to finely tuberculate R. perplexus

24b Lateral sepals glabrous; culms roughly tuberculate R. verrucosus

23b Spikelets about 8 mm long; perianth 5-6 mm long; lateral sepals glabrous R. patens

14b Spikelets 2-many-flowered:

25a Perianth distinctly exserted beyond the apex of the bract:

26a Culms and sheaths coarsely tubercled:

27a Only culm bases tubercled; from the Slanghoek Mts R. montanus

27b Entire culms tubercled; from the Cedarberg and Giftberg R. tuberculatus

26b Culms and sheaths not coarsely tubercled:

28a Bracts finely lacerate-ciliate:

29a Flowers 3-4 mm long; branching dichotomous R. sejunctus

29b Flowers 1,5-2 mm long; branching subverticellate R. fourcadei

28b Bracts not lacerate-ciliate:

30a Lateral sepals villous-carinate:

31a Bracts acute, loosely imbricate R. galpinii

31b Bracts rounded, closely imbricate R. burchellii

30b Lateral sepals glabrous:

32a Perianth 3,5^1 mm long; spikelets spicate R. distans

32b Perianth 4,5-8 mm long; spikelets in panicles R. cymosus

25b Perianth apex not or very shortly exserted beyond the bract, sometimes visible on the side of

the bracts:

33a Lateral sepals glabrous:

34a Culms simple R. pedicellatus

34b Culms branching:

35a Lateral sepals carinate-lacerate R. fusiformis

35b Lateral sepals not carinate-lacerate:

36a Style bases not all adjacent:

37a From north of the Limpopo River; styles evenly spaced on the ovary:

38a From tropical Africa R. mahonii

38b From Madagascar R. madagascariensis

37b From south of the Limpopo River; two styles adjacent, one separate:

39a Spikelets less than 5 mm long R. harveyi

39b Spikelets 10-20 mm long R ■ ambiguus

36b Style bases all adjacent:

40a Perianth 10 mm long; spikelets usually solitary R. egregius

40b Perianth 4-5 mm long; spikelets several R. micans

33b Lateral sepals pilose to villous on the carina:

41a Bracts acuminate:

42a Culms roughly tuberculate R- scaber

42b Culms smooth to finely tuberculate:

43a Bracts gradually acuminate:

44a Bracts recurved R- echinatus

44b Bracts erect:

45a Spikelets numerous R- occultus

45b Spikelets 1-2:

Bothalia 15, 3 & 4 (1985)

439

46a Spikelets (l)2-flowered; bracts brown R. perseverans

46b Spikelets many-flowered; bracts pale, red-speckled R. rarus

43b Bracts abruptly acuminate:

47a Spikelets widely ovate or rotundate, with about 14 bracts visible; from the Cedar-

berg R. brunneus

47b Spikelets narrower with about 10 bracts visible, not from the Cedarberg

R. pachystachyus

41b Bracts rounded to acute:

48a Bracts rounded to very obtuse, muticous or with a short erect mucro, with a small band

of dark hollow cells at or near the margin:

49a Spikelets 1,5-2 cm long:

50a Bracts rotundate, apex rounded with a short, stout mucro, perianth 6-8 mm long;

spikelets obtuse R. bolusii

50b Bracts elliptic, apex obtuse, muticous, perianth 5 mm long; spikelets acute

R. strobilifer

49b Spikelets 1-1,5 cm long:

51a Anterior sepal villous; culms flattened to sulcate R. bifurcus

51b Anterior sepal rarely villous; culms terete:

52a Spikelets globose; brown portion of the bracts truncate R. nodosus

52b Spikelets ellipsoid to ovoid; brown portion of the bracts obtuse to rounded:

53a Bracts rounded; dark hollow cells sharply delimited from the rest of the bract

R. burchellii

53b Bracts acute; dark hollow cells grading into the rest of the bract R. praeacutus

48b Bracts not as above:

54a Culms much branched:

55a Spikelets 3-4 mm long R. harveyi

55b Spikelets much longer:

56a Bracts lacerate-ciliate R. multiflorus

56b Bracts not lacerate-ciliate:

57a Styles connate at the base R. involutus

57b Styles free at the base:

58a Spikelets several, rarely solitary; culms less than 1 mm in diameter at the

base R. communis

58b Spikelets 1 (2); if the culms are less than 2 mm in diameter then the plants

are from Natal:

59a Bracts and sheaths with hyaline shoulders R. ingens

59b Bracts and sheaths without hyaline shoulders R. galpinii

54b Culms simple or sparingly branched:

60a Bracts almost flat and with distinct dark purple-brown bands extending down both

sides from the mucro:

61a Perianth 2,5 mm long; lateral sepals acute R . bifidus

61b Perianth 3,5—4 mm long; lateral sepals obtuse R ■ exilis

60b Bracts tightly convoluted, or if flat, then not coloured as above:

62a Culms coarsely tubercled R. scaber

62b Culms smooth or finely tubercled:

63a Bracts straw-coloured, apices red-speckled; culm bases rhizomatous or sto-

loniferous:

64a Culms 3-5 mm in diameter; from the southern Cape Province R. vallis-simii

64b Culms less than 1 mm in diameter; from Zululand R. zuluensis

63b Bracts brown or reddish; culm bases aggregated:

65a Bracts 1-2 cm long, half- flattened, very loosely imbricate:

66a Spathes larger than the bracts, almost as tall as the spikelet

R. purpurascens

66b Spathes smaller than the bracts, much shorter than the spikelets:

67a Bracts cartilaginous, 1,5-2 cm long R- bifarius

67b Bracts papery, 1-1,3 cm long R- papyraceus

65b Bracts 0,3-1 cm long:

68a Sheaths loosely convoluted R- singulars

68b Sheaths tightly convoluted:

69a Spathes with awns as long as the body R ■ strictus

440

Bothalia 15, 3 & 4 (1985)

69b Spathes not awned:

70a Bracts subacute to rounded, apically cleft, apices reflexed

R. nuwebergensis

70b Bracts acute, erect, entire:

71a Flowers 5,5-6 mm long R. inveteratus

71b Flowers 4,5-5, 5 mm long:

72a Spikelets about 1 cm long; plants 20-40 cm tall; sheaths rounded

R. filiformis

72b Spikelets 1-1,5 cm long; plants 40-80 cm tall, sheaths acute

R. praeacutus

13b Ovary with a single, occasionally indehiscent, fertile locule:

73a Spikelets 1-2-flowered:

74a Lateral sepals glabrous:

75a Culms tuberculate:

76a Bracts about half as long as the flowers:

77a Plants spreading, often stoloniferous R. zwartbergensis

77b Plants tightly caespitose R. pumilus

76b Bracts at least as long as the flowers:

78a Bracts as long as the flowers, almost flat:

79a Spikelets 5 mm long R. secundus

79b Spikelets 2-3 mm long R. harveyi

78b Bracts overtopping and obscuring the flowers and closely rolled around them

R. versatilis

75b Culms rugulose to smooth:

80a Spathes shorter than the spikelets:

81a Bracts acute:

82a Seed colliculate; from the Riviersonderend Mountains R. colliculospermus

82b Seed nitid; from the Langeberg R. fragilis

81b Bracts obtuse:

83a Bracts without hyaline apices R. distans

83b Bracts with hyaline apices:

84a Flowers longer than the bracts; spikelets about 3 mm long R. implicatus

84b Flowers shorter than the bracts; spikelets 5-7 mm long:

85a Terminal branches flexuose R. arcuatus

85b Terminal branches straight R. degenerans

80b Spathes taller than the spikelets:

86a Inflorescence much branched R. secundus

86b Inflorescence almost simple R. decipiens

74b Lateral sepals pilose or villous R. debilis

73b Spikelets 2, several- or many-flowered:

87a Lateral sepals distinctly winged on the upper half of the median nerve:

88a Lateral sepals equal R. distichus

88b Lateral sepals unequal:

89a Lateral sepal wings lacerate, glabrous R. dodii

89b Lateral sepal wings villous or puberulous R. ejuncidus

87b Lateral sepals not winged:

90a Culms with dense, flattopped, whitish tubercles R. triticeus

90b Culms smooth or minutely tuberculate:

91a Spathes as tall as or taller than the spikelets:

92a Spathes 2-4 cm long R. dispar

92b Spathes 1-1,5 cm long R. inconspicuus

91b Spathes much shorter than the spikelets:

93a Lateral sepals glabrous:

94a Spikelets many in a panicle, 5-6-flowered R. festuciformis

94b Spikelets solitary or rarely geminate, 1-3-flowered R. leptostachyus

93b Lateral sepals villous-carinate:

95a Flowers 3 mm long R. pillansii

Bothalia 15, 3 & 4 (1985)

441

95b Flowers 4-6 mm long:

96a Flowers 6 mm long; spikelets solitary R. alticola

96b Flowers 4-5 mm long; spikelets usually several:

97a Plants rhizomatous; bracts acute R. sarocladus

97b Plants caespitose; bracts awned R. corneolus

Restio acockii, Pillans in Trans. R. Soc. S. Afr.

29 : 339 (1942). Type: Cape, 3318 (Cape Town):

Brackenfell (-DC), Acocks 4159 9 (BOL, lecto.!;

K!; S!); 4158 cf (BOL!; K!; S!).

Note

1. This very rare species occurs in seasonally wet

hollows on the sandy coastal forelands. It was first

recorded from the Brackenfell area, where it is now

probably extinct, but is also known from the Mal-

mesbury-Mamre area, where its survival will depend

on active conservation measures. It shares its distri-

bution range with Restio micans.

Restio alticola Pillans in J1 S. Afr. Bot. 18 : 101

(1952). Type: Cape, 3319 (Worcester): Witteberg,

gully on the south side (-CA), Esterhuysen 9502 9

(BOL, lecto.!; K!); 9502 cf (BOL!).

Note

1. R. alticola is known from the Slanghoek and the

Wemmershoek Mountains, where it occurs between

1 500 and 1 800 m on shale bands or on rocky slopes.

Restio ambiguus Mast, in FI. Cap. 7 : 96 (1897).

Type: Cape, without precise locality, Zeyher s.n. cf

(K, holo.!; B!).

Plants caespitose, tangled, 40-60 cm tall. Culms

solid, terete, smooth, branching, to 1 mm in diam.

Sheaths tightly convoluted, 10-20 mm long, obtuse,

coriaceous, the lower half dark red-brown, the up-

per half paler, decaying. Male inflorescence of 1-2

slender subulate acute 20-30 x 3 mm spikelets.

Spathe like the bracts, imbricate, texture like the

sheaths but more acute, 6-9 mm long. Flowers nu-

merous, shortly pedicellate, perianth 5 mm long. Se-

pals cartilaginous, glabrous, acute, the laterals con-

duplicate, 5 x 1,5 mm; anterior sepal 5x1 mm.

Petals papyraceous, acute, 4,5 x 1 mm. Anthers ex-

serted at anthesis, 2 mm long, finely mucronate. Pis-

tillode 0,3 mm long, 3-lobed. Female inflorescence

similar to the male, but with the spikelets 15-25 mm

long and the bracts spreading. All bracts are fertile.

Flowers shortly pedicellate, perianth 4,5 mm long.

Sepals are cartilaginous to coriaceous, glabrous,

finely acute, 4,5 X 1,5 mm, the anterior sepal flat,

laterals conduplicate. Petals papyraceous, acute,

membranous in the upper 2/3, 3x1 mm. Stami-

nodes 1,5 mm long. Ovary bilocular; styles 3, slen-

der, villous, free to the base. Fruit a (l)-2-locular

capsule. ‘Seed 1 mm long, ellipsoid, grey-black, ver-

rucose’ (Parker, ms).

This species occurs from the Cape Peninsula to the

mountains at Genadendal and south to Hermanus. It

generally occurs in marshy habitats.

Masters (1897) described this species only from

male material, and Pillans (1928) included the name

under Chondropetalum hookerianum. As the female

has not been described, I give the above description .

Restio arcuatus Mast, in A. DC., Monogr. Phan.

1 : 247 (1878); in FI. Cap. 7 : 69 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 230 (1928). Type: Cape,

3320 (Montagu): mountain peak near Swellendam

(-CD), Bure he 1 1 7306 cf (K, holo.!; K!; BOL!; P!).

Notes

1. Restio arcuatus appears to be a Langeberg en-

demic species. It is very closely related to R. distans,

which occurs in the Caledon and Stellenbosch moun-

tains. R. arcuatus can be distinguished from R. dis-

tans by the hyaline margins of the bracts and the

more flexuose ultimate branchlets.

2. There are two sheets of the type at K. However,

only one is annotated by Masters and this is regarded

as the holotype.

Restio aureolus Pillans in Trans. R. Soc. S. Afr.

30 : 246 (1945). Type: Cape, 3319 (Worcester):

Buffelshoek area (-AD/BC), Esterhuysen 8415 J

(BOL, holo.!; K!).

Notes

1. Pillans did not describe any male material. The

male inflorescence, spikelets and flowers are almost

identical to those of the female.

2. R. aureolus occurs above 1 500 m on the rocky

slopes and gullies of the higher mountains in the Hex

River Mountains and the peaks around Tulbagh. It

is clearly a high-altitude species.

Restio bifarius Mast, in J. Linn. Soc., Bot. 10 :

278 (1868); in A. DC., Monogr. Phan. 1 : 288 (1878);

in FI. Cap. 7 : 91 (1897); Pillans in Trans. R. Soc. S.

Afr. 16 : 258 (1928). Type: Cape, 3418 (Simons-

town): Nieuw Kloof, Houw Hoek Mountains (-BB),

Burchell 8068 ? p.p. (K, lecto.!; BOL!; P!); 8068 cf

p.p. (BOL!; K!; P!).

Notes

1. The type, Burchell 8068, is a mixed collection.

The other element is R. purpurascens.

2. Restio bifarius is fairly frequent on dry moun-

tain slopes between Jonkershoek and Caledon.

There is one dubious record from the Cape Penin-

sula.

Restio bifidus Thunb. in Phytogr. Bl. 1 : 7

(1803); Thunb., FI. Cap. edn 1, 331 (1811); edn

Schultes, 87 (1823); Kunth, Enum. PI. 3 : 409

(1841); Mast, in A. DC., Monogr. Phan. 1 : 284

(1878); in FI. Cap. 7: 89 (1897); Pillans in Trans. R.

Soc. S. Afr. 16 : 260 (1928); in Adamson & Salter,

FI. Cape Penins. 139 (1950). Type: Cape, without

precise locality, herb. Thunberg 23225 9 (UPS,

holo.!; H!; S!).

Restio pseudoleptocarpus Kunth. Enum. PI. 3 : 399 (1841).

Syntypes: Cape, 3318 (Cape Town): Table Mountain (-CD),

Drege 28 $ (K, lecto.!; BM!; BOL!; MO!; NY!; OXF!; P!); with-

out precise locality, Sieber 221 cf (BR!; H!; K!; MEL!; MO!; P! ) ;

Garnot s.n. (?).

442

Bothalia 15, 3 & 4 (1985)

Notes

1. There is some variation in the size of the spike-

lets, bracts and flowers, on the basis of which a vari-

ant with smaller flowers may be recognized from the

Kogelberg Mountains. This, however, needs critical

investigation.

2. R. bifidus occurs on sandstone-derived soils

from the Cape Peninsula to Jonkershoek and Her-

manus, over a wide altitude range. It appears to be

most common on sandy flats in mountains.

Restio bifurcus Nees ex Masters in J. Linn. Soc.,

Bot. 8 : 247 (1865); in A. DC., Monogr. Phan. 1 : 275

(1878); in FI. Cap. 7 : 83 (1897); Pillans in Trans. R.

Soc. S. Afr. 16 : 245 (1928); in Adamson & Salter,

FI. Cape Penins. 138 (1950). Syntypes: Cape, 3418

(Simonstown): Cape Flats (-BA), Ecklon 564c cf

(S!); without precise locality, Drege 364 9 (OXF,

lecto.!); Drege 1608 9 (K!; MO!); Ecklon & Zeyher

39 c f & $ (?); Zeyher 1011 (?); Drege 46 $ (?); Thom

49 Cf (K!); Harvey s.n. cf (?).

? Restio furcatus Nees ex Mast, in J. Linn. Soc., Bot. 8 : 242

(1865); in A. DC., Monogr. Phan. 1 : 275 (1878); in FI. Cap. 7 : 83

(1897). Type: Cape, without precise locality, Ecklon & Zeyher

s.n. cf (?).

Notes

1. The typification of this species presents prob-

lems, as I have located so few of the syntypes and as

the general similarity of this species to R. burchellii

demands careful typification. However, Drege 364 at

OXF fits the description, and is annotated by Mas-

ters and so is a good candidate as lectotype. There is

an excellent sheet at K, labelled ‘ Restio furcatus’ by

Masters, which includes a Thom collection (herb.

Hooker, cited as a syntype of R. bifurcus), and an

Ecklon & Zeyher collection which was acquired in

1865 from herb. Reichenbach, and so was not seen

by Masters before the publication of the names.

2. R. furcatus presents even more problems than

R. bifurcus, as I have not seen any type material of

the species. As Masters himself annotated material

of R. bifurcus as R. furcatus, this name probably be-

longs here.

3. R. bifurcus has been much confused with the R.

burchellii-praeacutus-bolusii group, but is distin-

guished by the wrinkled, irregularly sulcate culms

and the villous patch near the apex of the anterior

sepal (this also occurs in some collections of R. bur-

chellii). It occurs in low-lying sandy areas on the

Cape Peninsula, and reaches north as far as Malmes-

bury and south to Shaw’s Mountain near Caledon.

Restio bolusii Pillans in Trans. R. Soc. S. Afr.

16 : 247 (1928). Type: Cape, 3419 (Caledon): Ge-

nadendal (-BA), Bolus 7431 J (BOL, lecto.!); 7431

Cf (BOL!).

Notes

1. This species belongs to a complex including R.

insignis, R. burchellii, R. strobilifer, R. praeacutus

and R. bifurcus. Detailed studies are required to es-

tablish the correct species limits in this complex and

to determine the ranks of the taxa. R. bolusii and R.

insignis are very close, and I have not succeeded in

separating them on morphological grounds. Geo-

graphically, R. bolusii has been recorded as occur-

ring in the mountains from Caledon to Worcester

and Stellenbosch, whereas R. insignis occurs in the

Ceres and Clanwilliam divisions. R. bolusii can be

distinguished from R. strobilifer by its obtuse spike-

lets and from R. burchellii by its larger spikelets. R.

praeacutus is little understood, but I am distinguish-

ing it from R. burchellii on the bract shape.

2. In addition to the taxa discussed above, there

are many collections (largely from the southern

Cape mountains, reaching Worcester) that I have

not yet studied. I have no doubt that a careful study

of the complex will result in a different classification

of its constituents.

Restio brachiatus (Mast.) Pillans in Ann. Bolus

Herb. 3 : 85 (1921); in Trans. R. Soc. S. Afr. 16 : 241

(1928). Syntypes: Cape, 3319 (Worcester): Gy-

douwsberg, 1 850 m (-AB), Schlechter 10229 5 (B,

lecto.!; BM!; BOL!; BR!; K!; MO!; P!; S!); 10228 cf

(B!; BM!; BOL!; BR!; K!; MOL!; P!; S!).

Leptocarpus brachiatus Mast, in Bot. Jb. 29 : Beibl. 66 : 9

(1900).

Notes

1. In the key this species may be confused with

Restio patens, which is, however, quite different.

2. R. brachiatus is close to R. cymosus, from which

it is distinguished by its single-flowered spikelets.

This may be a trivial character.

3. R. brachiatus is known from the inland moun-

tains from Ceres to Meiringspoort in the Great

Swartberg, with some collections from the Paarl-

Worcester Mountains.

Restio brunneus Pillans in Trans. R. Soc. S.

Afr. 16 : 250 (1928). Syntypes: Cape, without pre-

cise locality, Marloth 4793 9 (BOL, lecto.!); 11795 cf

(BOL!). 3219 (Wuppertal): Cedarberg (-A/C),

Barnard in SAM 33082 cf (SAM!).

Notes

1. At K there is a Marloth 11793 $, ex BOL, an-

notated as a type by Pillans, but not cited in the pro-

tologue.

2. R. brunneus occurs in the Cedarberg and the

adjacent mountains and so is allopatric with the re-

lated R. pachystachyus.

Restio burchellii Pillans in Trans. R. Soc. S.

Afr. 29 : 340 (1942). Type: Cape, 3419 (Caledon):

Nieuweberg (-AA), Stokoe 3187 9 (BOL, lecto.!);

3187 cf (BOL!).

Notes

1. See R. bolusii for notes on the taxonomic status

of this species. It may be distinguished from R. bolu-

sii by the spikelets shorter than 1 cm, and with only

1-3 flowers. Note that old flowers often have the

perianth much exserted from the bracts. On some

specimens the anterior sepal is villous, thus blurring

the distinction from R. bifurcus.

2. R. burchellii is known from the mountains from

Villiersdorp to Hermanus, over a large altitudinal

range, from moist to dry conditions.

Bothalia 15, 3 & 4 (1985)

443

Restio capillaris Kunth, Enum. PL 3 : 405

(1841); Mast, in A. DC., Monogr. Phan. 1 : 264

(1878); in FI. Cap. 7 : 79 (1897); Pillans in Trans. R.

Soc. S. Afr. 16 : 239 (1928). Type: Cape, 3319 (Wor-

cester): Du Toits Kloof (-CA), Drege 339 9 (B,

lecto.!; BOL!; K!; P!); 339 cf (B!; BOL!; K!; P!).

Restio perplexus Kunth var. gracilis Mast, in A. DC., Monogr.

Phan. 1 : 287 (1878). Restio implexus Mast, in FI. Cap. 7 : 90

(1897), nom. nov. for R. perplexus var. gracilis. Type: Cape, 3320

(Montagu): foot of the Langeberg near Swellendam (-CD), Bur-

chell 7430 ? (K, lecto.!; P!); 7430 cf (K!).

Note

1. Masters (1878) quoted the ‘type’ number of R.

perplexus var. gracilis as Burchell 7420. This is prob-

ably an error, as the material at K of this species is

Burchell 7430. There are three sheets of this collec-

tion at K, one sheet has an original Burchell label

with locality and ‘ Restio implexus' written in Bur-

chell’s hand. This sheet is also determined as ‘Restio

perplexus Kunth (forma gracilis)' , and so is clearly

the best candidate as lectotype.

Restio colliculospermus Linder, sp. nov., a R.

fragili Esterhuysen spathis spiculis minoribus, semi-

nibus colliculosis, stylis ad infimum non approxima-

tis differt.

TYPE. — Cape, 3419 (Caledon): Pilaarskop,

near Lindeshof, 1 350 m (-BB), Esterhuysen 33341

9 (BOL, holo.!; C; K; L; M; MO; E; S; STE).

Plants tangled. Culms very slender, 0,2 mm in

diam., terete, branching, rugulose, basal portions of

culms rhizomatous, rooting. Sheaths convoluted, 5-7

mm long, body coriaceous, green to pale brown,

similar to the culms, acute, apices extended into a

stout, subclavate awn, flanked and much overtopped

by membranous shoulders. Male inflorescence of

1-3, 3-4 mm long single-flowered spikelets. Spathes

like the sheaths, about half as long as the spikelets.

Bracts chartaceous in the lower half, membranous

above, acute, 3 mm long. Flowers subsessile, per-

ianth 2,5-3 mm long. Sepals cartilaginous, acute,

deeply concave, glabrous, 2,5-3 x 0,6 mm, laterals

subcarinate in the upper half. Petals membranous

apically rounded and somewhat lacerated, 2,8 x 0,7

mm. Anthers exserted at anthesis, 1,8 mm long. Fe-

male inflorescence like the male. Sepals cartilagi-

nous, acute, deeply concave, glabrous, 3,8 x 1 mm.

Petals membranous to chartaceous, acute, 3 x 0,6

mm. Ovary with 1 fertile locule. Staminodes present.

Styles three, free, widely separated at the base. Fruit

a capsule; seed 1,3 mm long, white, ends rounded,

surface roughly colliculate.

Restio colliculospermus occurs in the Riverson-

derend and Langeberg Mountains, at about 1 350 m.

It is only known from the steep south-facing cliffs,

and forms a low dense undergrowth between the

bushes.

Superficially this species is very similar to R. fragi-

lis by its habit, the single-flowered spikelets, the uni-

locular ovary and the glabrous lateral sepals. But the

spathes are shorter than the spikelets and the seed is

colliculate. This latter character is very curious and

probably needs verification. A further character that

separates R. fragilis and R. colliculospermus is the

insertion of the styles on the ovary — style bases are

fused in R. fragilis, but are widely separated in R.

colliculospermus.

CAPE. — 3419 (Caledon): Pilaarskop, near Lindeshof,

1 350 m (-BB), Esterhuysen 33341 (BOL; C; K; L; M; MO; E; S;

STE). 3320 (Montagu): Langeberg, Goedgeloof Peak (-CD), Es-

terhuysen 33420 (BOL; K; MO; S).

Restio communis Pillans in Trans. R. Soc. S.

Afr. 16 : 264 (1928); in Adamson & Salter, FI. Cape

Penins. 139 (1950). Type: Cape, 3418 (Simonstown):

Silvermine Valley (-AB), Pillans 4183 9 (BOL,

lecto.!; K!); 4183 cf (BOL!); 4135 cf & $ (BOL!).

Notes

1. Restio communis, in the strict sense, is only

known from the Cape Peninsula, where it occurs oc-

casionally in marshes.

2. There are numerous collections of a species

very close to R. communis, but distinct by having

fewer flowers per spikelet and many more spikelets.

This occurs in the mountains from Riversdale to Ca-

ledon. The decision on the status of this taxon is de-

pendent on a better knowledge of the variation

within the Cape Peninsula populations of R. com-

munis.

Restio confusus Pillans in J1 S. Afr. Bot. 18 : 103

(1952). Type: Cape, 3319 (Worcester): Waaihoek

(-CB), Esterhuysen 8369 9 (BOL, lecto.!); 8369 cf

(BOL!).

Note

1. R. confusus occurs in swampy areas in the

mountains from Clanwilliam to Caledon, but has not

yet been recorded from the Cape Peninsula.

Restio corneolus Esterhuysen, sp. nov., R. saro-

clado Mast, affinis, a qua differt plantis caespitosis et

bracteis aristatis.

TYPE. — Cape, 3318 (Cape Town): Jonkers-

hoek, Victoria Peak, c. 1 500 m (-DD), Esterhuysen

33499 9 (BOL, holo.!; C; E; K; L; M; MO; S; STE).

Plants caespitose, tussocks 15-50 cm tall, culm

bases aggregated. Culms solid, terete, rather

sparsely branched, about 5 mm in diam., surface ru-

gulose to very finely tuberculate. Sheaths closely

convoluted, 1,5-3 cm long, body brown, coriaceous,

acute, the apex extended into a terete awn 5-3 of the

total length of the sheath, upper margins of the

sheaths very narrowly submembranous. Male inflor-

escence of 1-8 racemose, rarely racemose-panicu-

late, 1,5-2 cm long, many-flowered spikelets.

Spathes like the sheaths, 1,5 cm long, with about 5

mm long awns. Bracts more or less acute, the apex

extended into an aciculate awn, 6-12 mm long, con-

cave, cartilaginous with a narrow submembranous

margin, margin reddish, body straw-coloured, all

bracts fertile. Flowers on short, pilose pedicels, per-

ianth about 4 mm long. Sepals cartilaginous; lateral

sepals conduplicate, densely pilose along the carina,

mucronate; odd sepal 4x1 mm, subacute, glabrous,

flat. Petals membranous, 3,5 x 0,8 mm, subacute.

Anthers exserted at anthesis, about 2 mm long. Fe-

male inflorescence similar to the male, but bracts less

spreading, basal bract sterile. Flowers subsessile,

perianth about 4 mm long. Sepals cartilaginous, sub-

444

Bothalia 15, 3 & 4 (1985)

acute to rounded; lateral sepals carinate, unequal,

one much more villous and often taller than the

other one; odd sepal about 4x1 mm, almost flat,

glabrous. Petals chartaceous, rounded, 3,5 x 1 mm.

Staminodes 0,5 mm long. Ovary with one fertile lo-

cule. Styles three, free, very slender, 3-4 mm long,

sparsely villous, adjacent at the base. Fruit a cap-

sule, seed 1,6 x 0,9 mm, the ends truncate, the sides

somewhat compressed, ridge present as a fine

groove, surface smooth, nitid.

Restio corneolus occurs in the Hottentots Holland

Mountains, from Villiersdorp to the Betty’s Bay

area. The altitude range of the species is from

600-1 500 m. Most of the collections are from moist

sandy localities, though some collections, especially

from higher altitudes, may be from drier habitats.

Restio corneolus is very closely related to R. saro-

cladus Mast., with the same floral and vegetative

structures, even with the same type of oblique fe-

male flowers. However, in R. sarocladus the plants

have spreading rhizomes (as are found in R. vallis-

simius ) and the floral bracts are acute, but not

awned. R. sarocladus has a wider distribution range,

extending from the Cape Peninsula to Hermanus

and to Paarl.

The anatomy of this species is discussed by Cutler

(1969: 270) under 'Restio saroclados Mast.’. See also

Linder, 1984: 41-42 where it is included under

Group D. Linder (1984: 30) placed the seed type

into group lib.

CAPE. — 3318 (Cape Town): Jonkershoek, 1st and 2nd Ridge

Peaks, 1 350 m (-DD), Esterhuysen 34950 (BOL; K; L; M; MO;

S); Jonkershoek, Victoria Peak, c. 1 500 m (-DD), Esterhuysen

33499 (BOL; C; E; K; L; MO; S; STE). 3319 (Worcester): Paar-

dekop, between Fransch Hoek and Villiersdorp, 1 500 m (-CC),

Esterhuysen 35428 (BOL; K; M; MO; S); between the top of

Fransch Floek Pass and Paardekop, 900 m (-CC), Esterhuysen

25824 (BOL; K). 3418 (Simonstown): Jonkershoek, 'Panorama

Path' below Triplets, c. 900 m (-BB), Esterhuysen 33138 (BOL;

K; MO; S); Elgin area. Somersfontein (-BB), Esterhuysen 32746

(BOL; E; K; L; M; MO; S); Betty's Bay, base of Platteberg

(-BD), Esterhuysen s.n. (BOL; K); Louwsbos, between Kogel-

berg Peak and Rooiels (-BD), Esterhuysen 29012 (BOL; K; L; M;

MO; S). 3419 (Caledon): Nuweberg Forest Reserve (-AA), Es-

terhuysen 33518 (BOL; K; MO; S); Esterhuysen 31036 (BOL; C;

E; F; K; L; LD; M; MO; NBG; PRE; S; STE; UC; US; W); east

slopes of Landdrost Kop (-AA), Stokoe 2482 (BOL; K).

Restio cymosus (Mast.) Pillans in Ann. Bolus

Herb. 3 : 85 (1921); in Trans. R. Soc. S. Afr. 16 : 242

(1928). Syntypes: Cape, 3219 (Wuppertal): Cold

Bokkeveld, 1 200 m (-C), Schlechter 8865 cf (B,

lecto.!; BM!; BR!; K!; MO!; P!; S!; Z!); 8866 $

(BM!; BR!; K!; MO!; P!; S!; Z!); Ezelbank (-AC),

Drege 6 cf (B!; P!); without precise locality,

Schlechter 8864 9 (?)•

Leptocarpus cymosus Mast, in Bot. Jb. 29 Beibl. 66 : 10(1900).

Notes

1 . R. cymosus is very close to R. brachiatus and

they may be conspecific. They are parapatric and

can be distinguished by the number of flowers per

spikelet.

2. R. cymosus occurs in the arid inland mountains

at the edge of the Cape Floral Region from the Ka-

miesberg in Namaqualand to the Anysberg at Swel-

lendam.

Restio debilis Nees in Linnaea 5: 641 (1830);

Kunth, Enum. PI. 3: 412 (1841); Mast, in A. DC.,

Monogr. Phan. 1: 290 (1878); in FI. Cap. 7: 96

(1897); Pillans in Trans. R. Soc. S. Afr. 16: 236

(1928). Type: Cape, ? 3419 (Caledon): Knoblauch

(-B), Zeyher s.n. cf (B, lecto.!; BOL!; K!).

Restio suhulatus Mast, in J. Linn. Soc., Bot. 8: 248 (1865); in

A. DC.. Monogr. Phan. 1: 281 (1878); in FI. Cap. 7: 87 (1897).

Restio debilis Nees var. subulatus (Mast.) Pillans in Trans. R. Soc.

S. Afr. 16: 237 (1928). Type: Cape, 3419 (Caledon): Swartberg,

and around baths, 300-600 m, July (-AB), Ecklon & Zeyher s.n.

? (?)■

Hypolaena virgata Mast, in J. Linn. Soc., Bot. 10: 268 (1868);

in A.DC., Monogr. Phan. 1: 374 (1878); in FI. Cap. 7: 134 (1897).

Calorophorus virgatus (Mast.) Kuntze, Rev. Gen. 747 (1891).

Type: Cape, 3419 (Caledon): Baviaanskloof Mountains at Ge-

nadendal (-BA). Burchell 7817 cf (K, holo.l; BOL!).

Hypolaena browniana Mast, in FI. Cap. 7: 132 (1897). Restio

brownianus (Mast.) Pillans in Ann. Bolus Herb. 3: 85 (1921).

Mastersiella browniana (Mast.) Gilg-Ben. in Pflanzenfam. edn 2,

15a: 25 (1930). Type: Cape, 3419 (Caledon): Houw Hoek, 300 m

(-AA). MacOwan 1727 cf (K, holo.l; Bl; BOL!; P!; Z\).

Hypolaena subtilis Mast, in Bot. Jb. 29 Beibl. 66: 15 (1900).

Type: Cape, 3419 (Caledon): Houw Hoek, 750 m (-AA),

Schlechter 7553 cf (Z, lecto.!; BM!: BOL!; BR!; MO!; S!).

Notes

1. I have not succeeded in locating the type of R.

subulatus and am including the name here on the ba-

sis of the description.

2. Pillans (1928) distinguished his var. subulatus

on the basis of having larger flowers than the typical

variety. I have not yet studied the variation within R.

debilis in order to assess this variety critically. Pillans

included as synonyms under var. subulatus Hypo-

laena virgata and H. browniana.

3. Although Schlechter 7553 (the type of H. subti-

lis) includes both male and female material. Masters

only described the male and commented ‘femina la-

tet’.

4. Masters (1868, 1878) under H. virgata refers to

Burchell 7187. This is probably an error — the anno-

tated material in both BOL and K is Burchell 7817.

5. R. debilis occurs in the mountains between Ca-

ledon and Ceres, with the majority of the collections

from the mountains between Houw Hoek and Sir

Lowry's Pass.

Restio decipiens (N.E. Br.) Linder, comb. nov.

Hypolaena decipiens N.E. Br. in FI. Cap. 7 : 755 (1900); Pillans

in Trans. R. Soc. S. Afr. 16 : 398 (1928). Type: Cape, 3320 (Mon-

tagu): Zuurbraak Mountains, 1 200 m (-DC), Galpin 4792 cf (K,

holo.!; BOL!).

Note

1 . R. decipiens is known from the steep south-fac-

ing slopes of the Langeberg between Swellendam

and Riversdale.

Restio degenerans Pillans in Trans. R. Soc. S.

Afr. 30 : 246 (1945). Type: Cape, 3418 (Simons-

town): Somerset Sneeukop (-BB), Stokoe 5028b $

(BOL, lecto.!; K!); 5028b cf (BOL!).

Leptocarpus intermedins Pillans in Trans. R. Soc. S. Afr. 29 :

346 (1942). Type: Cape, 3319 (Worcester): Bailey's Peak, 1 200 m

(-CA), Esterhuysen 1621 (BOL, holo.!; K!).

Bothalia 15, 3 & 4 (1985)

445

Notes

1. The specific epithet ‘intermedius’ cannot be

transferred to Restio, as it has been used twice be-

fore in that genus (by Steudel and by Kunth).

2. This species is under R. intermedius (Pillans)

Linder’ in Fairall & Goldblatt (1984).

3. R. degenerans occurs most commonly in the

mountains between Sir Lowry’s Pass and Tulbagh,

with outliers from near Ceres and Hermanus. It usu-

ally grows in steep rocky slopes and cliffs.

Restio dispar Mast, in J. Linn. Soc., Bot. 8 : 246

(1865); in A. DC., Monogr. Phan. 1 : 274 (1878); in

FI. Cap. 7 : 82 (1897); Pillans in Trans. R. Soc. S.

Afr. 16 : 253 (1928); in Adamson & Salter, FI. Cape

Penins. 138 (1950). Type: Cape, 3419 (Caledon):

mountain ridges between Babylons Tower and Cale-

don, 300-600 m, August (-DD), Ecklon & Zeyher

s.n. cT (?).

Hypolaena conspicua Mast, in Bot. Jb. 29 Beibl. 66 : 15 (1900).

Restio conspicuus (Mast.) Pillans in Ann. Bolus Herb. 3 : 85

(1921). Syntypes: Cape, 3419 (Caledon): mountains at Genaden-

dal, 800 m (-BA), Schlechter 10294 9 (B, lecto. ! ; BM!; BR!;

MO!; P!; SI; Zl); 10293 c f (B! ; BM!; BR!; MO!; P!; S!; Z!).

Notes

1. I have not been able to find any type material

of Restio dispar. From the description, the name cer-

tainly fits here.

2. R. dispar is a very distinct species with its long

reddish spathes in the female inflorescences. It oc-

curs in the mountains between Bainskloof, the Cape

Peninsula and Bredasdorp. The most common

habitat appears to be along streams.

Restio distans Pillans in Trans. R. Soc. S. Afr.

30 : 247 (1945). Syntypes: Cape, 3418 (Simonstown):

Landdrost Kop (-BB), Stokoe 4017 9 (BOL, lecto.!;

K!); 4016 cf (BOL!; K!).

Note

1. Restio distans is known only from the moun-

tains around Jonkershoek and in the Nuweberg For-

est Reserve, generally at about 1 000 m.

Restio distichus Rottb., Descriptiones Planta-

rum Rariorum 11 (1772); Descriptionum et Iconum

Rariores 6 (1773); Kunth, Enum. PI. 3 : 409 (1841).

Type: Cape, without precise locality, Konig s.n. cf

(C, holo.!).

Thamnochortus distichus (Rottb.) Mast, in J. Linn. Soc., Bot.

10 : 233 (1868); in A. DC., Monogr. Phan. 1 ; 326 (1878); in FI.

Cap. 7 : 128 (1897). Staberoha disticha (Rottb.) Dur. & Schinz,

Consp. FI. Afr. 5 : 520 (1894). Leptocarpus distichus (Rottb.) Pil-

lans in Trans. R. Soc. S. Afr. 16 : 348 (1928); in Adamson &

Salter, FI. Cape Pensins. 150 (1950).

Restio punctulatus Nees ex Mast, in J. Linn. Soc., Bot. 8 : 242

(1865). Type: Cape, Pampoenkraal [prob. = 3318 (Cape Town);

Durbanville (-DC)], Zeyher 1737 $ (MEL. lecto.!; BOL!; K!;

P!).

Thamnochortus membranaceus Mast. in. Bot. Jb. 29 Beibl. 66 :

11 (1900). Syntypes: Cape, 3319 (Worcester): Skurfteberg near

Gydouw, 1 500 m (-AB), Schlechter 10012 $ (B, lecto.!; BM!;

BOL!; BR!; K!; MO!; P!; S!; Z!); 10011 cf (B!; BM!; BOL!; BR!;

K!; MO!; P!; S!; Z!).

Icon: Rottb., Descriptionum et Iconum Rariorest.2f.5 (1773).

Notes

1. Zeyher 1737 (the type of R. punctulatus) is a

mixed collection, with the male from a different

locality and not type material. Pampoenkraal is

probably an old name for Durbanville.

2. R. distichus ranges from the Great Swartberg

to the Cedarberg and the Cape Peninsula, but is ab-

sent from the Outeniqua and Tsitsikamma Moun-

tains. It grows on dry, often gravelly mountain

slopes between 450 and 1 500 m.

Restio dodii Pillans in Ann. Bolus Herb. 3 : 85

(1921); in Trans. R. Soc. S. Afr. 16 : 251 (1928); in

Adamson & Salter, FI. Cape Penins. 138 (1950).

Nom. nov. for Hypolaena membranacea Mast.

Hypolaena membranacea Mast, in Bot. Jb. 29 Beibl. 66 : 14

(1900). Syntypes: Cape, 3418 (Simonstown): near Cape Point,

200 m (-AD ). Schlechter 7302 cf (B, lecto.!; BM!; BOL!; BR!; K!;

MO!; P!; S!; Z!); Cape Peninsula (-AB/AD), Dod 2551 (BM!;

BOL!; K!); 3293 (BOL!). 3419 (Caledon): hills near Rhenoster-

kop, 30 m (-DD), Schlechter 10572 cf (B!; BM!; BOL!; BR!; K!;

MO!; P!; S!; Z!); 10573 9 (BM!; BOL!; BR!; K!; MO!; P!; S!;

Z!).

Restio dodii Pillans var. purpureus Pillans in Trans. R. Soc. S.

Afr. 16 : 252 (1928). Syntypes: Schlechter 10573 9 (BOL, lecto.!):

10572 cf . For details of locality and distribution, see above.

Notes

1. There is already a Restio membranaceus (of

Nees, = Elegia intermedius) , hence the nom. nov.

2. R. dodii var. purpureus is based on part of the

type of R. dodii. It is quite distinct and future study

will show at which rank this segregate should be re-

cognized. It can be distinguished by the reddish-pur-

ple, rather than straw-coloured, bracts.

3. Although the wings on the perianth are un-

equal, this inequality is not very pronounced.

4. The typical variety is only known from the

Cape Peninsula, whereas var. purpureus occurs on

the coastal flats near Bredasdorp. Both forms occur

near sea level, on marine derived or limestone soils.

Restio echinatus Kunth, Enum. PI. 3 : 384

(1841); Pillans, Trans. R. Soc. S. Afr. 16 ; 295

(1928). Type; Cape, 3319 (Worcester): Du Toits

Kloof, 900-1 200 m (-CA), Drege 49 cf (B, holo.!;

BM!; BOL!; K!; MO!; NY!; OXF!; P!).

Notes

1. Masters (1865, 1878, 1897) included this name

under Restio squarrosus Poir. (see Pillans 1928 :

295).

2. R. echinatus occurs in the mountains from

Bainskloof to Houw Hoek and Kogelberg and grows

over a wide altitudinal range.

Restio egregius Hochst. in Flora, Jena 28: 337

(1845); Mast, in A. DC., Monogr. Phan. 1 : 288

(1878); in FI. Cap. 7 : 92 (1897); Pillans in Trans. R.

Soc. S. Afr. 16 : 254 (1928); in Adamson & Salter,

FI. Cape Penins. 138 (1950). Syntypes: Cape, with-

out precise locality, Ludwig s.n. (OXF, lecto.!);

Drege s.n. (?).

Restio egregius Hochst. var. nutans Mast, in J. Linn. Soc., Bot.

8 : 245 (1865); in A. DC., Monogr. Phan. 1 : 288 (1878); in FI.

Cap. 7 : 92 (1897). Syntypes: Cape, 3418 (Simonstown): False

Bay (-AB), Robertson s.n. 9 (BM, lecto.!); s.n. cf (BM!). 3318

(Cape Town): Table Bay (-CD), Robertson s.n. (?); without pre-

cise locality, Grey s.n. (P!) ; Zeyher s.n. (?).

446

Bothalia 15, 3 & 4 (1985)

Notes

1. I have not been able to locate any annotated

type specimens of R. egregius. It is possible that the

Ludwig specimen at OXF is either an isotype, or was

seen by Hochstetter, so I designate that as a lecto-

type.

2. The var. nutans of Masters does not seem to be

distinct in any way from the typical form.

3. R. egregius is a distinctive species that occurs in

the mountains and in hilly places from Bredasdorp

to Villiersdorp and the Cape Peninsula. It is most

common below 1 000 m.

Restio ejuncidus Mast, in Bot. Jb. 29 Beibl. 66 :

2 (1900). Type: Cape, 3419 (Caledon): mountains at

Genadendal, 1 200 m (-BA), Schlechter 10321 9 (B,

lecto.!; B!; BM!; BOL!; BR!; MO!; P!; S!; Z!).

Leptocarpus ejuncidus (Mast.) Pillans in Trans. R. Soc. S. Afr.

16 : 348 (1928).

Notes

1. There is some male material under Schlechter

10321, but this does not constitute type material.

2. The female flowers are often shed enclosing

the ripe fruit. Although there are dehiscence lines

visible on the ovary, the fruit generally seems to be

indehiscent.

3. The distribution is from Porterville to Ge-

nadendal and the Langeberg at Tradouw Peak.

Restio exilis Mast, in Bot. Jb. 29 Beibl. 66 : 6

(1900); Pillans in Trans. R. Soc. S. Afr. 16 : 261

(1928). Syntypes: Cape, 3419 (Caledon): mountains

at Genadendal, 800 m (-BA), Schlechter 10291 9

(B, lecto.!; BM!; BOL!; BR!; K!; P!; S!; Z!); 10290

Cf (B; BM!; BOL!; BR!; K!; S!; Z!).

Notes

1. This species is very close to R. bifidus, from

which it differs in flower size, bract coloration and

bract shape. A good decision on the relationship be-

tween R. exilis and R. bifidus is dependent on a bet-

ter knowledge of the former. On present knowledge,

the two taxa are allopatric.

2. R. exilis is only known from the type collection.

Restio festuciformis Nees ex Mast, in J. Linn.

Soc., Bot. 8 : 248 (1865); in A. DC., Monogr. Phan.

1 : 281 (1878); in FI. Cap. 7 : 87 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 251 (1928), as festucaefor-

mis. Type: Cape, 3419 (Caledon): Grietjiesgat, May

(-AA), Ecklon & Zeyhers.n. 9 (BOL, lecto.!); s.n.

Cf (?)•

Leptocarpus parkeri Pillans in J1 S. Afr. Bot. 18 : 109 (1952).

Syntypes: Cape, 3418 (Simonstown): Pringle Bay (-BD), Parker

4522 9 (BOL, lecto.!; K!); 4523 cf (BOL!; Kl).

Notes

1. There is a Zeyher collection at B inscribed by

Nees as ‘Restio ischaemoides mihi Caledon Zwart-

berg’. Masters (1865) mentions R. ischaemoides

Nees as a nom. nud. A collection from the same

locality, but without inscription, is at Kew, and there

are two collections from MEL labelled ‘ Restio festu-

caeformis Nees’ by Sonder, also from Caledon

Swartberg, (see Fig. 4) and one from locality 59 (W

slopes of the Hottentots Holland Mountains). Con-

sequently, the only collection from the correct local-

ity is at BOL.

2. Pillans (1928) followed Masters (1865) in re-

porting the ovary to be bilocular and dehiscent.

However, all the material which I have investigated

is unilocular. Although the ovary has dehiscence

lines, it appears as if dehiscence only occurs occa-

sionally and usually the ovary is shed with the

flower.

3. R. festuciformis occurs in the hills and lower

mountains between Bredasdorp and Somerset West,

often growing along streamlines and in damp places.

Restio filicaulis Pillans in Trans. R. Soc. S. Afr.

30 : 249 (1945). Type: Cape, 3319 (Worcester):

Ceres, Slab Peak (-AD), Esterhuysen 6194 9 (BOL,

lecto.!; K!); 6194 cf (BOL!: K!).

Note

1. Restio filicaulis is only known from the type

and may be a depauperate form of R. pedicellatus or

R. aureolus.

Restio filiformis Poir. in Lam., Encycl. 6: 173

(1804); Mast, in FI. Cap. 7: 86 (1897); Pillans in

Trans. R. Soc. S. Afr. 16: 243 (1928); in Adamson &

Salter, FI. Cape Penins. 137 (1950). Type: Cape,

without precise locality, s.l. in herb. Poir. (P,

holo.!), See Fig 1.

Restio monostachyus Steud. in Flora, Jena 12: 133 (1829).

Type; Cape, 3318 (Cape Town): ‘In praeruptis montis tabularis,

Nov.’ (-CD), Ecklon 840 cf (MO, lecto.!).

Restio oligostachyus Kunth, Enum. PI. 3: 399 (1841). Type:

Cape, 3323 (Willowmore): Welgelegen (-CD), Drege 37 9 (K,

lecto.!; MO!; OXF!).

Restio garnotianus Kunth, Enum. PI. 3: 392 (1841); Mast, in

A. DC., Monogr. Phan. 1: 280 (1878). Syntypes: Cape, 3219

(Wuppertal): Ezelbank, 900-1 200 m (-AC), Drege 2473 cf (K,

lecto.!; OXF!; MO!; NY!; P!). 3318 (Worcester): Du Toits Kloof,

300-1 200 m (-CA), Drege 1628 cf (K!; MO!; NY!).

Craspedolepis verreauxii Steud., Syn. PI. Glum. 2: 264 (1855).

Type: Cape, without precise locality, Verreaux s.n. (P, holo.!;

P!).

Restio multicurvus N.E. Br. in FI. Cap. 7: 751 (1900). Type:

Cape, 3321 (Riversdale): mountains at Garcias Pass (-CC), Gal-

pin 4785 cf (K, holo.!; BOL!).

Restio garnotianus Kunth var. oligostachyus Mast, in A. DC.,

Monogr. Phan. 1: 281 (1878). Restio filiformis Poir. var. oligosta-

chyus (Mast.) Mast, in FI. Cap. 7: 86 (1897). Syntypes: Cape,

3418 (Simonstown): False Bay (-AB) Robertson s.n. 9 (?)■ 3318

(Cape Town): Table Mountain (-CD), Ecklon s.n. 9 (?)• 3323

(Willowmore): Welgelegen (-CD), Drege 37 9 (K, lecto.!;

OXF!); 1624 9 (NY!). 3418 (Simonstown): Simonstown (-AB),

Wright s.n. 9 (?)•

Restio garnotianus Kunth var. monostachyus Steud. ex Mast, in

A. DC., Monogr, Phan. 1: 281 (1878). Restio filiformis Poir.

var. monostachyus (Mast.) Mast, in FI. Cap. 7: 86 (1897). Syn-

types: Cape. 3318 (Cape Town): Table Mountain (-CD), Ecklon

840 cf (K, lecto.!; MO!;); 846 (MO!).

Notes

1. Drege 37 is the type of three taxa: Restio spinu-

losus ( = Platycaulos callistachyus), Restio oligosta-

chyus and R. filiformis var. oligostachyus Mast.

Bothalia 15, 3 & 4 (1985)

447

2. There are iwo sheets of Verreaux s.n. in Paris.

The sheet in herb. Steud. is selected as the holotype

of Craspedolepis verreauxii.

3. Although the two varieties which Masters de-

scribed in 1878 have the same names as earlier

species, Masters does not refer to these species.

Consequently, the varieties are regarded as new

names and not as new combinations.

4. Restio filiformis is a very widespread and com-

mon species, occurring from Clanwilliam to the

Cape Peninsula and to Riversdale. Although it is dif-

ficult to distinguish it in the key, the only species

with which it may be confused is R. inveteratus.

Restio fourcadei Pit Ians in Trans. R. Soc. S.

Afr. 16: 232 (1928). Type: Cape, 3324 (Steytler-

ville): mountain slopes above forest at Witte Els

Bosch (-CC) Fourcade 1221 $ (BOL, lecto.!; K!);

1221 cf (BOL!; K!).

Note

1. R. fourcadei occurs in the southern Cape Prov-

ince between George and Humansdorp. It often

grows in half-shade along streams. Specimens have

often been confused with R. scaberulus.

Restio fragilis Esterhuysen, sp. nov., a R. deci-

pienti (N.E. Br.) Linder, culmis multo tenuioribus

differt.

TYPE. — Cape, 3320 (Montagu): Langeberg,

along the 10 o’clock Peak path, 900 m (-CD), Ester-

huysen 33242 $ (BOL, holo.!; B; C; E; F; GRA; K;

L; LD; M; MO; NBG; NY; PRE; RSA; S; STE;

TCD; UC; US; W; WAG).

Plants caespitose or tangled, 20-40 cm tall, culm

bases aggregated. Culms slender, solid, terete,

branching, up to 1 mm in diam., rugulose. Sheaths

4-10 mm long, coriaceous portion green to brown,

acute with the apex extended into a terete awn of

variable length, flanked by membranous shoulders

which vary from shorter to taller than the awn. Male

inflorescence of 3-10 racemose, 3 mm long, single-

flowered spikelets. Spathe 4—5 x 1 mm, chartaceous,

acute, the body pale brown, the apex submembra-

nous and decaying, overtopping and enclosing the

spikelets. Bracts 2-3 mm long, chartaceous to mem-

branous, decaying. Flowers 3-3,5 mm long. Sepals

chartaceous, acute, 3 x 0,6 mm, concave, glabrous,

subequal, apices membranous, decaying. Petals

membranous, acute, 2 x 0,3 mm. Anthers exserted

at anthesis, 1,3 mm long. Female inflorescence of

1-2(3) single-flowered, 4-5 mm long, racemose

spikelets. Spathe chartaceous, submembranous in

the upper half, about 4 mm long, acute, decaying,

shorter than the spikelet. Bracts chartaceous, acute,

3 mm long. Flowers subsessile, 3,5 mm long. Sepals

chartaceous, somewhat cartilaginous along the mid-

rib, acute, 3,5 x 0,6 mm, deeply concave with the

laterals subcarinate. Petals subchartaceous to mem-

branous, acute, 2,5 x 0,5 mm. Staminodes 0,5 mm.

Ovary shortly sessile, with a single fertile locule.

Styles three, about 3 mm long, fused at the base into

a short, hollow stylopodium. Fruit a capsule, seed

1,2 x 0,8 mm, obtuse at both ends, with two conca-

vities on the flanks, nitid. Fig. 18.

Restio fragilis is known from the Langeberg be-

hind Swellendam, where it occurs in damp seepage

patches and in marshy places on the upper south-fac-

ing slopes, at an altitude of 900 m. Flowering occurs

in September.

This species belongs to the Restio decipiens com-

plex, which is delimited by the single-flowered male

and female spikelets, the spathes overtopping the

flowers and the smooth shiny seeds with two conca-

vities (but see Linder, 1984: 30). It differs from R.

decipiens in the much more slender culms, the pos-

sibly smaller flowers, and possibly the fused style-

bases and female perianth taller than the spathe.

CAPE. — 3320 (Montagu): Langeberg, along the 10 o’clock

Peak path, 900 m (-CD), Esterhuysen 33242 (B; BOL; C; E; F;

GRA; K; L; LD; M; MO; NBG; NY; PRE; RSA; S; STE; TCD;

UC; US; W; WAG); Esterhuysen 33476 (BOL; K; M; MO; S).

Restio fusiformis Pillans in J1 S. Afr. Bot. 18:

104 (1952). Type: Cape, 3418 (Simonstown): Kogel-

berg area (-BB), Stokoe 2595a $ (BOL, holo.!).

Note

1. R. fusiformis is known from the mountains be-

tween Cape Hangklip and Jonkershoek (Caledon

division). It occurs in marshy habitats, frequently

along streams, between 450 and 900 m.

Restio galpinii Pillans in Trans. R. Soc. S. Afr.

30: 250 (1945). Type: Natal, 2929 (Underberg): be-

tween Cathkin Peak and Giants Castle, 2 100-2 700

m (-AB/AD), Esterhuysen 8820 9 (BOL, lecto.!;

K!; NY!); 8820 cf (BOL!; K!; NY!).

Note

1. R. galpinii occurs in the Drakensberg between

Maclear and Bergville, between 2 100 and 2 700 m,

in steep grassland, frequently on basalt.

Restio harveyi Mast, in J. Linn. Soc., Bot. 8:

253 (1865); in A.DC., Monogr. Phan. 1: 287 (1878);

in FI. Cap. 7: 91 (1897); Pillans in Trans. R. Soc. S.

Afr. 16: 229 (1928); in Adamson & Salter, FI. Cape

Penins. 137 (1950). Type: Cape, 3318 (Cape Town):

near Cape Town (-CD), Harvey s.n. $ (TCD, holo.;

MEL!).

Icon: J. Linn. Soc., Bot. 8: 1. 15 (1865).

Notes

1. There is a drawing of the type at K! This indi-

cates the lateral sepals as being pilose on the carina,

a character which is absent in most specimens attri-

buted to this species.

2. R. harveyi occurs on shaly soil from the Cape

Peninsula to Stellenbosch and Bredasdorp, usually

near sea level.

Restio implicatus Esterhuysen, sp. nov., a R. ar-

cuato Mast, spiculis femineis unifloris, culmis rectis,

a R. intermedio (Pillans) Linder spiculis femineis

unifloris, minoribus (3-4 mm) recedit.

Bothalia 15, 3 & 4 (1985)

FIG. 18. — Restio fragilis Esterhuysen. a, habit; b, female plant, showing the roots borne on aggregated root

bases, x 0,8; c, detail of sheath, showing the rugulose culm and wide membranous margins on the sheath,

x 4; d, male culms, x 0,8; e, female spikelet with the bract shorter than the perianth, x 8; f, female

flower, note dehisced and curved capsule, x 8; g, ovary with styles and staminodes, note oblique body of

the ovary indicating a unilocular condition, x 8; h, male inflorescence, x 4; i, male flowers with exserted

anthers, x 8. (From Esterhuysen 33242.)

Bothalia 15, 3 & 4 (1985)

449

TYPE. — Cape, 3320 (Montagu): Langeberg,

summit of Lemoenshoek Peak (Grootberg) (-DD),

Esterhuysen 30880 $ (BOL, holo.!; B; BOL; C; E;

F; GRA; K; L; M; MO; NBG; NY; RSA; S; STE;

TCD; UC; US; W; WAG).

Plants tangled. Culms solid, very slender, terete,

branching, to 0,5 mm in diam., rugulose, sometimes

forming stolons and rooting at the nodes. Sheaths

rather loosely convoluted, 3-10 mm long, body co-

riaceous, brown, ridged, densely and finely tubercu-

late, acute with the apex extended into a short cla-

vate awn, membranous shoulders taller than the

awn. Male inflorescence of 1-3 very lax, 2-3-flow-

ered, 3^1 mm in diam. spikelets. Spathes similar to

the sheaths, 3,5 mm long, shorter than the spikelets.

Bracts 3x2 mm, acute, chartaceous with a membra-

nous margin, nearly as tall as the flowers. Flowers

very shortly pedicellate, perianth 2,5-3 mm long. Se-

pals subcartilaginous, acute 3 x 0,9 mm, concave,

papillate in the upper half, lateral sepals subcari-

nate. Petals membranous, rounded, 2-2,5 x 0,8-1

mm. Anthers exserted at anthesis, 1,8 mm long. Fe-

male inflorescence like the male, spikelets with three

bracts. Flower solitary, taller than the bracts, shortly

pedicellate, perianth 3,2 mm long. Sepals subcartila-

ginous with membranous apical margins, subacute;

lateral sepals conduplicate, especially in the apical

half; odd sepal shallowly concave, 3 x 1 mm. Petals

membranous, rounded, 2,6 x 1 mm. Staminodes 0,5

mm long. Ovary with a single fertile locule. Styles

three, free to the base, adjacent, densely villous.

Fruit a capsule, seed 1,3 x 1 mm, round in cross-sec-

tion, surface smooth, nitid.

Restio implicatus is known from two collections

from the summit of Lemoenshoek Peak (Grootberg)

in the Langeberg, at an altitude of 1 500-1 600 m. It

forms solid compact extensive cushions, which are

probably eaten by rock hyraxes. It occurs on the

steep south-facing slopes below the peak.

This species is clearly related to R. intermedius

(Pillans) Linder and R. arcuatus Mast, in the matted

habit, the loosely convoluted sheaths with hyaline

apices, the glabrous flowers and the unilocular

ovaries. The seed is similar to R. arcuatus (see

Linder, 1984: 30). It can be distinguished by the

much smaller single-flowered female spikelets. The

very distinctive clavate awns of R. implicatus also oc-

cur in R. arcuatus, but the culms of R. implicatus are

always straight.

CAPE. — 3320 (Montagu): Langeberg, steep upper south-fac-

ing slopes at Lemoenshoek Peak, 1 500-1 600 m (-DD), Ester-

huysen 30880 (B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG;

NY; RSA; S; STE; TCD; UC; US; W; WAG); Esterhuysen 29493

(BOL; C; E; F; K; L; LD; M; MO; S; STE; UC; US; W).

Restio inconspicuus Esterhuysen , sp. nov., a R.

triticeo Rottb. culmis laevibus, spathis spiculorum

elatioribus differt.

TYPE. — 3320 (Montagu): Lemoenshoek Peak

(Grootberg), north slopes, 1 400 m (-DD), Ester-

huysen 29496 $ (BOL, holo. !; E; K; L; M; MO; S).

Plants with diffuse, isolated shoots ascending from

spreading, long rhizomes. Culms solid, terete,

branching, to 2,5 mm in diam., surface smooth,

shiny, grading smoothly into the long spreading rhi-

zomes with imbricate sheaths. Sheaths closely convo-

luted, 1-2 cm long, acuminate to awned, body

brown, coriaceous, with a narrow membranous up-

per margin. Male inflorescence of 4—8 subimbricate,

racemose, 8-12 mm long spikelets. Spathes like the

sheaths, 1-2 cm long, overtopping the spikelets.

Bracts imbricate, acute, shortly piliferous, cartilagi-

nous with a submembranous margin, 5-8 mm long,

obscuring the flowers. Flowers 3-10 per spikelet,

shortly pedicellate, perianth 3-4 mm long. Sepals

cartilaginous; lateral sepals conduplicate, the upper

half of the carina pilose to villose, acute; odd sepal

rounded to subacute, glabrous, almost flat, 3-4 x 1

mm. Petals membranous to papyraceous, acute,

fused at the base with the bases of the filaments, 3 x

1 mm. Anthers exserted at anthesis, 1,4 mm long.

Pistillode minute. Female inflorescence similar to the

male, but with the spikelets usually fewer and larger.

Flowers 2-A per spikelet, pedicels about 1 mm long,

pilose, perianths 4-5 mm long. Sepals cartilaginous,

acute, 4,5 x 1,2 mm; lateral sepals deeply concave

to conduplicate, the upper half of the carina villous;

odd sepal more or less flat, glabrous. Petals cartilagi-

nous with membranous margins, acute, 3,5 x 1,2

mm. Staminodes about 0,7 mm long. Ovary with one

fertile locule. Styles three, free, from three widely

separated horns on the ovary, about 3 mm long,

densely villous. Fruit a capsule, seed 1,6 mm long,

triangular, without a ridge, rounded to truncate,

roughly colliculate.

Restio inconspicuus occurs on the Langeberg and

the Outeniqua Mountains between Swellendam and

Knysna. The altitudinal range is from 600 m to 1 400

m. It appears to occur scattered in the vegetation, at

lower altitudes on the cool wet south-facing slopes,

at higher altitudes on the hot dry north-facing

slopes. The habit of the plants makes them quite in-

conspicuous in the vegetation. Flowering occurs in

autumn, in April.

This species is allied to Restio triticeus Rottb., al-

though it lacks the dense tubercles on the culms that

are so typical of the R. triticeus group. The seed mor-

phology is the same (see Linder, 1984: 29), the floral

bracts are acute, cartilaginous, making the spikelets

look very similar, and the plants have a similar over-

all aspect. It is very distinct within the group with its

spreading rhizome (which also occurs in R. vallis-

simius Linder) and its shiny-smooth culms.

CAPE. — 3320 (Montagu): Langeberg, Goedgeloof Peak

(-CD), Esterhuysen 35902 (B; BOL; C; E; F; GRA; K; L; LD;

M; MO; NBG; NY; PRE; RSA; S; STE; TCD; UC; US; W;

WAG); Langeberg, Lemoenshoek Peak, 1 400 m (-DD), Ester-

huysen 29496 (BOL; E; K; L; M; MO; S); Langeberg,

Lemoenshoek Peak, 600-900 m, Esterhuysen 29497 (BOL; C; E;

K; L; M; MO; NBG; S). 3321 (Ladismith): Riversdale, slopes

between Crystal Kloof and Mozambique Kop, 600-1 050 m

(-CC), Esterhuysen 31766 (BOL; C; E; F; K; L; LD; M; MO; S;

STE). 3322 (Oudtshoorn): Cradocksberg (-CD), Esterhuysen

30139 (BOL; C; E; F; K; L; LD; M; MO; NBG; PRE; S; STE;

TCD; UC; W; WAG). 3323 (Knysna): mountain slopes, Letter-

ing area (-DD), Esterhuysen 35917 (BOL; K).

Restio ingens Esterhuysen, sp. nov., a R. com-

muni Pillans, plantis unispiculatis, validioribus, aris-

tis vaginarum tenuissimis differt.

TYPE. — 3419 (Caledon): Riversonderend

Mountains, Kanonkop, ledges below the summit to

the south side (-BA), Esterhuysen 31473 $ (BOL,

450

Bothalia 15, 3 & 4 (1985)

holo.; B; C; E; F; GRA; K; L; LD; M; MO; NBG;

NY; PRE; RSA; S; STE; TCD; W; WAG; UC;

US).

Plants caespitose or rarely tangled, to 1 m tall.

Culms solid, terete, branching, to 6 mm in diam. in

the middle, surfaces densely tuberculate. Sheaths

more or less closely convoluted, 0,8-2, 5 cm long,

body coriaceous, brown, obtuse, with a cylindrical,

1-5 (10) mm long awn, flanked by two membranous,

acute shoulders about as tall as the awn. Male inflo-

rescense of solitary, 1-3 x 0,7 cm, many-flowered

spikelets. Spathe like the sheaths, 12 mm long.

Bracts acute, 10-15 x 3 mm, chartaceous in the

lower half, membranous in the upper half, all fertile.

Flowers on 1 mm long pedicels, perianth about 4 mm

long. Sepals subacute to acute, cartilaginous, 3,5-4

x 1-1,3 mm; lateral sepals conduplicate, sparsely

villous to pilose on the carina; odd sepal flat, gla-

brous. Petals membranous, acute, 4x1 mm. Anth-

ers exserted at anthesis, 2,5 mm long. Pistillode min-

ute. Female inflorescence similar to the male. Spathe

17 mm long, like the sheaths. Bracts 1-2 cm long,

acute, piliferous, body chartaceous, submembra-

nous in the upper half, imbricate, all fertile. Flowers

several per spikelet, pedicels pilose, about 1 mm

long, perianth 4,5-6 mm long. Sepals cartilaginous,

subacute; lateral sepals conduplicate, pilose along

the carina; odd sepal shallowly concave, glabrous

.4, 5-5, 5 x 1,5 mm. Petals cartilaginous, the margins

membranous, acute, 4,7 x 1,8 mm. Staminodes 1,5

mm long. Ovary with two fertile locules. Styles

three, free, adjacent at the base, about 4 mm long,

villous. Fruits, capsule, seed 1,5 x 0,7 mm, rounded

with two deep concavities, surface smooth, often ni-

tid.

Restio ingens occurs in the Riviersonderend

Mountains between 900 and 1 500 m, mostly on the

steep upper south-facing slopes. Most of the records

are from rocky places: screes, ledges, base of cliffs.

Most of the plants form tall tussocks that stand out

above the vegetation, but some depauperate speci-

mens are said to be spreading underneath the

shrubby vegetation. The flowering time is in Janu-

ary.

R. ingens is related to R. communis Pillans

through the shiny seed (Group lib of Linder, 1984:

30), subchartaceous bracts, the many-flowered spi-

kelets and the tuberculate culms. R. communis is re-

stricted to the Cape Peninsula and always has several

spikelets and very stout awns on the sheaths. The

plants are also smaller. There is a related, as yet un-

named taxon that extends eastwards from the Penin-

sula, but this is distinguished by having numerous

few-flowered spikelets.

CAPE. — 3419 (Caledon): Riviersonderend Mountains, Kanon-

kop, ledges below the summit on the south side (-BA), Esterhuy-

sen (B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY;

PRE; RSA; S; STE; TCD; W; WAG; UC; US). Esterhuysen

35624 (BOL; C; E; K; L; M; MO; S); between Olifantsberg and

Genadendal (-BA), Esterhuysen 30922 (BOL; K; S); Pilaarkop,

south slopes (-BB), Esterhuysen 31395 (BOL; E; F; K; L; LD; M;

MO; NY; S; STE; W; WAG; UC; US); Esterhuysen 33346 (BOL;

C; E; F; K; L; LD; M; MO; NBG; NY; RSA; S; STE; W; WAG;

US; UC); Esterhuysen 32582 (BOL; K; M; MO; S); Esterhuysen

32581( BOL; K; S); near Lindeshof, slopes above the Krom River

(-BB), Esterhuysen 31571 (BOL; C; E; F; K; L; LD; M; MO;

NY; S; STE; UC).

Restio insignis Pillans in Trans. R. Soc. S.

Afr. 30: 251 (1945). Type: Cape, 3219 (Wuppertal):

Cedarberg, Langberg (-AC), Esterhuysen 7346 $

(BOL, lecto.!; K!); 7346 cf (BOL!; K!).

Note

1. I cannot distinguish this species from R. bolusii

and they are probably conspecific. This would ex-

tend the distribution area of R. bolusii into the Ce-

darberg.

Restio inveteratus Esterhuysen , sp. nov., a R.

filiformi Poir., floribus maioribus, plantis validiori-

bus et seminibus sublaevibus, rubellis, recedit.

TYPE. — Cape, 3319 (Worcester): Slanghoek

Needle (-CA), Esterhuysen 33979, (BOL, holo.!;

C; K; L; M; MO; NBG; S; STE).

Plants tangled, more or less caespitose. Culms

branching, up to 3 mm in diameter, terete, erect,

surface smooth or very obscurely tubercled. Sheaths

1-1,5 cm long, closely convoluted, acute with an aci-

culate, 3-4 mm long awn, coriaceous, brown with

fine yellow speckling, the upper margin indistinctly

submembranous. Male inflorescence of 4-20 race-

mose or paniculate-racemose spikelets, spikelets

8-10 X 3-4 mm, elliptical. Spathes and bracts simi-

lar, acute with a short awn, 4x4 mm, the body co-

riaceous, with a narrow hyaline margin and a more

distinct margin of submembranaceous cells. The

lowermost bracts are frequently sterile, the fertile

bracts are often up to 6 mm long and obscure the

flowers. The flowers are on 0,8 mm, slightly villous

pedicels, the perianth is about 4 mm long. Sepals are

cartilaginous, acute, the laterals are conduplicate

and densely villous on the carina, the odd sepals are

shallowly concave, 4 x 0,8 mm, somewhat villous

near the apex. Petals are membranous from a

slightly swollen base, acute, 3 x 0,7 mm. Anthers

exserted at anthesis, 2 mm long, with a small mucro.

Pistillode three-lobed with three free minute stylo-

dia. Female inflorescence of 1-5 racemose, rarely

paniculate-racemose spikelets, spikelets 10-18 x 4—7

mm, narrowly ovate. Bracts 7x4 mm, acute, mu-

cronate, obscuring the flowers, coriaceous with a

narrow, often lacerate margin present. Lower bracts

and spathe shorter, with relatively longer awns, ster-

ile. Flowers 3-8 per spikelet, seated on a 1 mm long,

somewhat villous pedicel, perianth 5,5-6 mm long.

Sepals acute, cartilaginous, laterals conduplicate,

densely villous on the carina, odd sepal shallowly

concave, 5-5,5 x 2 mm, somewhat villous on the

midrib near the apex. Petals chartaceous, similar to

the odd sepal. Staminodes 1,5 mm long. Ovary with

two fertile locules, fruit a capsule, styles free, three.

Seeds trigonous, reddish, 1,8 mm long, surface

faintly rough. Fig. 19.

Restio inveteratus is known from the Slanghoek

and Klein Drakenstein Mountains. It occurs on rock

ledges, screes and along stream banks, almost al-

ways in rocky areas, as large tufts wedged in the

rocks, between 1 000 and 1 650 m. It flowers in Au-

gust and September.

R. inveteratus is related to R. filiformis through

the spikelet structure, but is distinguished by the

larger flowers with the perianth twice the length of

Bothalia 15, 3 & 4 (1985)

451

FIGf 19' ReSti° inveteratus Esterhuysen. a, habit; b, culms, x 0,8; c, female inflorescence, x 0,8; d, female spikelet, x 5; e,

female flowers with densely villous lateral sepal carinas, odd sepal with a small villous patch, x 7; f, dissected female flower,

note staminodes, x 7; g, male inflorescence x 0,8; h, male spikelet, x 5; i, male flower with exserted anthers, x 7; i, dissected

male flower, x 7. (From Esterhuysen 33979.)

452

Bothalia 15, 3 & 4 (1985)

the ovary, the more robust plants and the almost

smooth, reddish seed.

CAPE. — 3319 (Worcester): Slanghoek Needle (-CA), Ester-

huysen 33979 (BOL; C; E; K; L; M; MO; NBG; STE); Hawequas

Mountains (-CA), Esterhuysen 35471 (BOL; C; E; F; K; L; LD;

M; MO; S; STE; UC); Esterhuysen 35472 (BOL; K); Upper Wel-

lington Sneeukop (-CA), Esterhuysen 32800 (BOL; K; MO; S);

Esterhuysen 32801 (BOL; C; E; K; L; M; MO; S; UC); Upper

Witte River Valley (-CA), Esterhuysen 34167a (BOL; K); Hael-

hoek Spitskop (-CC), Esterhuysen 30972 (B; BOL; C; E; F;

GRA; K; L; LD; M; MO; NBG; NY; PRE; RSA; S; STE; UC;

US; W; WAG); Baviaanskloof, near waterfall, 1 000 m (-CA),

Esterhuysen 31084 (BOL; C; E; F; K; L; LD; M; MO; S; STE;

UC); suurvlakte above Du Toits Kloof between Krom River Peak

and Krom River Dome, 1 050 m (-CA), Esterhuysen 31079

(BOL; E; K; L; M; MO); Esterhuysen 31991a (BOL; K; MO; S);

Haelhoek Sneeukop, on cliffs at the head of Donkerkloof (-CC),

Esterhuysen 34160 (BOL, K); Esterhuysen 31428 (BOL; C; E; K;

L; M; MO; S).

Restio involutus Pillans in J1 S. Afr. Bot. 18: 105

(1952). Syntypes: Cape, 3418 (Simonstown): Land-

drost Kop (-BB), Stokoe 2858 £ (BOL, lecto.!; K!);

2844 Cf (BOL!; K!).

Note

1. Restio involutus occurs on Landdrost Kop and

Somerset Sneeukop, on steep marshy slopes at

about 1 200 m.

Restio leptostachyus Kunth, Enum. PI. 407

(1841); Mast in A. DC., Monogr. Phan, 1: 262

(1878); in FI. Cap. 7: 78 (1897); Pillans in Trans. R.

Soc. S. Afr. 16: 265 (1928). Type: Cape, 3319 (Wor-

cester): Fransch Hoek (-CC), Drege 12 0” (B, lecto.!;

BOL!; K!; MO!; OXF!; P!).

Restio pusillus Pillans in Trans. R. Soc. S. Afr. 16; 266 (1928);

in Adamson & Salter, FI. Cape Penins. 140 (1950). Syntypes;

Cape, 3318 (Cape Town): Table Mountain, south slopes of the

upper table (-CD), Pillans 4898 § (BOL, lecto.!); 4898 cf (BOL!;

K!).

Notes

1 . Although Drege 12 includes both male and fe-

male material, Kunth described only the male and

noted ‘Femina deest’.

2. Pillans (1928) regarded the ovary of R. leptosta-

chyus as 2-celled and the fruit as 1-seeded by abor-

tion, while the ovary of R. pusillus is 1-celled. This

distinction is academic.

3. R. leptostachyus occurs on the Cape Peninsula

and on the higher mountains from Du Toits Kloof to

the Riviersonderend Mountains. The habitat gener-

ally appears to be on rocky slopes at higher altitudes.

Restio madagascariensis Cherm. in Bull. Soc.

bot. Fr. 69: 318 (1922); Humbert in FI. Madag. 34: 3

(1946). Syntypes: Central Madagascar, Mt Ibity,

1 800-2 300 m, Perrier de la Bdthie 2735 £ (P,

lecto.!); 2735 cf (P!); between Ambatomainy and

Itremo, 1 700 m, Perrier de la Bdthie 2524 (K!; P!);

Massif d'Andringitra, Perrier de la Bdthie 7211 (P!);

13632 (P!); 14412 (P!).

Restio madagascariensis Cherm. var. humbertii Cherm. in

Archs Bot. 4, Bull. mens. 7: 85 (1930); Humbert in FI. Madag. 34:

3 (1946). Type: Madagascar, Massif de l’Andohahelo, 1 800 m,

Humbert 6147 9 (P, holo.!; B!; NY!).

Icon: FI. Madag. 34: f.l, 3-15 (1946).

Notes

1. This species is very close to R. mahonii from

tropical Africa and the complex is in need of critical

study.

2. As delimited here, R. madagascariensis is en-

demic to Madagascar.

Restio mahonii (N.E.Br.) Pillans in Trans. R.

Soc. S. Afr. 30: 255 (1945).

Hypolaena mahonii N.E.Br. in FI. Trop. Afr. 8: 265 (1901);

Pillans in Trans. R. Soc. S. Afr. 16: 398 (1928). Type: Malawi,

Southern Province, Mt Zomba, Mahon s.n. cf (K, holo.!; B!;

BOL!).

Notes

1. This species is very close to R. madagascarien-

sis, and it also contains wide variation. The complex

is in need of critical study.

2. R. mahonii is common in the eastern Highlands

of Zimbabwe, and further north it is known from Mt

Mulanje and Mt Zomba in Malawi, Mt Kivu in Zaire

and the Uluguru Mts in Tanzania. It occurs in mar-

shy habitats above 1 800 m.

Restio micans Nees in Linnaea 5 : 649 (1830);

Mast, in FI. Cap. 7 : 98 (1897); Pillans in Trans. R.

Soc. S. Afr. 16 : 261 (1928). Type: Cape, 3318 (Cape

Town): Doornhoogte, August (-DC), Zeyher s.n. cf

(B, lect.!; B!; MO!; S!).

Thamnochortus micans (Nees) Kunth, Enum. PI. 3 ; 441

(1841). Restio bigeminus Nees ex Mast, in J. Linn. Soc., Bot. 8 :

246 (1865); in A. DC., Monogr. Phan. 1 : 289 (1878); in FI. Cap.

7: 91 (1897); nom. illeg. , superfluous name for R. micans Nees.

Notes

1. In B there are two sheets of R. micans. One is

labelled ‘Restio micans mihi’ in Nees’s hand. This I

regard as the lectotype. The other sheet in B and the

sheet in S are labelled ‘Restio bigeminus Nees’ in an-

other hand.

2. Masters saw two Zeyher sheets in herb. Sond.,

one annotated ‘Restio micans’ and the other 'Restio

bigeminus ’ and noted: “as neither accords well with

the description of R. micans in the ‘Linnaea’, I have

preferred to adopt the name R. bigeminus' As,

however, R. bigeminus of Masters is based on the

same type as R. micans, it is a superfluous name.

3. This rather curious species occurs in moist de-

pressions in sand near sea level on the Cape Flats,

extending northwards along the sandy coastal fore-

lands as far as Malmesbury.

Restio miser Kunth, Enum. PI. 3 : 392 (1841);

Mast, in A. DC., Monogr. Phan. 1 : 285 (1878); in

FI. Cap. 7 : 89 (1897); Pillans in Trans. R. Soc. S.

Afr. 16 ; 235 (1928). Type: Cape, 3319 (Worcester):

Du Toits Kloof (-CA), Drege 1627 £ (B, lecto.!;

BM!; BOL!; K!; MO!; NY!; OXF!; P!; S!).

Notes

1. Although the spikelets are 1-2-flowered, they

are clustered together, giving the impression of be-

ing many-flowered.

2. R. miser generally occurs above 1 000 m in the

mountains between Caledon and Clanwilliam, but

Bothalia 15, 3 & 4 (1985)

453

there are as yci no records from the Koue Bokke-

veld Mountains.

Restio montanus Esterhuysen, sp. nov., a R. tu-

berculato Pillans, culmis in dimidio supero rugosis et

bracteis maribus acuminatioribus differt.

TYPE. — Cape, 3319 (Worcester): Slanghoek

Needle (-CA), Esterhuysen 35256 $ (BOL, holo.!;

C; D; E; F; K; L; LD; M; MO; S; TCD; UC; US;

W).

Plants caespitose, about 60 cm tall, culm bases ag-

gregated. Culms to 1,2 mm in diam., solid, terete,

branching, obscurely tubercled to wrinkled above,

very coarsely tubercled with conical tubercles at the

base. Sheaths loosely convoluted, 1—1,5 cm long,

awn 2-3(4) mm long, with hyaline shoulders almost

as tall as the awn, body of the sheath pale brown,

striate, sometimes speckled, the lower bracts often

tuberculate. Male inflorescences with 1-4 racemose

spikelets, spikelets with several to numerous flow-

ers, 5-10 x 5 mm. Spathe like the sheaths, 4—6 mm

long, with the mucro and the hyaline shoulders con-

stituting half the total length. Bracts 3,5-4 mm long,

narrowly ovate, acuminate, concave, cartilaginous,

loosely imbricate, the lower two sterile. Flowers very

shortly pedicellate, the perianth 2-2,5 mm long. Se-

pals subcartilaginous, outer sepals conduplicate,

acute to obtuse, sparsely villous on the carina, rather

falcately curved, odd sepal flat, glabrous, obtuse, 2

x 0,9 mm. Petals hyaline, otherwise similar to the

odd sepal. Anthers exserted at anthesis, 1,6 mm

long. Pistillode three-lobed. Female inflorescence

similar to the male, spikelets with two to several

flowers, bracts slightly less acuminate, not com-

pletely obscuring the flowers. Flowers shortly ped-

icellate, perianth 3 mm long. Sepals cartilaginous,

midrib darker, margins more hyaline; outer sepals

conduplicate, sparsely villous along the carina; odd

sepal 3 x 1,5 mm, acute, slightly concave, slightly

scabrid along the midrib. Petals chartaceous, other-

wise similar to the odd sepal, marginally taller than

the sepals. Staminodes about 1 mm long. Ovary with

two fertile locules, fruit a capsule, styles three, free

to the base.

Restio montanus is known from the Slanghoek

Needle only, where it occurs between 1 200 and

1 350 m on the western and north-western slopes in

rock cracks in wet seepages. It is rare.

Restio montanus is very close to R. tuberculatus,

from which it can be distinguished by the tubercles

being restricted to the basal portion of the culms, by

the awn on the sheaths being less than 3 of the length

of the sheaths and by the male floral bracts being

more acuminate. R. tuberculatus is restricted to the

Cedarberg. It is possible to treat R. montanus as a

subspecies of R. tuberculatus. However, until more

is known about both taxa, it is considered simpler to

separate them as distinct species.

CAPE. — 3319 (Worcester): Slanghoek Needle (-CA), Ester-

huysen 33980 (B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG;

NY; PRE; RSA; S; STE; TCD; UC; US; W; WAG); Esterhuysen

35256 (BOL; C; D; E; F; K; L; LD; M; MO; S; TCD; UC; US;

W).

Restio multiflorus Spreng., Syst. Veg. 1 : 187

(1824); Kunth, Enum. PI. 3 : 412 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 279 (1878); in FI. Cap. 7 :

85 (1897); Pillans in Trans. R. Soc. S. Afr. 16 : 225

(1928); in Adamson & Salter, FI. Cape Penins. 136

(1950). Type: Cape, 3318 (Cape Town): by the

waterfall on Table Mountain (-CD), Ecklons.n. (K,

neo.!).

C alop sis neglecta Hochst. in Flora, Jena 28 : 338 (1845). Lepto-

carpus neglectus (Hochst.) Mast, in J. Linn. Soc., Bot. 10 : 225

(1868); in A. DC., Monogr. Phan. 1 : 334 (1878); in FI. Cap. 7 :

117 (1897). Type: Cape, 3318 (Cape Town): Paarlberg (-DB),

Drege 68 (B, lecto.!; K!; MO!).

Restio protractus Mast, in FI. Cap. 7 : 97 (1897). Type: Cape,

3318 (Cape Town): in a valley near Table Mountain, 800 m

(-CD), Bolus 4443 9 (K, holo.!; BOL!).

Restio multiflorus Spreng. var. tuberculatus Pillans in Trans. R.

Soc. S. Afr. 16 : 226 (1928). Type: Cape, 3318 (Cape Town):

Lions Rump (-CD), Pillans 4395 $ (BOL, lecto.!); 4395 c f

(BOL!).

Notes

1. Sprengel described this new species very inad-

equately and I have not been able to locate any ma-

terial that can with confidence be regarded as type

material. Consequently, the identity of the name is

obscure. However, the name has been in constant

use since it was described and every author from

Nees (1830) to Pillans (1928) cited the Ecklon collec-

tion from Table Mountain. In K there is a note by

Masters that this Ecklon collection is close to the

material annotated by Nees in Berlin. It would,

therefore, be a satisfactory neotype. Fig. 20.

2. R. multiflorus is somewhat variable, especially

as regards the relative length of the flowers and the

bracts. It may be found to grade into R. sejunctus

from the southern Cape.

3. There is a form with coarsely tubercled culm-

bases and more robust spikelets in the Hottentots

Holland Mountains. This may be a distinct taxon.

4. R. multiflorus occurs in the mountains of the

south-western Cape, from Piketberg to Bredasdorp.

The plants generally occur scattered in the vegeta-

tion, often in rocky places.

Restio nodosus Pillans in Trans. R. Soc. S. Afr.

30 : 252 (1945). Type: Cape, 3319 (Worcester):

Waaihoek (-CB), Esterhuysen 8366 $ (BOL,

lecto.!; K!); 8366 O' (BOL!; K!).

Note

1. R. nodosus is known from the Hex River Moun-

tains, where it generally occurs above 1 500 m.

Restio nuwebergensis Esterhuysen , sp. nov.,

species culmis vix ramosis a speciebus seminibus col-

liculosis bene distincta; a R. nodoso Pillans bracteis

acutis vaginis convolutis, a R. obscuro Pillans spicu-

lis femineis multifloris differt.

TYPE. — Cape, 3419 (Caledon): eastern slopes

of Landdrost Kop in the Nieuweberg Forest Re-

serve, 800-900 m (-AA), Esterhuysen 31960 $

(BOL, holo.!; C; E; K; L; M; MO; S; STE).

Plants caespitose, culm bases aggregated, tussocks

to 45 cm tall. Culms solid, terete, simple or more

rarely branched, about 1 mm in diam., surface

obscurely tuberculate. Sheaths about 1,5 cm long.

454

Bothalia 15, 3 & 4 (1985)

f

FIG. 20. — Neotype of Restio multiflorus Spreng. Note the draw-

ings and annotations by Masters.

closely convoluted, acuminate with the apex devel-

oped into a short awn, body coriaceous, grey, apical

| submembranous, reddish, decaying. Male inflor-

escence of 1-3 more or less oblong, 10-15 x 8 mm

rounded spikelets. Spathes like the sheaths, less than

1 cm long. Bracts acute, 8x4 mm, concave, lower §

cartilaginous, the upper 3 submembranous, the sub-

membranous portions usually deeply cleft or later

decayed, several sterile bracts at the base of the spi-

kelet. Flowers on short pedicels, perianth about 4

mm long. Sepals cartilaginous, acute; lateral sepals

conduplicate, very densely villous along the carina;

odd sepal more or less flat, glabrous. Petals membra-

nous, subacute, 3,2 x 1,2 mm. Anthers exserted at

anthesis, about 2,4 mm long. Pistillode minute,

three-lobed. Female inflorescence similar to male,

spikelets marginally bigger, bracts almost all fertile,

10-13 mm long, totally obscuring the flowers. Flow-

ers shortly pedicellate, perianth about 4-4,5 mm

long. Sepals as in male flowers, 4-4,5 mm long. Pe-

tals submembranous, acute, 3-3,5 mm long. Stami-

nodes 2 mm long. Ovary bilocular, styles three, free.

Fruit a capsule, seed 1,4 x 1,1 mm, ends truncate,

triangular in cross-section, ridge well developed,

surface grey, ribbed with faint cross-ribbing.

Restio nuwebergensis is known from two collec-

tions from the Nieuweberg Forest Reserve on the

eastern slopes of Landdrost Kop, at an altitude of

between 800-900 m. It grows in more or less marshy

places, among other species of Restionaceae and is

fairly inconspicuous.

This new species is clearly related to R. nodosus

Pillans and R. obscurus Pillans by the triangular

seed, the large oblong spikelets, the numerous fer-

tile female bracts per spikelet and the curiously

pitted appearance of the ‘submembranous’ bract

apices. It differs from R. nodosus in its more tightly

convoluted sheaths, the larger spikelets and the

acute, rather than obtuse, bracts. In most of these

characters, it approaches R. obscurus, but it differs

in the larger spikelets and the many-flowered female

spikelets. A detailed study of the seed by scanning

electron microscopy shows that the seed is ribbed,

with the ribs flat-topped, and the grooves between

the ribs with obscure cross-ribbing. In this it also dif-

fers from the seed-type of the R. nodosus group,

which is colliculate.

CAPE. — 3419 (Caledon): eastern slopes of Landdrost Kop,

800-900 m (-AA), Esterhuysen 31960 (BOL; C; E; K; L; M: MO;

S; STE); Esterhuysen 31638 (BOL; C; E; K; L; M; MO; S).

Restio obscurus Pillans in Trans. R. Soc. S. Afr.

29 : 341 (1942). Type: Cape, 3319 (Worcester): Se-

ven Sisters (-CA), Stokoe in BOL 22641 $ (BOL,

lecto.!; B!; K!); in BOL 22641 cf (B!; BOL!; K!).

Note

1. R. obscurus occurs at middle altitudes in the

Slanghoek and Wemmershoek Mountains.

Restio occultus (Mast.) Pillans in Trans. R. Soc.

S. Afr. 16 : 243 (1928). Type: Cape, 3319 (Worces-

ter): Du Toits Kloof (-CA), Drege 1612 cf (K,

holo.!; B!; P!).

Thamnochortus occultus Mast, in FI. Cap. 7 ; 130 (1897).

Hypolaena schlechteri Mast, in Bot. Jb. 29 Beibl. 66 : 15 (1900).

Restio schlechteri (Mast.) Pillans in Ann. Bolus Herb. 3 : 85

(1921). Syntvpes: Cape, 3219 (Wuppertal): Ezelbank. 1 200 m

(-AC), Schlechter 8834 9 (K, lecto.!; BM!; BOL!; BR!; P!; MO!;

S!; Z!); 8833 cf (BM!; BOL!; BR!; K!; MO!; P!; S!; Z!).

Notes

1. The type specimen of T. occultus ( Drege 1612)

is very poor and its determination is somewhat du-

bious.

2. R. occultus occurs frequently in the Cedarberg-

Cold Bokkeveld area, often on sandy flats and occa-

sionally as far south as Caledon.

Restio pachystachyus Kunth, Enum. PI. 3 : 399

(1841); Mast, in A. DC., Monogr. Phan. 1 : 283

(1878); in FI. Cap. 7 : 88 (1897); Pillans in Trans. R.

Soc. S. Afr. 16 : 249 (1928). Type: Cape, 3219

(Wuppertal): Wuppertal, 450-600 m (-A), Drege 43

9 (B, lect. ! ; K!; MO!; OXF!; P!).

? Restio furcatus Nees ex Mast, in J. Linn. Soc., Bot. 8 : 242

(1865); in A. DC., Monogr. Phan. 1 : 275 (1878); in FI. Cap. 7 : 83

(1897). Type: Cape, without precise locality, Ecklon & Zeyher

s.n. cf (?)•

Restio dimorphostachyus Mast, in Bot. Jb. 29 Beibl. 66 : 2

(1900). Syntvpes: Cape, 3319 (Worcester): Bainskloof, 700 m

(-CA), Schlechter 9164 9 (B. lecto.!; BOL!; BR!; K!; MO!; P!;

S!; Z!); 9164 cf (B!; BOL!; K!; P!; S!; Z!). 3419 (Caledon): Zwart-

berg (-AB/BA), Schlechter 9788 9 (BM!; BOL!; BR!; K!; S!);

9789 cf (B!; BM!: BOL!; BR!; K!; S!).

Notes

1. The identity of R. furcatus is dubious, as the

description is not detailed enough, and I have not

been able to trace any type material.

2. R. pachystachyus is most common in the Slang-

hoek Mountains, but there are records from the Hex

Bothalia 15, 3 & 4 (1985)

455

River Mountains and as far south as the Caledon

Swartberg.

Restio papyraceus Pillans in Trans. R. Soc. S.

Afr. 29 : 342 (1942). Type: Cape, 3321 (Ladismith):

Seven Weeks Poort Berg (-AD), Stokoe 1925 9

(BOL, lecto.!; K!); 1925 cf (BOL!).

Note

1. This species occurs above 1 800 m in the Klein

Swartberg and the Great Swartberg, generally on

rocky ledges.

Restio patens Mast, in FI. Cap. 7: 97 (1897); Pil-

lans in Trans. R. Soc. S. Afr. 16: 241 (1928). Type:

Cape, 3319 (Worcester): Tulbagh distr . , Winter-

hoeksberg. 1 950 m (-AA), Bolus 7495 cf (K, holo.!;

B!).

Restio intricatus Mast, in Bot. Jb. 29 Beibl. 66: 1 (1900). Type:

Cape, 3319 (Worcester): Koude Bokkeveld, Skurfteberge near

Gydouw. 1 500 m (-AB). Schlechter 10009 $ (B. lecto.!; BM!;

BOL!; BR!; K!; MO!; P!; S! ; Z!).

Notes

1 . This species looks like a large version of R. per-

plexus, but is probably a good species.

2. R. patens occurs in the mountains from the Ce-

darberg to the Hex River Mountains, as well as on

the Piketberg.

Restio peculiaris Esterhuysen , sp. nov., a R. tri-

ticeo Rottb. ovariis bilocularibus, vaginis in dimidio

superiore hyalinis differt.

TYPE. — Cape, 3320 (Montagu): Langeberg,

Goedgeloof Peak (-CD). Esterhuysen 34519 J

(BOL, holo.!; B; C; E; F; GRA; K; L; M; MO;

NBG; NY; PRE; RSA; S; STE; TCD; UC; US; W;

WAG).

Plants caespitose, to 45 cm tall, sometimes spread-

ing, culms aggregated at the base, rarely forming

short rhizomes. Culms solid, terete, branching, to 2

mm in diam., very densely tuberculate with flat-

topped tubercles. Sheaths tightly convoluted, 1,5-3

cm long, body brown, coriaceous, acute and ex-

tended into an awn almost as long as the body and

flanked by large membranous shoulders as tall as the

awn. Male inflorescence of I- several racemose spi-

kelets. Spathes like the sheaths, but 1-1,5 cm long.

Spikelets about 10 x 5 mm, obconical to elliptical,

many-flowered. Bracts 6-8 x 3,5 mm, acute, carti-

laginous, concave, apices hyaline, imbricate, obscur-

ing the flowers, lowest few bracts sterile, acuminate.

Flowers shortly pedicellate, perianth 4—5 mm long.

Sepals cartilaginous, lateral sepals conduplicate, Car-

ina sparsely pilose; odd sepal shallowly concave, ob-

tuse, 4—5 x 1-1,5 mm. Petals membranous, 4—1,5

mm, obtuse, fused at the base. Stamens three, anth-

ers exserted at anthesis, 2,3 mm long. Pistillode min-

ute, three-lobed. Female inflorescences with 1-sev-

eral spikelets. Spikelets about 8 mm long, with 1-3

flowers. Spathes as in male, overtopping the spike-

let, bracts as in male, but fewer. Flowers very shortly

pedicellate, perianth 4,5 mm long. Sepals cartilagi-

nous; lateral sepals conduplicate, acute, sparsely pi-

lose on the carina; odd sepal very shallowly concave,

acute, 4,5 x 1,7 mm. Petals papyraceous to membra-

nous, 3,5 x 1,6 mm, rounded. Staminodes 2 mm

long. Ovary with two fertile locules, styles three,

free. Fruit a capsule. Seeds 1,5 x 1 mm, obtuse,

ridged on one side, colliculate. Fig. 21.

Restio peculiaris occurs on the Langeberg at Swel-

lendam along rocky ridges between 1 350 and

1 650 m. Flowering probably occurs in January.

It is allied to the Restio triticeus group by the colli-

culate seed, the cartilaginous bracts and the finely

and densely tuberculate culms. Superficially, the

plants look very peculiar with their squat habit and

the large hyaline shoulders on the sheaths. For de-

tails of the seed see Linder (1984).

CAPE. — 3320 (Montagu): Langeberg, Goedgeloof Peak

(-CD), Esterhuysen 32921 (BOL; C; E; F; K; L; LD; M; MO;

NBG; PRE; S; STE; TCD; UC; US; W; WAG); Esterhuysen

33410 (BOL; C; E; F; K; L; LD; M; MO; S; UC); Esterhuysen

34519 (B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY;

PRE; RSA; S; STE; TCD; UC; US; W; WAG).

Restio pedicellatus Mast, in J. Linn. Soc., Bot.

8: 252 (1865); in A. DC., Monogr. Phan. 1: 286

(1878); in FI. Cap. 7: 90 (1897); Pillans in Trans. R.

Soc. S. Afr. 16: 262 (1928). Syntypes: Cape, 3319

(Worcester): Du Toits Kloof (-CA), Drege 82 $ (K,

lecto.!; B!; BM!; MO!; NY!; P!); 91 $ (K!); 1629 cf

(K!; MO!; NY!; OXF!; P!).

Restio sonderianus Mast, in J. Linn. Soc., Bot. 8: 252 (1865); in

A. DC., Monogr. Phan. 1: 285 (1878); in FI. Cap. 7: 89 (1897);

Pillans in Trans. R. Soc. S. Afr. 16: 295 (1928). Type: Cape, 3319

(Worcester): Du Toits Kloof (-CA), Drege 82 cf (K, holo!; B!;

BM!; OXF!; P!).

Notes

1. Both names are based on Drege 82: R. pedicel-

latus on the female and R. sonderianus on the male.

2. Although Pillans (1928) maintains R. sonderia-

nus as a distinct species, he notes that they are very

closely related, separated by a character ‘which,

however, may be a variation in a variable species.’

3. R. pedicellatus occurs in the mountains, usually

in marshy areas or on rock flushes, from the Cedar-

berg to the Piketberg, the Cape Peninsula and Cale-

don.

Restio perplexus Kunth, Enum. PI. 3; 406

(1841); Mast, in A. DC., Monogr. Phan. 1: 286

(1878); in FI. Cap. 7: 90 (1897); Pillans in Trans. R.

Soc. S. Afr. 16: 239 (1928); in Adamson & Salter,

FI. Cape Penins. 137 (1950). Syntypes: Cape, 3318

(Cape Town): Table Mountain (-CD), Drege 339a cf

(B, lecto.!; B!; K!; OXF!; MO!; NY!; P!; S!); with-

out precise locality, in herb. Willdenow. 18279 fol. 3

Cf (B!).

Restio scopula Mast, in FI. Cap. 7: 96 (1897). Type: Cape, 3318

(Cape Town): Table Mountain, 800 m (-CD), Bolus 4095b $ (K,

holo.!; B!; BOL!).

Restio trichocaulis Mast, in Bot. Jb. 29 Beibl. 66: 6 (1900).

Type: Ceres Div., Koue Bokkeveld, 1 050 m, Schlechter 8877 cf

(B. lecto.!; B!; BM!; BR!; H!; MO!; S!; Z!).

Notes

1. This species is very close to R. capillaris, and I

have keyed them out together. There is wide varia-

tion in the degree of robustness of the culms, with

the type of R. scopula representing the robust ex-

treme, and the type of R. perplexus typical of the

more gracile forms.

456

Bothalia 15, 3 & 4 (1985)

FIG. 21. — Restio peculiaris Esterhuysen. a, habit; b, female plant, rooting from aggregated culm bases, note loosely convolutec

sheaths with large lacerated hyaline shoulders, x 0,8; c, female spikelets, x 3; d, female bract, x 6; e, female flower, lateral

sepals sparsely pilose on the carina, x 6; f, ovary, style and staminodes, x 6;g, T.S. of bilocular capsule, showing aborted third

locule, x 8; h, male spikelets, note also tuberculate culms, x 3; i, bracts, x 6; j, male flower-with pilose-carinate lateral sepa s

and membranous petals, x 6. (From Esterhuysen 34519.)

Bothalia 15, 3 & 4 (1985)

457

2. R. perplexus is widely distributed and quite

common, from Riversdale to the Cape Peninsula

and to Ceres. It is restricted to the mountains, where

the plants are often found among rocks.

Restio perseverans Esterhuysen, sp. nov., R.

aureolo Pillans et R. bifurco Mast, aemulans, brac-

teis spicularum feminarum omnibus fertilibus dif-

fers

TYPE. — Cape, 3320 (Montagu): Langeberg,

Goedgeloof Peak (-CD), Esterhuysen 34518 $

(BOL, holo.!; C; E; K; L; M; MO; NBG; PRE; S;

STE; US; W; WAG).

Plants caespitose, sometimes forming dense

stands, 20-40 cm tall, culm bases aggregated. Culms

erect, solid, terete, smooth to obscurely tubercled,

about 1 mm in diameter. Sheaths closely convoluted,

1-1,5 cm long, brown to dark brown, coriaceous

with an obscure submembranous margin, obtuse

with a stout 1-2 mm long awn. Male inflorescence

with 1-3 sublax racemose spikelets. Spikelets sev-

eral-flowered, 10-15 x 5 mm. Spathes to 1 cm long,

shorter than the spikelets, coriaceous, brown, with a

submembranous margin, acute. Bracts 7-10 mm

long, acuminate, concave, brown, the upper margin

and especially the apex submembranous, the bases

decussate on the floral axis. Flowers with a 0,5 mm

long, shortly villous pedicel. Perianth 5, 5-6, 5 mm

long. Sepals cartilaginous, laterals conduplicate,

acute, the upper 3 of the carina densely villous, odd

sepal shallowly concave, acute, 5 x 1 mm. Petals

hyaline, membranaceous, otherwise like the odd se-

pal. Perianth lobes fused to a 1 mm long floral axis.

Anthers exserted at anthesis, 2 mm long. Pistillodes

with three widely separated lobes. Female inflores-

cence similar to male, except that the spikelets are

usually two-, rarely one-flowered. Sepals as in the

male flowers, petals papyracous with membranous

margins, acute, 4x1 mm. Staminodes 0,8 mm long.

Ovary with two fertile locules and three free styles.

Seed 2 x 1,5 mm, semi-ovate, pale brown, smooth,

under x 50 magnification finely and densely reticu-

late foveate.

Restio perseverans has been recorded from the

Langeberg between Swellendam and Riversdale. All

the collections except one are from the Clock Peaks

and Goedgeloof Peak, between 1 200 and 1 500 m,

from marshy areas. The other collection is from

about 400 m in Garcias Pass, also from a marshy

area. The Langeberg is not well known, and it is

likely that many more localities could be found.

This new species is probably related to Restio aur-

eolas Pillans in overall morphology, but the spikelet

structure is different — R. aureolas has several ster-

ile bracts in the female spikelet, but in R. perseve-

rans there are none. There is also some similarity to

R. bifurcus Pillans, which has a similar type of seed,

but R. bifurcus also has sterile female bracts, as well

as a villous odd sepal.

CAPE. — 3320 (Montagu): Langeberg, Eleven o’clock Peak

(-CD), Esterhuysen 33475 (BOL; E; K; L; M; MO; S); Lange-

berg, Goedgeloof Peak (-CD), Esterhuysen 3451 (BOL; C; E; K;

L; M; MO? NBG; PRE; S; STE; US; W; WAG); Esterhuysen

31414 (BOL; C; E; K; L; M; MO; S); Esterhuysen 31414a (BOL;

C; E; K; L; M; MO; S). 3321 (Ladismith): Garcias Pass, near the

Toll House (-CD), Esterhuysen 31763 (BOL; K; M; MO; S).

Restio pillansii Linder , nom. nov.

Leptocarpus stokoei Pillans in Trans. R. Soc. S.

Afr. 30: 263 (1945). Type: Cape, 3418 (Simons-

town): Somerset Sneeukop, 1 280 m (-BB), Stokoe

5028a $ (BOL, lecto.!; K!); 5028a cf (BOL!; K!).

Note

1. R. pillansii is known from a few localities:

Dwarsberg and Somerset Sneeukop near Jonkers-

hoek, Babylons Tower near Caledon and Goedge-

loof Peak near Swellendam. It is recorded from be-

tween 1 000 and 1 800 m on rocky, dry slopes.

Restio praeacutus Mast, in FI. Cap. 7: 84 (1897);

Pillans in Trans. R. Soc. S. Afr. 16: 246 (1928). Syn-

types: Cape, 3118 (Van Rhynsdorp): Giftberg

(-BC), Drege 88 $ (K, lecto.!; B!; BOL!). 3319

(Worcester): Du Toits Kloof (-CA), Drege 46 $

p.p. (K!); 1608 p.p. (K!; MO!; OXF!), see Fig. 9.

Note

1. This species is very closely related to several

other species (see R. bolusii for notes) and may not

be distinct.

Restio pulvinatus Esterhuysen, sp. nov., R. aur-

eolo Pillans, primo adspecto simillimus, sed semini-

bus laevibus et spiculis minoribus differt.

TYPE. — Cape, 3319 (Worcester):

Wemmershoek Peak, 1 700 m (-CC), Esterhuysen

35168 $ (BOL, holo.!; C; E; F; K; L; LD; M; MO;

NBG; S; TCD; UC; US; W).

Plants forming dense cushions, 20 to 30 cm tall.

Culms solid, terete, wrinkled, branching, up to 1

mm in diam. , surface wrinkled, lower sections of the

culms often rooting at the nodes. Sheaths closely

convoluted, 3-6 mm long, acuminate, coriaceous,

pale brown, with an indistinct submembranous mar-

gin. Male inflorescence with solitary or rarely two,

3-6 mm long, elliptical or obconical spikelets.

Spathes like the sheaths, 3-4 mm long. Bracts 3 mm

long, obtuse, mucronate, lower 3- 3 cartilagious, up-

per \ - § submembranous. Flowers several per spike-

let, shortly pedicellate, perianth 3,5 mm long. Sepals

cartilaginous, acute; lateral sepals conduplicate, up-

per half of the carina densely villous, somewhat

curved falcately; odd sepal shallowly concave, gla-

brous, 2,6 x 0,6 mm. Petals membranous, acute, 2,5

x 0,8 mm. Anthers exserted at anthesis. Pistillode

minute, three-lobed. Female inflorescence like the

male, except that the spikelets are marginally bigger

and usually single, rarely two-flowered, spikelets

with about 5 sterile bracts. Flowers very shortly ped-

icellate, perianth 3,5 mm long. Sepals cartilaginous,

acute; lateral sepals conduplicate, carina densely vil-

lous in the upper half; odd sepal shallowly concave,

3,3 x 1 mm, villous along the midrib near the apex.

Petals 3 x 1,6 mm, acute, membranous at the mar-

gins, centre part thickened. Staminodes to 2 mm

Fong. Ovary with two fertile locules, occasionally

one. Styles three, free. Fruit a capsule, seed triangu-

lar, ridged, truncate, reddish, surface smooth, 1,1 x

0,8 mm.

Restio pulvinatus has been recorded from the

Slanghoek and the Klein Drakenstein and the

458

Bothalia 15, 3 & 4 (1985)

Wemmershoek Mountains, between 1 350 and 1 900

m. The collections are all from rock ledges and at the

base of cliffs. The species probably flowers in Oc-

tober.

The affinities of this species are difficult to deter-

mine. The habit is similar to that found in R. patens.

The spikelets are like those of R. aureolas, but the

seed surfaces are smooth, whereas they are collicu-

late in R. aureolas and the bract apices are different.

The species with similar seed to R. pulvinatus

(Group lib. Tinder 1984: 30) can all be readily dis-

tinguished by bract, spikelet or sheath characters.

CAPE. — 3319 (Worcester): Du Toits Peak (-CA), Esterhuy-

sen 31635 (BOL; K); Wemmershoek Peak (-CC), Esterhuysen

29624 (BOL; K; S); Esterhuysen 32098 (BOL; K); Esterhuysen

35168 (BOL; C; E; F; K; L;'LD; M; MO; NBG; S; TCD; UC;

US; W); Kaaimans Peak, south side (-CD), Esterhuysen 32130

(BOL; K); Esterhuysen 32377 (BOL; C; E; F; K; L; LD; M; MO;

NBG; PRE; S; STE; TCD; UC; US; W). 3418 (Simonstown):

Langklippiekloof, above Somerset West (-BB), Esterhuysen

28988 (BOL; C; E; K; L; M; MO; S; STE). 3419 (Caledon): Vic-

toria Peak (-AA), Esterhuysen 33757 (BOL; C; E; F; K; L; LD;

M; MO; S; STE; UC); Esterhuysen 29289 (BOL; C; E; K; L; LD;

M; MO; S; STE; UC).

Restio pumilus Esterhuysen, sp. nov., a R.

zwartbergensi Pillans plantis caespitosis, spiculis

masculis plurifloris differt.

TYPE. — Cape, 3319 (Worcester): Jonaskop,

north slopes, 900 m (-DC), Esterhuysen 32694 $

(BOL, holo.!; B; C; E; F; K; L; LD; M; MO; NBG;

NY; PRE; S; STE; TCD; UC; US; W; WAG).

Plants caespitose, 1 5—25 cm tall, culm bases aggre-

gated. Culms solid, terete, branching, to 0,7 mm in

diam., densely and irregularly tuberculate. Sheaths

closely convoluted, 10-15 mm long, body brown, co-

riaceous, obtuse, about § of the total sheath length,

with a very slender awn to 6 mm long, but soon de-

caying and two acute, membranous shoulders about

4 mm tall. Older sheaths decayed, only the coria-

ceous brown part persistent. Male inflorescence of

2-7 very lax, 2-3-flowered, 3-4 mm long spikelets.

Spathe 4-5 mm long, chartaceous, the apical half

membranous, acute, decaying. Bracts all fertile,

1,5-2 mm long, acute, chartaceous, membranous

margins decaying, about half as tall as the perianth.

Flowers very shortly stipitate, perianth 2-2,5 mm

long. Sepals cartilaginous, subacute to rounded,

2-2,5 mm long; lateral sepals deeply concave, cari-

nate, glabrous; odd sepal shallowly concave. Petals

membranous, rounded, marginally smaller than the

odd sepal. Anthers exserted at anthesis. Pistillode

minute. Female inflorescence like the male, spikelets

1-2-flowered. Bracts about four, with the lower

bracts sterile, about 1,5 mm long, obtuse. Flowers

pedicellate, perianth 2,5-3 mm long. Sepals cartila-

ginous, subacute to rounded, 2,3-3 mm long; lateral

sepals deeply concave, glabrous, carinate; odd sepal

shallowly concave. Petals chartaceous, obtuse to

rounded, 2 mm long. Staminodes about 1 mm long.

Ovary with 1-2 fertile locules. Styles three, free to

the base, about 2 mm long, densely villous. Fruit a

capsule, seed 1,2 mm long, rather obscurely collicu-

late.

Restio pumilus is known from between sea level

and 1 300 m in the Caledon area, from Kleinmond to

Wolfkop and Jonaskop in the Riviersonderend

Mountains. The habitat appears to be well-drained

slopes. The plants probably flower in September and

seed is released in November.

This species is very similar to R. zwartbergensis

Pillans, with which it is sympatric. The most striking

difference is in the tufted habit of the plants. It is

possible that the several florets in the male spikelet

will also appear as a significant character, but more

collections are required of both species to show that

these characters are correlated.

CAPE. — 3319 (Worcester): Wolfkop, Boschjesveld Moun-

tains, 1 230 m (-CD), Esterhuysen 32065 (B; BOL; C; E; F;

GRA; K; L; LD; M; MO; NBG_; NY; PRE; RSA; S; STE; TCD;

UC; US; W; WAG); Jonaskop, north slopes (-DC). Esterhuysen

32694 (B; BOL; C; E; F; K; L; LD; M; MO; NBG; NY; PRE; S;

STE; TCD; UC; US; W; WAG). 3418 (Simonstown): Arieskraal,

steep slopes above the Klein Palmiet (-BB). Esterhuysen 32750

(BOL; K; M; MO; S); mountains above the bridge over the Pal-

miet River at Kleinmond (-BD), Esterhuysen 32542a (BOL; K).

3419 (Caledon): ‘SilverstreanT, between Elandskloof and Wolf-

kop (-BC/DA). Esterhuysen 33712 (BOL; C; E; K; L; M; MO;

S).

Restio purpurascens Nees ex Mast, in J. Linn.

Soc., Bot. 8: 249 (1865); in A. DC., Monogr. Phan.

1: 283 (1878); in FI. Cap. 7: 88 (1897); Pillans in

Trans. R. Soc. S. Afr. 16: 257 (1928). Syntypes:

Cape, 3418 (Simonstown): Sir Lowry’s Pass, May

(-BB), Ecklon & Zeyher s.n. 9 (B, lecto.!; BOL!;

BR!; C!; K!; MEL!; MO!; P!; S!; Z!); without pre-

cise locality. Find s.n. cf (?).

Notes

1. In many of the European herbaria the collec-

tions of Ecklon & Zeyher ‘59.5’ have printed labels

indicating 'Restio purpurascens N. ab E.’, but in B

this is written in Nees's hand. The sheet in K was

only acquired from Reichenbach in 1865, so Masters

probably did not see it before R. purpurascens was

published.

2. R. purpurascens is known from the mountains

between Caledon, Stellenbosch and the Cape Penin-

sula, with one record from Bainskloof.

Restio quadratus Mast, in J. Linn. Soc., Bot. 10:

277 (1868); in A. DC., Monogr. Phan. 1: 293 (1878);

in FI. Cap. 7: 95 (1897); Pillans in Trans. R. Soc. S.

Afr. 16: 223 (1928); in Adamson & Salter, FI. Cape

Penins. 136 (1950). Syntypes: Cape, 3318 (Cape

Town): between Cape Town and Newlands (-CD),

Bure he 1 1 408 cf (K, lecto.!; BOL!; P!); see Fig 8;

Table Mountain (-CD), Ecklon s.n. cf (?); Drege

167a cf (BOL!; K!). 3419 (Caledon): Baviaanskloof

Mountains at Genadendal (-BA), Burchell 7778 cf

(BOL!); without precise locality, Ecklon 166 9 (?);

Zeyher 540 c f (?); Pappe 96 9 (K!); Niven s.n. 9

(BM!).

Notes

1. According to Masters, several of the types are

from Herb. Sonder, but I have not found them at

MEL.

2. The fruits are rarely dehiscent on the plants, but

are shed with the flowers. However, the fruit walls

are thin and dehiscence might occur after the flowers

are shed.

3. There are some collections which appear to be

intermediate between this species and R. tetragonus,

Bothalia 15, 3 & 4 (1985)

459

and careful fieldwork is needed to resolve the status

of R. quadratus.

Restio quinqucfarius Nees in Linnaea 5: 639

(1830); Kunth, Enum. PI. 3: 411 (1841); Mast, in A.

DC., Monogr. Phan. 1: 278 (1878); in FI. Cap. 7: 82

(1897); Pillans in Trans. R. Soc. S. Afr. 16: 233

(1928); in Adamson & Salter, FI. Cape Penins. 137

(1950). Syntypes: Cape, 3318 (Cape Town): ‘In

piano montis tabularis, Aprili’ (-CD), Ecklon s.n.

(?); Cape Flats, August (-DC), Ecklon s.n. (K,

lecto.!; BOL!; BR!; C!; MO!; P!; Z!).

Restio xyridioides Kunth, Enum. PI. 3: 397 (1841). Syntypes:

Cape, 3219 (Wuppertal): between Ezelbank and Dwars Rivier

(-AC), Drege 35 cf (B, holo.!; BM!; BOL!; K!; MO!; NY!; OXF!;

P!; S ! ) , without precise locality, Lalande s.n. (?).

Notes

1. Nees refers to two localities for the material on

which he based R. quinquefarius , but I have been

able to locate only material from the Cape Flats. The

other locality (summit of Table Mountain) is rather

unlikely. Most of the type collections in European

herbaria have printed labels reading 'Restio quinque-

farius N. ab E. 78.8’. The month of August given as

the collecting date, conflicts with the date given in

the protologue. However, in the absence of any

other likely type material and given the printed la-

bels on the collections, I suspect that this may be a

printing error.

2. This very distinctive species occurs on the sandy

flats between Cape Point and Clanwilliam. Most of

the collections are from near sea level, but in the Ce-

darberg, some populations occur at 1 000 m. The

habitat is generally dry.

Restio rarus Esterhuysen , sp. nov., a R. lepto-

stachyo ovariis bilocularibus, plantis validioribus dif-

fert.

TYPE. - — Cape, 3321 (Ladismith): Swartberg,

Toverkop, at shelter, 1 650 m (-AC), Esterhuysen

33785 $ (BOL, holo.!; B; C; E; F; GRA; K; L; LD;

M; MO; NBG; NY; PRE; RSA; S; STE; TCD; UC;

US; W; WAG).

Plants caespitose, 20-40 cm tall, culm bases aggre-

gated. Culms solid, terete, branching, ascending, to

about 0,5 mm in diam., surface wrinkled and tuber-

culate. Sheaths closely convoluted, 0,8-1, 6 cm long,

body pale brown, coriaceous, rounded to acute, with

a cylindrical awm to 3,5 mm long, flanked by two

rounded membranous lobes which are usually § the

length of the awn. Male inflorescence of a solitary,

10-15 x 5-10 mm, oblong to elliptical spikelet.

Spathes 5-9 mm long, similar to sheaths. Bracts paler

than the spathes, subcartilaginous, up to 10 x 2 mm,

acuminate with up to 2 mm long awns, red-speckled

and submembranous near the apices, all fertile,

bracts largely obscuring the flowers. Flowers very

shortly pedicellate, perianth 3-3,3 mm long. Sepals

cartilaginous; lateral sepals conduplicate, acute,

sparsely villous along the carina; odd sepal shallowly

concave, subacute, apex submembranous and decay-

ing, 3 x 0,8 mm. Petals membranous, acute, 2,5 x

0,8 mm. Anthers 1,7 mm long, pistillode minute. Fe-

male inflorescence like the male. Flowers shortly

pedicellate, perianth 3,5-4 mm long. Sepals cartila-

ginous, acute to finely mucronate; lateral sepals con-

duplicate, carina pilose; odd sepal shallowly con-

cave, glabrous. Petals subcartilaginous to membra-

nous, acute, 2,5-3 x 1 mm. Staminodes less than 1

mm long. Ovary with two fertile locules, styles

three, free, adjacent at the base, about 4 mm long,

villous. Fruit a capsule, seed 1 x 0,6 mm, ends round

to obtuse, round in cross-section, slightly flattened

at one side, smooth and nitid.

Restio rarus is known only from below the Tover-

kop behind Ladismith in the Little Swartberg. It

grows at about 1 650 m in marshy spots and seep-

ages, or on wet rock flushes.

It is clearly allied to the Restio leptostachyus group

by the smooth seed, the very narrowly lanceolate

acuminate bracts, the habit of the plants and the

overall spikelet appearance. It is the only species in

the group with bilocular ovaries.

CAPE. — 3321 (Ladismith): Swartberg, slopes below Tover-

kop, 1 650 m (-AC), Esterhuysen 33192 (BOL; C; E; K; L; M;

MO; S; STE); Esterhuysen 33785 (B; BOL; C; E; F; GRA; K; L;

LD; M; MO; NBG; NY; PRE; RSA; S; STE; TCD; US; UC; W;

WAG); Esterhuysen 33893 (BOL; C; E; K; L; M; MO; S; STE).

Restio rupicola Esterhuysen , sp. nov., a R. bra-

chiato seminibus colliculosis et bracteis obtusis, a R.

bifurco spiculis femineis unifloris recedit.

TYPE. — Cape, 3319 (Worcester): Matroos-

berg, 1 200 m (-BC), Esterhuysen 34348 9 (BOL,

holo.!; B; C; E; F; K; L; LD; M; MO; NBG; NY;

PRE; S; STE; TCD; UC; US; W; WAG).

Plants caespitose, tussocks to 1 m tall, culms as-

cending from a 6 mm thick, short, creeping rhizome.

Culms solid, terete, branching, to 3 mm in diam. at

the base, smooth, shiny, finely punctate. Sheaths

0,8-2 cm long, body coriaceous, brown, finely

golden speckled, acuminate with the apex continued

into an aciculate, up to 5 mm long awn, body with a

narrow membranous margin. Male inflorescence of

2-10 racemose or paniculate-racemose, elliptical,

acute, 6-9 x 3 mm spikelets. Spathes acuminate, 2,5

x 2,5 mm. Bracts 4 x 3 mm, obtuse, concave, carti-

laginous, the apical half submembranous, decaying,

lower 5 bracts marginally smaller. Flowers almost

sessile, perianth 3-3,5 mm long. Sepals cartilagi-

nous, acute; lateral sepals conduplicate, densely vil-

lous along the carina, especially on the upper half;

odd sepal almost flat, with a small villous patch near

the apex of the midrib, 3,5 x 1 mm. Petals membra-

nous, acute, 3 x 0,9 mm. Anthers exserted at anthe-

sis. Pistillode minute, three-lobed. Female inflores-

cence of 1-5 racemose, 10 x 3,5 mm, single-flowered

spikelets. Spathe and bracts as in male, to 4 mm

long, reaching up half the length of the mature

flower. In the younger flowers less of the flower is

exposed above the bracts. Flower sessile, perianth

about 7 mm long. Sepals cartilaginous, acute, sub-

equal, concave, 7x3 mm, lateral sepals subcari-

nate, villous in upper half of the carina. Petals carti-

laginous, acute, 5,5 x 2,3 mm. Staminodes 1,2 mm

long. Ovary with three locules, not always all three

fertile. Styles three, free to base, adjacent, very

densely villous. Fruit a capsule, seed triangular in

cross-section, with the ends truncate, surface gray,

irregularly tuberculate-colliculate, 2 x 1,2 mm.

460

Bothalia 15, 3 & 4 (1985)

Restio rupicola is recorded from a single locality

on the Matroosberg. It is locally plentiful at 1 200 m

in crevices and cracks in rugged rock on the southern

flank of the mountain, flowering in September.

This species is allied to the Restio bifurcus — R.

bolusii group in its seed morphology (triangular -

colliculate, see Linder, 1984: 29) and the cell types

of the bract apices. It is the only species in the group

with single-flowered female spikelets. There are

rather few species in Restio that have female flowers

that are solitary in the spikelet and overtop the

bracts — the only one that looks similar is R. bra-

chiatus. R. brae hiatus, however, has smooth, ellip-

soid seed and is allied to the R. occultus group.

Note that the appearance of the perianth changes

with the age of the flowers. The obvious female

flowers are older, with almost mature seed and the

lateral sepals at this stage have lost the carina and

the beard.

CAPE. — 3319 (Worcester): Matroosberg, 1 200 m (-BC), Es-

terhuysen 34348 ( B; BOL; C; E; F; K; L; LD; M; MO; NBG; NY;

PRE; S; STE; TCD; UC; US; W; WAG); Esterhuysen 35420 (B;

BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY; PRE; RSA;

S; STE; TCD; UC; US; W; WAG).

Restio sarocladus Mast, in A. DC., Monogr.

Phan. 1: 291 (1878); in FI. Cap. 7: 93 (1897); Pillans

in Trans. R. Soc. S. Afr. 16: 252 (1928); in Adamson

& Salter, FI. Cape Penins. 138 (1950). Syntypes:

Cape, 3318 (Cape Town): Table Mountain (-CD),

B lire he 1 1 572 $ (K, lecto.!; BOL!); 605 $ (BOL!;

K!).

Restio tabularis Pillans in Ann. Bolus Herb. 3: 82 (1921). Type:

Cape, 3318 (Cape Town): Table Mountain (-CD), Pillans 3904 $

(BOL. lecto.!); 3904 cf (BOL!).

Notes

1. Pillans (1928, 1950) referred to this species as

‘R. sarcocladus’ which is better Greek than ‘ R . saro-

cladus’. However, such a correction is not allowed

under Art. 73 of the ICBN (1983).

2. This species is easily confused with R. corneo-

lus.

3. R. sarocladus is known from the Cape Penin-

sula and from the coastal mountains of the Caledon

district. It has an altitude range from sea level to

1 000 m.

Restio scaber Mast, in Bot. Jb. 29 Beibl. 66: 1

(1900); Pillans in Trans. R. Soc. S. Afr. 16: 264

(1928). Type: Cape, 3419 (Caledon): Zwarteberg,

600 m (-AB/BA), Schlechter 10366 cf (B, lecto.!;

BOL!; K!; P!; Z!).

Note

1. This rather distinct species is only known from

the type collection.

Restio scaberulus N.E.Br. in FI. Cap. 7 : 751

(1900); Pillans in Trans. R. Soc. S. Afr. 16 : 231

(1928). Type: Cape, 3321 (Ladysmith): Garcias

Pass, 600 m (-CC), Galpin 4789 9 (K, lecto.!; B!);

4789 Cf (B!; BOL!; K!).

Notes

1 . This species is very close to R. stokoei, which

occurs in the Caledon-Stellenbosch Mountains. Crit-

ical analysis of the differential characters (flower size

and the relative sheath/awn length) may show con-

tinuous variation.

2. R. scaberulus occurs in the Langeberg, the Ou-

teniqua Mountains and the Tsitsikamma Mountains.

Restio secundus ( Pillans ) Linder, comb. nov.

Leptocarpus secundus Pillans in Trans. R. Soc. S. Afr. 29: 348

(1942). Type: Cape, 3320 (Montagu): south slopes of the Swellen-

dam Mountains (-CD), Esterhuysen 4798 $ (BOL, lecto.!; K!);

4798 cf (BOL!; K!).

Notes

1 . This species is not well understood. It is near R.

decipiens and probably also near R. pillansii.

2. It is known from the Langeberg between Swel-

lendam and Riversdale.

Restio sejunctus Mast, in FI. Cap. 7: 97 (1897);

Pillans in Trans. R. Soc. S. Afr. 16: 226 (1928).

Type: Natal, without precise locality, Bolton s.n. cf

(BM, holo., K!).

Notes

1. I have not been able to find the type specimen

at BM, where Masters reports it to be. At K is a frag-

ment, which is adequate for determination purposes.

2. This species is probably not distinct from R.

multiflorus , of which it is possibly a southern and

eastern extension.

3. R. sejunctus has a distribution range from Paarl

to the Natal Drakensberg. It generally occurs be-

tween 900 and 1 800 m on rocky slopes.

Restio similis Pillans in Ann. Bolus Herb. 3: 82

(1921); in Trans. R. Soc. S. Afr. 16: 234 (1928). Syn-

types: Cape, 3419 (Caledon): Houw Hoek Moun-

tain, 750 m (-AA), Schlechter 7382 9 (BOL, lecto.!;

BR!; S!); 5483 cf (B!; BOL!; Z!).

Note

1. R. similis ranges from the coastal hills and

mountains in the Bredasdorp and Caledon areas to

the mountains around Mitchell’s Pass near Ceres,

where it may reach 1 500 m.

Restio singularis Esterhuysen , sp. nov., a omni-

bus speciebus cognitis seminibus laevibus vaginis

subconvolutis et in dimidio superiore membranaceis

diversa.

TYPE. — Cape 3319 (Worcester): Witteberg

above Du Toits Kloof, 1 800 m (-CA), Esterhuysen

30959 9 (BOL, holo.!; E; K; L; M!; MO; S; STE).

Plants forming large dense mats. Culms solid, te-

rete, up to 2 mm in diam., branching, surface finely

tuberculate, lowest portions of culms spreading on

the ground, rooting at the nodes. Sheaths loosely

convoluted, 0,5-2 cm long, coriaceous portion con-

colorous brown, acuminate, extended into a pointed

apex about 3 of the total length of the sheath,

flanked by large, rounded membranous shoulders,

almost as tall as the apex. Male inflorescence of

1 (2), 5-7 x 5 mm spikelets. Spathe like the sheaths,

as tall as or slightly taller than the spikelets. Bracts

rounded-acuminate, 5-6 x 2,5 mm, the upper 3

membranous, the lower § coriaceous, partially

Bothalia 15, 3 & 4 (1985)

461

obscuring the flowers. Flowers several per spikelet,

shortly pedicellate, perianth 3-3,5 mm long. Sepals

cartilaginous; lateral sepals acute, conduplicate,

sparsely villous on the carina; odd sepal flat, obtuse,

glabrous, 3x1 mm. Petals membranous, 2, 5-3, 5 x

0,8-1 mm, obtuse. Anthers exserted at anthesis, 1,8

mm long, mucronate. Pistillode minute. Female in-

florescence of 1 (2) more or less globose, 1-3-flow-

ered, 4—8 mm in diam. spikelets. Spathes and bracts

similar, acute to rounded, membranous apical part

soon decaying, about 9 mm long. Flowers shortly

pedicellate, partially exposed, perianth 3, 5-4, 5 mm

long. Sepals cartilaginous, subacute; lateral sepals

conduplicate, carina sparsely villous in the middle;

odd sepal flat, glabrous, 3, 5-4, 5 x 1-1,2 mm. Stami-

nodes about 1 mm long. Ovary with two fertile and

one rudimentary locule. Styles free, adjacent, three,

about 3 mm long, very densely villous. Fruit a cap-

sule, seeds triangular, ends rounded, ridge weakly

developed, surface smooth, some roughness visible

at x 50, 1,1-1, 3 x 0,8 mm.

Restio singulars occurs on the rugged peaks in the

Slanghoek and Wemmershoek Mountains, above

1 650 m. The recorded habitat is on ledges and be-

low cliffs. No material has been collected at anthesis,

so the flowering time is not known.

I have not been able to determine the affinities of

this new species. On the basis of seed surface,

Linder (1984: 30) placed it in Group lib, but it does

not share any other characters with this group. It ap-

pears to be, as the specific epithet indicates, a rather

unique species.

CAPE. — 3319 (Worcester): Witteberg above Du Toits Kloof

(-CA), Esterhuysen 30959 (BOL; C; E; K; L; M; MO; S; STE);

Esterhuysen 34067 (BOL; K; MO; S); Esterhuysen 35244 (BOL;

K; MO; S); Esterhuysen 27606 (BOL; K; MO; S); plateau SE of

Du Toits Peak (-CC), Esterhuysen 30997 (BOL; K; M; MO; S);

Du Toits Peak (-CC), Esterhuysen 29893 (BOL; K; S); Esterhuy-

sen 29039 (BOL; C; E; K; L; LD; M; MO; S; UC); Esterhuysen

30571 (BOL; C; E; K; L; M; MO; S); Wemmershoek Mts, Win-

terberg (-CC), Esterhuysen 28231 (BOL; C; E; K; L; LD; M;

MO; NBG; PRE; S; STE; UC).

Restio stereocaulis Mast, in Bot. Jb. 29 Beibl.

66 : 1 (1900); Pillans in Trans. R. Soc. S. Afr. 16 :

235 (1928). Syntypes: Cape, 3319 (Worcester):

mountains at Fransch Hoek, 1 060 m (-CC), Sch-

lechter 9302 cf (B. lecto.!; BM!; BOL!; BR!; K! ;

MO!; P!; S!; Z!); 9303 $ (BM!; BOL!; BR!; K!;

MO!; P!; S!; Z!); without precise locality, Thom

1024 $ (K!).

Notes

1. R. stereocaulis is superficially similar to Platy-

caulos, due to its compressed culms. Detailed ana-

tomical studies show that it is a Restio. Morphologi-

cally, the sheaths are not green and with stout awns,

as is typical of Platycaulos.

2. This species occurs at higher altitudes in the

mountains between Fransch Hoek and Ceres.

Restio stokoei Pillans in Trans. R. Soc. S. Afr.

16 : 231 (1928). Syntypes: Cape, 3319 (Worcester):

Wilde Paarde Berg (-CD), Stokoe in BOL 17670 $

(BOL, lecto.!; K!); in BOL 17670 cf (BOL!; K!).

Notes

1. This species is close to R. scaberulus, with

which it is allopatric. It can be distinguished by the

larger flowers and the shorter sheath awns, but these

characters may break down.

2. R. stokoei occurs in the Caledon-Stellenbosch

Mountains, the Riviersonderend Mountains and the

Langeberg west of Swellendam.

Restio strictus N.E.Br. in FI. Cap. 7 : 752

(1900); Pillans in Trans. R. Soc. S. Afr. 16 : 228

(1928). Type: Cape, 3322 (Oudtshoorn): Cradock

Berg, 900 m (-CD), Galpin 4795 9 (K, lecto.!; B!;

BOL!); 4795 cf (B!; BOL!; K!), see Fig. 11.

Note

1. R. strictus has a distribution range that extends

along the mountains from Villiersdorp to George. It

occurs in marshy spots between 900 and 1 500 m.

Restio strobilifer Kunth, Enum. PI. 3 : 398

(1841); Mast, in A. DC., Monogr. Phan. 1 : 282

(1878); in FI. Cap. 7 : 88 (1897); Pillans in Trans. R.

Soc. S. Afr. 16 : 248 (1928). Type: Cape, 3219

(Wuppertal): Ezelbank (-AC), Drege 2474 9 (B,

lecto.!; BM!; K!; MEL!; MO!; OXF!; P!); 2474 cf

(B!; BOL!; K!; MEL!; MO!; OXF!; P!).

Restio sparsus Mast, in Bot. Jb. 29 Beibl. 66 : 2 (1900). Type:

Cape, 3319 (Worcester): mountains at Mitchell’s Pass, 450 m

(-AD), Schlechter 9957(3 (K, lecto.!; BM!; BOL!; BR!; MO!; P!;

S!; Z!).

Notes

1. See R. bolusii for comments on the taxonomic

status of this species. R. strobilifer is the oldest name

in the complex.

2. R. strobilifer occurs in the mountains from the

Cedarberg to the Hex River Mountains, generally

between 1 000 and 1 500 m.

Restio subtilis Nees ex Mast, in J. Linn. Soc.,

Bot. 8 : 251 (1865); in A. DC., Monogr. Phan. 1 : 282

(1878); in FI. Cap. 7 : 87 (1897); Pillans in Trans. R.

Soc. S. Afr. 16 : 237 (1928). Type: Cape, 3419 (Cale-

don): mountains near Grietjiesgat, June (-AA),

Ecklon & Zeyher s.n. $ (B, lecto.!; BOL!; K!; MO!;

P!; S!).

Notes

1. At B there are two sheets annotated by Nees.

The one is 'Restio subtilis 9 , Stellenbosch, Palmiet’,

and the other gives the locality as ‘Worcester, Tul-

bagh’. The former is the lectotype. However, Mas-

ters did not annotate the sheet. The collection from

Tulbagh is quite different and belongs to Ischyrolep-

sis tenuissima.

2. R. subtilis is known from the mountains be-

tween Caledon and Bainskloof.

Restio tetragonus Thunb., Diss. Restio 17

(1788); Thunb., FI. Cap. edn 1 : 331 (1811); edn

Schultes, 87 (1823); Kunth, Enum. PI. 3 : 401

(1841); Mast, in A. DC., Monogr. Phan. 1 : 273

(1878); in FI. Cap. 7 : 81 (1897); Pillans in Trans. R.

Soc. S. Afr. 16 : 223 (1928); in Adamson & Salter,

FI. Cape Penins. 136 (1950). Syntypes: Cape, with-

462

Bothalia 15, 3 & 4 (1985)

out precise locality, Thunberg in herb. Thunb. 23252

(UPS!); 23253 (UPS, lecto.!); 23254 (UPS!).

Notes

1. For notes on the taxonomy of this species, see

R. quadratus.

2. R. tetragonus occurs along the foothills of the

coastal mountains between Cape Town and

Humansdorp. It could well be restricted to shaly or

clayey soils.

Restio triticeus Rottb., Descriptiones Plantarum

Rariorum 11 (1772); Descriptionum et Iconum Ra-

riores 7 (1773); Thunb., FI. Cap. edn 1 : 333 (1811);

edn Schultes, 87 (1823); Mast, in A. DC., Monogr.

Phan. 1 : 277 (1878); in FI. Cap. 7 : 84 (1897); Pillans

in Trans. R. Soc. S. Afr. 16 : 226 (1928); in Adam-

son & Salter, FI. Cape Penins. 136 (1950). Type:

Cape, without precise locality, Konig s.n. cf (C,

holo.!).

Calopsis triticea (Rottb.) Kunth, Enum. PI. 3 : 424 (1841).

Restio pannosus Mast, in J. Linn. Soc., Bot. 8 : 244 (1865); in

A. DC., Monogr. Phan. 1 : 278 (1878); in FI. Cap. 7 : 85 (1897).

Syntypes: Cape, 3318 (Cape Town): mountains at Cape Town

(-CD), Zeyher 1742 cf (MEL, lecto.!; K!; MO!; P!; Z!), see Fig.

6; Cape Flats, Doornhoogte, July (-DC), Ecklon & Zeyher s.n. $

& Cf (BR!; K!; MEL!; P!; S!; Z!); Ecklon s.n. $ & cf (?).

Hypolaena bachmannii Mast, in Bot. Jb. 29 Beibl. 66 : 15

(1900). Type: Cape, Pondoland, Bachmann 344 cf (B, holo.!).

Icon: Rottb. , Descriptionum et Iconum Rariores t.3f. 1 (1773).

Notes

1. The type material of R. pannosus at K is anno-

tated by Masters. There are two sheets of Zeyher

1742 at K, one is annotated ‘R. pannosus’ by Masters

(this is the lectotype) and the other is annotated ‘R.

triticeus’, also by Masters!

2. Although the sheet of Bachmann 344 at B is not

in a good condition, it is clearly R. triticeus, and is

named ‘R. bachmannii’ in Masters’s hand.

3. R. triticeus ranges from Cape Town to the

Transkei and is particularly common in the southern

Cape in places where there are few other Restiona-

ceae.

Restio tuberculatus Pillans in Ann. Bolus Flerb.

3 : 146 (1922); in Trans. R. Soc. S. Afr. 16 : 263

(1928). Nom. nov.

Leptocarpus divaricatus Mast, in Bot. Jb. 29 Beibl. 66 : 9

(1900). Syntypes: Cape, 3118 (Van Rhynsdorp): Koude Berg, 800

m (-DC), Schlechter 8776 9 (B, lecto.!; BM!; BOL!; BR!; K!;

MO!; P ! ; S ! ; Z ! ; ) ; 8777 cf (B!;BM!; BOL! ; BR!; K!; MO!; P! ; S! ;

Z!).

Notes

1. The numbers of the type collections are rever-

sed on some sheets, so that the male is 8776 and the

female 8777. The isolectotypes are always the female

material, irrespective of the number.

2. R. tuberculatus occurs on the arid north-east

side of the Cedarberg, extending northwards as far

as the Giftberg at Van Rhynsdorp.

Restio vallis-simius Linder, sp. nov., R. saro-

clado Mast, affinis, sed spathis bracteis minoribus,

inflorescentiis spiculis numerosis differt.

TYPE. — Cape, 3324 (Steytlerville): eastern

end of the Baviaanskloof Mountains, 1 000 m

(-BB), Linder 3256 $ (PRE, holo.! ; BOL; K; MO).

Plants with diffuse, isolated culms growing

through the vegetation, rising from spreading rhi-

zomes with very few aerial culms. Culms terete,

solid, branching, 50-100 cm long, to 5 mm in diam.,

surface smooth. Sheaths closely convoluted, to 2 cm

long, acuminate with an awn 2-3 mm long, upper

margins narrowly membranous, body coriaceous,

brown to grey. Male inflorescence of up to 12 race-

mose to racemose-paniculate, 1-1,5 cm long, many-

flowered spikelets. Spathes like the sheaths, 0,5-1

cm long. Bracts papyraceous to cartilaginous, with-

out membranous margins, pale straw-coloured and

somewhat red-speckled towards the apices, concave,

subacute, imbricate, obscuring the flowers, 0,5-1 cm

long. Flowers pedicellate, perianth about 3,5 mm

long. Sepals cartilaginous, rounded; lateral sepals

conduplicate, carina sparsely scabrid; odd sepal gla-

brous, flat. Petals membranous, thicker at the base,

about 3 mm long. Anthers exserted at anthesis, pis-

tillode minute. Female inflorescence of up to six

racemose to racemose-paniculate spikelets. Spathes

like the sheaths, about 6 mm long. Bracts subimbri-

cate, 5-10 mm long, cartilaginous, acute, concave,

straw-coloured below with the apical half speckled

red. Flowers on a 1 mm long pilose stipe, perianth

about 4 mm long. Sepals cartilaginous, subacute; lat-

eral sepals conduplicate, carina sparsely pilose; odd

sepal glabrous, flat. Petals membranous, flat,

slightly shorter than the sepals. Staminodes about 1

mm long. Ovary with two fertile locules. Styles

three, free, sparsely villous, free to the base. Fruit a

capsule, seed 2 mm long, triangular, the ends

rounded, keel present as a faint groove, surface

smooth, nitid.

Restio vallis-simius is known from two collections

from the eastern end of the Baviaanskloof Moun-

tains. It grows fairly commonly on the south-facing

slopes at about 1 000 m, up to the summit ridge. The

plants occur dispersed in the vegetation, and with its

long spreading rhizomes, can often become incon-

spicuous.

This species is allied to the R. sarocladus group in

the smooth triangular seed, the ridge developed as a

faint groove, in the cartilaginous bracts without a

membranous margin, the smooth culms and the

growth form, especially the rhizomes. But this new

species differs from R. sarocladus in the spathes

which are smaller than the bracts, the less acute

bracts, the large number of spikelets in each inflores-

cence and the regular flowers. It is rather far away,

geographically, from R. sarocladus, which occurs in

the Caledon, Cape Peninsula and Paarl areas.

CAPE. — 3324 (Steytlerville): eastern end of the Baviaans-

kloof Mountains, 1 000 m (-BB), Linder 3256 (BOL; K; PRE;

MO); Noel 8777 (BOL; GRA).

Restio verrucosus Esterhuysen, sp. nov., species

vaginis in dimidio superiore hyalinis a speciebus affi-

nis distincta; differt a R. harveyi ovariis bilocularis; a

R. arcuato culmis tuberculatis, seminibus collicula-

tis; a R. scaberulo seminibus colliculatis, floribus gla-

bris.

Bothalia 15, 3 & 4 (1985)

463

TYPE. — Cape, 3419 (Caledon): Houw Hoek

Mt. (-AA), Esterhuysen 31113 $ (BOL, holo.!; B;

C; E; F; K; L; LD; M; MO; NBG; NY; PRE; RSA;

S; STE; TCD; UC; US; W; WAG).

Plants caespitose, tussocks 15-30 cm tall. Culms

slender, solid, terete, branching, 0,5 mm in diam.,

roughly tuberculate. Sheaths closely convoluted,

5-10 mm long, body \ - f of total length, coriaceous,

acuminate, topped by broad hyaline shoulders about

half of the total sheath length, the apex of the awn is

also hyaline. Male inflorescence of 3-6 racemose, lax

spikelets. Spikelets about 5 mm long. Spathes similar

to the sheaths, 6-8 mm long, overtopping the spike-

lets. Bracts 3-6 mm long, acute, cartilaginous to pa-

pyraceous below, membranous in the upper half.

Flowers 1-3 per spikelet, shortly pedicellate, per-

ianth 2-2,5 mm long. Sepals subcartilaginous; lateral

sepals subcarinate, glabrous, acute; odd sepal flat,

2,5 x 0,5 mm, more obtuse than the lateral sepals.

Petals membranous, obtuse, 1,2-2 x 0,5 mm. Anth-

ers exserted at anthesis, 1,1 mm long. Pistillode min-

ute, three-lobed. Female inflorescence of 1 - several

lax, racemose spikelets. Spikelets about 5 mm long,

single-flowered. Spathe like the sheaths, about as tall

as the spikelet. Bracts 5-6 mm long, 1,5 mm wide,

acute, obscuring the flower, the upper § membra-

nous, one sterile bract present. Flowers shortly ped-

icellate, perianth 3 mm long. Sepals cartilaginous,

glabrous; lateral sepals subcarinate, acute; odd sepal

flat, subacute. Sepals papyraceous, somewhat

shorter than the sepals. Staminodes minute. Ovary

with two fertile locules and three free styles, fruit a

capsule. Seeds 1,3 x 0,8 mm, rounded at the ends,

flattened on the sides, surface grey, colliculate.

Restio verrucosus has been found on Houw Hoek

Mt only, where it grows in somewhat marshy soil on

the shale band, between 450 and 600 m. Flowering

probably occurs in August, seed release in Decem-

ber to January.

The affinities of this species are difficult. It is most

likely allied to the R. zwartbergense-R. arcuatus-R.

stokoei groups, and may be most closely related to

R. pumilus of the first group by the tuberculate

culms, glabrous flowers, inflorescence structure and

seed surface. It is distinct in its bilocular ovary, sin-

gle flowered spikelet and bracts with the upper half

membranous. The seeds are also more colliculate

than in the R. zwartbergense group.

CAPE. — 3419 (Caledon): Houw Hoek Mt. (-AA). Esterhuy-

sen 31113 (B; BOL; C; E; F; K; L; LD; M; MO; NBG; NY; PRE;

RSA; S; STE; TCD; UC; US; W; WAG); Esterhuysen 32781

(BOL; K; L; M; MO; S); Esterhuysen 35395 (BOL; C; E; K; L;

M; MO; S); Schlechter 7347 (BOL; K).

Restio versatilis Linder , nom. nov.

Hypolaena diffusa Mast, in FI. Cap. 7 ; 132 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 395 (1928); in Adamson & Salter, FI.

Cape Penins. 155 (1950). Mastersiella diffusa (Mast.) Gilg-Ben. in

Pflanzenfam. edn 2, 15a ; 25 (1930). Type: Cape, 3418 (Simons-

town): Nieuw Kloof (-BB), Burchell 8065 cf (K, holo.!; P!).

Notes

1. As there is already a Restio diffusus Spreng.

(1824), I call this species Restio versatilis.

2. The generic position of this species is difficult to

resolve. Superficially, it fitted into Pillans’s artificial

genus Hypolaena, but with the stricter generic defi-

nitions of Linder (1984), it can no longer be placed

into any of the genera into which Hypolaena has

been divided and the best position appears to be

with Restio.

3. This species occurs in the mountains in the Ca-

ledon Division.

4. Pillans (1928) reported that the ovary has two

styles, but in 1952 he corrected this, stating that his

1928 description was based on material of another

species and that the ovary had three styles. I have

found material with both two and three styles, al-

though the three-styled condition is much more com-

mon.

Restio zuluensis Linder, sp. nov., species insig-

nis petalis connatis dimidio inferiore et ovariis tricor-

nibus, ab aliis speciebus Restionis bene distincta.

TYPE. — Natal, 2732 (Ubombo): Vuzi Swamp,

Manzengwenya (-AA), Moll 6005 $ (BOL, holo.!;

K; MO; PRE; S).

Plants up to 40 cm tall, culms ascending from a

creeping, about 3 mm in diameter, rhizome. Culms

solid, terete, branching, about 0,5 mm in diameter,

the surface obscurely rugulose. Sheaths tightly con-

voluted, 7-12 mm long, obtuse to truncate with a 2

- 3 mm long aciculate awn, body pale brown with

fine red speckling, the upper margin narrowly hya-

line. Male inflorescence a single spikelet, 1 - 1,5 cm

long, 2-4 mm in diam, with 3-5 flowers. Bracts

acute, 8 - 10 x 2 - 3 mm, with a short awn, papyra-

ceous to cartilaginous, pale yellow with fine red

speckling. Flowers shortly pedicellate, perianth 3

mm long. Sepals cartilaginous; lateral sepals condu-

plicate, carina sparsely pilose, the apex extended

into a fine point rising between two rounded lobes;

odd sepal flat, obtuse, 2,5 x 0,9 mm. Petals mem-

branous, 2,2 x 0,6 mm. Anthers 1,6 mm long,

shortly mucronate. Pistillode minute, three-lobed.

Female inflorescence similar to male, but with 2-4

flowers. Flowers shortly pedicellate, perianth 3,5

mm long. Sepals similar to male flowers. Petals

membranous, 2 x 0,6 mm, obtuse, fused in the

lower ? to form a tube surrounding the base of the

ovary. Staminodes about 1 mm long. Ovary with two

fertile locules, styles three, free, produced from

three horns on the ovary. Fruit a capsule. Seed not

seen.

Restio zuluensis has been recorded from near sea

level in seasonally waterlogged swamps on the tropi-

cal east coast of South Africa. It flowers in Decem-

ber. It is the only species known in Africa from this

phyto-chorological region.

The affinities of this curious species are not clear.

Anatomically, it shows similarities to R. diffusa from

the Cape (Linder, 1984), but there are no similarities

in flower morphology to support this grouping. The

rhizome is very atypical of the genus Restio, so

maybe it is a very isolated species.

NATAL. — 2732 (Ubombo): Vuzi Swamp, Manzengwenya

(-AA), Moll 6005 (BOL; K; PRE); SE Mbazwana. Mkuzi

Swamps (-DC), Ward 8110 (PRE); 8111 (PRE). 2832 (Mtuba-

tuba): St Lucia Park (-AD), Ward 7330 (PRE).

464

Bothalia 15, 3 & 4 (1985)

Restio zwartbergensis Pillans in Trans. R. Soc.

S. Afr. 16: 229 (1928). Type: Cape, 3419 (Caledon):

Zwarteberg, 900 m (-AB/BA), Schlechter 5562 §

(BOL!;lecto.!;B!;K!;Z!);5562cf(BOL!;K!;Z!).

Notes

1. This species is near R. piimilus and might easily

be confused with it.

2. R. zwartbergensis is a montane species that oc-

curs between Caledon and Worcester, generally at

higher altitudes.

CALOPSIS

9. Calopsis Desv. in Annls Sci. nat., ser. 1, 13 : 44

(1828); Linder in Bothalia 15 : 65 (1984).

Note

1. I have included C. sparsa in the key, although

the species has not yet been described. As it is quite

distinct, I shall describe it at the earliest opportunity.

KEY TO THE SPECIES OF CALOPSIS

la Lateral sepals keeled:

2a Culms simple; styles 1 or 2:

3a Style solitary C. monostylis

3b Styles 2 C. esterhuyseniae

2b Culms branching; styles 3:

4a Spike lets many-flowered C. levynsiae

4b Spikelets 2-3-flowered:

5a Spikelets 10-15 mm long; perianth 5-6 mm long C. impolita

5b Spikelets 6-8 mm long; perianth 3-3,5 mm long C. hyalina

lb Lateral sepals not keeled:

6a Sheaths with the upper portion pale-membranous and deciduous:

7a Internodes in spikelets elongated, flowers exposed:

8a Perianth glabrous C. clandestina

8b Perianth villous C. pulchra

7b Internodes not elongated; floral bracts imbricate:

9a Inflorescence a much branched panicle, 15-20 cm long C. paniculata

9b Inflorescence much smaller:

10a Lateral sepals villous-carinate:

11a Bracts completely cartilaginous C. rigida

lib Bracts with broad hyaline-membranous shoulders:

12a Spikelets 10-15 mm long; tepals with the upper half hyaline C. sparsa

12b Spikelets 6-8 mm long; tepals completely cartilaginous C. aspera

10b Lateral sepals glabrous:

13a Perianth largely exposed above the bracts:

14a Sepals acute, somewhat keeled: perianth 2,75-3 mm long C. membranacea

14b Sepals obtuse, rounded at the back; perianth 2,5 mm long C. nudiflora

13b Perianth obscured by the bracts:

15a Bracts 3-5 mm long C. rigida

15b Bracts 8-15 mm long C. levynsiae

6b Sheaths without a wide membranous and deciduous upper margin:

16a Sheaths convoluted:

17a Spikelets 2-4 mm long C. gracilis

17b Spikelets 5-10 mm long:

18a Sheaths tightly convoluted, acute, awn straight:

19a Spikelets 8-12 mm long; from the inland mountains C. andreaeana

19b Spikelets 5-7 mm long; from the coastal area between Gordons Bay and Bredasdorp

C. fdiformis

18b Sheaths loosely convoluted, obtuse, awn often recurved C. adpressa

16b Sheaths almost flat, usually spreading:

20a Culms with white tubercles C. burchellii

20b Culms obscurely tubercled, wrinkled or smooth:

21a Lateral sepals densely villous:

22a Female spikelets single-flowered; culms much branched C. dura

22b Female spikelets several-flowered; culms somewhat branched C. viminea

Bothalia 15, 3 & 4 (1985)

465

C. fruticosa

21b Lateral sepals glabrous or minutely puberulous:

23a Awn on sheath about J the length of the sheath

23b Awn on sheath £4 of the length of the sheath:

24a Female spikelets aggregated

24b Female spikelets spaced

Calopsis adpressa Esterhuysen, sp. nov., a Cal-

opsis fruticosa (Mast.) Linder vaginis arete convolu-

tis, sepalis quam petalis brevioribus differt.

TYPE. — Cape, 3449 (Caledon); hills inland

from Pearly Beach (-CB), Esterhuysen 32977 9

(BOL, holo.!; E; K; L; M; MO; S).

Plants forming rather disorganized tussocks,

20-50 cm tall. Culms solid, terete, much branched

with the branches ascending, up to 2 mm in diameter

at the base, surface generally rugulose, rarely papil-

lose below. Sheaths more or less convoluted, 7-25

mm long, body coriaceous, margins chartaceous,

acute to obtuse, awn a quarter to half the length of

the sheath, slender, aciculate, often recurved. Male

inflorescence of 1-4, 5-15 x 2^1 mm spikelets.

Spathes similar to the sheaths, about half as long as

the spikelets. Bracts cartilaginous, acuminate, con-

cave, 3-5 x 2 mm. Flowers subpedicellate, perianth

2-3 mm long. Tepals chartaceous, subacute, 2-3 mm

long, lateral sepals conduplicate, glabrous, odd sepal

obsolete. Anthers exserted at anthesis, 1,5-2 mm

long. Female inflorescence of 1-6 several-flowered,

5-10 x 2-4 mm racemose to compound-racemose

spikelets. Spathes like the sheaths, as tall as the spi-

kelets. Bracts all fertile, coriaceous, acuminate with

the apex extended into a tapering awn, 6-8 x 4 mm.

Flowers subsessile, perianth 2,8-4 mm long. Sepals

rounded, chartaceous to cartilaginous, 2,5-3 x

0, 3-0,5 mm, laterals conduplicate, carina very

sparsely villous to glabrous. Petals chartaceous to

cartilaginous, rounded, spathulate, 2,8-4 mm long.

Staminodes minute. Ovary with 3 ribs, unilocular.

Styles 3, 3-4 mm long, from a small cap, slender and

villous. Fruit 3 mm long, unilocular nut.

C. adpressa occurs along the coast between Cape

Infanta and Cape Hangklip and is probably restrict-

ed to limestone-derived soils. All collections are

from below 300 m.

This species is very close to C. fruticosa with which

it is sympatric and which is also a limestone endemic.

The taxa can be separated by the sheaths, which are

spreading in C. fruticosa and pale in colour, whereas

they are convoluted and dark in colour in C. ad-

pressa (hence the name). In addition, in C. fruticosa

the sepals are taller than the petals, whereas the re-

verse holds for most of the collections of C. ad-

pressa, except for Esterhuysen 35638, where they are

of equal size.

C. adpressa is also similar to C. filiformis, but it

differs in the more pilose midribs of the lateral sepals

and the more acute sheaths.

CAPE. — 3418 (Simonstown): N base of Cape Hangklip peak

(-BD), Esterhuysen 35638 (BOL; K; M; MO; S). 3419 (Caledon):

hills inland from Pearly Beach (-CB), Esterhuysen 32969 (BOL;

K; MO; S); Esterhuysen 32977 (BOL; E; K; L; M; MO; S); Soet-

anysberg (-DB), Esterhuysen 34881 (BOL; K; M; MO; S); Cape

L’ Agulhas (-DD), Esterhuysen 29704 (BOL; C: E; F; K; L; M;

MO; NBG; NY; PRE;JLSA; S; STE; TCD; UC; US; W; WAG);

Esterhuysen 29704a (BOL; C; K; M; MO; S); Mierkraal (-DB),

.... C. muirii

C. marlothii

Schlechter 10531, 10532 (K; S). 3420 (Bredasdorp): Fransrietfon-

tein, near Cape Infanta (-BD), Esterhuysen 29358 (BOL; K; L;

M; MO; S); Northumberland Point (-CC), Acocks 22636 (K;

PRE).

Calopsis andreaeana (Pillans) Linder, comb.

nov.

Leptocarpus andreaeanus Pillans in Trans. R. Soc. S. Afr. 16 :

349 (1928). Type: Cape, 3323 (Willowmore): Aasvogelberg

(-AC), Andreae 975 9 (BOL. lecto.l); 975 (BOL!).

Note

1. C. andreaeana is an ‘arid fynbos’ species, oc-

curring in the Swartberg, from Ladismith to Willow-

more, in the dry mountains around Willowmore and

on the Rooiberg. It is probably restricted to sand-

stone or quartzitic soils (but see Esterhuysen 33799),

usually above 800 m, often up to 1 500 m.

Calopsis aspera (Mast.) Linder, comb. nov.

Hypolaena aspera Mast, in J. Linn. Soc., Bot. 10 : 264 (1868);

in A. DC., Monogr. Phan. 1 : 371 (1878); in FI. Cap. 7 : 133

(1897). Syntypes: Cape, 3418 (Simonstown): Nieuw Kloof, Houw

Hoek Mountains (-BB), Burchell 8069 c f (K, lecto.!; BOL!; K!;

P!); 8065 cf (K!; P!); west side of Hottentots Holland Mountains,

May (-BB), Ecklon & Zeyher s.n. O' (K!); without precise local-

ity, Thom 632 O' (K!). 3419 (Caledon): mountains at Grietjiesgat,

June 600-1 200 m (-AA), Ecklon & Zeyher s.n. O’ (K!); Palmiet-

rivier (-AA), Ecklon 951 O’ (B!).

Calorophus asper Kuntze, Rev. 2 : 747 (1891).

Leptocarpus asper (Mast.) Pillans in Trans. R. Soc. S. Afr. 16 :

345 (1928).

Notes

1. Burchell 8065 is the type of Restio versatilis.

2. There are two sheets of Burchell 8069 at K.

One sheet (here selected as lectotype) is named by

Masters.

3. C. aspera occurs between Somerset West and

Bot River and also on Shaws Mountain near Cale-

don. It grows from dunes at sea level to somewhat

marshy slopes at about 450 m.

Calopsis burchellii (Mast.) Linder, comb. nov.

Leptocarpus burchellii Mast, in J. Linn. Soc., Bot. 10 : 222

(1868); in A.DC., Monogr. Phan. 1 : 331 (1878); in FI. Cap. 7 :

117 (1897); Pillans in Trans. R. Soc. S. Afr. 16 : 351 (1928). Syn-

types: Cape, 3421 (Riversdale): between Vet River and Krom-

becks Rivier (-AA), Burchell 7185 9 (K, lecto.!; BOL!); 7185 cf

BOL!; K!). 3321 (Ladismith): foot of the Lange Berge near

Kampsche Berg (-CD), Burchell 7146 cf (BOL!; K!). 3322

(Oudtshoorn): George, Malgaten Rivier at Wolf Drift (-CD),

Burchell 6101 cf (BOL; K!).

Note

1. C. burchellii grows along the base of the moun-

tains from Swellendam to Humansdorp, between

100 and 400 m. It is probably restricted to soils de-

rived from sandstone.

Calopsis clandestina Esterhuysen, sp. nov., a C.

pulchra Esterhuysen culmis validioribus, floribus et

bracteis femineis paene aequantibus flores et brac-

teas masculos, perianthiis femineis glabris differt.

466

Bothalia 15. 3 & 4 (1985)

TYPE. — Cape, 3418 (Simonstown): mountain

above the Porter Reserve (-BD), Esterhuysen 34146

(BOL. holo. !; B; C; E; F; GRA; K; L; M; MO;

NBG; NY; PRE; RSA; S; STE; TCD; UC; US; W;

WAG).

Plants caespitose or forming masses, 20-30 cm

tall. Culms solid, terete, branching, to 1 mm in di-

ameter, surface densely tuberculate. Sheaths convo-

luted, 5-15 mm long, coriaceous portion green,

brown or grey, truncate to obtuse, f to | of the total

sheath length; awn cylindrical to erect, almost as

long as body; membranaceous lobes acute, about as

tall as the awn. Male inflorescence a solitary, very lax

spikelet. Spathe and bracts similar, 5-9 mm long,

acute, concave, the lower §-§ pale brown, cartilagi-

nous, the upper membranaceous. Internodes be-

tween the bracts about 3 of the length of the bract,

leaving the flowers exposed. Flowers subpedicellate,

perianth 3 mm long. Sepals chartaceous to subcarti-

laginous; lateral sepals conduplicate, acute, 3 mm

long; odd sepal flat, rounded, 3 x 0,4 mm. Petals

membranaceous, acute, 2,3 mm long, fused at the

base to an elongated floral axis bearing the inner

whorls. Anthers exserted at anthesis, 2 mm long. Pis-

tillode minute. Female inflorescence similar to the

male. Flowers subsessile, perianth 2, 2-2, 5 mm long.

Sepals subcartilaginous, acute; laterals obscurely

conduplicate, 2 mm long; odd sepal flat, 2 x 0,5

mm. Petals papyraceous with membranous margins,

subacute, 2 x 0,5 mm. Staminodes 0,5 mm long.

Ovary with 1 fertile locule. Fruit a unilocular nutlet,

round in cross-section, 1,2 mm long.

C. clandestina occurs only in the coastal moun-

tains from the Rooi Els River to the Klein River at

Hermanus, at altitudes from sea level up to 1 000 m.

The plants occur in seasonally wet areas, either in

marshes or in shallow sand over rock. In these habi-

tats they can be locally dominant. Some collections

are also from along streams. Flowering occurs in

about November.

C. clandestina is related to C. pulchra, with which

it shares the wide membranaceous margins on the

bracts, the long internodes in the spikelets and the

overall habit. These two species are related to C.

membranacea and C. aspera. C. clandestina is much

more robust than C. pulchra. Furthermore, it differs

in the glabrous flowers and in the female flowers and

bracts being almost the same size as the male flowers

and bracts. The two taxa are also allopatric.

CAPE. — 3418 (Simonstown): mountains above Betty’s Bay

(-BD), Esterhuysen 34603 (BOL; K; M; MO; S); Esterhuysen

34602 (BOL; K; S); Esterhuysen 35391 (BOL; K; S); mountain

above the Harold Porter Reserve (-BD), Esterhuysen 34146 (B;

BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY; PRE; RSA:

S; STE; TCD; UC; US; W; WAG); between Rooiels and

Louwsbos (-BD), Esterhuysen s.n. (BOL; K); near the top of the

Platteberg (-BD), Esterhuysen 34146a (BOL; K; S). 3419 (Cale-

don): Maanschynkop, S side of summit (-AD), Esterhuysen 31697

(BOL; K); Esterhuysen 33645a (BOL; K; S).

Calopsis dura Esterhuysen sp. nov., a C. vimi-

nea (Rottb.) Linder spiculis femineis unifloris, spa-

this minoribus, culmis multum ramosissimis differt.

TYPE. — 3219 (Wuppertal): South Cedarberg

Suurvlakte, between Kaffirkop and Loskop (-CA),

Esterhuysen 34251 (BOL, holo.!; B; C; E; F; GRA;

K; L; LD; M; MO; NBG; NY; PRE; RSA; S; STE;

TCD; UC; US; W; WAG).

Plants more or less caespitose, somewhat disor-

ganized, 15-30 cm tall. Culms solid, terete, branch-

ing profusely with the branches ascending, to 1,5

mm in diameter, surface finely papillose. Sheaths

spreading, 5-15 mm long, coriaceous with charta-

ceous margins, obtuse to acute with a slender acicu-

late, 1^1 mm long awn; sheaths on the spreading

culms larger than on the erect culms, awns relatively

smaller. Male inflorescence of 1-3 almost imbricate,

4—8 x 2-3 mm spikelets. Spathes somewhat smaller

than the bracts, acute to obtuse, body subcoria-

ceous, awn aciculate, almost as long as the body of

the spathe. Bracts cartilaginous to subcoriaceous,

margins chartaceous, acute, about 3 mm long, awn

aciculate, about 0,5 mm long, obscuring the flowers.

Flowers subpedicellate, perianth 2 mm long. Tepals

chartaceous to membranous, with fine red lines, sub-

acute, 5; lateral sepals conduplicate, carina sparsely

villous. Anthers exserted at anthesis, 1,8 mm long,

shortly mucronate. Female inflorescence of 1-8 more

or less clustered, 6x1 mm spikelets. Spathes similar

to males, about 3 mm long. Bracts three, the lower-

most minute, double-ribbed, the second about half

the length of the spikelet, the third enclosing the

perianth and tightly convoluted around the single

flower, coriaceous, acute, about 6 mm long. Flower

subpedicellate 4—5 mm long. Lateral sepals 4-5 x 0,5

mm, acute, conduplicate, carina sparsely long-vil-

lous, cartilaginous. Petals rounded, subspathulate,

membranous with red stripes, 3 mm long. Styles 3,

about 4 mm long, pilose, the basal 1 mm swollen,

forming a cap. Ovary unilocular. Fruit a unilocular

nutlet.

C. dura is known from the southern Cedarberg

and the northern Cold Bokkeveld, where it occurs

between 900 and 1 500 m. It is rather rare and local,

occuring on sandy flats or rocky slopes.

It is clearly related to the C. viminea complex in

the spreading spathes and the villous-carinate flow-

ers. However, this species is readily distinguished by

the much-branched, erect culms, the single-flowered

female spikelets and the very short female spathes.

CAPE. — 3219 (Wuppertal): south Cedarberg, Suurvlakte, be-

tween Kaffirkop and Loskop (-CA), Esterhuysen 34251 (B; BOL;

C; E; F: GRA: K: L; LD; M; MO; NBG; NY; PRE; RSA; S;

STE; TCD: UC; US; W; WAG); Schurweberg Peak, south of

Boboskloof (-CD), Esterhuysen 31914 (B; BOL; C; E; F; GRA;

K; L; LD; M; MO; NBG: NY; PRE; RSA; S; STE; TCD; UC;

US; W: WAG). 3319 (Ceres): slopes above Op-die-Berg (-AB),

Esterhuysen 35924 (BOL; K; MO; S).

Calopsis esterhuyseniae (Pillans) Linder , comb.

nov.

Leptocarpus esterhuyseniae Pillans in Trans. R. Soc. S. Afr. 30:

260 (1945). Type: Cape, 3319 (Worcester): Mitchell's Pass, Slab

Peak, 1 200 m (-AD). Esterhuysen 6211 9 (BOL, lecto.!); 6211 cf

(BOL!; K!).

Note

1. C. esterhuyseniae is largely a high-altitude

species, usually occurring between 1 200 and 1 800

m, from the South Cedarberg to the Witteberg at

Laingsburg, Leeurivier Peak near Swellendam and

the Riviersonderend Mountains. There are a few

collections from as low as 500 m. Habitat notes gen-

Bothalia 15, 3 & 4 (1985)

467

erally refer to stony slopes, rock ledges, talus slopes

or shale bands.

Calopsis filiformis (Mast.) Linder, comb. nov.

Hypolaena filiformis Mast, in J. Linn. Soc., Bot. 10: 267 (1868);

in A. DC., Monogr. Phan. 1: 372 (1878); in FI. Cap. 7: 134

(1897); Pillans in Trans. R. Soc. S. Afr. 16: 397 (1928). Caloro-

phus filiformis (Mast.) Kuntze, Rev. Gen. 747 (1894). Syntypes:

Cape, 3419 (Caledon): Houw Hoek Mountain, July (-AA),

Zeyher 4349 c f (K, lecto.!; BOLL BR!; MEL!; P! ); without pre-

cise locality. Thom 1031 cf ( K! ).

Notes

1. Both Zeyher 4349 and Thom 1031 are mixed

collections. As well as C. filiformis, there is also ma-

terial of Mastersiella digitata and Calopsis hyalina

under Zeyher 4349 and Mastersiella digitata under

Thom 1031.

2. Thom 1031 at K is rather similar to C. adpressa.

These two species are superficially similar, but can

be distinguished by the sheaths and by the degree of

villousness of the female flowers.

3. C. filiformis occurs on the lower mountain

slopes and plateaux between Gordons Bay and Ar-

niston (near Bredasdorp), between sea level and 300

m. It is recorded usually from dry gravelly or sandy

habitats.

Calopsis fruticosa (Mast.) Linder , comb. nov.

Lectocarpus fruticosus Mast, in Bot. Jb. 29 Beibl. 66: 9 (1900);

Pillans in Trans. R. Soc. S. Afr. 16: 352 (1928). Syntypes: Cape,

3419 (Caledon): in hills near the Ratel River, 10 m (-DA), Sch-

lechter 9718 $ (K, lecto.!; BM!; BR!; MO!; P!; S!;); 9717 cf (B!;

BM!; BR!: K!; MO!; P!; S!).

Notes

1. The lateral sepals are, contrary to Pillans (1928:

353), frequently somewhat puberulous to pilose on

the carina.

2. C. fruticosa is a coastal species that occurs on

limestones, marine sands, sandstones or even clay,

usually within one kilometre of the coast. It ranges

from the mouth of the Duivenhoksrivier (near Riv-

ersdale) to the Cape Peninsula.

Calopsis gracilis (Mast.) Linder, comb. nov.

Hypolaena gracilis Nees ex Mast, in J. Linn. Soc., Bot. 10:

266 (1868); in A. DC., Monogr. Phan. 1: 375 (1878); in FI. Cap.

7: 135 (1897). Calorophus gracilis (Mast.) Kuntze. Rev. Gen. 747

(1894). Leptocarpus gracilis (Mast.) Pillans in Ann. Bolus Herb.

3: 146 (1922); in Trans. R. Soc. S. Afr. 16: 347 (1928); in Adam-

son & Salter, FI. Cape Penins. 149 (1950). Syntypes: Cape, 3418

(Simonstown): Muizenberg (-AB), Zeyher 4347 cf (K, lecto.!;

MEL!; P!; S!; Z!); near Simonstown, May (-AB), Zeyher 1006 cf

(K!); Wright 500 cf (P!).

Leptocarpus ramosissimus Pillans in Trans. R. Soc. S. Afr. 30:

262 (1945). Type: Cape, 3418 (Simonstown): Karbonkelberg

(-AB), Kies 201 9 (NBG. holo.!; BOLL K!).

Notes

1. I have not found any material annotated in

Nees’s hand. There are several sheets of Zeyher 4347

in European herbaria that have printed labels

reading ‘Restio gracilis Nees’.

2. Zeyher 4347 at K is a small twig but, in the ab-

sence of material authenticated by Nees or Masters,

it should be adequate as a type.

3. C. gracilis is known only from the southern

Cape Peninsula, extending as far north as Con-

stantiaberg. It grows from sea level to 450 m, often

on stabilized sand, but also on mountain slopes.

4. Some more robust plants may have 2-3 spike-

lets.

Calopsis hyalina (Mast.) Linder , comb. nov.

Hypolaena hyalina Mast, in Bot. Jb. 29 Beibl. 66: 13 (1900).

Leptocarpus hyalinus (Mast.) Pillans in Trans. R. Soc. S. Afr. 16:

344 (1928). Mastersiella hyalina (Mast.) Gilg-Ben. in Pflanzen-

fam. edn 2, 15a: 25 (1930). Syntypes: Cape, 3419 (Caledon):

Koude Rivier, 200 m (-DA), Schlechter 10464 9 (K, lecto.!;

BOLL BR!; MEL!; MO!; PL S!; Z!;); 10463 cf (BL BOL!; BR!;

MEL!; MO!; PL SL, Z!).

Notes

1. Schlechter 10463 in B is annotated ‘an H. impo-

lita forma minor'.

2. C. hyalina grows on low mountains and the

foothills of higher mountains in the Bredasdorp and

Caledon divisions, but with a collection from Rivers-

dale and from Tulbagh respectively. It rarely occurs

above 500 m.

Calopsis impolita (Kunth) Linder , comb. nov.

Restio impolitus Kunth. Enum. PI. 3: 404 (1841); Mast, in J.

Linn. Soc., Bot. 8: 249 (1865). Hypolaena impolita (Kunth) Mast,

in J. Linn. Soc., Bot. 10: 264 (1868); in A. DC., Monogr. Phan. 1:

370 (1878); in FI. Cap. 7: 131 (1897). Leptocarpus impolitus

(Kunth) Pillans in Trans. R. Soc. S. Afr. 16: 343 (1928). Type:

Cape, 3218 (Clanwilliam): between Langevlei and Heerenloge-

ment, below 150 m, July (-BC), Drege67 cf (B, holo.!; BM!; K!;

MEL!; MO!; NY!; OXF!; P!; S!).

Icon: Mason, W. Cape Sandveld Flow, t.3 f.6 (1972).

Notes

1. The spikelets of the type specimens are some-

what smaller than is typical for this species, but the

bract and spikelet structure agrees exactly. The type

is from the arid, northern extreme of the distribution

range and this might account for the smaller, prob-

ably depauperate, spikelets.

2. C. impolita is a ‘sandveld’ species, growing in

the Recent Sands of the coastal forelands from the

Cape Flats to the Olifants River. The populations

occur in dry sand.

Calopsis levynsiae (Pillans) Linder , comb. nov.

Leptocarpus levynsiae Pillans in Trans. R. Soc. S. Afr. 29: 346

(1942). Type: Cape, 3219 (Wuppertal): Katbakkies, Swartrug-

gens, 1 200 m (-DC), Levyns 1845 9 (BOL, lecto.!); 1845 cf

('BOL!).

Note

1 . C. levynsiae occurs on the arid mountains form-

ing the northern edge of the Koue Bokkeveld: the

Katbakkies and Swartruggens Mountains, generally

between 900 and 1 200 m.

Calopsis marlothii (Pillans) Linder , comb. nov.

Leptocarpus marlothii Pillans in Trans. R. Soc. S. Afr. 16: 353

(1928). Type: Cape, 3319 (Worcester): Vaalkloof, near Karoo-

poort (-B), Marloth 9102 9 (BOL, holo.!; B!; K!).

Note

1. C. marlothii occurs on rocky dry mountainsides,

between 1 000 and 1 800 m, generally on dry inland

mountains. It ranges from the Kamiesberg in Nama-

qualand to Worcester and is also recorded from the

Anysberg near Ladismith.

468

Bothalia 15, 3 & 4 (1985)

FIG. 22. — Calopsis pulchra Esterhuysen. a, habit; b, female culm, rooting at the base, x 0,8; c, detail of sheath with broad

membranous margins, x 3; d, female spikelet, note flexuose rachis, x 6; e, female bract with the upper M membra-

nous, x 12; f, female flower, villous on the outside, X 12; g, female lateral sepal and staminodes, X 12; h, female sepal,

x 12; i, female petal x 12; j, ovary with three ribs and widely spaced styles, x 12; k, male inflorescence, x 0,8; 1, male

spikelet, x 6; m, male bracts, x 12; n, male flower, glabrous x 12; o, lateral sepal, x 12; p, male odd sepal, x 12; q,

male petal, x 12. (From Linder 3080.)

Bothalia 15, 3 & 4 (1985)

469

Calopsis membranacea (Pillans) Linder, comb.

nov.

Hypolaena burchellii Mast, in J. Linn. Soc., Bot. 10 : 268

(1868); in A. DC., Monogr. Phan. 1 : 374 (1878); in FI. Cap. 7 :

134 (1897). Calorophus burchellii (Mast.) Kuntze, Rev. Gen. 747

(1891). Leptocarpus membranaceus Pillans in Trans. R. Soc. S.

Afr. 16 : 346 (1928); in Adamson & Salter, FI. Cape Penins. 149

(1950): nom. nov. for H. burchellii Mast. Syntypes: Cape, 3418

(Simonstown): Nieuw Kloof, Houw Hoek Mountains (-BB), Bur-

chell 8116 cf (BOL!; K!). 3419 (Caledon): Baviaanskloof Moun

tains near Genadendal (-BA), Burchell 7894a c f (BOL!; K!; P!);

7632 cf (K, lecto.!; BOL!).

Notes

1. Burchell 7894 is a mixed collection, with the

male being C. membranacea and the female Tham-

nochortus gracilis.

2. I have chosen Burchell 7632 as lectotype, as this

collection at K is annotated by both Masters and Pil-

lans.

3. C. membranacea ranges from the Cape Penin-

sula to Tulbagh, Hermanus and along the Rivier-

sonderend and Langeberg to George. It generally

occurs in mountains between 300 and 1 000 m, often

on steep slopes in dense vegetation, rarely in marshy

areas.

Calopsis monostylis (Pillans) Linder, comb.

nov.

Leptocarpus monostylis Pillans in J1 S. Afr. Bot. 18 : 108

(1952). Type: Cape, 3321 (Ladismith): south slopes of the Lange-

berg near Riversdale (-CC), Esterhuysen 16993 $ (BOL, lecto.!);

16993 cf (BOL!).

Note

1. This curious species grows in the Langeberg be-

tween Swellendam and Riversdale, between 900 and

1 350 m. Most of the collections refer to a marshy

habitat.

Calopsis muirii (Pillans) Linder, comb. nov.

Leptocarpus muirii Pillans in Trans. R. Soc. S. Afr. 16 : 353

(1928). Syntypes: Cape, 3321 (Ladismith): north end of Garcias

Pass (-CC). Muir 3179 $ (BOL, lecto.!); 3178 cf (BOL!).

Notes

1. This species may be difficult to distinguish from

C. marlothii.

2. It is known from both the northern (on sand-

stone) and the southern (on shale) bases of the

Langeberg at Garcias Pass.

Calopsis nudiflora (Pillans) Linder, comb. nov.

Leptocarpus nudiflorus Pillans in Trans. R. Soc. S. Afr. 30 : 261

(1945). Type: Cape, 3418 (Simonstown): Somerset Sneeukop,

1 200 m (-BB), Esterhuysen 8230 $ (BOL, lecto.!; K!); 8230 cf

(BOL!; K!).

Note

1. C. nudiflora grows in the mountains from

Jonkershoek to Kogelberg, at altitudes between 900

and 1 500 m. The habitat is usually recorded as mar-

shy or peaty, sometimes as streambanks or wet

ledges.

Calopsis paniculata (Rottb.) Desv. in Annls Sci.

nat., ser. 1.13 ; 44 (1828); Kunth, Enum. PI. 3 : 421

(1841). Type: Cape, without precise locality, Konig

s.n. cf (C, holo.!; BM!).

Restio paniculatus Rottb., Descriptiones Plantarum Rariorum

10 (1772); Descriptionum et Iconum Rariores 4 (1773); Thunb.,

FI. Cap. edn 1, 339 (1811); edn Schultes, 89 (1823). Leptocarpus

paniculatus (Rottb.) Mast, in J. Linn. Soc., Bot. 10 : 221 (1868);

in A. DC., Monogr. Phan. 1 : 330 (1878); in FI. Cap. 7 : 116

(1897); Pillans in Trans. R. Soc. S. Afr. 16 : 357 (1928); in Adam-

son & Salter, FI. Cape Penins. 150 (1950).

Restio ramiflorus Nees in Linnaea 5 : 644 (1830). Calopsis ram-

iflora (Nees) Kunth, Enum. PI. 3 : 423 (1841). Syntypes: Cape,

without precise locality, Restio thamnochortus’ in herb. Willd.

(B, lecto.!); Zeyher s.n., in herb. Giintheri (B!).

Restio pondoensis Mast, in Bot. Jb. 29 Beibl. 66 : 5 (1900).

Type: Cape, Pondoland, Bachmann 343 cf (B, holo.!).

Icon: Rottb., Descriptionum et Iconum Rariores t.2f.3( 1773).

Notes

1. Both syntypes of R. ramiflorus are at B. The

specimen in herb. Willd. is easy to locate and is cho-

sen as lectotype for this reason.

2. The type of Restio pondoensis is given by Mas-

ters as Bachmann 342, but the specimen in B, which

is clearly the type, is Bachmann 343.

3. C. paniculata grows along streams and less com-

monly in marshy areas, from the Cedarberg to Na-

tal, south of Durban. It is probably restricted to

sandstone or quartzitic soils and only rarely occurs

above 600 m.

Calopsis pulchra Esterhuysen, sp. nov., a C. as-

pera (Mast.) Linder internodiis bracteis jplo longio-

ribus, a C. clandestina Esterhuysen culmis magis ten-

uibus, floribus et bracteis femineis et bracteis mascu-

lis minoribus, ab ambabus perianthiis femineis pilo-

sis diversa.

TYPE. — Cape, 3419 (Caledon): along the

road from Elim to Stanford, 5| miles SSE of the

Papiesvlei turn-off (-DA), Esterhuysen 31255 $

(BOL, holo.!; B; BOL; C; E; F; GRA; K; L; LD;

M; MO; NBG; NY; PRE; RSA; S; STE; TCD; UC;

US; W; WAG).

Plants forming tangled tussocks, 15-50 cm tall.

Culms solid, terete, branching, slender, 0,5 mm in

diameter, somewhat flexuose, rugulose to obscurely

tuberculate. Sheaths convoluted, 4-12 mm long, co-

riaceous portion green when young, brown when

old, acute, the apex extended into a slender acicu-

late or stout cylindrical awn about as long as the

body; membranous lobes more or less acute, almost

as tall as the awn. Male inflorescence a solitary, very

lax spikelet. Bracts 4-7 x 1 mm, spreading at 45°,

concave, 3-10, lower cartilaginous, pale brown,

the upper 2-3 membranous, acute. Internodes about

half of the length of the bracts, exposing the flowers.

Flowers subsessile, perianth 2,5-3 mm long. Sepals

chartaceous to cartilaginous, 2,5-3 mm long, suba-

cute; lateral sepals conduplicate, glabrous to

sparsely pilose on the carina; odd sepal flat. Petals

membranous, similar to the odd sepal. Anthers ex-

serted at anthesis, 2 mm long. Female inflorescence

similar to the male, but the bracts are 1,5-5 mm

long, 3-8 per spikelet. Spathes sometimes present,

similar to the bracts. Flowers subpedicellate, per-

ianth 1,5 mm long. Tepals more or less chartaceous,

1,5 mm long, the backs, especially the margins and

midribs, sparsely pilose; sepals more carinate, more

acute than the petals, lateral sepals villous-carinate.

470

Bothalia 15, 3 & 4 (1985)

Staminodes about 0,5 mm long. Styles 3, slender, vil-

lous, widely separated at the base. Ovary unilocular,

trigonous. Fruit a unilocular nut, about 1,5-2 mm

long. Fig. 22.

C. pulchra occurs on the coastal shelf between

Struis Bay and the Koude Berge (Baard-

skeerdersbos). The altitudes are below 300 m and

the substrate is mostly a gravelly or sandy conglom-

erate. Flowering occurs in September. The males ap-

pear to be more conspicuous than the females.

This species is related to the C. membranacea-C.

aspera group by the large hyaline apex on the floral

bracts, the habit of the plants (particularly to that of

C. aspera ), i.e. erect culms often with several

branches from the same node, and the flower struc-

ture. However, C. pulchra and C. clandestina are

distinguished by having much elongated internodes

in the spikelets, so that the flowers are exposed. C.

pulchra is easily distinguished from C. clandestina

(of which it is probably a vicariant) by the female

flowers and bracts being half the size of the males,

the pilose female perianth and the more slender

culms.

CAPE. — 3419 (Caledon): inland from Pearly Beach (-CB),

Esterhuysen 32966 (BOL; K; M; MO; S); Pheasantshoek near

Viljoenshof (-DA), Esterhuysen 34458 (BOL; C; E; F; K; L; M;

MO; NBG; PRE; S; STE; TCD; UC); near Viljoenshof (-DA),

Linder 3080 (PRE); along the road from Elim to Stanford, 51

miles SSE from the Papiesvlei turn-off (-DA), Esterhuysen 31255

(B; BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY; PRE;

RSA; S; STE; TCD; UC; US; W; WAG). 3420 (Bredasdorp):

along the road from Bredasdorp to Cape L'Aghulhas, before

Zoetendalsvlei (-CA), Esterhuysen 31239 (BOL; C; E; F; K; L;

LD: M; MO: NBG; NY; PRE; S; STE; UC; US; W; WAG).

Calopsis rigida (Mast.) Linder , comb. nov.

Leptocarpus rigidus Mast, in Bot. Jb. 29 Beibl. 66 : 10 (1900);

Pillans in Trans. R. Soc. S. Afr. 16 : 350 (1928). Syntypes: Cape,

Ceres Div., Koue Bokkeveld, 1 200 m, Schlechter 8883 $ (K,

lecto. !; BM!; BOL!; BR!; K!;P!;S!;Z!); 8882 cf(B!;BM!; BOL!;

BR!; K!; P!; S!; Z!).

Notes

1. The lateral sepals vary from minutely puberu-

lous along the carina to distinctly villous.

2. C. rigida ranges from the Koue Bokkeveld

along the inland mountains (Hex River Mountains,

Keeromsberg, Witteberg, Swartberg, Anysberg) to

Meiringspoort. It occurs on steep rocky slopes and

shale bands between 600 and 1 500 m.

Calopsis rigorata (Mast.) Linder, comb. nov.

Leptocarpus rigoratus Mast, in Bot. Jb. 29 Beibl. 66 : 9 (1900);

Pillans in Trans. R. Soc. S. Afr. 16 : 352 (1928); in Adamson &

Salter, FI. Cape Penins. 150 (1950). Type: Cape, 3419 (Caledon):

in hills near Vogelvlei, near Elim, 80 m (-DB), Schlechter 10490

$ (B, lect.!; BOL!; K!; MO!; P!; S!; Z!).

Leptocarpus rigoratus Mast. var. simulans Pillans in Trans. R.

Soc. S. Afr. 16 : 352 (1928). Type: Cape, 3418 (Simonstown):

between Strand and Eerste River (-BB), Pillans 4909 $ (BOL,

holo.!; K!).

Note

1 . This species is probably not distinct from C. vi-

minea. The key in Pillans (1928) does not adequately

separate the taxa.

Calopsis viminea (Rottb.) Linder, comb. nov.

Restio vimineus Rottb., Descriptiones Plantarum Rariorum 10

(1772); Descriptionum et Iconum Rariores 4 (1773). Leptocarpus

vimineus (Rottb.) Pillans in Trans. R. Soc. S. Afr. 16 : 354 (1928);

in Adamson & Salter, FI. Cape Penins. 150 (1950). Type: De-

scriptionum et Iconum Rariores t. 2 fig. 1 (1773) (Iconotype).

Restio pauciflorus Poir. in Lam., Encycl. 6 : 168 (1804); Kunth,

Enum. PI. 3 : 412 (1841); Mast, in A. DC., Monogr. Phan. 1 : 235

(1878); in FI. Cap. 7 : 66 (1897). Type: Cape, without precise

locality, in herb. Lam. (P, holo.!).

Calopsis festucacea Kunth, Enum. PI. 3 : 425 (1841). Leptocar-

pus festucaceus (Kunth) Mast, in Bot. Jb. 29 Beibl. 66 : 8 (1900).

Type: Cape, 3218 (Clanwilliam): plains between Jakkals Rivier

and Klipfontein (-BA), Drege 2481 $ (B, holo.!; K!; MO!; NY!;

OXF!; P!).

Calopsis hirtella Kunth, Enum. PI. 3 : 426 (1841). Leptocarpus

peronatus (Kunth) Mast. var. hirtellus (Kunth) Mast, in J. Linn.

Soc., Bot. 10 : 224 (1868); in A.DC., Monogr. Phan. 1 : 334

(1878); in FI. Cap. 7 : 119 (1897). Leptocarpus vimineus (Rottb.)

Pillans var. hirtellus (Kunth) Pillans in Trans. R. Soc. S. Afr. 16 :

356 (1928). Type: Cape, 3219 (Wuppertal): between Grasberg Ri-

vier and Watervals Rivier (-AB), Drege 2500 cf (B, lecto.!; BM!;

K!; MO!; OXF!; P!).

Calopsis oxylepis Kunth, Enum. PI. 3 : 427 (1841). Leptocarpus

oxylepis (Kunth) Mast, in J. Linn. Soc., Bot. 10 : 223 (1868); in

A.DC., Monogr. Phan 1 : 332 (1878); in FI. Cap. 7 : 118 (1897).

Syntypes: Cape, 3219 (Wuppertal): sandy flats near Ezelbank

(-AC), Drege 39 $ (B, lecto.!; K!; MEL!; MO!; NY!; P!; S!);

2501 $ (P!; MEL!; MO!; NY!).

Calopsis peronata Kunth, Enum. PI. 3 : 426 (1841). Leptocar-

pus peronatus (Kunth) Mast, in J. Linn. Soc., Bot. 10 : 224

(1868); in A.DC., Monogr. Phan. 1 : 333 (1878); in FI. Cap. 7 :

118 (1897). Syntypes: Cape, 3219 (Wuppertal): between Grasberg

Rivier and Watervals Rivier (-AB), Drege 2499 $ (B, holo.!; B!;

BM!; K!; MEL!; MO!; OXF!; S!). Ceres Division, Bokkeveld

mountains at Uien Valley, Drege 1623 cf (B!; BM!; BOL!; K! ; P! ;

OXF!).

? Restio ecklonii Mast, in J. Linn. Soc., Bot. 8 : 236 (1865); in

A.DC., Monogr. Phan. 1 : 250 (1878); in FI. Cap. 7 : 75 (1897).

Syntypes: Cape, Caledon Div., between Bot River and Swart-

berg, August, Ecklon & Zeyher s.n. cf (?); without precise local-

ity, Ecklon 85 cf (?); Drege 2497 cf (B, lecto!).

Restio incurvatus Pillans in Trans. R. Soc. S. Afr. 16 : 355

(1928), non Nees in Linnaea 5 : 642 (1830), nom. illeg. Leptocar-

pus incurvatus Pillans in Trans. R. Soc. S. Afr. 16 : 355 (1928),

non Mast, in J. Linn. Soc., Bot. 10 : 233 (1868), nom. illeg. Cal-

opsis incun’ata Pillans in Trans. R. Soc. S. Afr. 16 : 355 (1928),

non Kunth, Enum. PI. 3 : 427 (1841), nom. illeg.

leones: Rottb., Descriptionum et Iconum Rariores t.2 f.l

(1773). Mason, W. Cape Sandveld Flow, t.3 f.5 (1950).

Notes

1. I have not found any Konig specimens of C. vi-

minea, consequently the plate published by Rott-

boell must serve as an iconotype.

2. Kunth (1841) appears to have given every

Drege specimen from the arid sandy flats north of

the Cedarberg a different name. Some of the collec-

tions annotated by Kunth are still extant at B (i.e.

Drege 2481, 2499), whereas in others the sheets at B

were acquired later (i.e. Drege 2500, ex herb.

Liibeck, Drege 39, ex herb. Altona). In these cases,

the sheets at B are designated as the lectotypes.

3. Drege 2500, the type of C. hirtella, is very vari-

able with regard to the presence of pubescence. As

the culms are explicitly stated to be pubescent in the

protologue, only the elements with pubescence can

be regarded as types.

4. C. peronata is based on two elements. Drege

2499 is C. viminea, whereas Drege 1623 is the type of

Ischyrolepis fraterna. Vegetatively, these two taxa

Bothalia 15, 3 & 4 (1985)

471

are very similar and are easily confused. Kunth que-

ried Drege 1623 as belonging to C. peronata and this

supports the lectotypification of Drege 2499.

5. The identity of R. ecklonii is dubious. Pillans

(1928) placed it in the synonymy of lschyrolepis sie-

beri, but did not cite any of the type material. Drege

2497 in B is clearly C. viminea, but Masters com-

mented: ‘Drege’s 2497, of which I have only seen a

small fragment, seems very like the above, but the

spikelets are much larger’.

6. Kunth (1841) and Masters (1878, 1897) have

misconstrued R. incurvatus Thunb. and have re-

garded it as being a Leptocarpus (i.e. Calopsis ). Pil-

lans (1828) corrected this error, but in doing so pub-

lished three illegitimate names.

7. I am interpreting this species here in a wide

sense, including all variation in one name. A de-

tailed study of the variation within populations and

between them, is essential to place any subdivision

of the species on a sound footing.

8. C. viminea ranges from the Kamiesberg to

Laingsburg, the Cape Peninsula and Bredasdorp.

There are a few collections from further east, as far

as Port Elizabeth. The species ranges from sea level

to 1 500 m, from a wide range of habitats.

THAMNOCHORTUS

10. Thamnochortus Berg., Descr. PI. Cap. 353

(1767); Linder in Bothalia 15 : 65 (1984).

Note

1. Care has to be taken with the interpretation of

branches in couplet 6, as most species develop sterile

branches on the culms the year after flowering,

whereas only a few have them concomitant with

flowering.

KEY TO THE SPECIES OF THAMNOCHORTUS

la Culms more or less velvety pubescent:

2a Rhizomes well developed; wings of the flowers well exserted from behind the bracts T. fruticosus

2b Rhizomes absent; wings of the flowers usually obscured by the bracts:

3a Perianth broadly obovate, culms usually completely covered in sheaths; plants to 30 cm tall

, T. acuminatus

3b Perianth elliptical to orbicular; culms at least partially exposed; plants 50-100 cm tall:

4a Sterile shoots with long fine silky branchlets; female flowers hardly exposed T. cinereus

4b Sterile shoots with short rigid glabrous branchlets; female flowers often exposed T. rigidus

lb Culms glabrous:

5a Perianth longer than wide:

6a Fertile culms with branches while flowering:

7a Branches usually fertile T. levynsiae

7b Branches usually sterile:

8a Perianth 2,5 mm long T. obtusus

8b Perianth at least 3 mm long:

9a Perianth 3,5-4 mm long; bracts entirely chartaceus and pale in the upper half T. pellucidus

9b Perianth 4-6 mm long; bracts not as above:

10a Leaf-sheaths on sterile branchlets with an apiculate blade; from Knysna to Natal T. glaber

10b Sheaths on sterile branchlets with a muticous blade; from the SW Cape T. guthrieae

6b Fertile culms simple while flowering:

11a Bracts mostly cartilaginous, brown:

12a Bracts aristate:

13a Perianth 3,5-5 mm long T. nutans

13b Perianth 6-7 mm long T. pulcher

12b Bracts muticous:

14a Stipe of the flowers with distinctly exposed strips between the decurrent sepal bases:

15a Bracts acuminate; perianth 3,5—4 mm long; lateral sepals narrowed at the apex....T. fraternus

15b Bracts acute; perianth 4,5-6 mm long; lateral sepals scarcely narrowed at the apex

T. paniculatus

14b Stipe with one or no exposed strips:

16a Perianth 2-3 mm long:

17a Plants caespitose; 50-80 cm tall T. pluristachyus

17b Plants with long spreading rhizomes, 20-40 cm tall T. obtusus

16b Perianth 4—6 mm long:

18a Bracts pale brown with reddish streaks T. guthrieae

18b Bracts concolorous dark brown:

19a Perianth 5-6 mm long; keel on sepals broadest at or above the middle T. muirii

19b Perianth 4 mm long; keel on the sepals widest above the middle T. ellipticus

472

Bothalia 15, 3 & 4 (1985)

lib Bracts mostly membranous, hyaline:

20a Spikelets 1(2):

21a Culms finely tuberculate T. scabridus

21b Culms smooth:

22a Perianth 7 mm long; bracts entirely hyaline T. papyraceus

22b Perianth 4-5 mm long; bracts at least partially brown:

23a Plants rhizomatous; flowers broadly elliptical T. schlechteri

23b Plants caespitose; flowers suborbicular T. stokoei

20b Spikelets several:

24a Bracts aristate T. dumosus

24b Bracts muticous:

25a Flowers 3,5-4 mm long; bracts largely hyaline T. pellucidus

25b Flowers 4-5 mm long; bracts only with narrow hyaline margins T. guthrieae

5b Perianth at least as wide as long:

26a Lateral sepals with a wing at least 2 mm wide T. platypteris

26b Lateral sepal wing 0,75-2 mm wide:

27a Culms stout, at least below; inflorescence mostly exceeding 10 cm, often much branched;

28a Perianth generally obscured by the bracts T. insignis

28b Perianth conspicuously protruding from the sides of the bracts:

29a Spikelets ovate, obovate or rotundate, light brown T. erectus

29b Spikelets elliptic-oblong to ovate-oblong, dark brown T. spicigerus

27b Culms slender, inflorescence usually less than 10 cm long, rarely branched:

30a Culms generally sulcate throughout when dried; nut visible above petals, as long as sepals

T. bachmannii

30b Culms usually terete; nut obscured by petals, shorter than lateral sepals:

31a Spikelets mostly solitary T. platypteris

31b Spikelets several to many:

32a Fertile culms with branches:

33a Branches sterile, from near the base of the culms:

34a Bracts with broad membranous margins, finely awned; perianth 3 mm long.. T. arenarius

34b Bracts almost completely brown, muticous; perianth 3-4 mm long T. sporadicus

33b Branches fertile T. gracilis

32b Fertile culms simple:

35a Bracts 1-1,5 cm long; wings 1,25-1,5 mm wide T. stokoei

35b Bracts 0,7-1 cm long; wings of sepals 0,75-1 mm wide:

36a Plants without rhizomes; culms always smooth T. lucens

36b Plants rhizomatous; culms smooth or punctate:

37a Culms punctate, clustered on rhizomes T. punctatus

37b Culms smooth, evenly spaced on the rhizomes:

38a Bracts with broad membranous margins, finely awned; perianth 3 mm long

T. arenarius

38b Bracts almost completely brown, muticous; perianth 3^4 mm long T. sporadicus

Thamnochortus acuminatus Pillans in Trans. R.

Soc. S. Afr. 29: 349 (1942): Syntypes: Cape, 3319

(Worcester); Ceres, Roodeberg (-BC), Esterhuysen

1506 ? (BOL, lecto.!; K!); 1505 cf (BOL!; K!).

Notes

1. T. acuminatus is superficially similar to T. fruti-

cosus, but is distinct in its caespitose habit.

2. T. acuminatus is a montane species, growing be-

tween 1 200 and 2 000 m in the mountains from the

Hex River Valley to Pakhuis Pass above Clanwil-

liam.

Thamnochortus arenarius Esterhuysen , sp.

nov., a T. lucenti (Poir.) Linder rhizomatibus prae-

sentibus, a T. obtuso Pillans rhizomatibus brevibus,

bracteis mucronatis, floribus 2,5-3 mm longis, a T.

sporadico Pillans longitudine tepalorum equalium

differt.

TYPE. — Cape, 3418 (Simonstown): Betty’s

Bay, on sandy slopes at the base of the mountain

(-BD), Esterhuysen 34674 9 (BOL, holo.!; C; E; F;

K; L; LD; M; MO; NBG; NY; S; STE; UC; US;

W).

Plants 30-80 cm tall. Rhizomes 2-4 mm in di-

ameter, culms closely arranged, scales nitid, reddish

dark brown, obtuse, imbricate, about 5 mm long.

Culms solid, terete; fertile culms simple at anthesis,

up to 2(3) mm in diameter, smooth; sterile culms ris-

ing either from the rhizome or from the fertile culms

of the year before, up to 15 cm long, flexuose, much

branched. Sheaths on the main axes tightly convo-

Bothalia 15, 3 & 4 (1985)

473

luted, 3-5 cm long, coriaceous at the base, charta-

ceous to membranous above, acute, piliferous, the

apical portion soon decaying and becoming lacer-

ated; sheaths on the sterile culms green, to 10 (17)

mm long, the sheathing portion 2-3 mm long, acute,

with membranous, soon lacerated shoulders to 7 mm

long, awn stout, acute, often curved, to 7 (15) mm

long. Male inflorescence 3-6 cm long from 1-3

nodes. Spathes like the sheaths, somewhat shorter.

Inflorescence branches several, either simple or

branched, flattened. Spikelets 10-15, pendulous,

10-15 x 6 mm. Spathe similar to the bracts, but

somewhat longer. Bracts 4-6 x 2-3 mm, acute, fi-

nely mucronate, chartaceous with wide hyaline mar-

gins, obscuring the flowers. Flowers shortly pedicel-

late, perianth 3 mm long. Tepals chartaceous to

membranous, acute; lateral sepals conduplicate, car-

inate, mucronate; odd sepal 3 x 0,6 mm; petals 2,5

x 1 mm. Anthers not exserted at anthesis, 1,5 mm

long. Female inflorescence 2-10 cm long. Spikelets

2-10, erect, 10-20 x 6 mm, spathes and bracts as in

the males, wings of flowers exserted at the side of

the bracts. Flowers subsessile, perianth 2,5-3 mm

long, 3-3,5 mm wide, transversely broadly elliptical.

Tepals chartaceous to membranous, margins widely

hyaline, rounded; lateral sepals 2,3 mm long, condu-

plicate, wings 0,7 mm wide; odd sepal 2 x 0,5 mm;

petals ovate, 2 x 1,2 mm. Style solitary, 2 mm long,

densely villous on one side, ovary unilocular. Fruit a

thin-walled ellipsoid nutlet, dispersed with the per-

ianth.

T. arenarius occurs on the Cape Peninsula and

along the coast from Cape Hangklip to Hermanus,

from sea level to 450 m. The collections are all from

sandy places.

This species is related to T. lucens and its allies. It

can be separated from T. lucens and T. sporadicus

by the smooth shiny nutlet, free or easily separated

from the perianth and by the smaller flowers with

tepals equal in length. It differs from T. obtusus in

the larger flowers and spikelets, the acute mucronate

bracts and the rhizomes with closely, never distantly

arranged culms.

CAPE. — 3418 (Simonstown): Steenberg Plateau, along the

road to the ranger’s house (-AB), Esterhuysen 31959a (BOL, K,

S); Muizenberg Mountain (-AB), Esterhuysen 35925 (B; BOL; C;

E; F; GRA; K; L; LD; M; MO; PRE; RSA; S; UC; WAG); Es-

terhuysen 35213 (BOL, K); Vlakkenberg (-AB), Esterhuysen

33<§67(BOL; K; MO; S); Cape Point Reserve (-AB), Esterhuysen

31071 (B; BOL; C; E; F; K; L; M; MO; S; STE; UC); Betty’s Bay

(-BD), Esterhuysen 34674 (BOL; C; E; F; K; L; LD; M; MO;

NBG; NY; S; STE; UC; US; W); Esterhuysen 32156 (BOL; K; L;

M; MO; S); Esterhuysen 34306 (BOL; K). 3419 (Caledon); Her-

manus, along the Rotary Drive (-AC), Esterhuysen 34978 (BOL;

C; E; F; K; L; LD; M; MO; S).

Thamnochortus bachmannii Mast, in Annin, na-

turh. Mus. Wien 15 : 11 (1900); Pillans in Trans. R.

Soc. S. Afr. 16 : 374 (1928). Syntypes: Cape, 3318

(Cape Town): vicinity of Hopefield (-AB), Bach-

mann 1773 $ (B, lecto.!; Z!); Olifantsriver, Penther

386 (?); 398 (?).

Thamnochortus sulcatus Mast, in Bot. Jb. 29 Beibl. 66 : 11

(1900). Type: Cape, 3218 (Clanwilliam): Brakfontein (-BD),

Ecklon & Zeyher 76 $ (B, lecto.!); 76 cf (B ! ) .

Notes

1. The material which Masters based his T. bach-

mannii on is from W, but it was destroyed during the

war. Consequently the material from B can only be

lectotype material.

2. The sheet of Ecklon & Zeyher in B is annotated

‘Restio (Thamnochortus) sulcatus’ by Nees, and

‘ Thamnochortus sulcatus Mast, ex Nees’ by Masters.

The material is rather depauperate and especially

the male spikelets are smaller than usual, but the fe-

male material is clearly the same as T. bachmannii.

3. T. bachmannii is an ‘arid sandy’ species, that

ranges from Malmesbury to northern Namaqualand

and also occurs in the Worcester-Wolseley Valley.

In the south the species occurs on the coastal fore-

lands or the sandy valley bottoms, but in Namaqua-

land it reaches up to 1 200 m. It grows both on sandy

or gravelly soils.

Thamnochortus cinereus Linder, sp. nov., a T.

fruticoso Berg, rhizomatibus nullis, a T. rigido Es-

terhuysen culmis sterilibus longis, gracilibus, villosis,

ab ambabus alis florum feminorum non exsertis re-

cedit.

TYPE. — Cape, 3322 (Oudtshoorn): Montagu

Pass, S base, between bridge and Toll-house (-CD),

Linder 3002 $ (PRE, holo.!; BOL!; K!).

Thamnochortus argenteus Pillans in Trans. R. Soc. S. Afr. 16 :

362 (1928) non Kunth (1841), nom. illeg., later homonym.

Notes

1. T. argenteus (Thunb.) Kunth is based on Restio

argenteus Thunb., which is the basionym of

Hypodiscus argenteus (Thunb.) Mast. Pillans

(1928) attempted to save the name by stating 'T. ar-

genteus, Kunth . . . excl. syn. omni’. Since this ex-

cludes the basionym of T. argenteus (Thunb.)

Kunth, Pillans thereby described T. argenteus Pillans

as a new species, which is then a later homonym.

2. ‘Restio scariosus’ in herb. Thunberg is T. cine-

reus, but in the publication of the name Thunberg

cited the earlier T. fruticosus in the synonymy,

thereby rendering Restio scariosus Thunb. illegiti-

mate. Brown (1810) suggested the combination

Thamnochortus scariosus, without actually making

it. The combination has to be attributed to Sprengel.

3. T. cinereus grows in the wet mountains from

Swellendam to Humansdorp and does not occur in

the inland ranges. It grows from sea level to about

1 000 m, usually at lower altitudes, in well-drained

habitats.

Thamnochortus dumosus Mast, in Bot. Jb. 29

Beibl. 66 : 11 (1900); Pillans in Trans. R. Soc. S.

Afr. 16 : 373 (1928). Syntypes: Cape, 3318 (Cape

Town): Kasteelspoort (-CD), Dod 1367 cf (BM!;

K!). 3319 (Worcester): mountains near Saron

(-AA), Schlechter 10656 cf (BM!; BR!; MO!;P!;S!;

Z!). 3419 (Caledon): Houw Hoek, 600 m (-AA),

Schlechter 7786 $ (B, lecto.!; BM!; BOL!; BR!; K!;

MO!; P! ; S! ; Z!); 7787 cf (BOL!; BR!; K!; MO!; P!;

S!; Z!); Zwarteberg, near Sandfontein (?-BA),

Schlechter 10348 cf (BR!; MO!; P!; S!; Z!); 10349 $

(B!; BR!; MO!; P!; S!; Z!).

Thamnochortus canescens Mast, in Bot. Jb. 29 Beibl. 66 : 12

(1900). Syntypes; Cape, 3419 (Caledon): hills at Koude Rivier,

200 m (-DA), Schlechter 10453 $ (B, lecto.!; BM!; BOL!; BR!;

474

Bothalia 15, 3 & 4 (1985)

K!; MO!; P!;S!;Z!); 10452 cf (B!; BM!; BOL!; BR!; K!; MO!; P!;

S!; Z!); without precise locality, Schlechter 10462 cf (?).

Notes

1. Dod 1367 cf is actually T. lucens.

2. T. dumosus occurs on the lower slopes and flats

from the Steenbras River to Cape Agulhas. All col-

lections are from below 600 m. Populations are re-

corded from sandy or slightly marshy habitats.

Thamnochortus ellipticus Pillans in Trans. R.

Soc. S. Afr. 16 : 368 (1928). Type: Cape, 3321 (La-

dismith): Phisantefontein, north slopes of the

Langeberg (-CC), Muir 3187 $ (BOL, holo.!; K!).

Notes

1 . The type specimen at K has no flowers.

2. T. ellipticus is only known from the type local-

ity. It may be a depauperate form of T. muirii, but it

is more likely a quite distinct species.

Thamnochortus erectus (Thunb.) Mast, in J.

Linn. Soc., Bot. 14 : 419 (1874); in A. DC., Monogr.

Phan. 1 : 321 (1878); in FI. Cap. 7 : 125 (1897); Pil-

lans in Trans. R. Soc. S. Afr. 16 : 380 (1928); in Ad-

amson & Salter, FI. Cape Penins. 152 (1950). Type:

Cape, without precise locality, in herb. Thunberg.

23234 cf (UPS, holo.!).

Restio erectus Thunb., Diss. Restio 14 : (1788); Thunb., FI.

Cap. edn 1 : 325 (1811); edn Schultes, 85 (1823).

Restio elongatus Thunb. in Phytogr. Bl. 1 : 7 (1803); Thunb.,

FI. Cap. edn 1, 318 (1811); edn Schultes, 83 (1823); Kunth,

Enum. PI. 3 : 408 (1841). Thamnochortus elongatus (Thunb.)

Mast, in J. Linn. Soc., Bot. 10 : 226 (1868); in A. DC., Monogr.

Phan. 1 : 320 (1878); in FI. Cap. 7 : 123 (1897). Type: Cape, with-

out precise locality, in herb. Thunberg. 23233 $ (UPS, holo.!).

Thamnochortus floribundus Kunth. Enum. PI. 3 : 435 (1841);

Mast, in A. DC., Monogr. Phan. 1 : 328 (1878); in FI. Cap. 7 : 130

(1897). Type: Cape, 3418 (Simonstown): False Bay (-AB), Rey-

naud s.ndcf (B. holo.!; BOL!; K!; P!).

? Thamnochortus scirpiformis Mast, in J. Linn. Soc., Bot. 10 :

228 (1868); in A. DC., Monogr. Phan. 1 : 318 (1878); in FI. Cap. 7:

121 (1897). Type: Cape. 3418 (Simonstown): Cape dunes

(-AB/BA), Ecklon & Zeyher s.n. 9 (?)•

Thamnochortus burchellii Mast, in A. DC., Monogr. Phan. 1 :

322 (1878); in FI. Cap. 7 : 126 (1897). Syntypes: Cape, 3422 (Mos-

selbay): between Great and Little Brak Rivers (-AA), Burchell

6169 $ (K, lecto.!); 6169 cf (BOL!; K!); without precise locality,

Drege 129 C f & 9 (B!; P!).

Thamnochortus caricinus Mast, in A. DC., Monogr. Phan. 1 :

327 (1878); in FI. Cap. 7 : 129 (1897). Staberoha caricina (Mast.)

Dur. & Schinz, Consp. FI. Afr. 5 : 520 (1894). Syntypes: Cape,

without precise locality, Masson s.n. J (BM. lecto.!); Thom 900 cf

(K!).

Icon: A. DC., Monogr. Phan. 1 : t.2 f. 16-27; t.5 f.8 (1878).

Notes

1. I have not been able to locate any material of

the type of T. scirpiformis , but from the description

it belongs here.

2. T. erectus occurs on the hills and flats of the

coastal forelands from Darling (near Malmesbury)

to Knysna. Most of the collections are from sandy

places, usually from the flats, occasionally from val-

ley bottoms. Only a few collections are from moun-

tain slopes.

Thamnochortus fraternus Pillans in Trans. R.

Soc. S. Afr. 16 : 370 (1928). Type: Cape, 3418 (Si-

monstown): near the mouth of the Eerste River

(-BB), Pillans 4904 $ (BOL, lecto.!; K!); 4904 cf

(BOL!; K!).

Notes

1. This species is very close to T. paniculatus.

When originally described, the two species were

allopatric, with T. paniculatus known from the Bre-

dasdorp Flats, whereas T. fraternus occurs around

False Bay. However, later collections show that the

complex extends right around the coastline from

Cape Point to Cape Agulhas, and critical studies

may well show that the two taxa intergrade.

2. T. fraternus is a limestone and coastal dune

species, that occurs around the coast of False Bay,

with some collections from the Caledon and Bredas-

dorp coasts.

Thamnochortus fruticosus Berg., Descr. PL

Cap. 353 (1767); Mast, in A. DC., Monogr. Phan. 1 :

316 (1878); in FI. Cap. 7 : 122 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 361 (1928); in Adamson

& Salter, FI. Cape Penins. 151 (1950). Type: Cape,

without precise locality, comm. Grubb (SBT,

holo.!).

Restio scariosus Thunb., Diss. Restio 15 (1788); Thunb., FI.

Cap. edn 1, 327 (1811); edn Schultes 86 (1823); nom. illeg., su-

perfluous name for T. fruticosus Berg. Thamnochortus scariosus

(Thunb.) Spreng., Syst. Veg. 1 : 188 (1824); Kunth, Enum. PI. 3 :

430 (1841).

Icon: Berg., Descr. PI. Cap. t.5 f.8 (1767).

Notes

1. Thunberg (1788) cited T. fruticosus in the syno-

nymy of his Restio scariosus, thereby rendering his

name illegitimate. The specimen in herb. Thunb. la-

belled 'Restio scariosus’ is, in fact, T. cinereus

Linder!

2. T. fruticosus occurs between Caledon, the Cape

Peninsula and Ceres, generally in localities below

600 m.

Thamnochortus glaber (Mast.) Pillans in Trans.

R. Soc. S. Afr. 16 : 363 (1928). Syntypes: Cape, 3423

(Knysna): Melville (-AA), Burchell 5548 cf (K,

lecto.!; B!); 5462 cf (K!).

Thamnochortus fruticosus Berg. var. glaber Mast, in J. Linn.

Soc., Bot. 10 : 229 (1868); in A. DC., Monogr. Phan. 1 : 317

(1878); in FI. Cap. 7 : 122 (1897).

Notes

1. T. glaber grows on the coastal plateau from

Knysna to Port St Johns. It often occurs in somewhat

damp places in grassland or grassy fynbos. West of

Port Elizabeth it may become very common on

heavily grazed grassland.

2. Although there is both male and female ma-

terial under Burchell 5548, Masters (1868) only cited

the male material.

Thamnochortus gracilis Mast, in A. DC.,

Monogr. Phan. 1 : 327 (1878); in FI. Cap. 7 : 129

(1897); Pillans in Trans. R. Soc. S. Afr. 16 : 379

(1928); in Adamson & Salter, FI. Cape Penins. 152

(1950). Type: Cape, 3419 (Caledon): Baviaanskloof

Mountains near Genadendal (-BA), Burchell 7894

$ (K, holo.!; BOL!).

Staberoha gracilis (Mast.) Dur. & Schinz, Consp. FI. Afr. 5 :

521 (1894).

Bothalia 15, 3 & 4 (1985)

475

Notes

1. Burchell 7894 cf is the type of Hypolaena bur-

chellii (= Calopsis membranacea).

2. T. gracilis ranges from the Cape Peninsula tc

Fransch Hoek, Genadendal and Hermanus. Collec-

tions range from sea level to 1 000 m, both from dry

as well as from cool and moist mountain slopes, from

amongst rocks or from sandy places.

Thamnochortus guthrieae Pillans in Trans. R.

Soc. S. Afr. 16 : 371 ( 1928). Type: Cape, 3419 (Cale-

don): Swartberg (-AB/BA), Guthrie in BOL 17215

$ (BOL, lecto. !; K!); in BOL 17215 cf (BOL!; K!).

Thamnochortus lewisiae Pillans in J1 S. Afr. Bot. 18 : 110

(1952). Type: Cape, Robertson Division, Bushmans River, Lewis

in BOL 24763 ? (BOL, lecto.!); in BOL 24763 cf (BOL!).

Thamnochortus nervosus Pillans in J1 S. Afr. Bot. 18 : 111

(1952). Type: Cape, 3419 (Caledon): Brandfontein (-DB). Ester-

huysen 19014 $ (BOL, lecto.!; K!); 19014 cf (BOL!; K!).

Thamnochortus phpnosus Pillans in J1 S. Afr. Bot. 18 : 115

(1952). Type: Cape, 3418 (Simonstown): slopes above Llandudno

(-AB), Esterhuvsen 18604 9 (BOL, lecto.!; BM!; K!; MO!);

18604 cf (BM!; BOL!; K!; MO!).

Notes

1. I am using T. guthrieae here in a wide sense and

it includes some variation on the production of ster-

ile culms, on bract texture and flower size. It may be

necessary to include T. pellucidus here as well, but

this would make it a very variable species indeed.

2. T. guthrieae occurs in hard gravelly to sandy

soils from sea level to about 600 m, from Malmes-

bury to Bredasdorp. It is often locally common.

Thamnochortus insignis Mast, in Gdnrs’ Chron.

Ser. 3, 25 : 242 (1899); Pillans in Trans. R. Soc. S.

Afr. 16 : 382 (1928). Syntypes: Cape, Riversdale,

Rust 406 $ (B, lecto.!; K!); 399a (B!); 407 (?).

Icon: Gdnrs’ Chron., ser. 3, 25 : 251 f.93 (1899).

Notes

1. The two sheets of the type in B are annotated

by Masters, whereas at K there are only scraps in a

capsule. Clearly the material in B has to be the lecto-

type.

2. T. insignis occurs naturally on the coastal fore-

lands of the Bredasdorp and Riversdale divisions, on

limestone hills or on sandy flats. However, its use as

thatching reed has resulted in its introduction along

the coast from Betty’s Bay to Cape St Francis.

Thamnochortus levynsiae Pillans in Trans. R.

Soc. S. Afr. 16 : 364 (1928); in Adamson & Salter,

FI. Cape Penins. 151 (1950). Type: Cape, 3418 (Sim-

onstown): south-east slopes of Kalk Bay Mountain

(-AB), Levyns 1404 $ (BOL, lecto.!); 1404 cf

(BOL!).

Note

1. T. levynsiae occurs on steep, rocky slopes or on

ledges between 450 and 900 m on the Cape Penin-

sula.

Thamnochortus lucens (Poir.) Linder, comb.

nov.

Restio lucens Poir. in Lam., Encycl. 6 : 169 (1804); Kunth,

Enum. PL 3 : 414 (1841); Mast, in A. DC., Monogr. Phan. 1 : 296

(1878); in FI. Cap. 7 : 98 (1897). Type: Cape, without precise

locality, in herb. Lamarck (P, holo.; K, microfiche!).

Restio thamnochortus Thunb., Diss. Restio 309 (1788); Thunb.,

FI. Cap. edn 1, 328 (1811): edn Schultes, 86 (1823). Type: Rottb.,

Descriptionum et Iconum Rariores t.l f.l (1773), (Iconotype!).

Thamnochortus bromoides Kunth, Enum. PI. 3 : 432 (1841).

Type: Cape, without precise locality, Sieber 225 cf (S, lecto.!;

BR!; BM!; P!).

Thamnochortus ecklonianus Kunth, Enum. PI. 3 : 430 (1841).

Syntypes: Cape, without precise locality, Sieber 114 9 (K, lecto.!;

H!; MO!; S!); Ecklon s.n. (?).

Thamnochortus dichotomus Mast, in J. Linn. Soc., Bot. 10 :

229 (1868); in A. DC., Monogr. Phan. 1 : 318 (1878); in FI. Cap.

7 : 124 (1897); Pillans in Trans. R. Soc. S. Afr. 16 : 377 (1928); in

Adamson & Salter, FI. Cape Penins. 152 (1950), nom. illeg. , later

homonym for T. dichotomus (L.) Spreng.

Thamnochortus papillosus Pillans in J1 S. Afr. Bot. 18 : 112

(1952). Type: Cape, 3319 (Worcester): Limietberg (-CA), Ester-

huysen 1632 9 (BOL, lecto.!; K!); 1632 cf (BOL!; K!).

Restio lucens Pior. var. minor Mast, in A. DC., Monogr. Phan.

1 : 296 (1878); in FI. Cap. 7 : 98 (1897). Type: Cape, without

precise locality, in herb. Lamarck (P).

Thamnochortus dichotomus Mast. var. hyalinus Pillans in

Trans. R. Soc. S. Afr. 16 : 379 (1928). Syntypes: Cape, 3321 (La-

dismith): Langeberg, Phisantefontein (-CD), Muir 3192 9 (BOL,

lecto.!); 3191 cf (BOL!).

leones: Rottb., Descriptionum et Iconum Rariores, t.l f.l

(1773). Rice & Compton, Wild Flow. Cape G.H. t.250 f.l, 2

(1950).

Notes

1. Thamnochortus dichotomus (L.) Spreng. is

based on Restio dichotomus L., which is a superflu-

ous name for Schoenus capensis L. ( = lschyrolepis

capensis (L.) Linder). Masters (1868) attempted to

circumvent this problem by citing 'Restio dichotomus

Rottb. . . . nec Linn. — Thamnochortus dichotomus

Kunth . . . (excl. syn.),’ but this establishes T. di-

chotomus Mast, as a new species, which is a later

homonym.

2. Thunberg (1788) was aware that the ‘ Restio di-

chotomus’ of Rottboell was not the same species as

that of Linnaeus, although Rottboell referred to Lin-

naeus in the synonymy. Thunberg called 'Restio di-

chotomus’ of Rottboell Restio thamnochortus. This

is the earliest legitimate epithet for the species.

However, because of Art. 23.4 of the ICBN (1983),

this epithet cannot be transferred to Thamnochor-

tus.

3. T. lucens is part of a complex including T.

bachmannii, T. punctatus and T. arenarius. Al-

though I have little doubt that these do all constitute

distinct species, the characters by which they are dis-

tinguished may be difficult to observe on some speci-

mens.

4. There is some variation within the species, es-

pecially in spikelet length and shape and in bract

colour. The var. hyalinus of Pillans may also be dis-

tinct, as its distinguishing features seem to be

unique.

5. T. lucens is known from Tulbagh to Rivier-

sonderend, Hermanus and the Cape Peninsula, and

up the west coast as far as Saldanha. Within this area

it is common between sea level and about 800 m,

usually on dry gravelly slopes. There is also one col-

lection from near Calvinia and the locality for var.

hyalinus, near Riversdale.

476

Bothalia 15, 3 & 4 (1985)

Thamnochortus muirii Pillans in Trans. R. Soc.

S. Afr. 16 : 366 (1928). Type: Cape, Riversdale Divi-

sion, Botteliersfontein, Muir 1444 $ (BOL, lecto.!);

1444 cf (BOL!).

Note

1. T. muirii occurs on the coastal forelands from

Riversdale to Mosselbay, probably restricted to

limestone shelves.

Thamnochortus nutans (Thunb.) Pillans in

Trans. R. Soc. S. Afr. 16 : 365 (1928); in Adamson

& Salter, FI. Cape Penins. 152 (1950). Type: Cape,

without precise locality, herb. Thunberg. 23239 cf

(UPS, holo.!).

Restio nutans Thunb. in Phytogr. Bl. 1 : 7 (1803); Thunb., FI.

Cap. edn 1, 321 (1811); edn Schultes, 84 (1823); Kunth, Enum.

PI. 3 ; 413 (1841).

Thamnochortus consanguineus Kunth, Enum. PI. 3 : 432

(1841). Cape, without precise locality, Garnot s.n. cf (P, lecto.!;

Kl).

Notes

1. The determination of this species from male

material is difficult, but the determinations of the

types above are probably correct.

2. Masters (1878, 1897) treated this species under

the name T. elongatus.

3. T. nutans is known only from the Cape Penin-

sula, where it generally occurs above 600 m, i.e. in

the south-east cloud belt.

Thamnochortus obtusus Pillans in J1 S. Afr.

Bot. 18 : 112 (1952). Type: Cape, 3420 (Bredas-

dorp): Bontebok Park (-AB/BA), Maguire 836 $

(BOL, lecto.!); 836 cf (BOL!).

Note

1. T. obtusus occurs in dry sandy habitats along

the coastal forelands below 300 m, from Malmes-

bury to Bredasdorp.

Thamnochortus paniculatus Mast, in Bot. Jb. 29

Beibl. 66 : 12 (1900); Pillans in Trans. R. Soc. S.

Afr. 16 : 369 (1928). Syntypes: Cape, 3419 (Cale-

don): hills near Mierkraal (-DA), Schlechter 10512

$ (B, lecto.!; BM!; BOL! BR!; K!;MO!;P!; S!; Z!);

10511 cf (B!;BM!;BOL!;BR!;K!;MO!;P!;S!;Z!).

Notes

1. T. paniculatus is rather closely related to T. fra-

ternus and it may not be possible to separate these

two taxa. For further notes see under T. fraternus.

2. This species appears to be a ‘limestone en-

demic,’ which grows below about 150 m on lime-

stone hills and shelves on the coastal flats of Bredas-

dorp and Riversdale divisions.

Thamnochortus papyraceus Pillans in Trans. R.

Soc. S. Afr. 29 : 350 (1942). Type: Cape, 3321 (La-

dismith): Seven Weeks Poort Mountains (-AD),

Stokoe in BOL 19091 $ (BOL, lecto.!; K!); in BOL

19091 cf (BOL!; K!)

Note

1. T. papyraceus occurs in the Little Swartberg

between I 800 and 2 100 m.

Thamnochortus pellucidus Pillans in J1 S. Afr.

Bot. 18 : 113 (1952). Type: Cape, 3419 (Caledon):

Brandfontein (-DA), Esterhuysen 19055 £ (BOL,

lecto.!; K!); 19055 cf (BOL!; K!).

Notes

1. This species may be difficult to distinguish from

T. guthrieae, especially from the form represented

by the type of T. nervosus.

2. T. pellucidus occurs on the coastal forelands

between Gordons Bay and Cape Agulhas, usually

below 150 m. All collections are from sandy areas.

Thamnochortus platypteris Kunth, Enum. PI.

3 : 429 (1841); Mast, in A. DC., Monogr. Phan. 1 :

322 (1878); in FI. Cap. 7 : 125 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 374 (1928). Syntypes:

Cape, 3118 (Van Rhynsdorp): Giftberge (-BC),

Dr'ege 139 $ (B, lecto.!; BM!; BOL!; K!; MO!; NY!;

OXF!; P!; S!). 3219 (Wuppertal): between Grasberg

River and Watervals Rivier (-AA), Drege 2512 9

(BM!; K!; MEL!; MO!; OXF!; P!).

Thamnochortus comptonii Pillans in Trans. R. Soc. S. Afr. 30 ;

264 (1945). Type: 3219 (Wuppertal): Cedarberg, Middleberg Pla-

teau (-AC), Esterhuysen 7193 9 (BOL, lecto.!; K!); 7193 cf

(BOL!; K!).

Notes

1. This is a variable species, with the northern col-

lections matching the type of T. platypteris, whereas

the southern collections agree with the type of T.

comptonii. There are many collections, especially

from the Cedarberg, which are intermediate be-

tween the two concepts, so the best solution appears

to be to recognize a single variable species.

2. The southern populations are rather close to T.

schlechteri, but may be distinguished by their caespi-

tose habit.

3. T. platypteris ranges from Nieuwoudtville

through the Giftberg, Cedarberg and Bokkeveld

Mountains to Laingsburg. Populations occur above

1 000 m, usually in dry sandy areas, sometimes in

localities which are wet in winter.

Thamnochortus pluristachyus Mast, in Bot. Jb.

29 Beibl. 66 : 12 (1900); Pillans in Trans. R. Soc. S.

Afr. 16 : 371 (1928). Syntypes: Cape, 3419 (Cale-

don); hills near Mierkraal (-DB), Schlechter 10524 cf

(B!; BOL!; BR!; K!; MO!; P!; S!; Z!); 10525 $ (K,

lecto.!; BM!; BOL!; BR!; MO!; P!; S!; Z!); hills at

Koude River (-DA), Schlechter 10461 C f (BM!;

BOL!; BR!; MO!; P!; Z!); 10462 $ (BM!; BR!;

MO!; P!; Z!).

Notes

1. Schlechter 10461 and 10562 are mixed collec-

tions that also contain material that are the types of

T. canescens (=T. dumosus) and Restio macowanii

(=R. fruticosus). The sheet at K contains drawings

and analysis by Masters that fit the protologue, so I

have chosen this material as lectotype.

2. This species is known from the limestones on

the Bredasdorp coastal plateau. It has been collected

only rarely and is not well known.

Bothalia 15, 3 & 4 (1985)

477

Thamnochortus pulcher Pillans in Trans. R.

Soc. S. Afr. 29 : 350 (1942). Type: Cape, 3418 (Sim-

onstown): Landdrost Kop (-BB), Esterhuysen 3593

$ (BOL, lecto.!; K!); 3593 C f (BOL!; K!).

Note

1. T. pulcher occurs on dry, often stony slopes be-

tween 300 and 1 200 m in the mountains between

Villiersdorp and Bredasdorp.

Thamnochortus punctatus Pillans in Trans. R.

Soc. S. Afr. 16 : 376 (1928); in Adamson & Salter,

FI. Cape Penins. 152 (1950). Type: Cape, 3418 (Sim-

onstown): Retreat (-AB), Pillans 4902 £ (BOL,

lecto.!; K!); 4902 cf (BOL!; K!).

Notes

1. Although the plants are rhizomatous, the

culms are closely clustered, with the rhizomes con-

necting the clusters. Often on herbarium specimens

the rhizomes are not visible.

2. T. punctatus is a ‘sandveld’ species, that ranges

from the Cape Peninsula to Nieuwoudtville. The

majority of the collections are from dry sandy flats

(where this species is quite common), with few col-

lections from mountains at Nieuwoudtville and on

the southern Cape Peninsula.

Thamnochortus rigidus Esterhuysen , sp. nov., a

T. cinereo Linder alis florum manifestis, a T. fruti-

coso Berg, rhizomatibus absentibus, ab ambabus

ramulis sterilibus obtusis, rigidis, sine pilis sericeis

recedit.

TYPE. — Cape, 3321 (Ladismith): Brande-

wynskuil, above Bobuffelskloof (-AD), Esterhuysen

33902 9 (BOL, holo.!; C; E; K; L; M; MO; S).

Plants caespitose, non-rhizomatous, tussocks up

to 1 m tall. Culms solid, terete, to 3 mm in diameter,

pubescent, simple during flowering, developing clus-

ters of sterile, much branched, flexuose culms, to 15

cm long, at the nodes during the first year after flow-

ering. Sheaths dimorphic; on the fertile culms closely

convoluted, 3-6 cm long, acute, piliferous, coria-

ceous at the base, the rest chartaceous to membra-

nous and soon decaying; on the sterile culms the

sheath is 1-1,5 mm long, but the awn is stout, rigid,

somewhat curved, 3 mm long, hyaline shoulders are

about half as tall as the awn. Male inflorescence

10-20 cm long, with numerous spikelets pendent

from several nodes. Spathes like the sheaths, some-

what shorter. Inflorescence branches 1-several per

node, often branched, with 5-10 pendulous, 10-25 x

5 mm spikelets. Bracts chartaceous, brown with hya-

line margins, acute, 7-10 x 2-3 mm. Flowers subpe-

dicellate, perianth 3,5-4 mm long. Tepals charta-

ceous, subequal, acute, 3,5 x 0,8 mm; lateral sepals

conduplicate, very narrowly carinate, mucronate;

petals somewhat smaller than the sepals. Anthers not

exserted at anthesis, 2 mm long, mucronate. Female

inflorescence similar in size to the males, but with

fewer, rigidly erect, 10-20 x 10-15 mm spikelets.

Bracts same as in the males, flowers with the wings

exposed at the sides of the bracts. Flowers shortly

pedicellate, perianth 4-5 mm long, 4,5-6 mm wide,

broadly ovate in outline. Tepals chartaceous to carti-

laginous; laterals acute, 4-5 mm long, conduplicate,

mucronate, wings 1,6-2 mm wide, decurrent on the

pedicel; odd sepal acute, 4,5 x 1 mm. Petals like the

odd sepal. Style solitary, about 3 mm long, sparsely

villous. Ovary unilocular. Fruit a thin-walled, ellip-

soid nutlet, held and dispersed with the perianth.

T. rigidus occurs along the arid inland fringe of the

Cape Fold Mountains, along the arid margins of the

'fynbos’, from Karoopoort to the Grootriver

Heights. The altitudinal range is from 500-1 900 m.

The plants generally occur as tussocks on stony

slopes often growing out of rock crevices.

To date, this species has been confused with T.

argenteus. It clearly belongs to the T. argenteus - T.

fruticosus group in the overall habit, but especially in

the pubescent culms. The differences between these

species are summarized in Table 1.

CAPE. — 3319 (Worcester): on slopes above Karoopoort

(-BA), Esterhuysen 30453 (BOL, K); on the steep rocky slopes

between Sandhills and Orchard (-DA), Esterhuysen 33189 (BOL;

K). 3320 (Montagu): Touws River (-AC); Van Niekerk 712

(BOL; K). 3321 (Ladismith): Brandewynskuil, above Bobuffels-

kloof (-AD), Esterhuysen 33902 (BOL; C; E; K; L; M; MO; S);

Seven Weeks Poort (-AD), Esterhuysen 24810 (BOL; K). 3322

(Oudtshoorn): Prince Albert distict (-AC), Acocks 18267 (K;

PRE); near Klaarstroom (-BC), Drege 132b (K). 3324 (Steytler-

ville): Groot River Heights (-AA), Linder 2892 (PRE).

Thamnochortus scabridus Pillans in Trans. R.

Soc. S. Afr. 16: 367 (1928). Syntypes: Cape, 3319

(Worcester): sandstone hill between Eendracht and

Triangle (-BD), Levyns 396 9 (BOL, lecto. l);395 cf

(BOL!).

Note

1 . T. scabridus occurs on the arid margin of the

Cape Floral Region from the quartzite hills at

Touwsriver to the Swartruggens Mountains north of

Ceres.

TABLE 1. — Comparison of Thamnochortus fruticosus, T. rigidus and T. argenteus

478

Bothalia 15, 3 & 4 (1985)

Thamnochortus schlechteri Pillans in Trans. R.

Soc. S. Afr. 16: 372 (1928). Syntypes: Cape, 3219

(Wuppertal): Koue Bokkeveld Tafelberg, 1 870 m

(-CD), Schlechter 10089 $ (BOL, lecto.!; BR!; K!;

MO!; P!; S!; Z!); 10088 cf (BM!; BOL!; BR!; K!;

MO!; P!; S!; Z!).

Notes

1. Pillans (1928) labelled his species as a nom.

nov. for T. bachmannii as in Bot. Jb. 29 Beih. 66,

rather than in Annin, naturh. Mus. Wien , because,

in the former publication (which is later in time than

the latter), Masters included two Schlechter collec-

tions which are quite different from the three collec-

tions cited in the protologue of the T. bachmannii.

However, it would appear to be simpler to treat T.

schlechteri as a sp. nov. based on the two Schlechter

collections, and to interpret Master’s inclusion of

Schlechter 10089 and 10088 as an error of identifica-

tion.

2. T. schlechteri is close to T. platypteris , but can

be separated by the narrower flowers and by the

presence of rhizomes.

3. T. schlechteri occurs in the mountains from the

Cedarberg to Ceres, in rather arid conditions, be-

tween 1 200 and 1 800 m.

Thamnochortus spicigerus (Thunb.) Spreng.,

Syst. Veg. 1: 187 (1824); Kunth, Enum. PI. 3: 440

(1841); Mast, in A. DC., Monogr. Phan. 1: 314

(1878); in FI. Cap. 7: 121 (1897); Pillans in Trans. R.

Soc. S. Afr. 16: 382 (1928); in Adamson & Salter,

FI. Cape Penins. 153 (1950). Syntypes: Cape, with-

out precise locality, in herb. Thunberg. 23247 (UPS,

lecto.!; C!; S!); 23246 cf (UPS!; C!; S!).

Restio spicigerus Thunb., Diss. Restio 11 (1788); Thunb., FI.

Cap. edn 1, 321 (1811); edn Schultes, 84 (1823).

? Thamnochortus striatus Hochst. in Flora, Jena 28: 339 (1845);

Mast, in A. DC., Monogr. Phan. 1; 328 (1878); in FI. Cap. 7; 130

(1897). Type: Cape, 3418 (Simonstown): Cape Flats (-BA),

Krauss s.n. (?).

Thamnochortus maximus Kuntze. Rev. Gen. 3: 330 (1898).

Type: Cape, 3318 (Cape Town): Diepriver Station (-DC), Kuntze

s.n. $ (NY, holo.!; Kl; B!; Z!); s.n. c f (B!; K!; Z!).

leones: Gdnrs’Chron., ser. 3, 25: 249 f .92 (1899). Mason, W.,

Cape Sandveld Flow. 4, 3 f.2, 3 (1972).

Notes

1. R. Brown (1810) suggested that Restio spicige-

rus should be transferred to Thamnochortus, but the

actual combination was made by Sprengel.

2. I have not been able to locate the type of T.

striatus.

3. T. spicigerus occurs on the Cape Flats and the

‘Sandveld’ from Somerset West to Hopefield and

Langebaan.

Thamnochortus sporadicus Pillans in J1 S. Afr.

Bot. 18: 116 (1952). Type: Cape, 3318 (Cape Town):

lower slopes of Window Stream (-CD), Esterhuysen

17297 $ (BOL, lecto.!; B!; K!); 17297 cf(B!; BOL!;

K!).

Thamnochortus muticus Pillans in J1 S. Afr. Bot. 18: 1 10 (1952).

Type: Cape, Robertson Division, Bushmans River, Lewis in BOL

24803 ? (BOL, holo.!).

Thamnochortus piketbergensis Pillans in J1 S. Afr. Bot. 18: 114

(1952). Type: Cape, 3218 (Clanwilliam): Piketberg Mountain

(-D), Pillans 7574 $ (BOL, lecto.!; K!); 7574 cf (BOL!; K!).

Thamnochortus similis Pillans in J1 S. Afr. Bot. 18: 116 (1952).

Type: Cape, 3319 (Worcester): Hex River Mountains, Kleinberg

(-AC), Esterhuysen 9322 $ (BOL, lecto!); 9322 cf (BOL!).

Note

1. T. sporadicus ranges from the Riviersonderend

Mountains to the Cape Peninsula, Ceres and the Pi-

ketberg. It occurs in sandy to stony soils between sea

level and 1 000 m.

Thamnochortus stokoei Pillans in Trans. R. Soc.

S. Afr. 16: 375 (1928). Type: Cape, 3319 (Worces-

ter): Onklaarberg (-CD), Stokoe 1214 $ (BOL,

holo.!).

Note

1 . T. stokoei is a montane species that occurs be-

tween 1 200 and 1 800 m on rocky slopes of the

Slanghoek, Wemmershoek, Hottentots Holland and

Riviersonderend Mountains.

RHODOCOMA

11. Rhodocoma Nees in Lindl., Nat. Syst. Bot.

edn 2, 450 (1836), Linder in Bothalia 15 : 65 (1984).

KEY TO THE SPECIES OF RHODOCOMA

la Culms with very many verticellately arranged short

branches R. capensis

lb Fertile culms simple:

2a Culms 0,3-1 m tall, 1-2 mm in diam. at the middle,

rarely with sterile branches R. fruticosa

2b Culms 1-3 m tall, 3-5 mm in diam. at the middle,

usually with tufts of sterile branches at the nodes

R. gigantea

Rhodocoma capensis Nees ex Steud., Syn. PI.

Glum. 2 : 249 (1855). Type: Cape, 3325 (Port Eliza-

beth): Swartkopsrivier (-DC), Ecklon s.n. $ (K,

lecto.!; MEL!).

Restio rhodocoma Mast, in J. Linn. Soc., Bot. 10 : 275 (1868);

in A. DC., Monogr. Phan. 1 : 294 (1878); in FI. Cap. 7 : 93 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 221 (1928). Nom. illeg. , su-

perfluous name for R. capensis.

Notes

1. Masters (1868) cites in his protologue ‘Rhodo-

coma equisetum . . . Steud. Synops. 2 p. 248.’ This

is incorrect, as the specific epithet mentioned by

Steudel is R. capensis , but it still renders Master’s

name superfluous and illegitimate.

2. The typification of R. capensis is difficult. Steu-

del (1855) referred to the name as ‘ex Nees’ and

there is no specimen of the species in Steudel’s her-

barium in P. As Nees did not seem to have had ac-

cess to Drege collections, it is likely to have been an

Ecklon or an Ecklon & Zeyher collection. There is

an Ecklon collection in several herbaria, annotated

‘Rhodocoma equisetum Nees’ — these I presume are

the isotypes of the material which Steudel saw, so I

selected the sheet at K as lectotype.

3. Common names for this species are ‘kanet,’

‘katstertkanet’ or iitjieskanet.’

4. R. capensis occurs along the arid inland mar-

gins of the Cape Fold Mountains, from the Cedar-

Bothalia 15, 3 & 4 (1985)

479

berg to Grahamstown. It is restricted to sandstones

or quartzites.

Rhodocoma fruticosa (Thunb.) Linder, comb.

nov.

Restio fruticosus Thunb., Diss. Restio 16 (1788); Thunb., FI.

Cap. edn 1, 329 (1811); edn Schultes, 86 (1823); Kunth, Enum.

PI. 3 : 413 (1841); Pillans in Trans. R. Soc. S. Afr. 16 : 218 (1928);

in Adamson & Salter, FI. Cape Penins. 136 (1950). Type: Cape,

without precise locality, Thunberg in herb. Thunb. 23235 $ (UPS,

holo.!).

Thamnochortus modestus Kunth, Enum. PI. 3 : 434 (1841).

Leptocarpus modestus (Kunth) Mast, in J. Linn. Soc., Bot. 10 :

225 (1868); in A. DC., Monogr. Phan. 1 : 337 (1878); in FI. Cap.

7 : 119 (1897). Type: Cape, Great Swartberg, Drege 138b cf (B,

holo.!; BM!; K!; MO!; OXF!; P!).

Restio macowanii Pillans in Ann. Bolus Herb. 3 : 80 (1921).

Syntypes: 3419 (Caledon): Koude Rivier, 200 m (-DA),

Schlechter 10462 $ (BOL, lecto.!; K!; S!); 10461 c f (BOL!; K!;

S!). 3322 (Oudtshoorn): Swartberg Pass (-AC), Tugwell in BOL

14288 (BOL!; K!). 3323 (Willowmore): Uniondale Road (-AC),

Paterson 3013 2 (BOL!; K!); Uniondale (-CA), Paterson 3009

(BOL!). 3324 (Steydtlerville): Steydtlerville (-AD), Paterson 19

Cf (BOL!). 3326 (Grahamstown): Grahamstown (-BC), White 34

(BOL!); Daley & Sole 36 (BOL!); MacOwan 1361 (BOL!;

MEL!).

Notes

1. N.E. Brown noted on the sheet at K that Drege

138b is from ‘Table Mountain and Devil’s Peak,’ in-

formation probably gleaned from the Liibeck Her-

barium. However, Pillans noted that the sheet at B is

from the Great Swartberg and that the latter locality

is more likely. I agree with Pillans.

2. This species is widespread, ranging from Cape

Town to the Natal Drakensberg on sandstones and

quartzites. It is particularly common in the southern

Cape Province, on the arid north-eastern edge of the

Cape Flora.

Rhodocoma gigantea (Kunth) Linder, comb.

nov.

Thamnochortus giganteus Kunth, Enum. PI. 3 : 435 (1841);

Mast, in A. DC., Monogr. Phan. 1 : 315 (1878); in FI. Cap. 7 : 121

(1897). Restio giganteus (Kunth) N.E.Br. in FI. Cap. 7 : 755

(1900); Pillans in Trans. R. Soc. S. Afr. 16 : 220 (1928). Type:

Humansdorp Div., Kromme River below 300 m, Drege 2 cf (B,

lecto.!; K!; OXF!; P!).

Restio foliosus N.E.Br. in FI. Cap. 7 : 753 (1900); Pillans in

Trans. R. Soc. S. Afr. 16 : 220 (1928). Type: Cape, 3321 (Ladi-

smith): Garcias Pass, 300 m (-CC), Galpin 4783 2 (K, lecto.!; B!;

BOL!); 4783 cf (B!; BOL!; K!).

Restio comosus N.E.Br. in FI. Cap. 7 : 753 (1900). Type: Cape,

3423 (Knysna): The Glebe, 200 m (-AA), Galpin 4783 (K, holo.!;

B!; BOL!).

Notes

1. There is a size difference between R. giganteus

and R. foliosus, but this appears to be more of the

nature of a cline. It is not likely that the two names

can be kept distinct.

2. R. gigantea occurs on the wet sea-side of the

Langeberg, Outeniqua Mountains and Tsitsikamma

Mountains. It is often associated with tall dense fyn-

bos.

CERATOCARYUM

12. Ceratocaryum Nees in Lindl., Nat. Syst.

Bot. edn 2, 451 (1836); Linder in Bothalia 15 : 65

(1984).

KEY TO THE SPECIES OF CERATOCARYUM

la Perianth sessile:

2a Culms widely fistular in all parts C. fistulosum

2b Culms solid or narrowly fistular in parts C. argenteum

lb Perianth stipitate:

3a Spathes 5-8 cm long C. decipiens

3b Spathes 3-3,5 cm long:

4a Tepals orbicular in outline; nut smooth... C. xerophilum

4b Tepals deltoid or oblong; nut rugose C. fimbriatum

Ceratocaryum argenteum Nees ex Kunth,

Enum. PI. 3 : 483 (1841); Mast, in A. DC., Monogr.

Phan. 1 : 390 (1878); in FI. Cap. 7 ; 147 (1897).

Type: ‘Restio argenteus Thg.’ in herb. Willdenow

18273 fol. 2 (B, holo.!).

Willdenowia argentea (Kunth) Hieron. in Pflanzenfam. 2,4 : 4

(1888); Pillans in Trans. R. Soc. S. Afr. 16 : 421 (1928).

Notes

1. Nees (1836) erected the genus Ceratocaryum

and noted ‘Male R. argenteus herb. Willd. (nec.

Thg.), female ab Ecklonio nuper inventa.’ This does

not constitute a valid combination (Article 33.1).

When Kunth made the valid combination (1841), he

only referred to the specimen in herb. Willdenow,

which is then the type.

2. C. argenteum ranges from Albertinia to Paarl,

from near sea level to 1 100 m. It occurs in sandy or

stony soils, in dry to somewhat wet habitats.

3. At MEL there are three sheets of an Ecklon &

Zeyher collection from 56.5, labelled ‘59’ in herb.

Sonder. One has a printed label ‘Ceratocaryum

speciosum’ while the other two are so named in Son-

der’s hand.

Ceratocaryum decipiens (N.E.Br.) Linder,

comb. nov.

Willdenowia decipiens N.E.Br. in FI. Cap. 7 : 758 (1900). Type:

Cape, 3321 (Ladismith): mountains at Garcias Pass, 300 m (-CC),

Galpin 4828 2 (K, lecto.!; B!; BOL!); 4828 cf (B!; BOL!; K!).

Notes

1. Pillans (1928) erroneously included W. deci-

piens under W. fimbriata. The description of W. fim-

briata in Pillans (1928) applies to C. decipiens.

2. C. decipiens is closely related to C. fimbriatum

and C. xerophilum. For a summary of the differ-

ences among these taxa see Table 2.

3. C. decipiens is known from the Langeberg, the

Klein Swartberg and the Caledon Mountains as far

west as the Nuweberg. It occurs on well-drained

mountain slopes between 300 and 1 200 m.

Ceratocaryum fimbriatum (Kunth) Linder.,

comb. nov.

Willdenowia fimbriata Kunth, Enum. PI. 3 : 455 (1841); Mast,

in A. DC., Monogr. Phan. 1 : 394 (1878); in FI. Cap. 7 : 147

(1897). Type: Cape, 3319 (Worcester): Du Toits Kloof (-CA),

Drege 1635b 2 (B, holo.!; K!; MO!; OXF!; P!), see Fig. 3.

Willdenowia esterhuyseniae Pillans in Trans. R. Soc. S. Afr. 30 :

265 (1945). Type: Cape, 3319 (Worcester): Slab Peak, Mitchells

Pass, 1 200 m (-AD), Esterhuysen 6212 2 (BOL, lecto.!; K!);

6212 Cf (BOL!; K!).

480

Bothalia 15, 3 & 4 (1985)

TABLE 2. — Summary of the differences between Ceratocaryum decipiens, C. fimbriatum and C. xerophilum

Notes

1. Pillans (1928) misunderstood W. fimbriata

Kunth as being identical to W. decipiens and in 1945

published the name W. esterhuyseniae for it. This

may have been caused by a capsule containing a nut

in BOL, labelled ‘Drege 1635b,’ which is W. deci-

piens and not the same as Drege 1635b at K.

2. C. fimbriatum is closely related to C. decipiens

and C. xerophilum. The differences are summarized

in Table 2.

3. C. fimbriatum occurs between 900 and 1 800 m

in the Slanghoek and Hex River Mountains and in

the mountains around Mitchell's Pass. It tends to oc-

cur around marshy areas and in some better drained

areas.

Ceratocaryum fistulosum Mast, in J. Linn. Soc.,

Bot. 10 : 274 (1868); in A. DC., Monogr. Phan. 1 :

391 (1878); in FI. Cap. 7 : 148 (1897). Type: Cape,

3321 (Ladismith): on the Kampsche Berg (-CD),

Burchell 7095 $ (K, lecto.!; BOL!); 7095 cf (BOL!;

K!).

Willdenowia fistidosa (Mast.) Pillans in Trans. R. Soc. S. Afr.

16 : 422 (1928).

leones: J. Linn. Soc., Bot. 10 : t.8E (1868). A. DC., Monogr.

Phan. 1 : t.4 f.16-20; t.5 f.18 (1878).

Note

1. C. fistulosum occurs on the south-facing slopes

of the Langeberg between Riversdale and Swellen-

dam, above 750 m and has also been recorded from

the Kogelberg.

Ceratocaryum xerophilum (Pillans) Linder,

comb. nov.

Willdenowia xerophila Pillans in J1 S. Afr. Bot. 18 : 122 (1952).

Type: Cape, 3320 (Montagu): south slopes of the Anysberg

(-DA), Esrerluiysen 17088 $ (BOL, lecto.!); 17088 C f (BOL!).

Notes

1. C. xerophilum is closely related to C. fimbria-

tum and C. decipiens. For the differences among the

taxa see Table 2.

2. C. xerophilum occurs in the arid inland moun-

tains; Matroosberg, the Klein Swartberg and the

Anysberg. It is recorded from steep rocky slopes be-

tween 1 000 and 2 000 m.

CANNOMOIS

13. Cannomois Desv. in Annls Sci. nat., ser. 1,

13 :43 (1828); Linder in Bothalia 15 : 65 (1984).

Cucullifera Nees in Lindl., Nat. Syst. Bot. edn 2,

451 (1836).

Mesanthus Nees in Lindl., Nat. Syst. Bot. edn 2,

451 (1836).

Note

The species limits in this genus are very difficult

and there are three Taxa’ which I have not yet for-

mally named.

KEY TO THE SPECIES OF CANNOMOIS

la Culms much branched C. virgata

lb Culms simple or sparsely branched:

2a Bracts and spathes long-acuminate; female spikelets

single; male spikelets globular;

3a Tepals longer than fruit C. nitida

3b Tepals shorter than the fruit:

4a Tepals § of the length of the fruit; from the

Worcester-Ceres area C. nitida var. (?)

4b Tepals ( of the length of the fruit; from the

Cedarberg C. aristata

2b Bracts and spathes acute to rarely acuminate;

female spikelets usually several:

5a Perianth as long as the nut; rhizome very

short C. congesta

5b Perianth shorter than the nut; rhizome

usually creeping;

6a Perianth about j of the length of the nut;

from above 1 500 m in the Cedarberg

C. congesta var. (?)

6b Perianth about ( of the length of the nut:

7a Spathes dark nitid; from the southern Cape

Province C. scirpoides

7b Spathes pale brown to rarely dark brown;

from the western Cape C. parriflora

Cannomois aristata Mast, in Bot. Jb. 29 Beibl.

66 : 19 (1900); Pillans in Trans. R. Soc. S. Afr. 16 ;

417 (1928). Syntypes: Cape, 3118 (Van Rhynsdorp):

Koude Berg, among rocks (-DC), Schlechter 8753 ^

(B, lecto.!; BM!; BR!; K!; MO!; P!; S!; Z!); 8752 cf

(B!; BM!; BR!; K!; MO!; P!; S!; Z!).

Notes

1 . This species is rather close to C. nitida, of which

it may merely be a northern extension. This would

mean a cline in the relative lengths of the nuts to the

perianth from the south to the north.

2. C. aristata is known from the Cedarberg, where

it occurs in fairly arid conditions amongst rocks be-

tween 1 000 and 1 300 m.

Cannomois congesta Mast, in FI. Cap. 7 : 143

(1897). Syntypes: Cape, 3419 (Caledon): Donker-

hoek Mountain (-AB), Burchell 7960 § (K, lecto.!;

BOL!; K!); Zwartberg, August (-AB/BA), Ecklon

Bothalia 15, 3 & 4 (1985)

481

c& Zeyhers.n. cf & $ (B!; BOL!; BR!; K!; MEL!;

MO!). 3219 (Wuppertal): Ezelsbank (-AC), Drege

2508 c? & 9 (K!; P!). 3319 (Worcester): Tulbagh,

Nieuwe Kloof (-AC), MacOwan 1682 cf (BM!; K!;

P!).

Cannomois scirpoides (Kunth) Mast. var. primosii Pillans in

Trans. R. Soc. S. Afr. 16 : 419 (1928). Syntypes: Cape, 3219

(Wuppertal): Cedarberg (-AC/CA), Primos in herb. Marloth.

11802 5 (BOL, lecto.!); 11796 cf (BOL!).

Cannomois scirpoides (Kunth) Mast. var. minor Pillans in

Trans. R. Soc. S. Afr. 16 : 419 (1928). Type: Cape, 3319 (Worces-

ter): between Villiersdorp and Fransch Hoek (-CC), Bolus 7485

$ (BOL, holo.l).

Notes

1. There are two sheets of Burchell 7960 at K. One

(7960a) is annotated by Masters and is here selected

as the lectotype.

2. Pillans (1928 : 418) confused C. scirpoides with

this species. The description which he gives under C.

scirpoides applies to this species.

3. Above 1 500 m in the Cedarberg there is a va-

riation that is quite distinct. It may be of hybrid ori-

gin with C. parviflora being the other parent. It dif-

fers from C. congesta in having a shorter perianth

and in being rhizomatous. This variant includes the

type of var. primosii of Pillans. More detailed stu-

dies are required to determine the correct taxonomic

status of this form.

4. C. congesta occurs between Van Rhynsdorp

and Caledon between 300 and 1 500 m on dry rocky

slopes.

Cannomois nitida (Mast.) Pillans in Trans. R.

Soc. S. Afr. 16 : 418 (1928). Type: Cape, 3319 (Wor-

cester): Winterhoeksberg, Nov., (-AA), Ecklon s.n.

O (B, holo.!; B!; MEL!; NY!; Z!).

Hvpodiscus nitidus Nees ex Mast, in J. Linn. Soc., Bot. 10 : 259

(1868); in A. DC., Monogr. Phan. 1 : 383 (1878); in FI. Cap. 7 :

136 (1897).

Notes

1. In B there are two sheets of this collection. The

one is annotated ‘herb. Sonder 48', and the other is

annotated ‘Distr. Tulbagh, Winterhoeksberg,

800-5 000 ft, Restio (an Hypodiscus) nitidus.' The

latter is regarded as the holotype. The sheet in Z has

a handwritten label ‘ Hypodiscus nitidus N. ab E.\

See also Fig. 5.

2. There are two forms within this species, varying

in the relative length of the tepals to the nut. This

suggests a north-south cline, which might include C.

aristata as the northern extreme.

3. This is a high-altitude species, with collections

ranging from (1 000) 1 500-1 800 m, generally from

dry, well-drained slopes. The distribution area of C.

nitida is from the Cold Bokkeveld (near Ceres) to

the great Swartberg east of Meiringspoort.

Cannomois parviflora (Thunb.) Pillans in

Trans. R. Soc. S. Afr. 16 : 415 (1928). Type: Cape,

without precise locality, in herb. Thunb. 23241 cf

(UPS, holo.!).

Restio parviflorus Thunb.. Diss. Restio 13 (1788); Thunb., FI.

Cap. edn 1 : 325 (1811); edn Schultes, 85 (1823). Elegia parviflora

(Thunb.) Kunth, Enum. PI. 3 : 467 (1841); Mast, in A. DC., Mon-

ogr. Phan. 1 : 353 (1878); in FI. Cap. 7 : 113 (1897).

Willdenowia compressa Thunb. in K. svenska VetenskAkad.

Handl. 11 : 31 (1790); Thunb., FI. Cap. edn 1, 315 (1811); edn

Schultes, 82 (1823); Kunth, Enum. PI. 3 : 456 (1841). Syntypes:

Cape, without precise locality, in herb. Thunberg. 23217 $ (UPS,

lecto.!), 23218 9 (UPS!).

Mesanthus macrocarpus Nees in Lindl., Introd. Nat. Syst., edn

2, 451 (1836), Kunth, Enum. PI. 3 : 485 (1841), nom. illeg. , su-

perfluous, cites Restio elegans Poir. and Willdenowia compressa

Thunb. (lectotype) in the synonymy.

Restio acuminatus Kunth, Enum. PI. 3 : 413 (1841), nom. illeg.,

non. Restio acuminatus Thunb. (1788). Type: Cape, without

locality, in herb. Thunb. 23221 (UPS, holo.!).

Cannomois acuminata (Kunth) Pillans in Trans. R. Soc. S. Afr.

16 : 413 (1928), nom. illeg., superfluous.

Thamnochortus strictus Kunth, Enum. PI. 3 : 438 (1841). Syn-

types: Cape, 3219 (Wuppertal): between Grasberg Rivier and

Watervals Rivier (-AB), Drege 101 cf (K, lecto.!; BM!; MO!; P!).

3419 (Caledon): Genadendal (-BA), Drege 1642 cf (K!; OXF!;

P!).

Cannomois simplex Kunth, Enum. PI. 3 : 448 (1841); Mast, in

A. DC., Monogr. Phan. 1 : 363 (1878); in FI. Cap. 7 : 142 (1897).

Type: Cape, 3219 (Wuppertal): between Grasberg Rivier and

Watervals Rivier (-AB), Drege 2514 9 (B, lecto.!; B!; K!; MO!;

NY!; OXF!; P!; S!).

Cannomois complanatus Mast, in Bot. Jb. 29 Beibl. 66 : 19

(1900). Syntypes: Cape, 3218 (Clanwilliam): Piketberg, 450 m

(-D), Schlechter 7913 9 (B, lecto.!; BM!; BOL!; BR!; K!; MO!;

P!; S!; Z!); 7912 cf (B!; BM!; BOL!; BR!; K!; MO!; P!; S!; Z!).

Cannomois schlechteri Mast, in Bot. Jb. 29 Beibl. 66 : 18

(1900). Syntypes: Cape, 3318 (Cape Town): Blouberg, amongst

rocks (-CB/CD), Schlechter 8468 9 (B, lecto.!; BM!; K!; MO!;

P!; S!; Z!); 8467 cf (B!; BM!; K!; MO!; P!; S!; Z!).

Cannomois spicatus Mast, in Bot. Jb. 29 Beibl. 66 : 19 (1900).

Syntypes: Cape, 3319 (Worcester): mountains at Villiersdorp

(-CC/CD), Schlechter 99 13 cf (B, lecto. !; BM!; K!; SI; Z!); 9914 9

(BM!; BR!; K!; P!; S!; Z!).

leones: K. svenska VetenskAkad. Handl. 11 t.3 (1790).

A. DC., Monogr. Phan. 1 : t.3 f. 33-41; t.5 f. 15 (1878).

Notes

1 . Thamnochortus strictus is based on two conflict-

ing elements. Drege 101 is a Cannomois species,

whereas Drege 1642 is Hypodiscus argenteus. As

Kunth (1841: 438) noted ‘Descript, juxta specimen

sub. no. 10T, I have lectotypified Drege 101.

2. Both male and female material occurs under

Drege 2514, the type of C. simplex. However, the

protologue only refers to female material. There are

three sheets of this collection in B: two are female

(one from Altona, the other from Liibeck), whereas

the third is male and is from herb. Berol. So it is

presumed that the holotype is lost.

3. Restio acuminatus Thunb. is an excellent de-

scription of a Cannomois based on good material in

herb. Thunb. Unfortunately, Thunberg (1788) cited

Chondropetalum nudum Rottboell in the synonymy,

thereby rendering the name illegitimate. Nees (1830

: 652) noted this and Kunth (1841) described R. acu-

minatus, but excluded Chondropetalum nudum,

thereby publishing a new name, which is illegitimate

as it is a later homonym of R. acuminatus Thunb.

However, when Pillans transferred this name to

Cannomois (1928), it is no longer a later homonym,

but is now superfluous.

4. Hochstetter (1845) suggested in a footnote that

Thamnochortus strictus is a Cannomois, but he failed

to make the actual combination.

5. There is still doubt about the correct delimita-

tion of this species. It is rather variable and may con-

482

Bothalia 15, 3 & 4 (1985)

tain several segregates. The male bracts vary from

rounded (in the type of C. pan’iflorus ) to shortly

acuminate (in the types of T. strictus and C. compla-

natus). The majority of the collections are rhizoma-

tous, but some plants are caespitose. The culms are

generally simple, but in some plants they are

sparsely branched (i.e. the types of W. compressa

and C. complanatus). There is also variation in the

female perianth shape [on which Pillans (1928) sub-

divided the species]. The best solution appears to be

to include all the variation in one species.

6. C. parviflora is common and widespread from

Nieuwoudtville to Swellendam and Houw Hoek. It

generally occurs in dry sandy soils, either on the

coastal forelands or in the mountains, between 100

and 1 300 m.

Cannomois scirpoides (Kunth) Mast, in J. Linn.

Soc., Bot. 10 : 236 (1868); in A. DC., Monogr. Phan.

1 : 362 (1878); in FI. Cap. 7 : 142 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 418 (1928). Type: Cape,

3323 (Willowmore): Welgelegen, in the Lange

Kloof, 600 m (-CD), Drege 2023 cf (B, holo.!; B!;

BM!; K!; MO!; NY!; OXF!).

Thamnochortus scirpoides Kunth, Enum. PI. 3 : 438 (1841).

Cannomois dregei Pillans in Trans. R. Soc. S. Afr. 16 : 416

(1928), nom. illeg. , superfluous name for C. scirpoides (Kunth)

Mast.

Notes

1. In B there are two sheets of Drege 2023. One

sheet is from herb. Meyer., and the other is probably

herb. Berol. and is annotated in Kunth’s hand. This

is the holotype.

2. Pillans (1928) appears not to have realized that

Cannomois scirpoides was based on Thamnochortus

scirpoides of Kunth. Consequently, he published C.

dregei as a nom. nov. for T. scirpoides and misap-

plied C. scirpoides (Kunth) Mast, to C. congesta.

3. Hochstetter (1845) suggested that T. scirpoides

be transferred to Cannomois , but he failed to make

the combination.

4. It may not be possible to separate C. scirpoides

from C. parviflora. The two taxa are allopatric and

morphological distinctions are the colour of the

spathes (not very good) and the position of the guard

cells in the stomata (see Linder, 1984), which is

based, to date, on an inadequate sample.

5. C. scirpoides occurs widespread in the moun-

tains of the southern Cape Province from Worcester

to Uitenhage, but is absent from the wet south-fac-

ing slopes of the coastal mountains east of Rivers-

dale. The altitudinal range is between 300 and 1 600

m. Most collections are from well-drained slopes.

Cannomois virgata (Rottb.) Steud., Syn. PI.

Glum. 2 : 263 (1855); Mast, in A. DC., Monogr.

Phan. 1 : 361 (1878); Pillans in Trans. R. Soc. S. Afr.

16 : 411 (1928); in Adamson & Salter, FI. Cape

Penins. 157 (1950). Type: Cape, without precise

locality, Konig s.n. cf (C, holo.!).

Restio virgatus Rottb., Descriptiones Plantarum Rariorum 10

(1772); Thunb., FI. Cap. edn 1, 338 (1811); edn Schultes, 89

(1823). Thamnochortus virgalus (Rottb.) Kunth, Enum. PI. 3 :

436 (1841).

Restio scopa Thunb.. Diss. Restio 19 (1788); Thunb., FI. Cap.

edn 1, 337 (1811), edn Schultes, 88 (1823). Type: Cape, without

precise locality, in herb. Thunb. 23244 (UPS, holo.!).

Restio elegans Poir. in Lam., Encycl. 6 : 171 (1804). Type:

Cape, without precise locality, herb. Poiret in herb. Moquin-Tan-

don (P, holo.!).

Elegia paniculata Pers., Syn. PI. 2 : 607 (1807), nom. illeg., su-

perfluous name for R. elegans Poir.

Cannomois cephalotes Desv. in Annls Sci. nat., ser. 1, 13 : 43

(1828); Kunth, Enum. PI. 3 : 447 (1841); Mast, in FI. Cap. 7 : 141

(1897). Type: Annls Sci. nat., 1, 13, t. 3 fig. 1 (1828) (iconotype).

Thamnochortus robustus Kunth, Enum. PI. 3 : 436 (1841).

Type: Cape, 3319 (Worcester): Du Toits Kloof (-CA), Drege

1606 cf (B, lecto.!; B!; BM!; K!; MEL!; MO!; P!).

leones: Rottb., Descriptionum et Iconum Rariores t.l f.2

(1773). Annls Sci. nat., ser. 1, 13 : t.3 f.l (1828).

Notes

1. There is a twig in P, labelled ‘Cannomois cepha-

lotes Beauv.’ in Poiret’s hand, which could be the

type of C. cephalotes. However, it is more certain to

use Desvaux’s illustration as an iconotype.

2. Hochstetter (1845) suggested that T. virgatus

(Rottb.) Kunth and T. robustus Kunth belong to

Cannomois, but he failed to make the necessary

combinations.

3. This widespread species is rather variable in its

habit. In the Swartberg Mountains and the dry

north-facing slopes of the Outeniqua Mountains it

grows into a 2-3 m tall, bamboo-like plant, whereas

on the wetter south-facing slopes of the coastal

ranges it grows to about 1 m tall. Plants from the

western Cape are intermediate. I have not been able

to locate any characters which can consistently sep-

arate all the forms.

4. On the Great Swartberg there are large popula-

tions of a plant which is morphologically interme-

diate between C. virgata and C. scirpoides and which

is probably of hybrid origin. It also occupies an inter-

mediate habitat — slightly damp to dry sandy pla-

teaux. In this study it keys out under C. virgata, but

can be separated from this species by the tepals,

which are only a third as long as the nut, instead of

being at least half as long as the nut.

5. C. virgata is ubiquitous in the mountains of the

Cape Floral Region, reaching from Uitenhage to

Nieuwoudtville. In drier areas it grows along streams

and in seepages, whereas in wetter areas it extends

onto the open mountain sides. The altitudinal range

is from near sea level to 1 800 m.

NEVILLEA

Nevillea Esterhuysen & Linder in Bothalia 15 :

66 (1984).

KEY TO THE SPECIES OF NEVILLEA

la Female spikelets taller than the spathes; bracts subacu-

minate; male spikelets ellipsoid, 10-12 mm long

N. singularis

lb Female spikelets obscured by the spathes; bracts acute;

male spikelets oblong, 10-20 mm long N. obtusissima

Nevillea obtusissima (Steud.) Linder in Bothalia

15: 66 (1984). Type: Cape, without precise locality,

Drege 22 cf (K, lecto.!; BM!; BOL!; MO!; NY!;

OXF!; P!).

Bothalia 15, 3 & 4 (1985)

483

FIG. 23. — Nevillea singularis Esterhuysen. a, habit; b, culm bases with a very compacted rhizome, note closely convoluted sheaths

and short awns, x 0,8; c, male inflorescence, x 0,8; d, female inflorescence, x 0,8; e, male spikelet, x 2; f, male bract, X 5; g,

female spikelets with several flowers, taller than the spathes, x 2; h, male flowers with exserted anthers, x 5; i, female flower

with curved wings, x 5; j, dissection of female flower, note membranous petals, minute staminodes, and styles, x 5. (From

Esterhuysen 35130.)

484

Bothalia 15, 3 & 4 (1985)

Restio obtusissimus Steud., Syn. PI. Glum. 2: 252 (1855); Mast,

in A. DC., Monogr. Phan. 1: 296 (1878); in FI. Cap. 7: 93 (1897);

Pillans in Trans. R. Soc. S. Afr. 16: 294 (1928).

Notes

1. There is no material of this species in herb.

Steudel at P, so I have lectotypified the material at

K.

2. N. obtusissima occurs in the mountains from

Tulbagh to Hermanus, usually in marshy habitats,

between 300 and 1 000 m.

Nevillea singularis Esterhuysen , sp. nov., a N.

obtusissima (Steud.) Linder spiculis masculis ellip-

soideis, spathis quam spiculis femineis minoribus,

bracteis femineis subacuminatis, sepalis femineis co-

riaceis differt.

TYPE. - — 3419 (Caledon): Kanonkop, in an exten-

sive marsh at the base of the shale band, 1 050 m

(-BA), Esterhuysen 35130 $ (BOL, holo.!; C; E; F;

K; L; LD; M; MO; NBG; S; STE; TCD; UC; US;

W; WAG).

Plants caespitose, 20-60 cm tall. Rhizome about

6-8 mm thick, culm bases densely aggregated.

Culms very narrowly fistular, terete, simple, stout,

to 4 mm in diameter, surface smooth. Sheaths usu-

ally 2 per culm, closely convoluted, 15-25 mm long,

obtuse, with an up to 5 mm long cylindrical acute

awn, body coriaceous, green, margins very narrowly

membranous, apex brown. Male inflorescence of 1-8

racemose to compound racemose, 10-12 x 5 mm,

elliptical, obtuse spikelets. Spathes ovate, acute,

generally shorter than the spikelets, coriaceous with

chartaceous margins, awn subulate, to 7 mm long.

Bracts imbricate, cartilaginous, obtuse, 4x3 mm,

obscuring the flowers. Flowers subpedicellate, 4-5

mm long. Sepals cartilaginous; lateral sepals

rounded, conduplicate, carina 0,8 mm wide; odd se-

pal flat, 4,5 x 1,2 mm, subacute. Petals membra-

nous, similar in shape to the odd sepal. Anthers ex-

serted at anthesis, obscurely mucronate. Female in-

florescence with 1-4, lax, 15-20 x 4 mm, several-

flowered spikelets. Spathes as in the males. Bracts

coriaceous, shortly acuminate, 7-15 x 6 mm, shortly

awned, concolorous, lowermost bracts sterile. Flow-

ers subsessile, 6,5-7 mm long. Sepals cartilaginous,

acute; laterals conduplicate, 6,5-7 mm long, carina

0,6-1 mm wide; odd sepal flat, 6,5 x 2 mm. Petals

membranous, 6x2 mm, acute to obtuse. Stami-

nodes minute. Styles 2, slender, to 15 mm long, free,

swollen at the base and seated on a stylopodium.

Ovary bilocular, fruit a nutlet. Fig. 23.

N. singularis is known from Kanonkop, near Ge-

nadendal, where it occurs in marshes between 1 000

and 1 200 m. This species is related to N. obtusis-

sima, but is easily distinguished by the smaller, ellip-

soid male spikelets, the female spikelets taller than

their spathes, the subacuminate female bracts and

the coriaceous sepals of the female flowers. The two

taxa are also allopatric, separated by the Rivier-

sonderend Valley.

The anatomy of this species is reported on by

Linder (1984: 66) under the name ‘Restio nr. obtusis-

simus’.

CAPE. — 3419 (Caledon): Kanonkop below the summit,

1 000-1 200 m, (-BA), Esterhuysen 35130 (BOL; C; E; F; K; L;

LD; M; MO; NBG; S; STE; TCD; UC; US; W; WAG); Ester-

huysen 31474 (BOL; K; L; M; MO; S). Esterhysen 34956 (B;

BOL; C; E; F; K; L; LD; M; MO; NGB; PRE;' RSA; S; STE;

UC; US; W); Esterhuysen 35623 (BOL; K; S).

HYDROPHILUS

15. Hydrophilus Finder in Bothalia 15: 66

(1984).

Hydrophilus rattrayi (Pillans) Under in Botha-

lia 15: 66 (1984). Syntypes: Cape, 3319 (Worcester):

Wilde Paarde Berg (-CD), Stokoe 2781 $ (BOL,

lecto.!; K!). 3319 (Worcester): plateau on Fontein-

tjiesberg (-CB), Stokoe 2982 cf (BOL!; K!).

Leptocarpus rattrayi Pillans in Trans. R. Soc. S. Afr. 29: 347

(1942).

Note

1. H. rattrayi has a distribution range from the Ce-

darberg to the Swartberg, with one record from the

Hogsback Mountains near Alice. It is absent from

the Cape Peninsula and from the Outeniqua-Tsitsi-

kamma Mountains. The plants occur in swampy

habitats in sandstone or quartzitic mountains be-

tween 900 and 1 500 m.

ANTHOCHORTUS

16. Anthochortus Nees in Lindl. , Nat. Syst. Bot.

edn 2, 451 (1836); Linder in Bothalia 15: 66 (1984).

Phyllocomos Mast, in Bot. Jb. 29 Beibl. 66: 19

(1900).

KEY TO THE SPECIES OF ANTHOCHORTUS

la Culms distinctly striate to four-angled:

2a Spikelets stipitate; spikelets lateral; from the Cape

Peninsula A. capensis

2b Spikelets sessile, terminal; from the Cape mountains

excluding the Cape Peninsula A. ecklonii

lb Culms smooth to tuberculate, not striate:

3a Spikelets aggregated into terminal heads A. insignis

3b Spikelets solitary:

4a Culms compressed; sheaths with large leaf-

like blades A. graminifolius

4b Culms terete or rarely compressed near the

base; only basal sheaths sometimes with

leaf-like blades:

5a Bracts almost entirely chartaceous; apices not

recurved A. laxiflorus

5b Bracts cartilaginous with distinct mem-

branous margins; long-acuminate; apices

recurved A. crinalis

Anthochortus capensis Esterhuysen , sp. nov.,

inflorescentiis lateralibus, spiculis pedicellatis, brac-

teis duobus hyalinis, a congeneribus diversa.

TYPE. — Cape, 3418 (Simonstown): Muizen-

berg Mountain, about half a mile along the road

above the waterfall (-AB), Esterhuysen 35597 £

(BOL, holo.!; C; E; K; L; LD; M; MO; S; STE;

UC).

Plants tangled. Culms slender, solid, terete, to 0,5

mm in diameter, branching, branches straight, as-

cending, green with a fine reddish mottling. Sheaths

closely convoluted, green or brown, 15-25 mm long,

body obtuse to acute with a narrowly membranous

margin, awn straight, aciculate, of the total

sheath length. Male inflorescence with 1-3 nodes.

Bothalia 15, 3 & 4 (1985)

485

486

Bothalia 15, 3 & 4 (1985)

Spathes 15-30 mm long, piliferous, papyraceous,

pale brown with fine maroon speckling and a narrow

submembranous margin. Bracts narrowly lanceo-

late, overtopping the flowers, acute, papyraceous.

Flowers 2-4 per spathe on a flowering branch with

internodes almost as long as the flowers (a very lax

spikelet ?); flowers shortly pedicellate, about 3 mm

long. Tepals subcartilaginous to papyraceous, 3 x

0,5 mm, acute, the tepals somewhat larger and

firmer than the petals, glabrous. Anthers 1,7 mm

long, slender, the filaments not elongating at anthe-

sis. Female inflorescence a solitary, lateral spikelet,

with the culm continuing up to 5 times the length of

the spikelet. Spathe 10-15 mm long, slender, acute,

about 1,5 mm wide, chartaceous, pale brown with

red mottling. Spikelet 4-5 mm long, single-flowered,

pedicellate. Bracts 2, membranous, acuminate 4x1

mm, at the base enveloping the pedicel of the

flower. Flower on a 0,5-0, 8 mm long pedicel, per-

ianth 2,5-3 mm long. Tepals subequal, chartaceous,

subacute, 2,5 x 0,8 mm, concave, petals marginally

smaller than the sepals. Staminodes minute, styles 2,

densely villous, free to the base. Ovary unilocular.

Fruit a cylindrical, unilocular nutlet, 2,2 x 0,8 mm.

Fig. 24.

A. capensis is known from Muizenberg Mountain

and the Cape Point Reserve, where it occurs on mar-

shy streambanks or in marshes, up to 300 m. It prob-

ably flowers in October.

This very curious species is clearly referable to the

genus Anthochortus because of the thin-walled nuts,

the habit of the plants, the stoloniferous culm-bases

and the structure of the flowers. Linder (1984: 46,

Table 4) reported on the anatomy, under the name

‘ Hypolaena capensis' , showing that it is rather cu-

rious. The effectively lateral spikelets and the two

hyaline bracts and the stipitate spikelets easily distin-

guish this species from all others in the genus.

CAPE. — 3418 (Simonstown): Muizenberg Mountains, about

half a mile along the road above the waterfall (-AB), Esterhuysen

35597 (BOL; C; E; K; L; LD; M; MO; S; STE; UC); in the Witse-

nia Marsh, towards Steenberg Plateau (-AB), Esterhuysen 34774

(BOL; K); along stream on Muizenberg plateau flowing towards

W, and at Nellie’s Pool (-AB). Esterhuysen 35185a (BOL; K; S);

in a marsh at the WNW base of Paulsberg (-AD), Esterhuysen

35806 (BOL; C; E; K; L; M; MO; S; UC).

Anthochortus crinalis (Mast.) Linder, comb.

nov.

Restio crinalis Mast, in J. Linn. Soc., Bot. 8; 229 (1865); in A.

DC., Monogr. Phan. 1; 239 (1878); in FI. Cap. 7: 72 (1897); Pil-

lans in Trans. R. Soc. S. Afr. 16: 296 (1928). Hypolaena crinalis

(Mast.) Pillans in J1 S. Afr. Bot. 18: 119 (1952). Type: Cape, 3319

(Worcester): Du Toits Kloof (-CA), Drege 11 cf (K, holo,!; B!;

BOL!; P!).

Restio oblongus Mast, in J. Linn. Soc., Bot. 8: 229 (1865); in A.

DC., Monogr. Phan. 1: 239 (1878); in FI. Cap. 7: 71 (1897); Pil-

lans in Trans. R. Soc. S. Afr. 16: 296 (1928). Type: Cape, without

precise locality, Drege 65 cf (K, holo.!; B ! ; BOL!; P!).

Hypolaena tabularis Pillans in Trans. R. Soc. S. Afr. 16: 394

(1928); in Adamson & Salter, FI. Cape Penins. 155 (1950). Type:

Cape, 3318 (Cape Town): Table Mountain, Echo Valley (-CD),

Pillans 4143 $ (BOL, holo.!; K!).

Notes

1. The culms vary from tuberculate to smooth.

2. A. crinalis occurs in marshy places, in shallow

soil over rock or on damp rock ledges in the moun-

tains from Riversdale to Bainskloof (near Welling-

ton) and the Cape Peninsula. The altitude ranges be-

tween 600 and 1 800 m.

Anthochortus ecklonii Nees in Lindl., Nat.

Syst. Bot., edn 2, 451 (1836); Kunth, Enum. PI. 3:

486 (1841); Mast, in J. Linn. Soc., Bot. 10: 274

(1868); in A. DC., Monogr. Phan. 1: 398 (1878); in

FI. Cap. 7: 148 (1897); Pillans in Trans. R. Soc. S.

Afr. 16: 432 (1928). Type: Cape, 3320 (Montagu):

Mountains near Swellendam (-CD), Ecklon &

Zeyher s.n. cf (B, lecto.!).

Willdenowia ecklonii (Nees) Dur. & Schinz, Consp. FI. Afr. 5:

525 (1894), non (Nees) Kunth (1841).

Hypolaena tenuis Mast, in J. Linn. Soc., Bot. 10: 265 (1868); in

A. DC., Monogr. Phan. 1: 372 (1878); in FI. Cap. 7: 134 (1897).

Calorophus tenuis (Mast.) Kuntze, Rev. Gen. 747 (1891). Syn-

types: Cape, 3320 (Montagu): summit of a mountain peak near

Swellendam (-CD), Burchell 7360 cf (B!; BOL!; K!; P!). 3321

(Ladismith): lower part of the Lange Berg near Kampsche Berg

(-CD), Burchell 7028 9 (K, lecto. f; BOL!; P!).

Notes

1. There are two collections of Ecklon & Zeyher

from locality 70 in B. The first sheet is herb. Sonder

50, ‘A. ecklonii b floribus minoribus’, which I desig-

nate as lectotype, and the second sheet is annotated

‘var a, floribus majoribus’. It is evident that both

sheets were seen and annotated by Nees. The same

collection at K is female and so is not part of the

type.

2. A. ecklonii occurs in swampy habitats between

600 and 1 400 m between the Winterhoek Mountains

near Uitenhage and the Cold Bokkeveld near Ceres.

It has not been recorded from the Cape Peninsula,

the Piketberg or the Cedarberg.

Anthochortus graminifolius (Kunth) Linder ,

comb. nov.

Restio graminifolius Kunth, Enum. PI. 3: 407 (1841); Mast, in

A. DC., Monogr. Phan. 1: 265 (1878); in FI. Cap. 7: 77 (1897).

Hypolaena graminifolia (Kunth) Pillans in Trans. R. Soc. S.

Afr. 16: 396 (1928). Type: Cape, 3319 (Worcester): French Hoek

Mountains, 300-600 m (-CD), Drege 2021 cf (B, holo.!; B!; BM!;

BOL!; K!; MEL!; MO!; NY!; OXF!; P!; S!).

Hypolaena foliosa Mast, in Bot. Jb. 29 Beibl. 66: 14 (1900).

Mastersiella foliosa (Mast.) Gilg-Ben. in Pflanzenfam. edn 2,

15a: 25 (1930). Syntypes: Cape, 3419 (Caledon): mountains near

Houw Hoek, 900 m ( — AA), Schlechter 7347 cf ( B ! ; BOL!; BR!;

K!; MO!; P! ; S! ; Z!). 3318 (Cape Town): Table Mountain, lower

plateau (-CD), Dod 3403 C f (K, lecto.!; BOL!).

Notes

1. Drege 2021 is a mixed collection, including ma-

terial of A. graminifolius and Restio depauperatus

Kunth. Superficially the two species are rather simi-

lar, but A. graminifolius is distinct on account of the

flattened awn on the sheaths.

2. There are two sheets of Drege 2021 in B, but

only one (here selected as lectotype) is annotated by

Kunth.

3. Hypolaena foliosa is based on two discordant

elements. Schlechter 7347 has tuberculate, terete

culms and sheaths without large foliose awns. This

agrees with A. laxiflorus (‘Culmus . . . punctulato-

exasperatus’, Nees, 1830: 663). Dod 3403 has

smooth, compressed culms and large foliose awns on

the sheaths. This agrees with the protologue of the

Bothalia 15, 3 & 4 (1985)

487

species (‘ramulis patentibus glabris filiformis com-

pressis’, vaginis . . . longe foliaceo-ligulatis’. Masters,

1900: 14). Contrary to the implicit lectotypification

in Pillans (1928), I have selected Dod 3403 as lecto-

type.

4. A. graminifolius occurs in marshy habitats from

Babylons Tower (near Caledon) to the Cape Penin-

sula and to Villiersdorp. The altitudinal range of the

species is from 450 to 1 300 m.

Anthochortus insignis (Mast.) Linder, comb.

nov.

Phyllocomos insignis Mast, in Bot. Jb. 29 Beibl. 66: 20 (1900);

Pillans in Trans. R. Soc. S. Afr. 16: 433 (1928). Type: Cape, 3319

(Worcester): mountains above Mitchells Pass, 450 m (-AD),

Schlechter 9956 9 (B, lecto.l; BM!; BOL!; BR!; Kl; MO!; SI; Z!);

9956 cf (Bl; BOL!; BM!; BR!; K!; MO!; S!; Z!).

Notes

1. The type collection is hermaphrodite, but sub-

sequent collections are dioecious. The type collec-

tion is probably monstrous.

2. A. insignis is known from the mountains around

Ceres, between 450 and 1 000 m. It extends from the

foothills of the Hex River Mountains at Ezelsfontein

to Visgat and Groen to the north.

Anthochortus laxiflorus (Nees) Linder , comb.

nov.

Hypolaena laxiflora Nees in Linnaea 5: 663 (1830); Kunth,

Enum. PI. 3: 451 (1841); Mast, in A. DC., Monogr. Phan. 1: 369

(1878); in FI. Cap. 7: 133 (1897); Pillans in Trans. R. Soc. S. Afr.

16: 395 (1928); in Adamson & Salter, FI. Cape Penins. 155 (1950).

Calorophus laxiflorus (Nees) Kuntze, Rev. Gen. 747 (1891). Mas-

tersiella laxiflora (Nees) Gilg-Ben. in Pflanzenfam. edn 2,15a: 25

(1930). Type: Cape, 3318 (Cape town): Table Mountain (-CD),

Ecklon 843 $ (BOL, lecto.!; MO!); 843 cf (BOL!; MO!).

Hypolaena stokoei Pillans in Trans. R. Soc. S. Afr. 29: 352

(1942). Syntypes: Cape, 3418 (Simonstown): Hottentots Holland

Mountains (-BB), Stokoe 7621 9 (BOL, lecto.!; K!); 7620 cf

(BOL!; K!).

Note

1. A. laxiflorus occurs on the Cape Peninsula and

the Hottentots Holland Mountains.

MASTERSIELLA

17. Mastersiella Gilg-Ben. in Pflanzenfam. edn

2,15a: 25 (1930); Linder in Bothalia 15: 66 (1984).

KEY TO THE SPECIES OF MASTERSIELLA

la Female flowers sessile; male spikelets oblong M. digitata

lb Female flowers pedicellate; male spikelets globose:

2a Culms terete; spikelets reddish; petals minute

M. purpurea

2b Culms somewhat compressed near the apices;

spikelets green; petals almost as long as the

sepals M. spathulata

Mastersiella digitata (Thunb.) Gilg-Ben. in Pf-

lanzenfam. edn 2,15a: 25 (1930). Syntypes: Cape,

3318 (Simonstown): Hottentots Holland Mountains

(-BB), in herb. Thunb. 23232 0* (UPS, lecto.!);

23231 cf (UPS!).

Restio digitatus Thunb., Diss. Restio 18 (1788); Thunb., FI.

Cap. edn 1, 335 (1811); edn Schultes, 88 (1823); Kunth, Enum.

PL 3: 410 (1841). Calorophus digitatus (Thunb.) Kuntze, Rev.

Gen. 2: 747 (1891). Hypolaena digitata (Thunb.) Pillans in Trans.

R. Soc. S. Afr. 16: 391 (1928); in Adamson & Salter, FI. Cape

Penins. 154 (1950).

Hypolaena eckloniana Nees ex Mast, in J. Linn. Soc., Bot. 10:

263 (1868); in A. DC., Monogr. Phan. 1: 369 (1878); in FI. Cap.

7: 132 (1897). Syntypes: Cape, 3419 (Caledon): Houw Hoek

Mountains, July (-AA), Zeyher 4348 cf (K, lecto.!; BOL!; P! ; S! ;

Z!;); Baviaanskloof Mountains near Genadendal (-BA), Burchell

7844 cf (BOL!; K!); 7897 cf (?)• 3418 (Simonstown): Sir Lowry’s

Pass (-BB), Burchell 8267 cf (K!); Hottentots Holland kloof

(-BB), Ecklon 950 9 (?) ; Muizenberg (-AB), Zeyher 4348b cf &

9 (?); Simonstown (-AB), Ecklon 997 9 (?) ; without precise

locality, Masson s.n. cf (?); Thom 1031 cf (K!); Thom 632 cf (K!);

Grey s.n. cf (OXF!); Sieber 228 cf (OXF!).

Hypolaena incerta Mast, in FI. Cap. 7: 133 (1897). Syntypes:

Cape, 3419 (Caledon): Baviaanskloof Mountains near Genaden-

dal (-BA), Burchell 7844 cf (K, lecto.! K! ) . 3418 (Simonstown):

Sir Lowry’s Pass (-BB), Burchell 8267 cf (K!); without precise

locality, Thom 1031 cf ( K! ).

Icon: J. Linn. Soc., Bot. 10: t.7a (1868).

Notes

1. In the protologue of H. eckloniana Masters

cites ‘Burchell 7897' without locality. This number is

repeated in the Monogr. Phan., but in FI. Cap. ‘ Bur-

chell 7989’ is mentioned and not Burchell 7897. The

latter number is at K, but I have not been able to

find the former number. It may be a printing error.

2. Masters (1897) treated H. incerta as a segregate

of H. eckloniana , based on bract shape. All three

specimens on which H. incerta is based are named

H. incerta (in pencil) and H. eckloniana (in ink) by

Masters.

3. M. digitata occurs between the Potberg (near

Bredasdorp) the Cape Peninsula and Villiersdorp.

Collections are generally from dry rocky slopes, usu-

ally from below 400 m. There are some collections

from the Baviaanskloof and Fransch Hoek Moun-

tains from as high as 1 200 m.

Mastersiella purpurea (Pillans) Linder , comb.

nov.

Hypolaena purpurea Pillans in Trans. R. Soc. S. Afr. 16: 393

(1928). Type: Cape, 3323 (Willowmore): hills near Avontuur

(-CA/CC), Fourcade2898 9 (BOL, lecto. ! ; K !);2898 cf (BOL!).

Notes

1. There is some variation in the length of the stipe

of the female flower and in some collections the stipe

approaches that of M. spathulata.

2. This species is allopatric with M. digitata.

3. M. purpurea is centred in the southern Cape,

along the Swartberge from Ladismith to Willow-

more, the Kouga and Baviaanskloof Mountains and

occasionally along the coastal mountains from Swel-

lendam to Knysna. It occurs from dry to mesic con-

ditions in well-drained, usually rocky soils, between

400 and 1 500 m.

Mastersiella spathulata (Pillans) Linder, comb.

nov.

Hypolaena spathulata Pillans in Trans. R. Soc. S. Afr. 30: 264

(1945). Type: Cape, 3323 (Willowmore): Kouga Mountains, peak

east of Smutsberg, 1 200 m (-DA), Esterhuysen 6982 9 (BOL,

lecto.!; K!); 6982 cf (BOL!; K!).

Notes

1 . This species has a curious distribution. It occurs

above 1 200 m in the Kouga and Tsitsikamma Moun-

488

Bothalia 15, 3 & 4 (1985)

tains, at about 1 000 m in the Outeniqua Mountains,

300 m in Garcias Pass near Riversdale, and on the

limestone hills and gravelly slopes near sea level on

the Bredasdorp coastal plateau.

HYPODISCUS

18. Hypodiscus Nees in Lindl., Nat. Syst. Bot. edn

2, 450 (1836); Linder in Bothalia 15: 67 (1984).

Lepidantlms Nees in Linnaea 5: 665 (1830).

Leucoploeus Nees in Lindl., Nat. Syst. Bot. edn 2,

450 (1836).

Boeckhia Kunth, Enum. PI. 3: 448 (1841).

Note

The exact ornamentation of the nut is very import-

ant for identification and this may be difficult to de-

termine from a young flower.

Hypodiscus albo-aristatus (Nees) Mast, in J.

Linn. Soc., Bot. 8 : 255 (1865); in A. DC., Monogr.

Phan. 1 : 382 (1878); in FI. Cap. 7 : 140 (1897); Pil-

lans in Trans. R. Soc. S. Afr. 16 : 402 (1928); in Ad-

amson & Salter, FI. Cape Penins. 156 (1950). Type:

Cape, without precise locality, Ecklon s.n. cf in herb,

cl. Giintheri (?). 3325 (Port Elizabeth): Van Sta-

densberg (-CC), Ecklon & Zeyher 816 cf (K, neot.!;

MEL!; NBG!).

Restio albo-aristatus Nees in Linnaea 5 : 635 (1830); Kunth,

Enum. PI. 3 : 407 (1841).

Hypodiscus duplicatus Hochst. in Flora, Jena 28 : 338 (1845).

Syntypes: Cape, without precise locality, Krauss s.n. (?); Ecklon

113 (?).

Hypodiscus oliverianus Mast, in J. Linn. Soc., Bot. 10 : 254

(1868); in A. DC., Monogr. Phan. 1 : 381 (1878); in FI. Cap. 7 :

137 (1897). Hypodiscus albo-aristatus (Nees) Mast. var. oliveria-

nus (Mast.) Pillans in Trans. R. Soc. S. Afr. 16 : 404 (1928). Syn-

types: Cape, 3418 (Simonstown); Nieuw Kloof, Houw Hoek

Mountains (-BB), Burcheil 8118 2 (K, lecto.!; BOL!); 8118 cf

(BOL!; K!).

Hypodiscus capitatus Mast, in J. Bot., Lond. 1901: 402 (1901).

Syntypes: Cape, 3418 (Simonstown): near Hout Bay (-AB), Dod

2645 2 (K, lecto.!; BM!; BOL!); 2644 c f (BM!; BOL!; K!); with-

out precise locality, Thom 1060 cf (K!).

Notes

1. From the information given in the protologue,

it is almost impossible to determine which Ecklon

collection is the type. Consequently, I have desig-

nated a neotype, which is cited by Masters and which

appears to fit the diagnosis.

KEY TO THE SPECIES OF HYPODISCUS

la Culms more or less striate:

2a Culms compressed

2b Culms terete:

3a Perianth absent

3b Perianth present:

4a Rhizomes creeping; culms widely spaced on the rhizomes

4b Rhizomes short; culms closely aggregated:

5a Styles fused below:

6a Nut sessile, base rounded

6b Nut elevated above the perianth on a pedicel; base concave:

7a Nut 10 mm long

7b Nut 3-5 mm long

5b Styles free to the base

lb Culms smooth:

8a Culms compressed (often only slightly):

9a Nut smooth, about 2 mm long

9b Nut obscurely lobed, about 5 mm long

8b Culms terete:

10a Spikelets entirely obscured by the spathes

10b Spikelets exposed above or on the side of the spathes:

11a Nut with a smooth cap

lib Nut with a variously warty or lobed cap:

12a Nut irregularly tubercled over the entire surface

12b Nut surface only partly tubercled:

13a Nut with a warty cap

13b Nut with a cap of teeth or fleshy lobes:

14a Nut with linear fleshy lobes

14b Nut-cap with 1-3 series of teeth:

15a Inflorescence of 1-3 erect spikelets; spathes dark with pale margins

15b Inflorescence with 3-8 spikelets borne at 45°; spathes concolorous, reddish

H. willdenowia

... H. argenteus

..H. procurrens

... H. neesii

H. sulcatus

. . H. rigidus

. H. striatus

... H. squamosus

H. rugosus

H. montanus

H. aristatus

H. rugosus

II. laevigatus

H. synchroolepis

H. albo-aristatus

H. alternans

Bothalia 15, 3 & 4 (1985)

489

2. I have not succeeded in tracing either of the

types of H. duplicatus, but from the description it

probably belongs here.

3. The var. oliverianus differs only by its yellow

spathes and bracts, a form of variation also encoun-

tered in H. aristatus.

4. H. albo-aristatus occurs on well-drained moun-

tain slopes between Humansdorp, Oudtshoorn, Tul-

bagh and the Cape Peninsula. It is found on both wet

coastal mountains (Tsitsikamma, Langeberg) and

dry inland mountains (Swartberg, Witteberg). The

altitudinal range is from 100 to 1 600 m.

Hypodiscus alternans Pillans in Trans. R. Soc.

S. Afr. 29 : 353 (1942). Type: Cape, ? 3419 (Cale-

don); between Palmiet River and Kleinmond

(-AC), Stokoe 1342 cf (BOL, holo.!; B!; K!).

Note

1. H. alternans is known from the mountains be-

tween the Bot River and Gordons Bay. It occurs be-

low 300 m on well-drained sandy to stony, rarely

clayey, slopes.

Hypodiscus argenteus (Thunb.) Mast, in J.

Linn. Soc., Bot. 10 : 261 (1868); in A. DC., Monogr.

Phan. 1 : 383 (1878); in FI. Cap. 7 : 137 (1897); Pil-

lans in Trans. R. Soc. S. Afr. 16 : 405 (1928). Type:

Cape, without precise locality, in herb. Thunberg.

23222 Cf (UPS, holo.!; S!).

Restio argenteus Thunb., Diss. Restio 14 (1788); Thunb., FI.

Cap. edn 1, 326 (1811); edn Schultes, 85 (1823). Thamnochor-

tus argenteus (Thunb.) Kunth, Enum. PI. 3 : 432 (1841); Mast, in

FI. Cap. 7 ; 123 (1897).

Leucoploeus argenteus Nees ex Mast, in J. Linn. Soc., Bot. 10 :

262 (1868), nom. nud.

Willdenowia simplex N.E.Br. in FI. Cap. 7 ; 756 (1900). Type:

Cape, 3321 (Ladismith): Garcias Pass, 300 m (-CC), Galpin

4817a 9 (K, holo.!).

Note

1 . The distribution range of H. argenteus is from

Riversdale to the North Cedarberg and the Cape

Peninsula. It generally occurs on dry and well-

drained stony slopes between 100 and 1 200 m.

Hypodiscus aristatus (Thunb.) Krauss in Flora,

Jena 28 : 338 (1845); in A. DC., Monogr. Phan. 1 :

380 (1878); in FI. Cap. 7 : 137 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 400 (1928); in Adamson

& Salter, FI. Cape Penins. 156 (1950). Syntypes:

Cape, without precise locality, in herb. Thunb.

23223 cf (UPS, lecto.!); 23224 $ (UPS!).

Restio aristatus Thunb., Diss. Restio 10 (1788); Thunb., FI.

Cap. edn 1, 318 (1811); edn Schultes, 83 (1823); Kunth, Enum.

PI. 3 ; 383 (1841).

Hypodiscus protractus Mast, in Bot. Jb. 29 Beibl. 66 : 17

(1900). Hypodiscus aristatus (Thunb.) Krauss var. protractus

(Mast.) Pillans in Trans. R. Soc. S. Afr. 16 : 402 (1928). Syntypes:

Cape, 3418 (Simonstown): Sir Lowry's Pass, 150 m (-BB),

Schlechter 7795 $ (B, lecto.!; BM!; BOL!; K!; MO!; P!; S!; Z!);

7794 cf (B!; BM!; BOL!; K!; MO!; P!; S!; Z!).

Hypodiscus aristatus (Thunb.) Krauss var. bicolor Mast, in J.

Linn. Soc., Bot. 10 : 253 (1868); in A. DC., Monogr. Phan. 1 : 381

(1878); in FI. Cap. 7 : 138 (1897). Syntypes: Cape, 3218 (Clanwil-

liam): between Piquenierskloof and Markuskraal (-DB), Drege

2509 9 (K, lecto!; MO!); Brakfontein, June (-BD), Ecklon &

Zeyher s.n. cf & 9 (MEL!). 3419 (Caledon): Grietjiesgat, April

(-AA ). Ecklon & Zeyher s.n. 9 ( K! ; MO!; NY!); Donkerhoek

Mountain (-AB), Burchell 7963 9 ( K ! ) ; Baviaanskloof Mountain

(-BA), Burchell 7594 9 (K!).

Notes

1. Thunberg based his description on two sheets.

Herb. Thunb. 23223 is male and agrees with the cur-

rent concept of H. aristatus. Herb. Thunb. 23224 is

female and is H. albo-aristatus. Consistent with cur-

rent usage, the former sheet is lectotypified.

2. When Nees described the genus Hypodiscus

(1836), he noted below the description ‘Restio arista-

tus L. et plures sp. confinis’, but he never actually

made any combinations. Kunth (1841) did not

uphold the genus, consequently the valid combina-

tion was made by Krauss in 1845.

3. The two varieties have been proposed on the

basis of the paler spathes and bracts, but I do not

think that this is a good character, as it occurs in par-

allel in H. albo-aristatus.

4. H. aristatus is almost ubiquitous in the Cape

Floral Region, growing on well-drained sandy soils

from near sea level to 1 500 m, from Humansdorp to

the Cedarberg.

Hypodiscus laevigatus (Kunth) Linder , comb.

nov.

Boeckhia laevigata Kunth, Enum. PI. 3 : 450 (1841). Type:

Cape, 3218 (Clanwilliam): Piketberg, 450-900 m (-D), Drege

2478 9 (K, lecto.!; MO!; OXF!; P!).

Dovea binata Steud., Syn. PI. Glum. 2 : 248 (1855). Hypodiscus

binatus (Steud.) Mast, in J. Linn. Soc., Bot. 10 : 258 (1868); in

A. DC., Monogr. Phan. 1 : 388 (1878); in FI. Cap. 7 : 138 (1897);

Pillans in Trans. R. Soc. S. Afr. 16 : 404 (1928). Type: Cape, 3218

(Clanwilliam): Piketberg, 450-900 m (-D), Drege 2477 cf (P,

holo.!; B!; K!; MO!; NY!; OXF!; P!; S!).

Notes

1. It is strange that Pillans (1928) did not realize

that the correct name for this species is H. laeviga-

tus, as he cited the type under H. binatus. The types

of Dovea binata ( Drege 2477) and Boeckhia laevigata

(Drege 2478) are the male and female of the same

collection from the Piketberg. Drege often matched

the male and female of the same species in conse-

cutive numbers, but the taxonomists in Europe took

little notice of this, often placing. the two numbers in

different species or even genera.

2. In P there are three sheets of Drege 2477 . The

holotype is in 'herb. SteudeP.

3. H. laevigatus occurs between the South Cedar-

berg and Riversdale and between Kleinmond and

the Witteberg. It is most common on the drier inland

mountains. It grows on stony to sandy well-drained

slopes between sea level and 1 500 m.

Hypodiscus montanus Esterhuysen, sp. nov.,

species insignis vaginis laxe convolutis, floribus mas-

culis solitariis, petalis femineis sepalis longioribus,

ab aliis speciebus nobis notis bene distincta.

TYPE. — Cape, 3320 (Montagu): Goedgeloof

Peak (Misty Point), on E slopes towards the summit

(-CD), Esterhuysen 34521 $ (BOL, holo.!; C; E; F;

K; L; LD; M; MO; S; STE; TCD; UC; US; W).

Plants caespitose. Rhizome 2-3 mm in diameter,

culms arising close together from the upper surface.

490

Bothalia 15, 3 & 4 (1985)

scales loosely convoluted, obtuse to acute, 6-8 mm

long. Culms solid, terete or slightly compressed,

simple, about 1 mm in diameter, very obscurely sul-

cate. Sheaths loosely convoluted, 20-35 mm long,

cartilaginous to chartaceous, body yellow to pale

brown with a hyaline margin, acute, awn piliferous,

to 5 mm long, sheath solitary on the culms, but with

several imbricate sheaths at the base of the culm,

grading into the scales. Male inflorescence about 3

cm long. Spathes 2-4, cartilaginous, 20 x 8 mm,

acute, awn piliferous, about 2 mm long, body

brown, margins yellow, completely obscuring the

spikelets. Inflorescence branches shorter than the

spathes, spathellae chartaceous, conduplicate, lin-

ear, overtopping the flowers. Bracts like the spathel-

lae, as tall as the flowers. Flowers solitary, subpedi-

cellate, perianth 5-6 mm long. Sepals chartaceous,

5-6 x 0,6 mm, very acute, somewhat conduplicate.

Petals more membranous than the sepals, acute, 4,5

x 1,5-2 mm. Anthers 3,5-4 mm, very shortly mucro-

nate, dehiscing inside the perianth. Pistillode min-

ute. Female inflorescence similar to the male, but fe-

male spikelets are solitary in each spathe. Spikelets

stipitate, 7-8 x 3 mm. Bracts 3-5 mm long, 4, acute

to truncate, conduplicate to flat, awned to emucro-

nate. Flower solitary, sessile, perianth 5 mm long.

Sepals chartaceous, truncate, upper margin serru-

late, subspathulate, 4 x 2,5 mm. Petals more mem-

branous than the sepals, subspathulate, truncate,

apically serrulate-emarginate, 5 x 2,5 mm. Styles 2,

4^-5 mm long, free to the base, the lower portion

slender, the upper stigmatic half 0,5 mm in di-

ameter, plumose with the stigmatic hairs much divi-

ded. Ovary unilocular, cap on the ovary apically cra-

teriform. Fruit a nitid black unilocular spindle-

shaped, 5 x 2,5 mm nutlet.

H. montanus is known only from the slopes to the

east of Goedgeloof Peak. It occurs between 1 500

and 1 700 m, in shallow soil among rocks. The flow-

ering time is in summer, probably in February.

This very curious species is superficially interme-

diate between the genera Elegia and Hypodiscus.

The solitary male flowers with the anthers dehiscing

inside the flowers and the loosely convoluted sheaths

are typical of several species in Elegia, whereas the

female spikelet, with its sterile bracts and plumose

styles, is more typical of the situation in Hypodiscus.

The anatomy clearly indicates that this species is a

Hypodiscus (Linder, 1984: Table 4), but it occupies

a very isolated position. It, and H. squamosus, are

the only two taxa in the genus in which the perianth

equals the nutlet and both have plumose styles. It is

possible that they may be related, despite the nu-

merous differences between them.

CAPE. — 3320 (Montagu): below the summit of Goedgeloof

Peak, to the east side (-CD), Esterhuysen 32920 (B; BOL; C; E;

F; GRA; K; L; LD; M; MO; NBG; PRE; RSA; S; STE; TCD;

UC; US; W; WAG); Esterhuysen 33419 (BOL; E; K; L; M; MO;

S); Esterhuysen 34521 (BOL; C; E; F; K; L; M; MO; S; STE;

TCD; UC; US; W).

Hypodiscus neesii Mast, in J. Linn. Soc., Bot.

10 : 260 (1868); in A. DC., Monogr. Phan. 1 : 384

(1878); in FI. Cap. 7 : 138 (1897); Pillans in Trans.

R. Soc. S. Afr. 16 : 407 (1928). Type: Cape, 3218

(Clanwilliam): near Brak Fontein, June (-DB), Eck-

lon s.n. $ (K, lecto.!; BOL!; MEL!; NY!; S!; Z!);

s.n. cf (BOL!; K!; NY!; S!; Z!).

Hypodiscus eximius Mast, in Bot. Jb. 29 Beibl. 66 : 17 (1900).

Syntypes: Cape, 3219 (Wuppertal): Ezelbank (-AC), Schlechter

8821 $ (B,lecto.!;BM!;BOL!;BR!;K!;MO!;P!;S!;Z!);SS20cf

(B!; BOL!; BR!; K!; MO!; P!; S!; Z!).

Notes

1. The type material of H. neesii in S, MEL, and

Z is annotated ‘ Leucoploeus striatus N. ab E’. Eck-

lon noted on several sheets that the common name

of this species was ‘Steenbok Riet’.

2. H. neesii occurs in dry fynbos on sandstone de-

rived soils from the North Cedarberg to Tulbagh and

to Laingsburg. The altitudinal range of the species is

between 300 and 1 500 m.

Hypodiscus procurrens Esterhuysen, sp. nov.,

H. willdenowia (Nees) Mast, affinis, culmis striatis,

rhizomatibus repentibus, sed culmis teretibus, stylis

ad infimum connatis.

TYPE. — Cape, 3419 (Caledon): near Pearly

Beach (-CB), Esterhuysen 33000 $ (BOL, holo.!; C;

E; K; L; LD; M; MO; S).

Plants subcaespitose, spreading from rhizomes,

15-20 cm tall. Rhizomes spreading, 2-3 mm in di-

ameter, rarely branching, scales imbricate, acute,

5-8 mm long. Culms solid, simple, erect, arising

about 5-10 mm apart on alternating sides of the rhi-

zome, to 1 mm in diameter, terete, striate. Sheaths

closely convoluted, 10—15 mm long, cartilaginous,

pale green, mottled towards the apex, rounded with

a narrow membranous margin, awn slender acicu-

late, about 15 mm long; at base of culms the sheaths

are shorter and imbricate, and grade into the scales.

Male inflorescence a solitary, 10 x 2-3 mm spikelet.

Spathe 7,5 mm long, awn 2,5 mm long, brown and

gold mottled, acute. Bracts all fertile, similar to the

spathe but progressively smaller, concolorous. Flow-

ers subpedicellate, perianth 6 mm long. Sepals 4,5 x

0,5 mm, the lateral sepals are conduplicate, charta-

ceous, hyaline at the base, reddish-mottled in the

upper §. Petals 6 x 1,5 mm, texture and colours the

same as the sepals. Anther 3 mm long, much ex-

serted at anthesis. Female inflorescence like the

male, spikelet with a single flower. Spathe 5,5 mm

long, acute, awn 2 mm long, brown with golden

mottling. Bracts 3, all similar to the spathe, but more

acuminate, with the inner bracts smaller. Flower

subpedicellate. Tepals subequal, oblong, 1,5 mm

long, imbricate, apex truncate to obtuse-lacerate.

Styles two, about 6 mm long, the lower 1,5 mm

united, stigmatic areas densely bearded. Ovary cap

hemispherical, smooth, 1,2 mm in diameter. Fruit a

pear-shaped nut, seated on a 1 mm pedicel, 5 x 3,5

mm, surface pitted, perianth persistent, charta-

ceous, obtuse, imbricate at the base. Fig. 25.

H. procurrens occurs on the Bredasdorp-Rivers-

dale coastal shelf, in limestone-derived sand. The

species appears to be locally abundant.

This species is clearly related to H. willdenowia in

its creeping rhizomes and striate culms, but it can be

readily distinguished by the terete culms and fused

styles. From the other species in the Hypodiscus

group with striate culms it can be separated by the

Bothalia 15, 3 & 4 (1985)

FIG. 25. — Hypodiscus procurrens Esterhuysen. a, male plants with long creeping rhizomes, x 1; b, male spikelet, x 8; c, male

flower, with exserted anthers and lateral sepals shorter than the petals, x 8; d, female culms, x 1; e, female spikelet with

solitary flower, x 8; f, female flower at anthesis, note stout bilobed style, long-papillate on the inside, x 8; g, nutlet, sessile

with a pitted surface, x 8. (From Esterhuysen 33000.)

492

Bothalia 15, 3 & 4 (1985)

creeping rhizomes. For notes on the anatomy see

Linder, 1984: Table 4.

CAPE. — 3419 (Caledon): near Pearly Beach (-CB), Esterhuy-

sen 33000 (BOL; C; E; K; L; LD; M; MO; S). 3420 (Bredasdorp):

between Zoetendalsvallei and Struis Bay (-CA), Esterhuysen

31243a (BOL). 3421 (Riversdale): 6.7 miles SE of Vermaaklik-

heid P.O. (-AC), Acocks 23246 (K; PRE).

Hypodiscus rigidus Mast, in Bot. Jb. 29 Beibl.

66 : 18 (1900). Syntypes: Cape, 3419 (Caledon): hills

near Mierkraal, 75 m (-DB), Schlechter 10515 $ (B,

lecto.!; BM!; BOL!; BR!; K!; MO!; P!; S!; Z!);

10514 cf (B!;BM!;BOL!;BR!;K!;MO!;P!;S!;Z!).

Notes

1. Pillans (1928 : 408) included this species in H.

striatus. It is indeed very closely related to H. striatus

and may be regarded as a limestone segregate. Es-

terhuysen distributed material of this species under

the name ‘ Hypodiscus maritimus Esterhuysen’. It is

readily distinguished from H. striatus by the spathes

and the fused styles.

2. H. rigidus is known from the Bredasdorp coast,

between Cape Infanta and the Soetanysberg.

Hypodiscus rugosus Mast, in J. Linn. Soc., Bot.

10 : 255 (1868); in A. DC., Monogr. Phan. 1 : 386

(1878); in FI. Cap. 7 : 140 (1897). Syntypes: Cape,

Swellendam distr. , Voormansbosch, Oct., Zeyher

4336 $ (K, lecto.! MEL!); without precise locality,

Ecklon & Zeyher 85 9 (?).

Hypodiscus tristachyus Mast, in Bot. Jb. 29 Beibl. 66 : 17

(1900). Syntypes: Cape, ‘Hottentots Holland Berg und bei Con-

stantia’, Zeyher s.n. $ (EL lecto.!). 3419 (Caledon): mountains at

Caledon (-AB), Zeyher s.n. cf (MEL!); without precise locality,

Ludwig s.n. (7).

Hypodiscus zeyheri Mast, in Bot. Jb. 29 Beibl. 66 : 17 (1900).

Type: Cape, 34(8 (Simonstown): Hottentots Holland (-BB),

Zeyher s.n. $ (B, holo.!; MEL!).

Hypodiscus paludosus Pillans in Trans. R. Soc. S. Afr. 29 : 354

(1942); in Adamson & Salter, FI. Cape Penins. 156 (1950). Type:

Cape, 3418 (Simonstown): Steenberg (-AB), Pillans 4179 cf

(BOL, holo.!; K!).

Hypodiscus parkeri Pillans in J1 S. Afr. Bot. 18 : 121 (1952).

Syntypes: Cape, 3418 (Simonstown): Somerset West (-BB), Par-

ker 3520 $ (BOL, lecto.!; MO!); 3519 cf (BOL!; MO!).

Notes

1. The types of H. tristachyus and H. zeyheri in B

are both annotated in Masters’s hand (see Fig. 2).

Presumably the missing type specimens of H. trista-

chyus mentioned above were also there, but were

lost during the war. Pillans (1928) included H. trista-

chyus, H. zeyheri and H. rugosus under H. albo-aris-

tatus.

2. There is some variation in the degree of com-

pression of the culms and on some specimens the

culms appear terete.

3. H. rugosus occurs below 450 m on the flats and

hills of the coastal forelands between Malmesbury

and Riversdale. Collections are recorded from sandy

or clayey soils, usually from localities wet in winter

and dry in summer.

Hypodiscus squamosus Esterhuysen, sp. nov.,

species msignis spiculis parvulis, combinatione spa-

tharum membranacearum cum culmis compressis ab

aliis speciebus nobis notis bene distincta.

TYPE. — Cape, 3419 (Caledon): Galgeberg,

slopes above the neck of the pass (-BA), Esterhuy-

sen 34540 9 (BOL, holo!; B; C; E; F; GRA; K; L;

LD; M; MO; NBG; NY; PRE; RSA; S; STE; TCD;

UC; US; W; WAG).

Plants caespitose, 20-60 cm tall. Rhizomes as-

cending, possibly above the ground, about 2 mm in

diameter, scales imbricate, loosely convoluted, red-

dish with hyaline margins, acute, 8x3 mm, awn aci-

culate, 2 mm long. Culms solid, compressed, simple,

to about 1 mm in width, surface smooth to obscurely

rugulose. Sheaths closely convoluted, 25-35 mm

long, produced occasionally, chartaceous with a

membranous margin, apex produced into a pilifer-

ous to 8 mm long awn. Male inflorescence compound

racemose, 2-6 x 0,8 cm. Spathes up to 40 x 2 mm,

chartaceous, acute, like the sheaths, Spathellae acu-

minate, membranous with reddish midveins, up to

10 mm long, overtopping the spikelets. Spikelets

subglobular, 4-15 flowered, 2-4 mm in diameter.

Bracts acuminate, concave, obscuring the flowers,

cartilaginous, 1,5-3 x 1 mm, imbricate. Flowers sub-

sessile, perianth 1,5-2 mm long. Tepals acute, sube-

qual, chartaceous, 1,5-2 x 0,6 mm; lateral sepals

conduplicate, carinate, mucronate, subspathulate.

Anthers exserted at anthesis, 1-1,2 mm long, mucro

0,3 mm long. Female inflorescence compound race-

mose, 3-5 x 0,5 cm, not very visible. Spathes like

the sheaths, but more hyaline. Inflorescence

branches somewhat branched, with 2-5 spikelets per

spathe. Spathellae hyaline, often with a reddish mid-

rib, acute, with a long piliferous awn, overtopping

the spikelet. Spikelets 3-5 mm long, single-flowered.

Bracts and spathe similar, obtuse to truncate, coria-

ceous, 3 mm long. Flowers subpedicellate, perianth

2,5-3 mm long. Sepals like the bracts, 3 x 1,8 mm;

petals 2 x 1,2 mm, truncate, chartaceous. Styles 2,

free to the base, 3-4 mm long, the apical half is vil-

lous. Ovary unilocular. Fruit a nitid black nutlet, 2 x

2 mm, with a persistent perianth the same size as the

nutlet.

H. squamosus occurs on the Riviersonderend

Mountains and the Bosjesveld Mountains. The pop-

ulations are at altitudes between 900 and 1 300 m,

growing amongst rocks in gravelly sand. The plants

appear to flower in summer, during February.

This is a very distinctive species, unique in the

genus by the combination of compressed culms,

globose nuts and the curious aerial rhizomes. The

flowers are much smaller than is usual in Hypo dis-

cus. There may be some affinity to H. montanus, but

it only shares the aerial rhizomes with this species.

There may also be an affinity to H. neesii, by the

hyaline spathes and very small and numerous male

spikelets.

CAPE. — 3319 (Worcester): Wolfieskop, NE of Villiersdorp in

the Bosjesveld Mountains, on summit (-CD), Esterhuysen 32064

(BOL; C; E; F; K; L; LD; M; MO; S; TCD; UC; US). 3419 (Ca-

ledon): Galgeberg, at neck of pass (-BA), Esterhuysen 34540 (B;

BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; PRE; RSA; S;

STE; TCD; UC; US; W; WAG).

Hypodiscus striatus (Kunth) Mast, in J. Linn.

Soc., Bot. 10 : 258 (1868); in A. DC., Monogr. Phan.

1 : 385 (1878); in FI. Cap. 7 : 139 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 408 (1928). Syntypes:

Cape, 3323 (Willowmore): Wagenbooms Rivier

(-CD), Drege 2022 $ (B, lecto.!; OXF!; MEL!; P!);

2022 cf (B!; K!; MEL!; MO!; OXF!; P!). 3218 (Clan-

william): at Groenrivier and Watervals Rivier

(-DB), Drege 2488 $ (P!; S!). 3019 (Loeriesfon-

tein): Ezelsfontein, Kamiesberge, 1 200 - 1 500 m

(-CD), Drege 2479 cf (K!; MEL!; MO!; OXF!; P!;

S!); 2480 $ (B!; K!; MO!; NY!; OXF!; P!).

Boeckhia striata Kunth, Enum. Pl. 3 : 449 (1841).

Bothalia 15, 3 & 4 (1985)

493

Hypodiscus gracilis Mast, in Bot. Jb. 29 Beibl. 66 : 16 (1900).

Syntypes: Cape, Ylandsrivier, March, Zeyher 105 c f (B, Iecto.!;

Z!); Kromrivier, Drege s.n. (?).

Notes

1. In B there is a sheet of Zeyher 105 labelled ‘Hy-

podiscus gracilis’. The locality ‘Ylandsrivier’ could

well be the Elandsrivier near Uitenhage.

2. H. striatus occurs in dry to arid fynbos, grassy

fynbos, or rarely renosterveld. The distribution

range is from Port Elizabeth to the Kamiesberg in

Namaqualand. It is absent from the more mesic

coastal mountains and the Cape Peninsula. The alti-

tude range is from 100 to 1 500 m.

Hypodiscus sulcatus Pillans in Trans. R. Soc. S.

Afr. 29 : 354 (1942). Type: Cape, 3319 (Worcester);

Gouronna (-CB), Esterhuysen 3726 $ (BOL,

lecto.!; K!); 3726 cf (BOL!;K!).

Note

1. H. sulcatus is an arid fynbos species, known

from the arid inland margins of the Cape sandstone

mountains in the Worcester and Ceres divisions.

Hypodiscus synchroolepis (Steud.) Mast, in J.

Linn. Soc., Bot. 8 : 255 (1865); in A. DC., Monogr.

Phan. 1 : 387 (1878); in FI. Cap. 7 : 139 (1897); Pil-

lans in Trans. R. Soc. S. Afr. 16 : 405 (1928). Syn-

types: Cape, 3325 (Port Elizabeth): Van Stadens-

berg, 300 - 900 m (-CC), Zeyher 4334 9 (P, lecto.!;

BOL!; BR!; K!; MEL!; P!; Z!); 4334 cf (BOL!; BR!;

K!; MEL!; P!; Z!).

Restio synchroolepis Steud., Syn. PI. Glum. 2 : 250 (1855).

Icon: J. Linn. Soc., Bot. 10 t. 7 C (1868).

Notes

1. There is a sheet of the type collection in P from

‘herb. Steudel’, this is clearly the lectotype.

2. H. synchroolepis is restricted to the southern

Cape Province. It grows between 600 and 1 500 m in

the coastal mountains between Uitenhage (Cocks-

comb) and Garcias Pass (Riversdale), as well as in

the Swartberg.

Hypodiscus willdenowia (Nees) Mast, in J. Linn.

Soc., Bot. 10 : 259 (1868); in A. DC., Monogr. Phan.

1 : 389 (1878); in FI. Cap. 7 : 136 (1897); Pillans in

Trans. R. Soc. S. Afr. 16 : 409 (1928); in Adamson

& Salter, FI. Cape Penins. 156 (1950). Type: ‘Willde-

nowia striata Spr.’ in herb. Willd. (B, holo.!).

Lepidanthus willdenowia Nees in Linnaea 5 : 665 (1830).

Restio sulcatus Kunth, Enum. PI. 3 : 404 (1841), nom. illeg.,

superfluous name for L. willdenowia Nees.

Icon: A. DC., Monogr. Phan. 1 : t.4, f.21-29; t.5, f. 19 (1878).

Notes

1. According to Nees (1830) the type is in ‘herb,

cl. Guntheri' and is labelled ‘Willdenowia striata

Spr'. In herb. Willdenow is a sheet which fits this

perfectly and is most likely the same sheet. It can be

regarded as the holotype. There is also a sheet in the

general herbarium, collected by Bergius from Berg-

vliet, labelled ‘Lepidanthus willdenowia N. ab E.

Schlechtend. Lin. V p. 665’, but this might be a later

annotation.

2. H. willdenowia is almost always recorded from

sandy, often seasonally wet habitats. The distribu-

tion range is from Grahamstown to Ceres. In the

southern Cape it is almost restricted to the coastal

plateau, but in the western Cape it reaches altitudes

of 1 200 m.

WILLDENOWIA

19. Willdenowia Thunb. in K. svenska

VetenskAkad. Handl. 11: 28 (1790); Linder in Bo-

thalia 15: 67 (1984).

Nematanthus Nees in Linnaea 5: 661 (1830).

KEY TO THE SPECIES OF WILLDENOWIA

la Perianth sessile:

2a Culms branching:

3a Culms striate W. incurvata

3b Culms smooth:

4a Tepals 6, 2-2, 5 mm long; styles oblong W. affinis

4b Tepals 3, 0,5 mm long; styles filiform W. purpurea

2b Culms usually simple:

5a Nut smooth or pitted; culms stout, at least 2 mm in

diam W. stokoei

5b Nut ornamented, not smooth or pitted; culms

slender, about 1 mm in diam.:

6a Plants rhizomatous; nut about 8 mm long ..W. humilis

6b Plants not rhizomatous; nut about 5 mm long

W. rugosa

lb Perianth stipitate:

7a Culms striate or sulcate:

8a Perianth stipe lobed; perianth 1,5 mm long ....W. bolusii

8b Perianth stipe entire, terete; perianth 3-4 mm

long W. sulcata

7b Culms smooth:

9a Tepals not touching at the margins W. glomerata

9b Tepals imbricate:

10a Tepals narrowed towards the base W. arescens

10b Tepals widened towards the base W. teres

Willdenowia affinis Pillans in Trans. R. Soc. S.

Afr. 16: 425 (1928); in Adamson & Salter, FI. Cape

Penins. 158 (1950). Type: Cape, 3318 (Cape Town):

north slopes of Table Mountain (-CD), Pillans 2940

9 (BOL, lecto.!; K!); 2940 cf (BOL!; K!).

Note

1. This species is only known from the type collec-

tion from Table Mountain.

Willdenowia arescens Kunth , Enum. PI. 3: 454

(1841); Mast, in A. DC., Monogr. Phan. 1: 393

(1878); in FI. Cap. 7: 145 (1897); Pillans in Trans. R.

Soc. S. Afr. 16: 425 (1928). Type: Cape, 3219 (Wup-

pertal): Cedarberg, 600-750 m (-AA/AC), Drege

2522 9 (B, lecto.!; BOL!; K!; MO!; NY!; OXF!;

P!); 2522 cf (B!; BOL!; K!; MO!; NY!; OXF!; P!).

Willdenowia brevis Nees ex Mast, in A. DC., Monogr. Phan. 1:

397 (1878); in FI. Cap. 7: 146 (1897). Type: Cape, 3318 (Cape

Town): banks of the Mosselbanks Rivier (-DB), Zeyher s.n. cf

(K, holo.!).

Notes

1. W. arescens is closely related to W. teres and W.

glomerata , with which it shares the smooth branch-

ing culms and pedicellate perianth, but it is readily

distinguished by the spathulate tepals.

2. Masters mentioned W. brevis as a nom. nud.

under W. teres in 1868 and in 1878 and 1897 listed it

with a description, with a note: ‘Perhaps a form of

W. teres’.

494

Bothalia 15, 3 & 4 (1985)

3. W. arescens occurs widespread in the western

Cape, from the Kamiesberg in Namaqualand to the

Hex River Valley and the Cape Flats. It grows in

sandy or rocky soils, with an altitude range from 50

m on the Malmesbury-Cape Flats coastal flats to

1 500 m in the inland mountains.

Willdenowia bolusii Pillans in Trans. R. Soc. S.

Afr. 16: 429 (1928). Syntypes: Cape, 3321 (Ladi-

smith): Phisantefontein, north slopes of the Lange-

berg (-CC), Muir 3194 $ (BOL, lecto.!; K l);3193 cf

(BOL!; K!).

Note

1. This species occurs in arid fynbos on sandstone-

derived soils, on the northern slopes of the Lange-

berg near Riversdale, in the Klein Swartberg and in

the Rooiberg. The altitude range is between 200 and

1 200 m.

Willdenowia glomerata (Thunb.) Linder , comb.

nov.

Restio glomeratus Thunb., Diss. Restio 18 (1788); Thunb., FI.

Cap. edn 1, 334 (1811); edn Schultes, 88 (1823); Kunth. Enum.

PI. 3: 414 (1841); Mast, in FI. Cap. 7: 98 (1897). Type: Cape,

without precise locality, in herb. Thunberg. 23236 cf (UPS,

holo.l).

Willdenowia lucaeana Kunth, Enum. PI. 3; 455 (1841); Mast, in

A. DC., Monogr. Phan. 1: 392 (1878); in FI. Cap. 7; 146 (1897);

Pillans in Trans. R. Soc. S. Afr. 16: 423 (1928); in Adamson &

Salter, FI. Cape Penins. 157 (1950). Type: Cape, 3219 (Wupper-

tal): Wuppertal, 600-900 m (-AC), Drege 2515a 9 (B, holo.l;

BOL!; Kl; OXF1; PI).

leones: J. Linn. Soc.. Bot. 10: t.8 D (1868). A. DC., Monogr.

Phan. 1 : t. 4 f.21-29; t. 5 f.19 (1878).

Notes

1. Thunberg 23236 is a monstrous specimen and as

such until 1978 the name based on it, Restio glomera-

tus Thunb., could be rejected (see Pillans, 1928:

425). However, with the change of the rules at the

Leningrad Conference, the name now has to be ap-

plied.

2. This species is closely related to W. arescens and

W. teres, but is distinct by its small, oblong, tepals

that do not touch at the margins.

3. W. glomerata is almost ubiquitous in the moun-

tains of the Cape Floristic Region, but has only

rarely been recorded from the southern Cape. Most

collections are from dry, stony slopes, with an altitu-

dinal range of 50 - 1 500 m.

Willdenowia humilis Nees ex Mast, in J. Linn.

Soc., Bot. 10: 272 (1868); in A. DC., Monogr. Phan.

1: 396 (1878); in FI. Cap. 7: 144 (1897); Pillans in

Trans. R. Soc. S. Afr. 16: 431 (1928); in Adamson &

Salter, FI. Cape Penins. 158 (1950). Type: Cape,

3318 (Cape Town): Doornhoogte, May (-DC), Eck-

lon 867 Cf (B, lecto.!; K!; MEL!; MO!; NY!; P!; S!;

Z!).

Hypodiscus dodii Mast, in J. Bot., Lond. 1901: 402 (1901).

Type: Cape, 3418 (Simonstown): Steenberg Plateau (-AB), Dod

2720 9 (K, holo.l).

Notes

1. Although Ecklon 867 in B is annotated in

Nees’s hand, there are no annotations by Masters, so

it is best to regard it as a lectotype, rather than as a

holotype. The sheet in K was only received in 1867

and has no Nees or Masters annotations.

2. W. humilis occurs in moist sand from the Cale-

don coast to the Cedarberg, from sea level to about

1 200 m.

Willdenowia incurvata (Thunb.) Linder, comb.

nov.

Restio incurvatus Thunb., Diss. Restio 18 (1788); Thunb., FI.

Cap. edn 1, 334 (1811); edn Schultes, 88 (1823). Calopsis incur-

vata (Thunb.) Kunth. Enum. PI. 3; 427 (1841), nom. illeg. , su-

perfluous. Leptocarpus incurvatus (Thunb.) Mast, in J. Linn.

Soc., Bot. 10: 223 (1868); A. DC., Monogr. Phan. 1: 332 (1878);

in FI. Cap. 7: 118 (1897), nom. illeg., superfluous name for R.

vimineus Rottb. Type; Cape, without precise locality, in herb.

Thunberg. 23238 cf (UPS, holo.l).

Willdenowia striata Thunb. in K. svenska VetenskAkad.

Handl. 11: 29 (1790); Thunb., FI. Cap. edn 1, 312 (1811); edn

Schultes, 82 (1823); Kunth, Enum. PI. 3: 453 (1841); Mast, in A.

DC., Monogr. Phan. 1: 394 (1878); in FI. Cap. 7: 144 (1897); Pil-

lans in Trans. R. Soc. S. Afr. 16: 428 (1928); in Adamson &

Salter. FI. Cape Penins. 158 (1950). Type: Cape, without precise

locality, in herb. Thunberg. 23219 9 (UPS, holo.l; SI).

? Nematanthus ecklonii Nees in Linnaea 5: 662 (1830). Willde-

nowia ecklonii (Nees) Kunth, Enum. PI. 3: 456 (1841). Type:

Cape. Plattekloof and Cape Flats, Ecklon s.n. (?).

? Willdenowia neglecta Steud., Syn. PI. Glum. 2: 263 (1855).

Type: Cape, without precise locality, Drege 2515b (?).

Willdenowia cuspidata Mast, in J. Linn. Soc., Bot. 10: 271

(1868); in A. DC., Monogr. Phan. 1: 396 (1878); in FI. Cap. 7:

144 (1897). Type: Cape, 3219 (Wuppertal): near Groen Rivier

and Waterfalls Rivier, 450-600 m (-?AA), Drege 2516 9 (K,

holo.l; Bl; MO!; OXF1; PI).

leones: K. svenska VetenskAkad. Handl. lit. 2 (1790). Mason,

W., Cape Sandveld Flow, t.2 f.8 (1972).

Notes

1. As in W. glomerata, the type of R. incurvatus is

monstrous and consequently Pillans (1928) did not

uphold the name.

2. There is a sheet in B labelled 'Nematanthus eck-

lonii' in Nees’s hand, but it is collected from Brak-

fontein in the Olifantsriver Valley. Nees also la-

belled it 'Willdenowia striata’ a name which he does

not mention in Linnaea 5 (1830).

3. The sheet of Drege 2516 at B is clearly not the

same collection as that at K and is W. arescens.

4. Calopsis or Leptocarpus incurvatus has been

misapplied by Kunth and Masters to Calopsis vimi-

nea.

5. W. incurvata occurs widespread in the western

Cape, from False Bay to Springbok in Namaqua-

land. It is most common and often dominant, on the

sandy coastal forelands (‘sandveld’) between Melk-

bos and the Olifants River, but also occurs occasion-

ally on sandy flats and the lower slopes of mountains

between Clanwilliam, the Hex River Valley and the

Cape Peninsula. The altitudinal range is from sea le-

vel to about 1 200 m.

Willdenowia purpurea Pillans in Trans. R. Soc.

S. Afr. 16: 430 (1928). Type: Cape, 3419 (Caledon):

Viljoenspas (-AA), Pillans 4811 $ (BOL, holo.!;

K!).

Notes

1. This species is superficially very similar to Ele-

gia prominens, but the anatomy identifies it clearly

as a Willdenowia.

Bothalia 15, 3 & 4 (1985)

495

2. W. purpurea occurs in sandy marshy areas along

Viljoens Pass and along the eastern base of the

Fransch Hoek Mountains.

Willdenowia rugosa Esterhuysen, sp. nov., a W.

humile Mast, culmis tuberculatis vel rugosis, gracili-

bus (0,5-1 mm diametro), rhizomatibus aeriis, inflo-

rescentiis omnibus partibus minoribus differt.

TYPE. — Cape, 3418 (Simonstown): on the

plateau at the base of Buffelstal Peak, above Stal-

berg (-BD), Esterhuysen 34606 C f (BOL, holo.!; B;

C; E; F; GRA; K; L; LD; M; MO; NBG; NY; PRE;

RSA; S; STE; TCD; UC; US; W; WAG).

Plants caespitose, 25-35 cm tall. Rhizomes aerial,

ascending, 1-2 mm in diameter, scales loosely con-

voluted, chartaceous with a membranous margin,

acute, piliferous, about 5 mm long. Culms solid, te-

rete, simple, up to 1 mm in diameter, surface irregu-

larly tuberculate to rugulose. Sheaths occasionally

present, closely convoluted, to 2 cm long, body co-

riaceous, rounded or obtuse, piliferous, with a hya-

line apex. Male inflorescence up to 4 cm long.

Spathes caducous, like the sheaths. Inflorescence

branches solitary per node, about 7 mm long. Flow-

ers not organized into spikelets, shortly pedicellate,

numerous, racemose. Bracts chartaceous-membra-

nous, linear, 5-7 x 0,3 mm. Tepals 3, linear, similar

to the bracts but 3,5 x 0,1 mm. Anthers the most

visible part of the perianth, 1-1,5 mm long, emucro-

nate. Female inflorescence more or less lateral, usu-

ally at 2 nodes, with the culm continuing its growth

some distance above the apical spikelet. Spathes

about 12-25 mm long, coriaceous, acute, pale

brown, margins membranous, especially at the apex,

awn piliferous, up to 5 mm long. Spikelet solitary,

enclosed within the spathe. Bracts 4, chartaceous,

7-10 x 1-2,5 mm, the outer two bracts much more

slender than the inner two bracts. Tepals charta-

ceous, acute to truncate, equal, imbricate, 5x2

mm. Styles 2, free to the base, to 15 mm long, ex-

serted above the spathe, the upper § somewhat vil-

lous. Fruit a cylindrical, 5,5-6 x 1,8 mm nutlet, ir-

regularly grooved to fluted in the apical 3, truncate.

Fig. 26.

W. rugosa occurs in the mountains between Sir

Lowry's Pass and Houw Hoek Pass. It has been col-

lected from sandy or somewhat marshy areas be-

tween 300 and 600 m. It probably flowers in May.

This species is closely related to the more wide-

spread W. humilis, with which it shares the highly

reduced male flowers, the cylindrical fluted nutlet

and the female perianth. It differs from W. humilis

in the more slender, tuberculate culms (less than 1

mm in diameter), the aerial rhizomes, the almost

complete absence of sheaths and in the male and fe-

male inflorescences being smaller in all parts.

CAPE. — 3418 (Simonstown): on the plateau at the base of

Buffelstal Peak, above Stalberg (-BD), Esterhuysen 34606 (B;

BOL; C; E; F; GRA; K; L; LD; M; MO; NBG; NY; PRE; RSA;

S; STE; TCD; UC; US; W; WAG). 3419 (Caledon): Highlands

For. Res., Elgin (-AA), Esterhuysen 35890 (B; BOL; C; E; F;

GRA; K; L; LD; M; MO; NBG;' NY; PRE; RSA; S; STE; UC;

US; W; WAG).

Willdenowia stokoei Pillans in Trans. R. Soc. S.

Afr. 29 : 355 (1942). Syntypes; Cape, 3319 (Worces-

ter): Ceres, summit of the Roodeberg, 2 100 m

(-BC), Esterhuysen 1489 $ (BOL, lecto.!; K!); 1490

Cf (BOL!; K!).

Notes

1 . Some plants have occasionally branching culms.

2. This is a montane species that occurs above

1 200 m in the mountains from the Cedarberg to the

Klein Swartberg. In general, it is restricted to arid

mountains, where it occurs in somewhat moist habi-

tats on shale bands and rocky slopes.

Willdenowia sulcata Mast, in J. Linn. Soc., Bot

10 : 270 (1868); in A. DC., Monogr. Phan. 1 : 395

(1878); in FI. Cap. 7 : 145 (1897); Pillans in Trans.

R. Soc. S. Afr. 16 : 427 (1928); in Adamson &

Salter, FI. Cape Penins. 158 (1950). Syntypes: Cape,

3318 (Cape Town): Cape Flats (-DC), Ecklon 930 cf

(B, lecto.!); Ecklon s.n. in herb. Sonder 9 (B!).

Notes

1. The collection at B does not have a number, but

it is from Doornhoogte and so is probably isotype

material, which can be lectotypified. The collection

is annotated by Nees as ‘Willdenowia striata var.

microstachyus’ .

2. There is one collection, geographically isolated

from the rest of the species, from the Swartberg east

of the Seven Weeks Poort, which differs marginally

from the species by having a much branched inflo-

rescence and slightly larger tepals.

3. W. sulcata occurs in the western Cape, from the

Cedarberg to Caledon and the Cape Peninsula.

Plants occur in either sandy or rocky soils, usually in

well-drained habitats. The altitudinal range is from

near sea level on the Malmesbury Flats to 1 800 m in

the Cedarberg.

Willdenowia teres Thunb. in K. svenska

VetenskAkad. Handl. 11 : 30 (1790); Thunb., FI.

Cap. edn 1, 314 (1811); edn Schultes, 82 (1823);

Kunth, Enum. PI. 3 : 452 (1841); Mast, in A. DC.,

Monogr. Phan. 1 : 392 (1878); in FI.. Cap. 7 : 146

(1897); Pillans in Trans. R. Soc. S. Afr. 16 : 426

(1928); in Adamson & Salter, FI. Cape Penins. 158

(1950). Type: Cape, without precise locality, in

herb. Thunberg. 23220 9 (UPS, holo.!; S!).

Willdenowia fraterna N.E.Br. in FI. Cap. 7 : 756 (1900). Type:

Cape, 3423 (Knysna): hillside near The Glebe (-AA), Galpin

4830 9 (K, holo.!; BOL!).

Willdenowia galpinii N.E.Br. in FI. Cap. 7 : 757 (1900). Syn-

types: Cape, 3321 (Ladismith): mountains at Garcias Pass, 300 m

(-CC), Galpin 4824 9 (K, lecto.!; BOL!); 4825 cf (BOL!; K!);

4831 Cf (BOL!; K!).

Willdenowia peninsularis N.E.Br. in FI. Cap. 7 : 757 (1900).

Syntypes: Cape, 3418 (Simonstown): Cape Peninsula, hills near

Kommetjie (-AB), Galpin 4822 9 (K, lecto.!; BOL!); 4832 C f

(BOL!; K!).

Icon: K. svenska VetenskAkad. Handl. 11 t.3 (1790).

Notes

1 . W. teres is closely related to W. arescens and W.

glomerata, but is easily distinguished by the imbri-

cate, suborbicular tepals.

2. W. teres is widespread in the Cape Floristic Re-

gion, occurring in the mountains and on the sandy

Bothalia 15, 3 & 4 (1985)

496

FIG. 26. - Willdenoma rugosa Esterhuysen. a, habit; b, male plant, showing ascei ndl"g“"a‘^ d male flower,

detail of male inflorescences, showing branches with lax^ arranged fema[e ’spike let, x 4; g,

note linear bract and hair-like tepals, x 9; e, female mflorcscence with 1-J later 4 styl« x 6. (From Ester-

female bract, x 5; h, female flower with mature, ornamented nut, membranous tepals a y

huysen 34606.)

Bothalia 15, 3 & 4 (1985)

497

coastal forelands from Uitenhage to the Kamiesberg

in Namaqualand. It has only rarely been recorded

from the mountains north of the Hex River Valley.

The altitudinal range is from sea level to 1 500 m.

EXCLUDED NAMES

Restio articulatus Retz. = Lepironia articulata

(Retz.) Domin. (Cy-

peraceae)

Restio squamosus Thunb. = Thesium sp. (Santala-

ceae)

Restio simplex Thunb. = An Australian species

of Restio.

NOMINA NON SATIS COGNITAE

Thamnochortus mastersii Gand.

ACKNOWLEDGEMENTS

Without the very rich collections and extensive

knowledge of Miss E. E. Esterhuysen this Conspec-

tus would not have been possible. I am also grateful

for the advice on nomenclatural matters, so freely

given by Dr R. K. Brummitt of Kew, for comments

by numerous colleagues, and to the numerous her-

baria and universities that have provided working fa-

cilities, or loaned their material for study. The work

was done at the Royal Botanic Gardens, Kew and at

the Bolus Herbarium, University of Cape Town.

UITTREKSEL

'n Oorsig van die spesies van Restionaceae van

Afrika word gegee. 318 spesies, volgens Linder se

generiese klassifikasie van 1984 gerangskik, word er-

ken. Vir elke spesie word gedetailleerde notas oor die

nomenklatuur en die tipifikasie, en kort notas oor die

taksonomie en verspreiding gegee. Sleutels tot die

genera en spesies word voorsien. V yf-en-vyftig nuwe

spesiename word gepubliseer, wat 51 nuwe spesies

verteenwoordig (Anthochortus capensis Esterhuy-

sen, Askidiosperma rugosum Esterhuysen, Calopsis

adpressa Esterhuysen, C. clandestina Esterhuysen.

C. dura Esterhuysen, C. pulchra Esterhuysen, Chon-

dropetalum decipiens Esterhuysen, Elegia atratiflora

Esterhuysen, E. caespitosa Esterhuysen, E. fucata

Esterhuysen, E. grandispicata Linder, Hypodiscus

montanus Esterhuysen, H. procurrens Esterhuysen,

H. squamosus Esterhuysen, Ischyrolepis affinis Es-

terhuysen, I. caespitosa Esterhuysen, I. curvibrac-

teata Esterhuysen, I. feminea Esterhuysen, I. ka-

rooica Esterhuysen, I. longiaristata Pillans ex

Linder, I. nana Esterhuysen, I. nubigena Esterhuy-

sen, I. papillosa Esterhuysen, I. pratensis Esterhuy-

sen, I. rivula Esterhuysen, I. sporadica Esterhuysen,

I. unispicata Linder, I. wittebergensis Esterhuysen,

Nevillea singularis Esterhuysen, Platycaulos acutus

Esterhuysen, Restio colliculospermus Linder, R.

corneolus Esterhuysen, R. fragilis Esterhuysen, R.

implicatus Esterhuysen, R. inconspicuus Esterhuy-

sen, R. ingens Esterhuysen, R. inveteratus Esterhuy-

sen, R. montanus Esterhuysen, R. nuwebergensis Es-

terhuysen, R. peculiaris Esterhuysen, R. perseverans

Esterhuysen, R. pillansii Linder, R. pulvinatus Es-

terhuysen, R. pumilus Esterhuysen, R. rarus Ester-

huysen, R. rupicola Esterhuysen, R. singularis Ester-

huysen, R. vallis-simius Linder, R. verrucosus Ester-

huysen, R. versatilis Linder, R. zuluensis Linder,

Staberoha ornata Esterhuysen, Thamnochortus cine-

reus Linder, T. arenarius Esterhuysen, T. rigidus Es-

terhuysen, Willdenowia rugosa Esterhuysen). Een

nuwe subspesie (Askidiosperma chartaceum (Pil-

lans, Linder subsp. alticolum Esterhuysen) word be-

skryf, en 84 nuwe kombinasies word gemaak. Veer-

tie n van die nuwe spesies word geillustreer. Uitge-

breide notas oor die taksonomiese geskiedenis van die

Restionaceae van Afrika, veral wat die tipifikasie van

die name betref, word gegee.

REFERENCES

BERGIUS, P. J., 1767. Descriptiones Plantarum ex Capite Bonae

Spei. Stockholm: Laur. Salvii.

BREYNE, J., 1678. Exoticarum aliarumque minus cognitarum

plantarum centuriae primae. Danzig.

BROWN, N.E., 1900. In ‘Addenda and Corrigenda’ to W. T.

Thiselton-Dyer, FI. Cap. 7. London: Lovell Reeve.

BROWN, R., 1810. Prodromus florae Novae Hollandiae. Lon-

don: Johnson.

BURMAN, N. L., 1768. Florae Capensis Prodromus, in Flora

Indica. Amsterdam: Schreuder.

DESV AUX, N. A. , 1828. Observationes sur quelque families des

plantes monocotyledones, d’apres les manuscripts de feu le

Baron Paliscot de Beauvois. Annls Sci. nat. ser. 1, 13 :

37-52.

DREGE, J. F., 1841. Zwei Pflanzengeographische Documente.

Beigabe zur Flora, Jena, 1843, Band 1.

DREGE, J. F., 1847. Standorter-Verzeichniss der von C. L.

Zeyher in Siidafrika gesammelten Pflanzen. Linnaea 19 :

583-598.

GANDOGER. M., 1913. L’Herbier africain de Sonder. Bull.

Soc. bot. Fr. 60 : 414-422.

GOLDBLATT, P., 1978. An analysis of the flora of southern

Africa: its characteristics, relationships, and origins. Ann.

Mo. bot. Gdn 65 : 369-436.

GUNN, M. & CODD, L. E., 1981. Botanical exploration of

Southern Africa. Cape Town: Balkema.

HOOKER, J. D., 1865. Report on the progress and condition of

the Royal Gardens at Kew. London : Eyre & Spottiswoode .

HOOKER, J. D., 1867. Report on the progress and condition of

the Royal Gardens at Kew. London: Eyre & Spottiswoode.

HOPPE, D. H., 1825. Botanische Notizen. Flora, Jena 8, 1 :

269-271.

KRAUSS, F., 1845. Pflanzen des Cap- und Natal-Landes, gesam-

melt und zusammengestellt von Dr Ferdinand Krauss.

Flora, Jena 28 : 337-344.

KUNTH, C. S., 1841. Enumeratio Plantarum, Vol. 3. Stuttgart:

Cotta.

LINDER, H. P., 1984. A phylogenetic classification of the genera

of the African Restionaceae. Bothalia 15 : 11-76.

MASTERS, M. T., 1865. Observations on the morphology and

anatomy of the genus Restio, Linn., together with an enu-

meration of the South African species. J. Linn. Soc., Bot.

8 : 211-255.

MASTERS, M. T., 1868. Synopsis of the South African Restia-

ceae. J. Linn. Soc., Bot. 10 : 209-279.

MASTERS, M. T., 1874. On the Restiaceae of Thunberg’s Her-

barium. J. Linn. Soc., Bot. 14 : 413^121.

MASTERS, M. T., 1878. Restiaceae. In A. L. P. P. de Candolle,

Monographiae Phanerogamarum 1 : 218-398. Paris: Mas-

son.

MASTERS, M. T., 1885. Supplementary notes on Restiaceae. J.

Linn. Soc., Bot. 21 : 574-594.

MASTERS, M. T., 1897. Restiaceae. In W. T. Thiselton-Dyer,

FI. Cap. 1 : 59-149. London: Reeve.

MASTERS, M. T., 1900. Restionaceae novae capenses herbarii

Berolinensis, imprimis Schlechterianae. Bot. Jb. 29 Beibl.

66 : 1-20.

NEES AB ESENBECK, C. G., 1830. Beitrag zur Kenntnis der

Familien der Restiaceen in Riicksicht auf Gattungen und

Arten. Linnaea 5 : 627-666.

NEES AB ESENBECK, C. G., 1836. In J. Lindley, A natural

system of Botany, edn 2. London: Longmans.

NORDENSTAM, B., 1980. Nomenclatural notes on South Afri-

can Cruciferae. Jl S. Afr. Bot. 46 : 251-263.

PILLANS, N. S., 1921. In H.M.L. Bolus, Novitates Africanae.

Ann. Bolus Herb. 3 : 80-85.

498

Bothalia 15, 3 & 4 (1985)

PILLANS, N. S., 1922. In H.M.L. Bolus, Novitates Africanae.

Ann. Bolus Herb. 3 : 123-146.

PILLANS, N. S., 1928. The African genera and species of Restio-

naceae. Trans. R. Soc. S. Afr. 16 : 207-440.

PILLANS, N. S., 1942. New species of South African Restiona-

ceae. Trans. R. Soc. S. Afr. 29 : 339-356.

PILLANS, N. S., 1945. New and hitherto imperfectly known

species of African Restionaceae. Trans. R. Soc. S. Afr. 30 :

245-266.

PILLANS, N. S., 1952. New and hirherto imperfectly known

species of South African Restionaceae. Jl S. Afr. Bot. 18 :

101-122.

PLUKENET, L., 1700. Almagesti Botanici Mantissa. London.

ROTTBOELL, C. F., 1772. Descriptiones Plantarum Rariorum

iconibus illustrandas programmate indicit. Copenhagen.

ROTTBOELL. C. F., 1773. Descriptionum et Iconum Rariores.

Copenhagen.

SCHELPE, E.A.C.L.E., 1967. The identity of Equisetum ca-

pense Burm. f. Jl S. Afr. Bot. 33 : 155-156.

SCHEUCHZER, J., 1775. Agrostographia. Zurich: Orell,

Gessner, Fuesli.

SPRENGEL,K., 1824. Systema Vegetabilium, Vol. 1. Gottingen.

STEUDEL, E. G.. 1829. Einige Bemerkungen uber Kap’schen

Restiaceen, Cyperaceen und Gramineen. Flora, Jena 12 :

129-140.

STEUDEL, E. G., 1850. Ueber den gegenwartigen Stand der

Synopsis plantarum und eine neue Gattung der Graser.

Flora, Jena 33 : 225-232.

STEUDEL, E. G., 1855. Synopsis Plantarum Glumacearum,

Vol. 2. Stuttgart: Metzler.

TAYLOR, H. C., 1978. Capensis. In M. J. A. Werger, Biogeo-

graphy and ecology of southern Africa. The Hague: Junk.

THLINBERG, C. P., 1788. Dissertatione de Restione. Gottingen:

Dieterich.

THUNBERG, C. P.. 1790. Beskrifning pS Wildenowia, et syn-

nerligt och nytt GrSs-Slag. K. svenska VetenskAkad.

Handl. 11 : 26-32.

THUNBERG, C. P., 1794. Prodromus Plantarum Capensium.

Upsala: Edman.

THUNBERG, C. P., 1803. Novae species plantarum Capensium

examinatae et descriptae. Phytogr. Bl. 1 : 1-34.

THUNBERG, C. P., 1811. Flora Capensis. Upsala: Edman.

THUNBERG, C. P., 1823. Flora Capensis, edn Schultes.

Stuttgart: Cotta.

WILLIAMS, I. J. M., 1972. A revision of the genus Leucaden-

dron (Proteaceae). Contrib. Bolus Herb. 3 : 1-425.

WOLLEY-DOD. A. H., 1901. New plants from the Cape Penin-

sula. J. Bot., Lond. 39 : 397^102.

ALPHABETICAL INDEX TO THE NAMES APPLIED TO THE AFRICAN RESTIONACEAE

Page

Anthochortus Nees 484

capensis Esterhuysen 484

crinalis (Mast.) Linder 486

ecklonii Nees 486

graminifolius (Kunth) Linder 486

insignis (Mast.) Linder 487

laxiflorus (Nees) Linder 487

Askidiosperma Steud 431

albo-aristatum (Pillans) Linder 431

andreaeanum (Pillans) Linder 431

capitatum Steud 431

chartaceum (Pillans) Linder 431

chartaceum (Pillans) Linder subsp. alticolum Esterhuysen 432

esterhuyseniae (Pillans) Linder 432

insigne (Pillans) Linder 432

longiflorum (Pillans) Linder 432

nitidum (Mast.) Linder 432

paniculatum (Mast.) Linder 432

rugosum Esterhuysen 432

Boeckhia Kunth 488

laevigata Kunth 489

striata Kunth 492

Calopsis Desv 464

adpressa Esterhuysen 465

andreaeana (Pillans) Linder 465

aspera (Mast.) Linder 465

burchellii (Mast.) Linder 465

clandestina Esterhuysen 465

dura Esterhuysen 466

esterhuyseniae (Pillans) Linder 466

festucacea Kunth 470

filiformis (Mast.) Linder 467

fruticosa (Mast.) Linder 467

gracilis (Mast.) Linder 467

hyalina (Mast.) Linder 467

hirtella Kunth 470

impolita (Kunth) Linder 467

incurvata Pillans 470

incurvata (Thunb.) Kunth 494

levynsiae (Pillans) Linder 467

marlothii (Pillans) Linder 467

membranacea (Pillans) Linder 469

monostylis (Pillans) Linder 469

muirii (Pillans) Linder 469

neglecta Hochst 453

nudiflora (Pillans) Linder 469

oxylepis Kunth 470

peronata Kunth 470

paniculata (Rottb.) Desv 469

pulchra Esterhuysen 469

ramiflora (Nees) Kunth 469

rigida (Mast.) Linder 470

rigorata (Mast.) Linder 470

triticea (Rottb.) Kunth 462

viminea (Rottb.) Linder 470

Calorophus Labill.

anceps (Mast.) Kuntze 436

asper Kuntze 465

burchellii (Mast.) Kuntze 469

digitatus (Thunb.) Kuntze 487

filiformis (Mast.) Kuntze 467

gracilis (Mast.) Kuntze 467

laxiflorus (Nees) Kuntze 487

tenuis (Mast.) Kuntze 486

virgatus (Mast.) Kuntze 444

Cannomois Desv 480

acuminata (Kunth) Pillans 481

aristata Mast 480

cephalotes Desv 482

complanatus Mast 481

congesta Mast 480

dregei Pillans 482

nitida (Mast.) Pillans 481

parviflora (Thunb.) Pillans 481

schlechteri Mast 481

scirpoides (Kunth) Mast 482

scirpoides Mast. var. minor Pillans 481

scirpoides Mast. var. primosii Pillans 481

simplex Kunth 481

spicatus Mast 481

virgata (Rottb.) Steud 482

Ceratocaryum Nees 479

argenteum Nees ex Kunth 479

decipiens (N.E.Br.) Linder 479

fimbriatum (Kunth) Linder 479

fistulosum Mast 480

xerophilum ( Pillans ) Linder 48O

Chondropetalum Rottb 427

acockii Pillans 428

aggregatum (Mast.) Pillans 428

albo-aristatum Pillans 431

andreaeanum Pillans 431

capitatum (Steud.) Pillans 431

chartaceum (Pillans) Pillans 431

decipiens Esterhuysen 428

deustum Rottb 429

Bothalia 15, 3 & 4 (1985)

499

ebracteatum (Kunth) Pillans

esterhuyseniae Pillans

hookerianumfMasr.) Pillans

insigne Pillans

longiflorum Pillans

macrocarpum (Kunth) Pillans

marlothii (Pillans) Pillans

microcarpum (Kunth) Pillans

mucronatum (Nees) Pillans

nitidum (Mast.) Pillans

nudum Rottb

paniculatum (Mast.) Pillans

rectum (Mast.) Pillans

tectorum ( L.f '.) Rafin

Craspedolepis Steud

verreauxii Steud

Cucullifera Nees

dura Nees

Dovea Kunth

aggregata Mast

binata Steud

bolusii Mast

chartacea Pillans

cylindrostachya Mast

ebracteata Kunth

hookeriana Mast

macrocarpa Kunth

marlothii Pillans

microcarpa Kunth

mucronata (Nees) Mast

nitida Mast

nuda (Rottb.) Pillans

paniculata Mast

racemosa (Poir.) Mast

recta Mast

rigens Mast

tectorum (L.f.) Mast

thyrsoidea Mast

Elegia L

acuminata Mast

altigena Pillans

amoena Pillans

asperiflora (Nees) Kunth

asperiflora (Nees) Kunth var. lacerata (Pillans) Pillans ....

atratiflora Esterhuysen

bella Pillans

caespitosa Esterhuysen

capensis (Burnt, f.) Schelpe

ciliata Mast

coleura Nees ex Mast

cuspidata Mast

deusta (Rottb.) Kunth

dregeana Kunth

elongata Mast

equisetacea (Mast.) Mast

esterhuyseniae Pillans

esterhuyseniae Pillans var. dispar Pillans

exilis Mast

extensa Pillans

fastigiata Mast

fenestrata Pillans

filacea Mast

fistulosa Kunth

fistulosa Kunth var. parviflora Pillans

fucata Esterhuysen

fusca N.E.Br

galpinii N.E.Br

glauca Mast

gracilis N.E.Br

grandis (Nees) Kunth

grandispicata Linder %

hutchinsonii Pillans

intermedia. (Steud.) Pillans j

juncea L

juncea L. var. geniculata Pillans

lacerata Pillans

membranacea (Nees) Kunth

mucronata (Nees) Kunth

muirii Pillans

neesii Mast 425

nuda (Rottb.) Kunth 430

obtusiflora Mast 426

panicoides Kunth 430

paniculata Pers 482

parviflora (Thunb.) Kunth 481

parviflora Pillans 423

parviflora Pillans var. filacea (Mast.) Pillans 423

parviflora Pillans var. rigida (Mast.) Pillans 426

pectinata Pillans 426

persistens Mast 425

prominens Pillans 425

propinqua (Nees) Kunth 425

propinqua (Nees) Kunth var. equisetacea Mast 422

propinqua (Nees) Kunth var. minor Mast 425

racemosa (Poir.) Pers 425

rehmannii Mast 423

rigida Mast 426

spathacea Mast 426

spathacea Mast. var. attenuata Pillans 426

squamosa Mast 426

stipularis Mast 426

stokoei Pillans 427

thyrsifera (Rottb.) Pers 427

thyrsoidea (Mast.) Pillans 427

vaginulata Mast 427

verreauxii Mast 427

verticillaris (L.f.) Kunth 421

Equisetum

capense Burm. f 421

Hydrophilus Linder 484

rattrayi (Pillans) Linder 484

Hypodiscus Nees 488

albo-aristatus (Nees) Mast 488

albo-aristatus (Nees) Mast. var. oliverianus (Mast.) Pillans 488

alternans Pillans 489

argenteus (Thunb.) Mast 489

aristatus (Thunb.) Krauss ....: 489

aristatus (Thunb.) Krauss var. bicolor Mast 489

aristatus (Thunb.) Krauss var. protractus (Mast.) Pillans . 489

binatus (Steud.) Mast 489

capitatus Mast 488

dodii Mast 494

duplicatus Hochst 488

eximius Mast 490

gracilis Mast 493

laevigatus (Kunth) Linder 489

montanus Esterhuysen 489

neesii Mast 490

nitidus Nees ex Mast 481

oliverianus Mast 488

paludosus Pillans 492

parkeri Pillans 492

procurrens Esterhuysen 490

protractus Mast 489

rigidus Mast 492

rugosus Mast 492

squamosus Esterhuysen 492

striatus (Kunth) Mast 492

sulcatus Pillans 493

synchroolepis (Steud.) Mast 493

tristachyus Mast 492

willdenowia (Nees) Mast 493

zeyheri Mast 492

Hypolaena R.Br.

anceps Mast 436

aspera Mast 465

bachmannii Mast 462

browniana Mast 444

burchellii Mast 469

conspicua Mast 445

crinalis (Mast.) Pillans 486

decipiens N.E. Br 444

diffusa Mast 463

digitata (Thunb.) Pillans 487

eckloniana Nees ex Mast 487

filiformis Mast 467

foliosa Mast 486

gracilis Nees ex Mast 467

429

432

429

432

431

429

430

432

430

432

430

437

446

480

430

431

428

489

429

431

430

429

431

429

430

432

430

432

426

430

429

430

427

418

427

420

426

420

426

421

420

421

422

429

420

430

422

421

422

423

426

423

420

423

424

425

420

424

430

425

500

Bothalia 15, 3 & 4 (1985)

grarninifolia (Kunth) Pillans 486

hyalina Mast 467

impolita (Kunth) Mast 467

incerta Mast 487

laxiflora Nees 487

mahonii N.E.Br 452

membranacea Mast 445

purpurea Pillans 487

schlechteri Mast 454

spathulata Pillans 487

stokoei Pillans 487

subtilis Mast 444

tabularis Pillans 486

tenuis Mast 486

tenuissima Pillans 415

virgata Mast 444

Ischyrolepis Steud 397

affinis Esterhuysen 401

arida (Pillans) Linder 402

caespitosa Esterhuysen 402

capensis (L.) Linder 402

cincinnata (Mast.) Linder 403

coactilis (Mast.) Linder 403

curvibracteata Esterhuysen 403

curviramis (Kunth) Linder 404

distracta (Mast.) Linder 404

duthieae (Pillans) Linder 404

eleocharis (Mast.) Linder 404

esterhuyseniae (Pillans) Linder 404

feminea Esterhuysen 404

fraterna (Kunth) Linder 405

fuscidula (Pillans) Linder 405

gaudichaudiana (Kunth) Linder 405

gossypina (Mast.) Linder 405

helenae (Mast.) Linder 406

hystrix (Mast.) Linder 406

karooica Esterhuysen 406

laniger (Kunth) Linder 406

leptoclados (Mast.) Linder 407

longiaristata Pillans ex Linder 407

macer (Kunth) Linder 407

marlothii (Pillans) Linder 407

monanthos (Mast.) Linder 407

nana Esterhuysen 409

nubigena Esterhuysen 409

ocreata (Kunth) Linder 410

paludosa (Pillans) Linder 410

papillosa Esterhuysen 410

pratensis Esterhuysen 411

pygmaea (Pillans) Linder 413

rivula Esterhuysen 413

rottboellioides (Kunth) Linder 413

sabulosa ( Pillans ) Linder 413

schoenoides (Kunth) Linder 413

setiger (Kunth) Linder 413

sieberi (Kunth) Linder 414

sporadica Esterhuysen 414

subverticellata Steud 415

tenuissima (Kunth) Linder 415

triflora (Rottb.) Linder 415

unispicata Linder 415

vilis (Kunth) Linder 417

virgea (Mast.) Linder 417

wallichii (Mast.) Linder 417

wittebergensis Esterhuysen 417

Lamprocaulis Mast.

grandis (Nees) Mast 424

neesii (Mast.) Mast 425

schlechteri Gilg-Ben 425

Lepidanthus Nees 488

willdenowia Nees 493

Leptocarpus R. Br.

andreaeanus Pillans 465

asper (Mast.) Pillans 465

brachiatus Mast 442

burchellii Mast 465

cymosus Mast 444

distichus (Rottb.) Pillans 445

divaricatus Mast 462

ejuncidus (Mast.) Pillans 446

esterhuyseniae Pillans 466

festucaceus (Kunth) Mast 470

fruticosus Mast 467

gracilis (Mast.) Pillans 467

hyalinus (Mast.) Pillans 467

impolitus (Kunth) Pillans 467

incurvatus Pillans 470

incurvatus (Thunb.) Mast 494

intermedins Pillans 444

levynsiae Pillans 467

marlothii Pillans 467

membranaceus Pillans 469

modestus (Kunth) Mast 479

monostylis Pillans 469

muirii Pillans 469

neglectus (Hochst.) Mast 453

nudiflorus Pillans 469

oxylepis (Kunth) Mast 470

paniculatus (Rottb.) Mast 469

parkeri Pillans 446

peronatus (Kunth) Mast 470

peronatus (Kunth) Mast. var. hirtellus (Kunth) Mast 470

ramosissimus Pillans 467

rattrayi Pillans 484

rigidus Mast 470

rigoratus Mast 470

rigoratus Mast. var. sirnulans Pillans 470

secundus Pillans 460

stokoei Pillans 457

vimineus (Rottb.) Pillans 470

vimineus (Rottb.) Pillans var. hirtellus (Kunth) Pillans .... 470

Leucoploeus Nees 488

argenteus Nees ex Mast 489

Mastersiella Gilg-Ben 487

browniana (Mast.) Gilg-Ben 444

diffusa (Mast.) Gilg-Ben 463

digitata (Thunb.) Gilg-Ben 487

foliosa (Mast.) Gilg-Ben 486

hyalina (Mast.) Gilg-Ben 467

laxiflora (Nees) Gilg-Ben 487

purpurea (Pillans) Linder 487

spathulata (Pillans) Linder 487

Mesanthus Nees 480

macrocarpus Nees 481

Nematanthus Nees 493

ecklonii Nees 494

Nevillea Esterhuysen & Linder 482

obtusissima (Steud.) Linder 482

singularis Esterhuysen 484

Phyllocomos Mast 484

insignis Mast 487

Platycaulos Linder 434

acutus Esterhuysen 434

anceps (Mast.) Linder 436

callistachyus (Kunth) Linder 436

cascadensis (Pillans) Linder 436

compressus (Rottb.) Linder 436

depauperatus (Kunth) Linder 436

major (Mast.) Linder 436

subcompressus (Pillans) Linder 437

Restio Rottb 437

acuminatus Thunb 430

acuminatus Kunth 481

acockii Pillans 441

albo-aristatus Nees 488

alticola Pillans 441

ambiguus Mast 441

ameles Steud 415

anceps (Mast.) Pillans 436

araneosus Mast 415

arcuatus Mast 441

argenteus Thunb 489

aridus Pillans 402

aristatus Thunb 489

aspericaidis Pillans 436

asperiflorus Nees 420

Bothalia 15, 3 & 4 (1985)

501

aureolus Pillans

bifarius Mast

bifidus Thunb

bifurcus Nees ex Mast

bigeminus Nees ex Mast

bolusii Pillans

brachiatus (Mast.) Pillans

brownianus (Mast.) Pillans

brunneus Pillans

burchellii Pillans

callistachyus Kunth

capillaris Kunth

cascadensis Pillans

cernuus L.f

chondropetalum Nees

cincinnatus Mast

cirratus Mast

coactilis Mast

colliculospermus Linder

communis Pillans

comosus N.E. Br

compressus Rottb

compressus Rottb. var. major Mast

concolor Steud

confusus Pillans

consimilis Mast

conspicuus (Mast.) Pillans

corneolus Esterhuysen

crinalis Mast

curviramis Kunth

cuspidatus Thunb

cymosus (Mast.) Pillans

debilis Nees

debilis Nees var. subulatus (Mast.) Pillans

decipiens (N.E. Br.) Linder

degenerans Pillans

depauperatus Kunth

dichotomus L

digitatus Thunb

dimorphostachyus Mast

dispar Mast

distachyos Rottb

distans Pillans

distichus Rottb

distractus Mast

divaricatus Mast

dodii Pillans

dodii Pillans var. purpureas Pillans

duthieae Pillans

echinatus Kunth

ecklonii Mast

egregius Hochst

egregius Hochst. var. nutans Mast

ejuncidus Mast

elatus Mast

elegans Poir

elegia Murray

eleocharis Nees ex Mast

elongatus Thunb

erectus Thunb

esterhuyseniae Pillans

exilis Mast

fasti giatus Nees ex Mast

ferruginosus Link ex Kunth

festuciformis Nees ex Mast, (‘festucaeformis’)

filicaulis Pillans

filiformis Poir

filiformis Poir. var. monostachyus (Mast.) Mast

filiformis Poir. var. oligostachyus (Mast.) Mast

foliosus N.E. Br

fourcadei Pillans

fragilis Esterhuysen

fraternus Kunth

fruticosus Thunb

fuirenoides Kunth

furcatus Nees ex Mast 442,

fuscidulus Pillans

fusiformis Pillans

galpinii Pillans

garnotianus Kunth

garnotianus Kunth var. monostachyus Steud. ex

Mast 446

garnotianus Kunth var. oligostachyus Mast 446

gaudichaudianus Kunth 405

gaudichaudianus Kunth var. luxurians Pillans 405

gaudichaudianus Kunth var. microstachyus Nees ex

Mast 405

giganteus (Kunth) N.E. Br 479

glomeratus Thunb 494

gossypinus Mast 405

graminifolius Kunth 486

grandis Spreng. ex Nees 424

harveyi Mast 447

helenae Mast 406

humilis Pillans 417

hystrix Mast 406

imbricatus Thunb 396

implexus Mast 443

implicatus Esterhuysen 447

impolitus Kunth 467

inconspicuus Esterhuysen 449

incurvatus Pillans 470

incurvatus Thunb 494

ingens Esterhuysen 449

insignis Pillans 450

intermedins Steud 424

intermedins Kunth 414

intricatus Mast 455

inveteratus Esterhuysen 450

involutus Pillans 452

junceus (L.) Nees 425

kunthii Steud 415

laniger Kunth 406

laniger Kunth var. distractus (Mast.) Pillans 404

leptoclados Mast 407

leptostachyus Kunth 452

luceanus Kunth 405

lucens Poir 475

lucens Poir. var. minor Mast 475

ludwigii Steud 415

macer Kunth 407

macowanii Pillans 479

madagascariensis Cherm 452

madagascariensis Cherm. var. humbertii Cherm 452

mahonii (N.E. Br.) Pillans 452

mayor (Mast.) Pillans 436

marlothii Pillans 407

marlothii Pillans var. parviflorus Pillans 407

mastersii F. Muell 436

membranaceus Nees 424

micans Nees 452

miser Kunth 452

monanthos Mast 407

monostachyus Steud 446

montanus Esterhuysen 453

mucronatus Nees 430

multicurvus N.E. Br 446

multiflorus Spreng 453

multiflorus Spreng. var. tuberculatus Pillans 453

neesii Mast 414

nodosus Pillans 453

nudus (Rottb.) Nees 430

nutans Thunb 476

nutans Steud 415

nuwebergensis Esterhuysen 453

oblongus Mast 486

obscurus Pillans 454

obtusissimus Steud 484

occultus (Mast.) Pillans 454

ocreatus Kunth 410

oligostachyus Kunth 446

pachystachyus Kunth 454

paludosus Pillans 410

paniculatus Rottb 469

pannosus Mast

papyraceus Pillans ...

parviflorus Thunb. ...

patens Mast

pauciflorus Poir

peculiaris Esterhuysen

pedicellatus Mast

441

442

452

442

444

442

436

443

436

396

429

403

405

403

443

479

436

443

414

445

443

486

404

402

444

436

402

487

454

445

396

445

404

414

445

404

445

470

445

446

405

482

425

404

474

404

446

436

405

446

479

447

405

479

414

454

405

447

446

502

Bothalia 15, 3 & 4 (1985)

penicillatus Mast 407

perplexus Kunth 455

perplexus Kunth var. gracilis Mast 443

perseverans Esterhuysen 457

pillansii Linder 457

polystachyus Kunth 436

pondoensis Mast 469

praeacutus Mast 457

praefixus Mast 436

procurrens Mast 406

productus Mast 406

propinquus Nees 425

protractus Mast 453

pseudoleptocarpus Kunth 444

punctulatus Nees ex Mast 445

pulvinatus Esterhuysen 457

pumilus Esterhuysen 458

purpurascens Nees ex Mast 458

pusillus Pillans 452

pycnostachyus Mast 444

pygmaeus Pillans 443

quadratus Mast 458

quinquefarius Nees 459

racemosus Poir 426

ramiflorus Nees 469

rams Esterhuysen 459

rhodocoma Mast 478

rottboellioides Kunth 443

rupicola Esterhuysen 459

sabulosus Pillans 443

sarocladus Mast 460

scaber Mast 460

scaberulus N.E. Br 460

scariosus Thunb 474

schlechteri (Mast.) Pillans 454

schoenoides Kunth 443

scopa Thunb 482

scoparius Kunth 444

scopula Mast 455

secundus (Pillans) Linder 460

sejunctus Mast 460

setiger Kunth 443

sieberi Kunth 444

sieberi Kunth var. schoenoides (Kunth) Pillans 443

sieberi Kunth var. venustulus (Kunth) Pillans 444

similis Pillans 460

singularis Esterhuysen 460

sonderianus Mast 455

sparsus Mast 464

spicifer Poir 397

spicigerus Thunb 478

spiculatus Mast 444

spinulosus Kunth 436

sprengelii Mast 402

squarrosus Poir 402

stereocaulis Mast 461

stokoei Pillans 461

strictus N.E. Br 461

strobilifer Kunth 461

subcompressus Pillans 437

subfalcatus Nees ex Mast 414

subtilis Nees ex Mast 461

subulatus Mast 444

subverticillatus (Steud.) Mast 415

sulcatus Kunth 493

synchroolepis Steud 493

tabularis Pillans 460

tectorum L.f 430

tenuissimus Kunth 415

tetragonus Thunb 461

tetrasepalus Steud 396

thamnochortus Thunb 475

thyrsifer Rottb 427

trichocaulis Mast 455

triflorus Rottb 415

triticeus Rottb 462

tuberculatus Pillans 462

umbellatus Thunb 396

vaginatus Thunb 397

vallis-simius Linder 462

venustulus Kunth 414

verrucosus Esterhuysen 462

versatilis Linder 463

verticillaris L.f 421

vilis Kunth 417

vimineus Rottb 470

virgatus Rottb 482

virgeus Mast 417

wallichii Mast 417

xyridioides Kunth 459

zuluensis Linder 463

zwartbergensis Pillans 464

Rhodocoma Nees 478

capensis Nees ex Steud 478

fruticosa (Thunb.) Linder 479

gigantea (Kunth) Linder 479

Schoenus L.

capensis L 402

Staberoha Kunth 396

aemula (Kunth) Pillans 396

banksii Pillans 396

caricina (Mast.) Dur. & Schinz 474

cernua (L.f.) Dur. & Schinz 396

distachyos (Rottb.) Kunth 396

disticha (Rottb.) Dur. & Schinz 445

gracilis (Mast.) Dur. & Schinz 474

imbricata (Thunb.) Kunth 396

imbricata (Thunb.) Kunth var. stenoptera (Kunth) Dur.

& Schinz 396

multispicula Pillans 396

ornata Esterhuysen 396

remota Pillans 397

stenoptera Kunth 396

stokoei Pillans 397

vaginata (Thunb.) Pillans 397

Thamnochortus Berg 471

acuminatus Pillans 472

aemulus Kunth 396

arenarius Esterhuysen 472

argenteus (Thunb.) Kunth 489

argenteus Pillans 473

bachmannii Mast 473

bromoides Kunth 475

burchellii Mast 474

canescens Mast 473

caricinus Mast 474

cernuus (L.f.) Kunth 396

cinereus Linder 473

comptonii Pillans 476

consanguineus Kunth 476

dichotomus (L.) Spreng 402

dichotomus Mast 475

dichotomus Pillans var. hyalinus Pillans 475

distichus (Rottb.) Mast 445

dumosus Mast 473

ecklonianus Kunth 475

ellipticus Pillans 474

elongatus (Thunb.) Mast 474

erectus (Thunb.) Mast 474

floribundus Kunth 474

fraternus Pillans 474

fruticosus Berg 474

fruticosus Berg. var. glaber Mast 474

giganteus Kunth 479

glaber (Mast.) Pillans 474

gracilis Mast 474

guthrieae Pillans 475

imbricatus (Thunb.) Mast 396

imbricatus (Thunb.) Mast. var. stenopterus (Kunth) Mast. 396

insignis Mast 475

levynsiae Pillans 475

lewisiae Pillans 475

lucens (Poir.) Linder 475

mastersii Gand 497

maximus Kuntze 478

membranaceus Mast 445

micans (Nees) Kunth 452

modestus Kunth 479

muirii Pillans 476

muticus Pillans 478

Bothalia 15, 3 & 4 (1985)

503

nervosus Pillans 475

nutans (Thunb.) Pillans 476

obtusus Pillans 476

occultus Mast 454

paniculatus Mast 476

papillosus Pillans 475

papyraceus Pillans 476

pellucidus Pillans 476

piketbergensis Pillans 478

platypteris Kunth 476

plumosus Pillans 475

pluristachyus Mast 476

pulcher Pillans 477

punctatus Pillans All

rigidus Esterhuysen All

robustus Kunth 482

scabridus Pillans All

scariosus (Thunb.) Spreng 474

schlechteri Pillans 478

scirpiformis Mast 474

scirpoides Kunth 482

similis Pillans 478

spicigerus (Thunb.) Spreng 478

sporadicus Pillans 478

stokoei Pillans 478

striatus Hochst 478

strictus Kunth 481

sulcatus Mast 473

umbellatus (Thunb.) Kunth 396

virgatus (Rottb.) Kunth 482

Willdenowia Thunb 493

affinis Pillans 493

arescens Kunth 493

argentea (Kunth) Hieron 479

bolusii Pillans 494

brevis Nees ex Mast 493

compressa Thunb 481

cuspidata Mast 494

decipiens N.E. Br 479

ecklonii (Nees) Kunth 494

ecklonii (Nees) Dur. & Schinz 486

esterhuyseniae Pillans 479

fimbriata Kunth 479

fistulosa (Mast.) Pillans 480

fraterna N.E. Br 495

galpinii N.N. Br 495

glomerata (Thunb.) Linder 494

humilis Nees ex Mast 494

incurvata (Thunb.) Linder 494

lucaeana Kunth 494

neglecta Steud 494

peninsularis N.E. Br 495

purpurea Pillans 494

rugosa Esterhuysen 495

simplex N.E. Br 489

stokoei Pillans 495

striata Thunb 494

sulcata Mast 495

teres Thunb 495

xerophila Pillans 480

Bothalia 15, 3 & 4: 505-530 (1985)

The genus Scleria in southern Africa

E. F. FRANKLIN HENNESSY*

Keywords: Scleria , Cyperaceae, classification, southern Africa

ABSTRACT

The 23 species of Scleria (Sclerieae, Caricoideae, Cyperaceae) in southern Africa are revised. Two subgenera

are recognized, Hypoporum with one section, Hypoporum, and Scleria with three sections, Scleria, Acriulus and

Schizolepis. The tribes Sclerieae and Bisboeckelereae are distinguished.

INTRODUCTION

Scleria Bergius is a pantropicai genus comprising

approximately 200 species, 23 of which have been

recorded in southern Africa. No general agreement

has been reached on the circumscription of the

genus, its subdivision, its tribal affiliation, or the sys-

tematic position of its tribe within the family Cypera-

ceae.

The number and delimitations of tribes in Cypera-

ceae varies, for example, Bentham (1883) recog-

nized six tribal units, Clarke (1908) seven, Holttum

(1948) six, Hutchinson (1959) seven, Koyama (1961)

six, Hooper (in Metcalfe, 1971) eight and Eiten

(1976) nine with two genera (Scleria and Dulichium)

not assigned to tribes because of uncertainty with re-

gard to their taxonomic position.

The main lines of evolutionary development have

been suggested by the grouping of tribes in catego-

ries of higher hierarchical level within which the

positioning of the tribes indicates putative phyloge-

netic relationships. Bentham (l.c.) grouped his six

tribes into two Series, Monoclines, with hermaphro-

dite flowers (Scirpeae, Hypolytreae, Rhynchospo-

reae) and Diclines with unisexual flowers (Cryptang-

ieae, Sclerieae, Cariceae). The four subfamilies re-

cognized by Clarke (l.c.) were Scirpo-Schoeneae

(Cypereae, Scirpeae, Schoeneae, Rynchosporeae,

all with hermaphrodite flowers); Mapaniae with a

single tribe diagnosed as having unisexual flowers in

an inflorescence with a terminal female flower and

basal male spikelets; Scleriae with a single tribe diag-

nosed as having unisexual flowers in monopodial bi-

sexual spikelets with a single basal female flower and

male flowers towards the apex or in monopodial uni-

sexual spikelets, the male multiflowered, the female

1-flowered; and Caricineae, also with a single tribe

having unisexual flowers in monopodial spikelets

with the female flowers enclosed in a utricle. Holt-

tum (l.c.) followed Bentham’s arrangement but

switched the position of Hypolytreae and Scirpeae in

Monoclines. Hutchinson (l.c.) failed to recognize

subfamilial rank thereby indicating his seven tribes

as representing separate evolutionary lines. Koyama

(1961) recognized four subfamilies, Mapanioideae

with a single tribe, Hypolytreae; Scirpoideae (Scir-

peae, Cypereae); Rhynchosporoideae (Rhyncho-

*Department of Botany, University of Durban-Westville, Private

Bag X54001, Durban 4000.

sporeae, Sclerieae); and Caricoideae (Cariceae). By

his placement of Mapanioideae as the most primitive

subfamily, he failed to recognize, as had Bentham

(1883), Pax (1886, 1887) and Holttum (1948) the

pseudanthial nature of the ultimate inflorescence

unit in this group.

Hooper (l.c.) proposed the acceptance of the se-

ven tribes recognized by Hutchinson with the addi-

tion of an eighth tribe for Dulichium arranged, with

slight modification in the sequence of some genera,

in the framework proposed by Clarke. Subfamily

Scirpoideae comprised Cypereae, Scirpeae, Rhyn-

chosporeae (Rhynchosporeae and Schoeneae of

Clarke) and Dulicheae; Mapanieae comprised a sin-

gle tribe, Hypolytreae; Caricoideae comprised Scle-

rieae (Scleriae, part 1 of Clarke), Cryptangieae

(Scleriae part 2 of Clarke) and Cariceae.

Eiten (l.c.) proposed a system of classification

based upon analysis of the branching patterns of the

ultimate branch orders of the inflorescence together

with the sex of the flowers. She recognized three

subfamilies, the arrangement of which differs from

that of all earlier systems in their sequence. By plac-

ing Mapanioideae last, she suggests that this is the

most specialized group in the family. Cyperoideae

( = Rhynchosporoideae) (Scirpeae, Cypereae, Rhyn-

chosporeae and the genus Dulichium) is distin-

guished as having true, bisexual flowers arranged in

true, racemosely-branched spikelets; Caricoideae

(Lagenocarpeae, Bisboeckelereae, Cariceae and the

genus Scleria ) as having true, always unisexual flow-

ers arranged in true, racemosely-branched spikelets;

and Mapanioideae (Mapanieae, Syntrinemeae,

Micropapyreae) having an inflorescence of one or

more pseudospikelets, each pseudospikelet made up

of pseudanthia of unisexual flowers borne race-

mosely on a rhachilla.

By its placement in the tribe Sclerieae (Nees,

1834) the distinctiveness of Scleria within Cypera-

ceae was early recognized. The unisexuality of its

flowers was also soon brought to attention (Ben-

tham, 1883). Both Pax (1886, 1887) and Eiten (1976)

placed the genus in subfamily Caricoideae but,

according to the former author, the subfamily was

diagnosed by ‘spike let with a terminal flower';

according to the latter author the ‘true, always uni-

sexual flowers are arranged in true, racemosely-

branched spikelets’. This reflects both the changing

diagnosis of categories with time and increasing

understanding and knowledge and, for Scleria, the

506

Bothalia 15, 3 & 4 (1985)

on-going conflict in intrepretation of the position of

the female flower in the bisexual (androgynaeceous)

spikelet.

Interpretation of spikelet morphology

Eiten’s recent work on South American plants has

brought yet further authority to the opinion held by

Nees (1842), Bentham (1883), Holttum (1948),

Koyama (1961) favouring the lateral positioning of

the female flower in the bisexual spikelet. Pax (l.c.),

Core (1936), Kern (1961, 1974), Schultze-Motel

(1964) and Koyama (1967, 1969) have interpreted

the bisexual spikelet of Scleria as comprising two

axis systems; a sympodial axis which terminates in a

female flower with a second, higher order axis sys-

tem bearing male flowers arising laterally from the

first axis.

Since the first appendage of a lateral branch is a

prophyll (Blaser, 1944), which is recognizable be-

cause of its position and its distinctive form then, if

the second interpretation were correct, a prophyll

would be present in the adaxial position near the

base of the axis which bears male flowers.

1 have found no evidence of a prophyll in such a

position in the bisexual spikelets of any species.

Kern (1961) illustrates (Fig. 1; I, II) prophylls in dia-

grams of spikelets of ‘bisexual Scleria species’. It

seems apparent that the diagrams were constructed

to support an interpretation for which no direct evi-

dence could be found and are hypothetical. Koyama

(1961, p. 50) stated unequivocally that the female

flower in the bisexual spikelet of Scleria gracillima is

truly axillary, citing evidence and illustrating his

findings in Fig. 3C. In a later publication (1969), he

included a prophyll in illustrations (Figs 5, 6 & 27)

which appear to be based on the 1961 drawing. How-

ever, I doubt whether the prophyll depicted in the

1969 publication really exists.

Since in the plants themselves no prophyll is inter-

posed between the proximal female part of the spi-

kelet and the distal male part, there is direct evi-

dence that the spikelet is a monopodial structure.

Further evidence obtained from analysis of branch-

ing patterns of the whole inflorescence also supports

this interpretation.

If the bisexual spikelet is interpreted as consisting

of two axes, with the female flower terminating the

first axis (thus sympodial) and the lateral (second)

axis bearing male flowers, then in the species which

have truly unisexual spikelets, lateral branches of

the axis which terminates in a female spikelet might

be expected to bear terminal male spikelets. Only

Scleria angusta, S. greigiifolia and S. poiformis

among southern African species have female spike-

lets which usually lack male rudiments and none of

their inflorescences have such a branching pattern. If

a lateral axis is developed in these species from an

axis with a terminal female spikelet it, too, termin-

ates in a female spikelet.

There is, therefore, no morphological evidence to

support the interpretation of the bisexual spikelet as

a double axis system with the female flower termin-

ating a sympodial axis from which the second axis

bearing male flowers arises laterally.

My findings, based on study of southern African

material have led me to conclude, independently of

Eiten (1976) who worked with Brazilian species, that

the bisexual spikelet of Scleria is a monopodial struc-

ture. Analytical diagrams of Scleria spikelet types

are given in Fig. 1, A-D.

Amended circumscription of tribe Sclerieae

The tribe Sclerieae comprising the genera Becque-

relia, Bisboeckelera, Calyptrocarya, Diplacrum (in-

cluding Pteroscleria ), and Scleria (including Acriu-

lus ), was not upheld by Eiten (1976), on the grounds

that whereas the ultimate inflorescence unit of four

of these genera is a compound axis system in which

the main axis which bears lateral, true, racemosely-

branched spikelets of true, male flowers terminates

in a pistil, the ultimate inflorescence unit of Scleria is

a simple axis which comprises a true, racemosely-

branched spikelet of true, unisexual flowers; that is,

in Scleria no branch system terminates in a pistil.

Accordingly, Scleria was excluded from the as-

semblage and the tribe Bisboeckelerieae was pro-

posed for the other four genera, since Bisboeckelera

is the earliest legitimate generic name in the new

tribe.

Scleria was not assigned by Eiten (l.c.) to any tribe

since, although two tribes, Rhynchosporeae and

Cariceae have spikelets with a branching pattern

similar to that of Scleria , the former has bisexual

flowers and the latter has the female flower included

in a utricle or a semiutricular prophyll.

Since my findings support Eiten’s view of the in-

trepretation of spikelet morphology in Scleria, it is

proposed that the tribe Sclerieae be maintained, and

that its circumscription be modified so that it in-

cludes (in the present state of our knowledge) only

the name genus.

Therefore in subfamily Caricoideae (Eiten, 1976)

which is diagnosed as having true, always unisexual

flowers in true, racemosely-branched spikelets, it is

proposed to distinguish the tribes Bisboeckelereae

and Sclerieae as follows:-

Ultimate inflorescence unit compound, comprising an axis

apparently terminating in a pistil, and lateral, true,

racemosely-branched spikelets of true male flowers

Bisboeckelereae ( Bisboeckelera , Becquerelia, Calyptroca-

rya, Diplacrum).

Ultimate inflorescence unit simple, comprising a true,

racemosely branched spikelet of true unisexual flow-

ers Sclerieae ( Scleria )

Tribus Sclerieae Nees emend. E. F. Franklin a

tribu Bisboeckelereae Mattf. in Diels inflorescentiae

monadate ultima simplici, ex spicula vera racemosa

floribus unisexualibus veris constanti dignoscendus.

Typus: Scleria Bergius.

Generic limits of Scleria

Although the distinctiveness of Scleria has long

been recognized, there is still some difference of

opinion with regard to the relationship of Scleria and

Diplacrum R. Brown (1810) and Scleria and Acriu-

lus Ridley (1884). Diplacrum is maintained as a sep-

Bothalia 15, 3 & 4 (1985)

507

arate genus by most cyperologists but Kern (1961,

1974), Koyama (1961) and Raymond (1966) in-

cluded it in Scleria. Eiten (1976) has demonstrated

that the fundamental branching patterns of the ulti-

mate inflorescence units of Scleria and Diplacrum

differ, so that, far from being congeneric, these taxa

must, on this basis, be assigned to different tribes.

The genus Acriulus was reduced to congenerity in

Scleria by Clarke (1902) and subsequently (1908) re-

stored by him to generic rank. In 1963 it was once

more reduced to synonymy by Kern, and the num-

ber of species was reduced from three to one (5.

greigiifolia) and the spelling of the specific epithet

was corrected. Although I do not agree with Kern’s

intrepretation of the inflorescence of Scleria, I agree

that inflorescence structure in Scleria and Acriulus is

fundamentally the same. The validity of his argu-

ment in favour of reducing Acriulus to congenerity

in Scleria is accepted and additional evidence in sup-

port of this course is offered.

FIG. 1. — Scleria , explanatory diagrams: A, androgynaceous; B, male; C, subandrogynaeceous; D,

female spikelets; p, prophyll; g, glume; E, female flower in axil of glume; F, male flower in

axil of glume; G, L/S ovary of subgenus Hypoporum; H, L/S ovary of subgenus; Scleria ; J-L,

raised surface-patterns on achenes; J, tuberculate; Jp, pattern profile; K, tuberculate-verru-

cose; Kp, pattern profile; L, trabeculate; Lp, pattern profile; M-O, depressed surface-pat-

terns on achenes; M, alveolate, Mp, pattern profile; N, lacunose; Np, pattern profile, O,

striate-lacunose; Op, pattern profile; P, achene of 5. nutans, subgenus Hypoporum, with tri-

gonous stipe without hypogynium; Q, achene of S. poiformis, subgenus Scleria, with obpyra-

midal stipe with hypogynium; h, hypogynium.

508

Bothalia 15, 3 & 4 (1985)

1. The type of habitat occupied by Acriulus is the

same as that occupied by some species of Scle-

ria, for example, S. poiformis.

2. The habit of Acriulus is like that of, for ex-

ample, S. poiformis.

3. The fundamental branching pattern of the inflo-

rescence of Acriulus and Scleria is the same.

4. Spikelet morphology of Acriulus and some

species of Scleria is the same.

5. Morphology of the flowers of Acriulus is funda-

mentally the same as that of Scleria.

6. Achene morphology of Acriulus and some

species of Scleria such as 5. melanomphala is

very similar, Acriulus being distinguished only

by its pronounced beak, a feature which may oc-

cur in species of Scleria not represented in

southern Africa.

7. The form of the hypogynium of Acriulus is simi-

lar to that of some species of Scleria, notably S.

melanomphala.

8 The achenes of Acriulus and Scleria are silicified

(Franklin, 1979) in the same manner, and to an

extent not known in any other genus in Cypera-

ceae.

9. Structure of the pericarp is fundamentally the

same in Acriulus and Scleria.

10. No anatomical feature of the root, rhizome or

lamina of Acriulus is not shared by one or more

species of Scleria.

11. Most anatomical features of the culm of Acriu-

lus are shared by one or more species of Scleria ,

the exception being the possession by Acriulus

of some (not all) amphivasal vascular bundles, a

feature which may be shared by some species of

Scleria from regions other than southern Africa.

Infrageneric (supraspecific) groups in Scleria

It has long been recognized that the genus com-

prises several natural assemblages which have vari-

ously been designated as subgenera, sections and se-

ries. Little agreement on the hierarchical levels of

infrageneric groups, or their delimitation has been

reached. Opinions as to whether evolution in the

genus has been monophyletic, diphyletic or polyphy-

letic vary.

Analysis of habitat preferences, habit, morpho-

logy and anatomy of southern African species has

led me to conclude that evolution in Scleria has been

diphyletic. Some of the evidence upon which this

conclusion is based is shown in Table I. Each of the

two co-lateral ‘lines’ comprises not a linear sequence

of taxa, but a far more complex dendritic pattern of

ascending, interlinked branches of different lengths.

Subgeneric rank is proposed for each of these lines,

a treatment first used by Clarke in 1894 but aban-

doned by him later.

Features of the subgenera

Subgenus Hypoporum (Nees) C.B.C1. comprises

slender, narrow-leaved, sometimes hairy plants of

open, seasonally dry, often temperate habitats. They

are either annuals, or have subterranean perennat-

ing organs and annual aerial parts. The evolution of

drought/cold escape mechanisms (completion of the

life-cycle in a season, or withdrawal of food reserves

into a protected underground organ and sometimes

also into enlarged culm-bases) has permitted mem-

bers of this subgenus to exploit a wider range of

habitats than is available to taxa without such escape

mechanisms, such as predominate in subgenus Scle-

ria.

Subgenus Scleria (Berg.) C.B.C1. comprises more-

or-less robust, usually broad-leaved, often evergreen

perennials, and, less often, annuals. Many taxa are

shade-dwellers and the majority are hygrophilous or

hydrophilous, are restricted to tropical and subtropi-

cal habitats, and, with few exceptions do not mani-

fest drought/cold escape-mechanisms. The annual

species occupy tropical habitats in areas where sea-

sonal drought may be experienced and it is suggested

that the annual habit is a drought escape-mechanism

in such species. Among southern African species in

this subgenus, only one is known which has annual

aerial parts and a perennial rhizome with swollen,

persistent, culm-bases. It is significant that this

species, S. transvaalensis, occurs at higher, more

temperate altitudes than other local species in the

subgenus.

The more robust habit and greater breadth of the

laminas of most members of subgenus Scleria is con-

sistent with the longer lifespan of the aerial parts of

these plants compared with that of members of sub-

genus Hypoporum. The shady habitats occupied by

some members of subgenus Scleria is also conducive

to increased stature and to increased breadth of

laminas. It is suggested that the lateral, pseudodorsi-

ventral wing of laminar tissue present in the species

with praemorse leaves is a modification which, by

increasing surface area and volume, may increase

the photosynthetic capacity of these shade-dwellers.

A feature of the laminas of shade-dwelling species is

the absence of stomata from the adaxial epidermis

(except a few in the pseudodorsiventral laminar ex-

tension when it is present), which is wholly bulliform

except where it overlies mechanical tissue in the lat-

eral ribs. The role of the bulliform cells is not

known, but I suggest that, since a wholly-bulliform

adaxial epidermis is present only in shade-tolerant

species, it may serve as a light-transmitting layer.

Epidermal cells of Scleria are a repository for silica.

Since the walls of bulliform cells are silicified and

their lumina often filled with silica, the uninter-

rupted layer is rigid, so that laminas which possess

such a layer are maintained in a fully-expanded

state, which may be an advantage in a shady habitat.

Thickness of laminas is greatest in lacustrine, he-

liophilic taxa of members of subgenus Scleria and is

an external manifestation of the development of an

extensive air-space system in the mesophyll. Such

laminas are amphistomatic, as are the laminas of all

southern African members of subgenus Hypoporum

and, as in Hypoporum, may lack intercostal bulli-

form cells in the adaxial epidermis or may have files

of bulliform cells alternating with files of smaller

cells in which stomata are present.

TABLE 1.— Distribution of some taxonomically useful characters among southern African species of Scleria

509-510

Not

praemorse

± branched ± simply

Praemorse glomerate glomerate

spicate spicate

inter-

rupted

panic-

ulate

rarely

X

androgyn-

aeceous,

and male

X

androgyn-

aeceous

subandro-

gynaeceous

and male

subandro-

gynaeceous

female and

male

Present

vascular bundles

female

and

male

Absent

non-buoyant

X

rarely

margin margin

smooth fimbriate

buoyant col-

" lateral

margin

ciliate

amphi-

vasal

peripheral

sclerenchyma ring

absent present

X

■ amphi-

stomatic

outer parenchyma sheath

usually

hypo-

stomatic

not

continuous

around

assoc, sc.

continuous

around

associated

sclerenchyma

Section Subgenus

Hypoporum

> Scleria

Acriulus

Schizolepis

Ophryoscleria

“Hypoporum

Scleria

Bothalia 15, 3 & 4 (1985)

511

Although the fundamental branching pattern of

the paniculate inflorescence of all species of Scleria

is the same, modification has taken place in two

ways, namely, by progressive contraction of all or

most ramuli leading to the ‘glomerate-spicate’ type

of inflorescence characteristic of subgenus Hypopo-

rum in which the bracts are (mostly) reduced, glumi-

form structures, and, by progressive contraction of

some ramuli together with progressive elongation of

others leading to the ‘interrupted-paniculate’ type of

inflorescence characteristic of subgenus Scleria, in

which most bracts are foliaceous. It is suggested that

in Hypoporum the branched glomerate-spicate type

of inflorescence such as is seen in S. woodii is less

specialized than the simply glomerate-spicate type

seen in S. aterrima. In subgenus Scleria it is postula-

ted that the greater the degree of elongation of prox-

imal internodes in the inflorescence, and the greater

the number of such elongated ramuli, the more

highly specialized the inflorescence. Based upon

these criteria, the southern African species in subge-

nus Scleria with the least specialized inflorescences

are S. lacustris and S. poiformis and those with the

most specialized are S. greigiifolia and S. angusta.

It is postulated that unisexual spikelets in Scleria

have been derived by reduction from bisexual (an-

drogynaeceous) spikelets (Fig. 1. A-D). Unisexual

female spikelets are unknown in subgenus Hypopo-

rum, which has bisexual spikelets and unisexual

male spikelets. It is suggested that the higher the ra-

tio of bisexual to male spikelets in the inflorescence,

the less specialized the inflorescence.

In subgenus Scleria, unisexual male and func-

tionally female spikeUU occur in an inflorescence

and, rarely, (S. lacustris) also some bisexual spike-

lets. It is suggested that species which consistently

produce some bisexual spikelets are more primitive

than those which consistently lack them, and that

species whose functionally female spikelets consist-

ently lack any vestigial male parts (Fig. ID) are

more advanced than those which have male rudi-

ments (Fig. 1C).

The hypogynium or ‘disc’ which is present on

some achenes is considered to be a new modification

of the stipe of the achene and not a vestigial struc-

ture (Fig. 1H, Q), therefore it is postulated that the

type of achene found in subgenus Hypoporum which

has a trigonous stipe lacking any distal elaboration as

a hypogynium (Fig. 1G, P), is primitive, and that the

type of achene found in subgenus Scleria which has

an obpyramidal stipe elaborated distally as a hypo-

gynium is derivative.

There is evidence that in Scleria, evolution of two

different types of plants, two different types of inflo-

rescence, different types of achene-bearing spikelets

and two different types of achene has occurred in

response to differences in habitat.

Glomerate-spicate inflorescences are characteris-

tic of taxa which occupy seasonally dry, grassland

habitats. The achene-bearing spikelets are bisexual,

mature simultaneously on the plant, and are often

held stiffly erect. There is no hypogynium devel-

oped, therefore the female glumes continue to clasp

the achene firmly even after silicification of the peri-

carp is complete and the achene becomes detached

from its pedicel. Achenes are not shed, but reach the

substratum only when the aerial parts of the plant

die back in winter. The inevitable result is that

achenes are mostly distributed in the immediate vi-

cinity of the parent plants, thereby ensuring that

when germination takes place the seedlings are in a

suitable habitat. Since germination is likely to occur

more-or-less simultaneously with the onset of fa-

vourable conditions in spring, shedding of achenes

over an extended period would confer no advantage

on such plants.

Taxa with interrupted-paniculate inflorescences

are, with few exceptions, evergreen hygro- or hydro-

philous plants of tropical or subtropical habitats.

The inflorescences, achene-bearing spikelets and

achenes of these plants have become modified so

that achenes mature progressively in an inflores-

cence, and can be shed instead of being retained on

the plants for the extended lifespan of the aerial

parts. By elongation of their proximal internodes the

partial panicles have become pendulous (mostly) so

that the spikelets hang upside-down. The achene-

bearing spikelets have lost, wholly or in part, the dis-

tal male part so that the solitary female flower ap-

pears to be terminal. The loss of the distal male part

of the spikelet and assumption of a pseudoterminal

position by the female flower remove mechanical

obstruction to the spreading of the female glumes as

the pistil matures. The presence of a hypogynium in-

creases the width of the achene towards its base and

forces the glumes to spread apart further than would

occur if no hypogynium were present. When the pro-

cess of silicification of the achene is complete and

the vascular supply is severed, the hypogynium

which is not wholly silicified, becomes desiccated

and shrinks, and the achene drops out of the inflo-

rescence. The advantages of such a system to taxa

occupying wet habitats in areas with little seasonal

temperature fluctuation are obvious.

Sections within the subgenera

The number of southern African species of Scleria

is small, representing only about 10% of the genus.

It has not been possible to accumulate sufficient evi-

dence from such a small sample to permit grouping

of the southern African species of subgenus Hypo-

porum into more than one section. Pending world-

wide revision of the genus a single section, Hypopo-

rum (Nees) Endlicher, is recognized, with the

characters of the subgenus.

Even with such a small sample, subdivision of sub-

genus Scleria as it is represented in southern Africa

is possible, based partly upon the presence or ab-

sence of male rudiments in the achene-bearing spi-

kelets. Three sections are recognized in our area and

a fourth is represented close to our border in Mo-

zambique.

To section Scleria (Bergius) Endlicher are as-

signed those taxa in which most functionally female

spikelets retain rudimentary distal male parts. The

least specialized condition is one in which there are

some fully androgynaeceous (bisexual) spikelets as

well as functionally female and male spikelets in the

512

Bothalia 15, 3 & 4 (1985)

same inflorescence e.g. S. lacustris. The hypogynia,

although morphologically different in different taxa,

all have entire margins.

This is the least homogeneous section in subgenus

Scleria. Its members occupy a greater variety of

habitats and show a wider range of morphological

diversity than do members of the more highly

specialized sections. Within the section it is possible

to discern groups of species which share similar mor-

phological and anatomical features. However, since

so few species are present in southern Africa cir-

cumscription of species groups will have to be de-

ferred pending world-wide revision of the genus.

Those taxa in subgenus Scleria, in which the func-

tionally female spikelets are (with rare exceptions)

without male rudiments, have been assigned to three

sections namely, Acriulus (Ridley) C.B.C1., Schizo-

lepis (Nees) C.B.C1. and Ophryoscleria (Nees)

C.B.C1.

Section Acriulus is monotypic, S. greigiifolia being

distinguished from members of section Scleria by its

strictly female spikelets; by its adaxially hirsute fe-

male glumes; by its long-beaked achene and by the

presence of amphivasal vascular bundles in the culm.

Its hypogynium is like that of some members of sec-

tion Scleria. Acriulus is maintained as a section pen-

ding world-wide revision of the genus which may

show that the characters enumerated are shared by

some members of section Scleria in which case S.

greigiifolia would have to be placed in that section

and the sectional diagnosis amended.

Members of sections Schizolepis and Ophryoscle-

ria occupy swamp-forest habitats, are alike in habit

and have praemorse leaves. The sections are distin-

guished by their hypogynia which, in Schizolepis

have fimbriate margins, and in Ophryoscleria are

corky, cupuliform, exceed the achene in width and

have ciliate margins.

Although there are hydrophilous taxa in all sec-

tions of subgenus Scleria, none but the members of

section Ophryoscleria have buoyant achenes. Buoy-

ancy of the fruit is attributable to the highly special-

ized type of hypogynium. The achene sinks in water

if the hypogynium is artifically removed. Population

spread is probably facilitated by buoyancy of the

fruit, therefore it is considered to be likely that taxa

with this type of morphologically distinctive hypogy-

nium which serves a biologically important function

not attained in any other section, have reached the

highest level of specialization in Scleria.

SCLERIA

Scleria Berg, in Vet. Akad. Handl. Stockh. 26:

142, t.4 (1765); Sw. in Prodr. 18 (1788); Endl., Gen.

PI. 112 (1836); Kunth, Enum. PI. 2: 339 (1837);

Boeck. in Vidensk. Medd. Dansk Naturh. Foren.

Kbh. 9-13: 150 (1869); Boeck. in Linnaea 38: 436

(1874); Boeck. in Flora 62: 569 (1879); Benth. &

Hook, f., Gen. PI. 3: 1070 (1883); C.B.C1. in FI.

Brit. Ind. 6: 685 (1894); C.B.C1. in FC 7: 293 (1898);

C.B.C1. in Urban, Symb. Antill. 2: 137 (1900);

C.B.C1. in FTA. 8: 493 (1902); C.B.C1. in Journ.

Misc. Inf. Kew add. ser. 8: 131 (1908); Schonland in

Mem. bot. Surv. S. Afr. 3: 64 (1922); Brain in Proc.

Rhod. scient. Assoc. 33: 51 (1934); Hutch, in FWTA

2: 491 (1936); Chermezon in Arch. Bot. Caen 7,

Mem. 2: 88 (1936); Core in Brittonia 2: 1 (1936);

Pierart in Lejeunia 13: 1 (1951); Phillips, Gen. edn

2, 158 (1951); Nelmes in Kew Bull. 10: 415 (1955);

Nelmes in Kew Bull. 11: 73 (1956); Kern in Blumea

11: 140 (1961); Koyama in J. Fac. Sci. Tokyo Univ.

(Bot.) 8: 134 (1961); Napper in J1 E. Africa nat.

Hist. Soc. 24: 23 (1964); Robinson in Kew Bull. 18;

487 (1966); Podlech in FSWA 51 (1967); Jacot Guil-

larmod in FI. Lesotho 132 (1971); Gordon-Gray in

Ross, FI. Natal 111 (1972); Kern in FI. Malesiana

722 (1974); Compton in FI. Swaziland 73 (1976);

Dyer, Gen. 2: 889 (1976). Type species: S. flagel-

lum-nigrorum Berg.

Acriulus Ridl. in J. Linn. Soc., Bot. 20: 336

(1883). For other synonymns see Benth. in Benth. &

Hook, f., Gen. PI. 3: 1071 (1883).

Monoecious annual herbs with fibrous roots, or

stout or slender monoecious perennial herbs with

short or long, horizontal, oblique or descending,

fleshy or woody rhizomes or with ± horizontal sub-

terranean soboles, or with both rhizomes and so-

boles. Culms nodose, solitary or ± tufted, erect or

scandent, trigonous or triquetrous, leafy towards the

base or throughout, smooth or more usually scabrid

on the angles, glabrous or hairy. Leaves 3-ranked,

narrowly to broadly linear with sheathing bases, ±

smooth to scabrid on the margins and the 3-5 princi-

pal ribs, glabrous or hairy, the lowermost represent-

ed by almost bladeless or bladeless sheaths; laminas

tapering smoothly towards apex or suddenly nar-

rowed at unequal distances on each side from the

apex (‘praemorse’); profile ± V-shaped, flanged V-

shaped or in the praemorse species flanged V-

shaped distally with additional lateral wings to the

flanges proximally; sheaths closed, sometimes

shortly 3-winged, the mouth truncate, concave, con-

vex or produced into a short tongue. Inflorescence

paniculate with a lax or compact terminal panicle

and usually one or more lateral panicles, with (rarely

without) foliaceous bracts, or branched or simply

glomerate-spicate with ± glumaceous bracts. Spike-

lets androgynaeceous (bisexual) or unisexual, the

functionally female spikelets sometimes subandrogy-

naeceous; androgynaeceous spikelets with one basal

or sub-basal, lateral female floret and 1 to several

upper male florets some of which may be sterile;

functionally female spikelets with 2-4 empty glumes

proximally and one sub-basal lateral female floret

and 1 to several empty glumes distally (subandrogy-

naeceous), or lacking sterile distal glumes; male spi-

kelets with 1-2(3) empty glumes proximally and sev-

eral to many male florets of which the distal few may

be sterile. Flowers unisexual, solitary in axils of spi-

rally arranged glumes. Male flower of (1) 2-3 sta-

mens; anthers bithecate, linear, often apiculate. Fe-

male flower consisting of a tricarpellary, unilocular

ovary with a terminal style branched above into 3

filiform stigmas, the style deciduous or, rarely, the

base persistent. Achenes ovoid, ellipsoid or subglo-

bose and obscurely or obtusely trigonous or strongly

trigonous, smooth or variously sculptured, glabrous

or hairy, with silicified pericarp, whitish, grey,

Bothalia 15, 3 & 4 (1985)

513

brown, purple or violet, lustrous or dull, borne on a

trigonous or obpyramidal stipe which is sometimes

expanded at the apex into a persistent, triangular,

trilobed, zoniform or cupulate hypogynium with en-

tire, fimbriate or ciliate margin.

A genus of ± 200 species distributed throughout

the tropics and subtropics. 23 indigenous species oc-

cur in our area, 3 of which may be endemic.

The generic name Selena is derived from the

Greek word scleros, meaning hard, in allusion to the

hard fruit, the pericarp of which is silicified.

The genus is divided into 2 subgenera and 5 or 6

sections, 5 of which are represented in Africa and 4

in southern Africa. The section Hymenolytrum

(Nees) Core which is endemic in South America may

not be distinct from sect. Scleria (Berg.) Endl. (Re-

fer Fig. 1).

KEY TO SECTIONS

Inflorescence glomerate-spicate, branched or simple, terminal; bracts glumiform or the lower subfoliaceous;

spikelets androgynaeceous or androgynaeceous and male; hypogynium absent Hypoporum (spp. 1-12)

Inflorescence paniculate, terminal and lateral or terminal (5. poiformis); bracts foliaceous (except 5. poi-

formis); spikelets androgynaeceous, subandrogynaeceous and male (S. lacustris), subandrogynaeceous,

female and male, or female and male; hypogynium present:

Hypogynium margin entire, glabrous:

Female glumes glabrous on adaxial surface Scleria (spp. 13-21)

Female glumes densely hirsute on adaxial surface Acriulus (sp. 22)

Hypogynium margin fimbriate or ciliate:

Margin of hypogynium fimbriate; hypogynium not cupuliform; style-base not persistent

Schizolepis (sp. 23)

Margin of hypogynium ciliate; hypogynium cupuliform, broader than the achene; style-base persistent

*Ophryoscleria

I. Subgenus Hypoporum (Nees) C.B.Cl. in

Hook, f., FI. Brit. Ind. 6: 685 (1894). Type species:

Hypoporum pergracile Nees.

Slender to very slender perennial or annual herbs

up to 1,5 m tall, the perennial species rhizomatous or

soboliferous or with both rhizomes and soboles.

Leaves evenly spaced along length or crowded to-

wards base of culm, (1) 2-9 mm broad, tapering

smoothly towards apex, glabrous or hairy, the ribs

and margins scaberulous or smooth. Inflorescence

terminal, glomerate-spicate, branched or un-

branched, with ± glumaceous bracts. Spikelets all bi-

sexual (androgynaeceous) or bisexual and male.

Achenes smooth or variously sculptured, glabrous,

the stipe trigonous. Hypogynium absent.

Section Hypoporum (Nees) Endl., Gen. PI. 112

(1836).

Hypoporum Nees in Journ. Edinb. Phil. Soc. 17:

266 (1834) et in Mart., FI. Bras. 2: 169 (1842). Type

species: Hypoporum pergracile Nees.

Characters of subgenus Hypoporum.

II. Subgenus Scleria (Berg.) C.B.Cl. in Hook,

f., FI. Brit. Ind. 6: 686 (1894). Type species: Scleria

flagellum-nigrorum Berg.

Plants herbaceous, tall (to 2,5 m), stout, rhizoma-

tous perennials or medium-sized (to 2 m) rhizoma-

tous perennials or medium-sized annuals. Leaves

evenly spaced along length or crowded towards base

of culm, 2-40 mm broad, tapering smoothly towards

apex or abruptly and unequally narrowed in the dis-

tal part, glabrous or hairy, the ribs and margins sca-

brid or scaberulous. Inflorescence paniculate, the

* Recorded from Mozambique but not yet recorded from FSA

area.

panicles lax or contracted, terminal or terminal and

lateral, with foliaceous bracts. Spikelets androgynae-

ceous (rarely), subandrogynaeceous, female and

male. Achenes smooth or variously sculptured, gla-

brous or hairy, the stipe obpyramidal. Hypogynium

present.

1. Section Scleria (Berg.) Endl., Gen. PI. 112

(1836).

Scleria Berg, in Vet. Akad. Handl. Stockh. 26:

142 (1765). Type species: Scleria flagellum-nigrorum

Berg.

Plants tall (to 2 m), stout (S. poiformis) or

medium-sized perennials, or medium-sized annuals

(S. lacustris, S. foliosa). Leaves evenly spaced along

length of culm or, (S. poiformis), crowded towards

base of culm, 2 mm broad (S. unguiculata) - 40 mm

broad (S. poiformis), usually tapering smoothly to-

wards apex. Inflorescence terminal (S. poiformis) or

terminal and lateral. Spikelets androgynaeceous

(rarely, 5. lacustris) subandrogynaeceous, female

and male. Female glumes glabrous on adaxial sur-

face. Achenes smooth or variously sculptured, gla-

brous or hairy, beakless or almost so. Hypogynium

strongly or obscurely trilobed, rarely zoniform (S.

melanomphala), the margin entire, glabrous.

2. Section Acriulus (Ridl.) C.B.Cl. in FTA 8:

495 (1902).

Acriulus Ridl. in J. Linn. Soc. Bot. 20: 336

(1883). Type species: Acriulus greigiifolius Ridl. as

greigifolius.

Plants tall (to 2 m), stout, perennial. Leaves

crowded towards base of culm, 5-12 mm broad, usu-

ally tapering smoothly towards apex. Inflorescence

lax, copious, terminal and lateral. Spikelets female

and male. Female glumes densely hirsute on adaxial

514

Bothalia 15, 3 & 4 (1985)

surface. Achenes smooth, glabrous, strongly beaked.

Hypogyniumzomiorm, the margin entire, glabrous.

3. Section Schizolepis (Nees) C.B.C1. in Hook,

f., FI. Brit. Ind. 6: 694 (1894).

Schizolepis Nees in Mart., FI. Bras. 2: 186 (1842).

Type species: Scleria latifolia Sw.

Plants tall (to 2,5 m), stout, perennial. Leaves

evenly spaced along length of culm, 6-16 mm broad,

usually abruptly and unequally narrowed towards

apex. Inflorescence terminal and lateral. Spikelets fe-

male and male. Female glumes glabrous on adaxial

surface. Achenes smooth, glabrous, beakless. Hypo-

gynium trilobed, the margin fimbriate.

5. Section Ophryoscleria (Nees) C.B.C1. in Ur-

ban, Symb. Antill. 2: 138 (1900).

Ophryoscleria Nees in Mart., FI. Bras. 2: 182

(1842). Type species: Scleria racemosa Poir.

Plants tall (to 2,5 m), stout, perennial. Leaves

evenly spaced along length of culm, 8-35 mm broad,

usually abruptly and unequally narrowed towards

apex. Inflorescence terminal and lateral. Spikelets fe-

male and male. Female glumes glabrous on adaxial

surface. Achenes smooth or variously sculptured,

glabrous or hairy, beakless or beaked, with persist-

ent style-base. Hypogynium cupulate, broader than

the achene, the margin ciliate.

KEY BASED ON VEGETATIVE AND FRUIT CHARACTERS

a Inflorescence in terminal and lateral panicles, or (in S. poiformis and, rarely, S. melanomphala) in a

solitary terminal panicle; bracts foliaceous (except 5. poiformis); hypogynium present:

Lamina usually abruptly narrowed towards apex (praemorse); hypogynium margin fimbriate

23. 5. angusta

Lamina tapering smoothly towards apex; hypogynium margin entire:

Leaves 20-40 mm broad; inflorescence terminal, without foliaceous bracts; hypogynium obscurely 3-

lobed; achene ± globose, smooth, glabrous, grey-brown or white if immature 21. 5. poiformis

Leaves less than 20 mm broad; inflorescence terminal and lateral, with foliaceous bracts; hypogynium

3-lobed or collar-like without defined lobes; achene smooth or patterned, glabrous or hairy:

b Achene hairy (at least proximally):

Achene hairy proximally, glabrous distally, smooth or faintly striate-lacunose 19. S. lagoensis

Achene hairy distally and proximally, patterned:

Achene very faintly reticulate-lacunose; male spikelets 7-9 mm long; lateral panicles single at

nodes 18. 5. achtenii

Achene distinctly tessellate-lacunose; male spikelets 3-5 mm long; lateral panicles 2^4(5) at

the nodes 17. S. unguiculata

bb Achene glabrous:

c Achene patterned:

Plant caespitose, without rhizome, annual; achene alveolate-lacunose 16. S. foliosa

Plant rhizomatous, perennial; achene reticulate-lacunose to tuberculate-lacunose:

Culm bases 3-4 mm in diameter, not or hardly swollen; foliaceous bracts of terminal panicle

ensiform, 3-4 mm broad 10 mm behind apex, exceeding inflorescence 15. S. natalensis

Culm bases 9-10 mm in diameter, swollen; foliaceous bract of terminal panicle subulate, 1-2

mm broad 10 mm behind apex, sometimes exceeding inflorescence ....14. S. transvaalensis

cc Achene smooth:

Plant without rhizome, annual, lacustrine; with adventitious roots at base of culm from several

nodes above the base; hypogynium very small, 3-lobed 13. 5. lacustris

Plant rhizomatous:

Panicles compact, spiciform, 1—2— (4); female glumes 7-11 mm long, glabrous adaxially;

achene avoid, beakless, grey with blackish apex 20. S. malanompliala

Panicles lax, many; female glumes 6-7 mm long, densely villous adaxially in distal

half; achene broadly ovoid, strongly beaked, pinkish-brown, sometimes with violet

blotches 22. 5. greigiifolia

aa Inflorescence terminal, glomerate-spicate, branched or simple; bracts glumaceous or the lowermost

sub-foliaceous; hypogynium absent:

Plants without propagative stems, annual; lamina profile flattened V-shaped 9. S. pergracilis

Plants with subterranean propagative stems:

Plants without rhizome, spreading by means of long, hard, horizontal, culm-like soboles with inter-

nodes 8-47 mm long; achene with a series of deep horizontal and vertical ridges at junction of

stipe and body 8. 5. sobolifer

Plant with true rhizome:

Rhizome soft, fleshy, strongly-scented, white or pink, tuberous, shrinking markedly soon after

removal from soil; inflorescence much-branched, the branches delicate; lamina profile

V-shaped 1- 5. woociii

Rhizome not as above; inflorescence simply glomerate-spicate or sparingly branched; lamina

profile flanged V-shaped:

d Rhizome descending:

Bothalia 15, 3 & 4 (1985)

515

Rhizome stout, woody; leaves crowded towards base of culm, 2-7 mm broad, the majority

short-bladed or bladeless; achene acutely trigonous, strongly reticulate-trabeculate,

8rey 6. 5. veseyfitzgeraldii

Rhizome very slender, terminating in a swollen tuber up to 1 cm long; leaves 1-2,5 mm broad;

achene subglobose, trabeculate-verrucose, grey, the trabeculae pale to bright reddish-

8°^ 10. S. dieterlenii

dd Rhizome ± horizontal:

Lulms distinctly bulbous and ± woody at the base 5. 5. bulbifera

Culms not, or only very slightly bulbous at the base:

e Culms clustered , rhizome with very short internodes ; soft, short soboles sometimes present:

Glomerules reflexed at maturity; inflorescence unbranched; glumes densely hirsute, the

hairs blackish; achene smooth 12. S. aterrima

Glomerules not reflexed; inflorescence unbranched or with few short basal branches;

glumes glabrous or sparsely ciliate; achene smooth or tuberculate or trabeculate

towards the apex 7. 5. dregeana

ee Culms arising in a ± straight series from a hard horizontal rhizome at least 2 mm thick:

Glomerules reflexed at maturity; inflorescence unbranched; glumes densely hirsute, the

hairs pale, reddich or blackish; achene smooth or lightly tuberculate 11. S. nutans

Glomerules not reflexed:

Spikelets 8-9 mm long; glumes hairy, the hairs pale; achene smooth, light brown with

black stipe 4 5 longispiculata

Spikelets less than 8 mm long; glumes glabrous or glabrescent; achene smooth or

tuberculate, stipe pale:

Inflorescence rhachis drooping, 60-250 mm long; spikelets 5-7(8) mm long; glumes

tawny, glabrous or glabrescent 2. S. welwitschii

Inflorescence rhachis stiffly erect, less than 170 mm long; spikelets 4-5 mm long;

glumes reddish-brown to dark brown, glabrous 3.5. rehmannii

KEY FOR FIELD USE

a Plants tall (to 2,5 m) and stout or medium sized; inflorescence paniculate:

Leaves abruptly and unequally narrowed towards the apices; shade-dwelling in coastal swamp-

forest; Transkei, Natal 23. 5. angusta

Leaves tapering smoothly towards apices:

Inflorescence terminal, without foliaceous bract; leaves ± 40 mm broad, thick and spongy proxi-

mally; forming dense stands in open coastal pans; Natal, north of Tugela River.... 21. 5. poiformis

Inflorescence terminal, or terminal and lateral, with foliaceous bracts:

b Achene hairy (at least proximally):

Achene hairy proximally, glabrous towards and on top; open damp habitats; known for FSA

area only from Swaziland 19. 5. lagoensis

Achene hairy proximally and on top:

Lateral panicles single at each node; male spikelets 7-9 mm long; achene hairs white; open

damp habitats; Natal coastbelt 18 5. achtenii

Lateral panicles 1-3 or more at each node; male spikelets 3-5 mm long; achene hairs golden;

open wet habitats; northern Botswana 17. 5. unguiculata

bb Achene glabrous:

Achene patterned:

Plants caespitose, without rhizome; achene smooth on top, patterned proximally; annual

in open, seasonally wet habitats inland; Swaziland, Transvaal, northern Namibia

16. 5. folios a

Plants rhizomatous, perennial:

Panicles lax, pale greenish-yellow; bracts ensiform, 3-4 mm broad in the distal 10 mm, over-

arching their panicles; culm bases not or hardly swollen; partly shaded streambanks in

margins of coastal forest; Natal and Transkei 15. 5. natalensis

Panicles dense, golden or reddish; bracts subulate, 1-2 mm broad in the distal 10 mm,

not conspicuously overarching their panicles; culm-bases swollen to c. 10 mm dia-

meter; open, damp habitats or semi-sheltered by banks or among rocks; northern and

eastern Transvaal, Swaziland and known from one locality (Nkandla) in Natal

14. 5. transvaalensis

Achene smooth:

Leaves evenly spaced along length of culm; panicles very compact, spike-like, dark reddish-

brown; achene ovoid, beakless, grey with black apex; in, or on periphery of open,

wet, frostfree habitats; Transkei, Natal, Transvaal, Swaziland, northern Botswana

20. 5 melanomphala

Leaves crowded towards base of culm; panicles lax, very copiously branched, dark reddish-

brown; achene ovoid, strongly beaked, light brown sometimes with violet blotches; in, or

516

Bothalia 15, 3 & 4 (1985)

on periphery of open, wet, frost-free habitats; Southern Natal and known from one local-

ity near Lake St. Lucia 22. S. greigiifolia

aa Plants medium sized to small (0,5 m), sometimes very slender; inflorescence glomerate-spicate,

branched or unbranched;

Plants without propagative stems; annuals:

Plants medium sized, aquatic, with floating roots at several basal nodes of culm; inflorescence stiffly

branched, the branches glomerate-spicate; bracts subfoliaceous; in rivers or lakes; northern

Botswana 13. S. lacustris

Plants slender, caespitose; leaf profile flattened V-shaped without lateral flanges, margins slightly

recurved; inflorescence simply glomerate-spicate or'with one short basal branch; seasonally

wet areas; near Dundee in Natal 9. 5. pergracilis

Plants with subterranean propagative stems:

Achene acutely trigonous; rhizome very short, ± vertical and difficult to discern; culm bases thickly

invested with numerous dry leaf-sheaths; leaves mostly short-bladed, crowded towards base of

culm; seasonal flood-plains; Caprivi Strip 6. 5. veseyfitzgeraldii

Achene ovoid to subglobose, obscurely trigonous:

Culm bases distinctly bulbous and woody; leaves subequally spaced along length of culm; inflores-

cence simply glomerate-spicate or with 1 - several basal branches; open, seasonally damp

grassland; Transkei, Natal, Transvaal, Swaziland 5. S. balbifera

Culm bases not, or only very slightly bulbous:

c Rhizome woody, at least 2 mm thick, elongate, ± horizontal, with culms arising at intervals from

it:

Glumes densely hairy:

Glomerules reflexed at maturity; hairs on glumes mostly reddish-black; leaves ± evenly

spaced along length of culm; open, permanent bogs; Transkei, Natal, Transvaal

11. S. nutans

Glomerules not reflexed at maturity; hairs on glumes pale; achene stipe black; open, damp

grassland in sandy soil; northern Botswana and northern Namibia 4. 5. longispiculata

Glumes glabrous or very sparsely hairy:

Inflorescence simply glomerate-spicate or sparingly branched, usually drooping; spikelets

dull straw-coloured, 5-8 mm long; open permanent bogs; Natal midlands and uplands,

Transvaal, Swaziland 2. S. welwitschii

Inflorescence simply glomerate-spicate or sparingly branched, stiffly erect; spikelets red-

dish-brown, 4-5 mm long; open, seasonally or permanently wet grassland; Transvaal,

north-east Namibia 3. S. rehmannii

cc Rhizome and culms not as above;

Glomerules reflexed at maturity; glumes densely hairy, the hairs purplish-black; leaves

crowded towards base of culm; plant apparently caespitose because rhizome very short;

sometimes with 1 - several soft fleshy soboles arising from culm-cluster; open, perma-

nently wet areas in sandy soil; Transkei, Natal, Transvaal 12. 5. aterrima

Glomerules not reflexed at maturity:

Perennating stem wholly or partly softly tuberous, swollen:

Rhizome descending, very slender becoming swollen and softly tuberous towards the tip;

slender, delicate, caespitose, strictly montane plants of open, seasonally wet habitats

eastern Cape, Transvaal, Natal, Lesotho 10. 5 dieterlenii

Rhizome horizontal or oblique, the internodes of young rhizomes pearly-white or pink,

swollen and softly tuberous, very strongly scented; lamina profile V-shaped without

lateral flanges; inflorescence copiously branched, the branches delicate; glomerules

of few spikelets; open, seasonally wet habitats or damp woodland in partial shade;

Transkei, Natal, O.F.S.. Transvaal, Lesotho, Swaziland 1. S. woodii

Perennating stem not swollen and tuberous:

Culms clustered, sometimes with 1 or more soft, terete soboles from

base of culm-cluster; rhizome very short; inflorescence simply glomerate-spicate or

sparingly branched towards base; open, permanently wet habitats; northern and

eastern Cape, Natal, Transvaal, Lesotho, Swaziland, northern Botswana. 7. 5. dregeana

Culms usually solitary, widely spaced, linked by hard, trigonous, red-speckled soboles

with internodes 8—47 mm long; inflorescence simply glomerate-spicate; strictly

coastal in open, seasonally wet habitats in sandy soil; Natal 8. S. sobolifer

1. Scleria woodii C.B.Cl. in FC 7:295 (1898) et

in FT A 8:501 (1902); Nelmes in Kew Bull. 10:428

(1955); Napper in Kew Bull. 25:443 (1971) non Ro-

binson (1966); [S. woodii sensu E.A. Robinson in

Kew Bull. 18:512 (1966), pro parte]. Type: South

Africa, Zululand, Wood 3994 (K, lecto.!; NH!;

BOL!).

Perennial. Rhizome 2-4 mm thick, fleshy, white

or pink, strongly scented; scales pink. Culms

0,25-0,75 mm tall, solitary or few, clustered, gla-

brous or glabrescent. Leaves 1-3 mm broad, gla-

brous or glabrescent; sheaths with truncate or con-

cave, glabrescent to densely hirsute mouths. Inflo-

rescence branched, delicate, 80-200 mm long. Glo-

Bothalia 15, 3 & 4 (1985)

517

merules 1—3— (4) per branch, sessile, of 1-6 spikelets.

Bracts shorter than or exceeding the glomerules,

with scabrid awn 1-10 mm long. Spikelets 2,5-5 mm

long. Glumes 2, 5-4, 5 mm long, glabrous, shortly

awned, pale red with darker red striae. Achene

ovoid to subglobose, 1,5-1, 8 x 1-1,5 mm, glabrous,

smooth, tuberculate or trabeculate, even on the

same plant, grey.

Perennial with annual aerial parts, widespread in

seasonally boggy open areas and damp woodland in

partial shade in the summer rainfall region of south-

ern Africa (Fig. 2) and to the north in Zimbabwe,

Zambia, Tanzania and Angola.

This species is distinguished from other southern

African representatives of subgenus Hypoporum

mainly by its strongly scented, soft, fleshy, tuberous,

pearly white or pink rhizome; by its narrowly Y--

shaped lamina profile; and by its delicate, often pro-

fusely-branched inflorescence.

Vouchers : P. A. Smith 2033; Acocks 11340; Kil-

lick 1222; Ward 8738; B. R. Roberts 3087.

FIG. 2. — Distribution map of Scleria woodii.

2. Scleria welwitschii C.B.Cl. in Dur. & Schinz,

Consp. FI. Afr. 5:675 (1895) et in FTA 8:501 (1902);

Nelmes in Kew Bull. 10:423 (1955); Robinson in

Kew Bull. 18:506 (1966); Compton, FI. Swaziland 74

(1976). Type : Angola, Welwitsch 7138 (BM,

lecto.!).

S. junciformis Welw. in Ridl. in Trans. Linn. Soc. ser. 2, Bot. 2:

168 (1884), nom illegit, non Thw. (1864). Type : as above.

Perennial. Rhizomes 3-4 mm thick, woody, red;

scales stramineous-reddish. Culms 0,30-1 m tall, vil-

lous to glabrescent, 5-7 mm distant. Leaves 2-3 mm

broad, glabrous to villous; sheath with mouth pro-

duced into a triangular or rounded, villous or gla-

brescent tongue 1-2 mm long. Inflorescence

branched or unbranched, (60)-150-250 mm long,

drooping. Glomerules ( 1 )— 3— 8 per branch, of

1—2— (6) sessile spikelets. Bracts shortly awned,

shorter than the glomerules. Spikelets 5-7-(8) mm

long. Glumes 3-5 mm long, glabrous or minutely his-

pidulous distally, shortly awned, pale stramineous

with faint reddish streaks. Achene ellipsoid to ovoid,

1,5-1 ,8 x 1-1,2 mm, glabrous, smooth, grey.

Perennial with annual aerial parts, occurring in

permanently boggy grassland areas at temperate alti-

tudes in summer rainfall region of southern Africa

(Fig. 3) and to the north in Zimbabwe, Malawi and

Angola.

This species is morphologically similar to S. reh-

mannii C.B.Cl. and S. longispiculata Nelmes. All

three species have ± horizontal, woody rhizomes;

that of S. welwitschii is reddish, those of S. rehman-

nii and S. longispiculata pale stramineous. The inflo-

rescences of S. welwitschii and S. rehmannii are sim-

ply glomerate-spicate or, in S. welwitschii and less

often in S. rehmannii with one or a few lateral

branches from the proximal glomerules, whereas

that of S. longispiculata is unbranched. The inflores-

cence of 5. welwitschii is lax and drooping, those of

S. rehmannii and S. longispiculata are rigid, ± erect.

Spikelets of 5. welwitschii are 5-8 mm long; of S.

rehmannii 4-5 mm long; of S. longispiculata 8-9 mm

long. The glumes of 5. welwitschii are fulvous-stra-

mineous or castaneous; of S. rehmannii wholly or

partly blackish-red; of S. longispiculata light casta-

neous. Achenes of S. welwitschii and 5. longispicu-

lata are smooth; those of S. rehmannii smooth or

lightly tuberculate. Achenes of S. welwitschii and S.

rehmannii are grey or light brown; of 5. longispicu-

lata brown with the stipe black. Achenes of S. wel-

witschii and S. rehmannii are ± 2 mm long; those of

S. longispiculata ± 4 mm long.

Vouchers : C.J. du Plessis 880; Gordon-Gray

6096; Smook 1058; Killick 1233; Edwards 1127.

FIG. 3. — Distribution map of Scleria welwitschii •; 5. rehman-

nnii A and S. longispiculata ■

3. Scleria rehmannii C.B.Cl. in FC 7: 295 (1898)

et in FTA 8: 501 (1902); Nelmes in Kew Bull. 10: 425

(1955); Robinson in Kew Bull. 18: 507 (1966); Pod-

lech in FSWA 165: 52 (1967). Type: Rehmann 5626

(K, holo.!; BOL!).

S. welwitschii var. tuberculata Cherm. in Arch. Bot. Caen 7: 13

(1936). Type: Ubangi Shari (Central African Republic), Tisserant

2922 (P, holo.!).

Perennial. Rhizome 3-6 mm thick, woody, yellow;

scales light brown. Culms 0,3-1, 5 m tall, glabrous,

5-10 mm distant. Leaves 1-3,5 mm broad, villous to

glabrescent; sheath with mouth produced into a tri-

518

Bothalia 15, 3 & 4 (1985)

angular, villous or glabrescent: tongue 1-2 mm long.

Inflorescence branched or unbranched,

40-120-(170) mm long, stiffly erect. Glomerules

(l)-2-12 per branch, of 1-6 sessile spikelets. Bracts

shortly awned, shorter than or equalling the glomer-

ules. Spikelets 4—5 mm long. Glumes 2-4,5 mm long,

glabrous, shortly awned, blackish-red or pale red

with darker streaks. Achene broadly ovoid or sub-

globose, 1,25-2 x 1-1,6 mm, glabrous, smooth or

tuberculate, grey or light brown.

Perennial with annual aerial parts, occurring in

seasonal or permanent bogs in open grassland.

Known in southern Africa only from Transvaal and

north-east Namibia near Rundu (Fig. 3). Wide-

spread in Zimbabwe, Zambia, Mozambique, Mal-

awi, Tanzania, Angola, Zaire and Central African

Republic.

Morphological characters which distinguish this

species from the closely allied species S. welwitschii

and S. longispiculata are discussed under S. welwits-

chii.

Vouchers: Vesey-Fitzgerald 1007 (NU); A.

Johnston 17 (NU); De Winter & Marais 5049.

4. Scleria longispiculata Nelmes in Kew Bull. 13:

150 (1958); Robinson in Kew Bull. 18: 506 (1966);

Podlech in FSWA 165: 52 (1967). Type: Tanzania,

Milne-Redhead & Taylor 9739 (K, lecto.!, sheet I).

Perennial. Rhizome 4-6 mm thick, woody, tawny;

scales light brown. Culms 0,45-1,05 m tall, glabrous

or sparsely villous, 5-20 mm distant. Leaves 2-4-(5)

mm broad, glabrous or villous; sheath with mouth

produced into a triangular or rounded, glabrous or

villous tongue 0,5-5 mm long. Inflorescence un-

branched, 40-130 mm long. Glomerules 4—8, of 1-5

sessile spikelets. Bracts awned, shorter than or ex-

ceeding the glomerules. Spikelets 8-9 mm long.

Glumes 4—7 mm long, hispidulous-pubescent,

awned, castaneous. Achene ovoid to broadly ovoid,

4—4,4 x 2-2,75 mm, glabrous, smooth, light brown

with three darker interangular stripes, the stipe

black.

Perennial with annual aerial parts, occurring in

well-drained sandy soil bordering wet grassland on

Kalahari sands. Known in southern Africa only from

north-east Namibia (Fig. 3), and elsewhere only

from Zambia and Tanzania.

Morphological characters which distinguish this

species from the closely allied species, S. welwitschii

and S. rehmannii are discussed under S. welwitschii.

Vouchers: Story 6467; De Winter 3915.

5. Scleria bulbifera Hochst. ex A. Rich., Tent.

FI. Abyss. 2: 510 (1851); C.B.C1. in FTA 8: 500

(1902); Kukenth. in Fedde, Repert Beih. 40: 530

(1938); Pierart in Lejeunia 13: 24, t.l, fig. 9 (1953) :

Nelmes in Kew Bull. 10: 438 (1955); Robinson in

Kew Bull. 18: 503 (1966); Compton, FI. Swaziland

74 (1976). Syntypes: Ethiopia, Schimper 1557 (BM;

K!); Quartin- Dillon & Petit s.n. (BM; K!).

S. atrosanguinea Hochst. ex Steud., Syn. PI. Glum. 2: 175

(1885). Type: Ethiopia, Schimper 327 (K, holo.!; BM!).

S. schweinfurthiana Boeck. in Flora 62: 570 (1879). Type: Su-

dan, Schweinfurth 2193 (K, holo.!).

S. buchananii Boeck., Cyper. Nov. 1: 33 (1888); C.B.C1. in FC

7: 295 (1898) et in FTA 8: 499 (1902). Syntypes: Malawi, Bucha-

nan 32 (K!); 1272 (K!).

S. verdickii De Wild, in Rev. Zool. Afr. 14 Suppl. Bot. 26

(1926). Type: Congo, Verdick 398 (BR, holo.!).

S. schliebenii Gross in Notizbl. Bot. Gart. Berlin 11: 657

(1932). Type: Tanzania, Schlieben 782 (Bt).

S. thomasii Pierart in Bull. Soc. Bot. Belg. 83: 405 (1951).

Type: Zaire, R. X. L. Thomas 1202 (BR, holo.!).

Perennial. Rhizome little more than the connec-

tive between swollen, contiguous culm-bases, or the

internodes longer and culm-bases intervallate; scales

light brown. Culms 0,12-1,10 m tall; bases swollen,

woody, up to 12 mm thick, glabrous or hairy above.

Leaves 1-5(9) mm broad, glabrous or hirsute; sheath

with mouth concave, truncate, or produced into a

short membranous tongue. Inflorescence un-

branched or branched, 20-200 mm long. Glomerules

3-17, of 1-12 sessile spikelets. Bracts awned, equal-

ling or up to twice the length of the glomerule, or the

lowermost subfoliaceous, up to 30 mm long. Spike-

lets 4—6,5 mm long. Glumes 2-5 mm long, glabrous

or hairy, awned, castaneous or dark reddish-brown.

Achene obovoid to subglobose, 1,6-2 x 1-1,8 mm,

glabrous, smooth or lightly or strongly tuberculate

or trabeculate, grey or light brown.

Perennial with annual aerial parts, occurring in

dry, or seasonally wet or permanently boggy open

grassland habitats. Widespread in the summer rain-

fall region of southern Africa (Fig. 4) and to the

north in tropical Africa and in Malagasy.

The morphology of the inflorescence of this

species varies from branched to simply glomerate-

spicate, with the glomerules comprising few or many

spikelets. The surface patterning of the achene is

also variable, with smooth, tuberculate or trabecu-

late achenes occurring, sometimes even on the same

plant. The same types of inflorescence and achene

may occur in several species from all of which S. bul-

bifera is distinguished by its bulbous, woody and sil-

icified culm bases which arise from a woody, hori-

zontal rhizome with (usually) very short internodes

so that adjacent culm bases are ± contiguous.

Vouchers: C. A. Smith 1341; Acocks 18794; R. P.

Ellis 3279; E. F. Hennessy 407 (NU; UD-W); Ruda-

tis 528 (STE).

6. Scleria veseyfitzgeraldii E. A. Robinson in

Kew Bull. 18: 503 fig. 3. (1966). Type: Robinson

4220 (K, holo.!; NU!).

Perennial. Rhizome descending, ± vertical, short,

woody. Culms clustered, to 1 m tall, the bases 3 mm

thick, invested with many dry leaf-sheaths. Leaves

2-7 mm broad, glabrescent to densely hirsute, most

crowded towards base of culm, bladeless or short-

bladed, with deeply concave mouths. Inflorescence

unbranched, rarely branched, 50-150 mm long. Glo-

merules dense, multispiculate. Bracts awned, equal-

ling or to twice the length of glomerules, the lower-

most subfoliaceous. Spikelets 4—6,5 mm long.

Glumes 4-6 mm long, glabrescent or hirsute, shortly

awned, castaneous or dark brown with green keels.

Achene acutely trigonous, broadly obovoid, 2 x

1,3-1 ,6 mm, glabrous, reticulate-trabeculate, grey.

Botnalia 15, 3 & 4 (1985)

519

FIG. 4. — Distribution map .of Scleria bulbifera A and S. vesey-

fitzgeraldii @.

Perennial with annual aerial parts, occurring in

open, seasonally inundated or saturated flood

plains. Known for FSA area only from the Caprivi

Strip in north east Namibia (Fig. 4) and elsewhere

from Zambia and Tanzania.

In its aerial parts this species bears a superficial

resemblance to some specimens of 5. bulbifera from

which it is distinguished by its short, descending rhi-

zome which is not apparent without sectioning; by its

culm bases not or hardly enlarged, thickly invested

with numerous dry leaf sheaths; by its leaves mostly

short-bladed or bladeless, crowded towards the base

of the culm, and by its acutely trigonous achenes.

Voucher: Killick & Leistner 3218.

7. Scleria dregeana Kunth, Enum. PI. 2: 354

(1837); C.B.C1. in FC 7: 295 (1898) et in FTA 8: 499

(1902); Nelmes in Kew Bull. 10: 426 (1955); Robin-

son in Kew Bull. 18: 510 (1966); Compton, FI. Swa-

ziland 74 (1976). Type: South Africa, Cap. b. spei,

Drege s.n. sub C.B.Cl. 3934 (Bt; K, lecto.!).

5. meyeriana Kunth l.c.:354 (1837); C.B.Cl. in FC 7: 294 (1898)

et in FTA 8: 498 (1902); Nelmes in Kew Bull. 10: 431 (1955).

Type: South Africa. Cap. b. spei, Drege s.n. sub C.B.Cl. 4363

(Bt).

S. holcoides Kunth l.c.:354 (1837); C.B.Cl. in FC 7: 296 (1898);

Nelmes in Kew Bull. 10: 427 (1955). Type: South Africa, Drege

s.n. sub C.B.Cl. 4381 (Bt; K, lecto.!).

S. caespitosa Welw. ex Ridl. in Trans. Linn. Soc. ser. 2 Bot. 2:

167 (1884). Type: Angola, Welwitschl\35 (BM, holo.; K; LISU).

S. setulosa Boeck., Cyper. Nov. 1: 33 (1888). Type: Malawi,

Buchanan 36 (K, holo.!).

Perennial with two different kinds of propagative

stems. Rhizome little more than the connective be-

tween ± contiguous culm bases; scales reddish-

brown. Soboles sometimes produced from culm

bases, soft, terete, 10 mm long, with collateral vas-

cular bundles. Culms 0,25-1 m tall, glabrous or gla-

brescent. Leaves 1-3 mm broad glabrescent or hir-

sute; mouth of sheath truncate, villous below. Inflo-

rescence unbranched or branched, 10-100 mm long.

Glomerules 1-11 per branch, of 1-8 sessile or subses-

sile spikelets. Bracts aristate, equalling the glome-

rules or the lowermost up to 35 mm long. Spikelets

4,5-6 mm long. Glumes 2,25-5 mm long, glabrous or

sparsely ciliate, shortly awned, blackish-red, red-

brown or pale with red streaks. Achene ovoid, ellip-

soid or subglobose, l,5-2.x 1,2 mm, glabrous,

smooth or tuberculate or trabeculate towards the

apex, smooth at junction of stipe and body, grey.

Perennial with annual aerial parts, widespread in

open, permanent bogs and wet streambanks in sum-

mer rainfall area of southern Africa (Fig. 5) and to

the north in Zimbabwe, Zambia, Malawi, Tanzania,

Angola and Zaire.

This is a morphologically variable species. Rhi-

zome internodes may be so short that the plants ap-

pear caespitose or they may be slightly longer so that

the culm bases are not contiguous. Subterranean

propagative stems (soboles) may arise from the culm

bases. Whereas rhizomes have amphivasal vascular

bundles, those of soboles are collateral. Inflores-

cences may be simply glomerate-spicate or sparingly

branched glomerate-spicate. Glumes may be dark or

pale. S. dregeana Kunth, 5. caespitosa Welw. ex

Ridl. and S. setulosa Boeck. were based on glabres-

cent plants with unbranched inflorescences and dark

glumes; S. meyeriana Kunth on a very hairy plant

with an unbranched inflorescence and pale glumes

and 5. holcoides Kunth on a hairy plant with a

branched inflorescence and pale glumes. The

achenes of the types are sparingly tuberculate dis-

tally and smooth proximally but gatherings of plants

with wholly smooth achenes ( Lubke 181), and of

plants with achenes which are tuberculate distally

and trabeculate proximally ( Galpin 9104) exist. Tu-

bercles, when present, are invariably best developed

near the apex of the achene.

This species is likely to be confused with S. sobo-

lifer E. F. Franklin from which it is distinguished

mainly by the presence of a true rhizome with am-

phivasal vascular bundles and by the absence of a

pattern of ridges, troughs and fine transverse bars

between the stipe and body of the achene.

Vouchers: P. A. Smith 2635; Rodin 3922; Dieter-

len 889; Moll 1424; R. A. Lubke 181.

Bothalia 15, 3 & 4 (1985)

6. - Scleria sobolifer. 1-5 from Ward 5218; 6,7 from Ward 8851. 1, habit, x 0,9; 2, portion of teal ‘ a”d ^

lamina and sheath, x 8,5; 3, plan of T/S mid-lamina, x 29,8; 4, androgynaeceous spikelet, x 13,6 5, androgynaeceous

spikelet with lower glumes removed to reveal achene , x 13,6; 6, achene, x 25 ,5 ; 7 , achene surface pattern from SEM, x 213.

Bothalia 15, 3 & 4 (1985)

521

8. Scleria sobolifer E. F. Franklin in Kew Bull.

38: 33 (1983). Type: South Africa, Natal, Ward 5128

(K, holo.!; PRE!; NH!; NU!; UD-W!).

Perennial. Rhizome none. Soboles trigonous,

1-1,5 mm thick, hard, whitish with wine-red

blotches; scales purple-red. Culms 0,18-1,01 m tall,

glabrous or glabrescent. Leaves 1,1-2, 6 mm broad,

glabrous adaxially, sparsely hirsute abaxially; mouth

of sheath truncate, hirsute. Inflorescence un-

branched, 20-65 mm long. Glomerules 2-6, of 2-6

sessile spikelets. Bracts glabrous, the margins

sparsely hirsute, shortly awned, pale stramineous

with red striae. Spikelets c.4 mm long. Glumes

1, 7-2,6 mm long, glabrous or sparsely hirsute or

midrib stramineous with red striae. Achene subglo-

bose, 1,5-1, 8 x 1,2 mm, glabrous, undulate-tuber-

culate, with a series of deep horizontal troughs sep-

arated by horizontal and vertical ridges at junction

of stipe and body, grey. (Fig. 6).

Perennial with annual aerial parts known only

from coastbelt of Natal in open, sandy, seasonally

damp areas (Fig. 7).

This perennial species lacks a true rhizome and

has elongate, subterranean propagative stems (so-

boles) with collateral vascular bundles. In its aerial

parts it resembles some species of S. dregeana in

having unbranched glomerate-spicate inflores-

cences. The achenes are tuberculate and differ from

the tuberculate achenes of some specimens of S. dre-

geana in that the distal tubercles are no more promi-

nent than the proximal ones, and in having a pattern

of ridges, troughs and fine transverse bars between

the stipe and body of the achene (Fig. 6,6). At high

magnification it is possible to discern a pattern of

fine raised, more-or-less horizontal ridges (Fig. 6,7)

on the crests of the tubercles created by the proud-

standing, silicified radial walls of groups of horizon-

tally elongated epidermal cells. Such a pattern is ab-

sent from the crests of the tubercles of the achene of

S. dregeana.

Vouchers: Strey 5136; Michelmore 44; Arnold 467;

Baijnath 126; Nicholson 1141.

9. Scleria pergracilis (Nees) Kunth, Enum. PI.

2: 354 (1837); C.B.C1. in FTA 8: 495 (1902); Hutch.

& Dalz., FWTA 2: 491 (1936); Pierart in Lejeunia

13: 20, t.4, figs 1,2 (1953); Nelmes in Kew Bull. 10:

445 (1955); Robinson in Kew Bull. 18: 494-5 (1966).

Type: India, Wallich 3406 (K, holo., as Hypoporum

pergracile Nees).

Hypoporum pergracile Nees in Edinb. New Phil. Journ. 17:267

(1837). Type: as above.

Scleria pergracilis var. brachystachys Nelmes in Kew Bull. 10:

446 (1955); Napper in J1 E. Africa nat. Hist. Soc. 24: 31 (1964);

Robinson in Kew Bull. 18: 494 (1966). Type: Zimbabwe, Brain

3710 (K, holo.!; PRE; SRGH).

Annual, caespitose. Culms 0,13-0,38 m tall, gla-

brous. Leaves 1-2 mm broad, glabrous or the

sheaths minutely pilose towards the truncate or con-

cave mouths. Inflorescence unbranched, 20-65 mm

long, or with one short basal branch bearing a single

glomerule. Glomerules 2—13 — (18) of 1-7 sessile or

subsessile spikelets. Bracts acuminate, shorter than

or slightly exceeding the spikelets. Spikelets 4-5 mm

FIG. 7. — Distribution map of Scleria sobolifer A; S. pergracilis

var. brachystachys # and S. dieterlenii ■

long. Glumes 2, 5-3, 5 mm long, glabrous, mucronu-

late, pale with red-brown striae, or wholly reddish-

brown. Achene subglobose, 1,4— 1,8 x 1,2-1, 6 mm,

glabrous, trabeculate to tuberculate-verrucose, grey

possibly with brown stipe (only immature achenes

seen).

Open seasonal bogs in neutral or acid soil. This

species is known in Southern Africa from a single

gathering (Pentz & Acocks 10277) made near Dun-

dee in Natal in 1944 (Fig. 7). The gathering was

identified as var. brachystachys by Nelmes in 1955.

Robinson (1966) expressed reservations about the

validity of maintaining var. brachystachys. I am un-

able to express an opinion since the species is un-

known to me in the field and I have seen very little

material. The species is widely distributed in Africa,

India, Sri Lanka and New Guinea and var.

brachystachys is recorded outside the FSA area from

Zimbabwe and Tanzania.

This annual species may be distinguished from the

perennial S. dieterlenii which it resembles in its aerial

parts, by the lack of a rhizome or any other peren-

nating stem, by its broadly V-shaped lamina-profile

(the lamina-profile of S. dieterlenii is flanged-V-

shaped); by its glabrous glumes and by its grey,

strongly tuberculate-verrucose achenes with the

crests of the tubercles lighter grey.

Voucher: Pentz & Acocks 10277 (PRE; NH).

10. Scleria dieterlenii Turrill in Bull. Misc. Inf.

Kew 1914: 20 (1914); Nelmes in Kew Bull. 10: 441

(1955); Napper in Kew Bull. 25: 443 (1971). Type:

Lesotho, Dieterlen 749 (K, holo.!; PRE!; NH!).

S. schweinfurthiana sensu Hutch. & Dalz., FWTA 2: 491 (1936)

pro parte, non Boeck. (1879).

S. flexuosa sensu Robinson in Kew Bull. 18: 505 (1966), pro

parte, non Boeck. (1888).

Perennial. Rhizome curved-descending, 0,5-1 mm

thick with terminal swollen tuber up to 10 x 4 mm;

scales pale, reddish-striate. Culms clustered,

0,15-0,37-(0,45) m tall, glabrous or sparsely villous.

Leaves 1-2,5 mm broad, glabrous or sparsely villous;

sheaths with truncate or concave, villous mouths. In-

522

Bothalia 15, 3 & 4 (1985)

florescence unbranched, rarely with one basal

branch, 20-80 mm long. Glomerules 3-8 of 1-8 ses-

sile spikelets. Bracts awned, shorter than or exceed-

ing the glomerules, the lowermost subfoliaceous,

25-40 mm long. Spikelets 4—5 mm long. Glumes

2,5-4 ,5 mm long, sparsely or densely villous, shortly

awned, orange-brown with red striae. Achene sub-

globose, 1,3-1 ,8 x 1-1,25 mm, glabrous, trabecu-

late-verrucose, grey with the trabeculae light red-

dish-gold.

Perennial with annual aerial parts, occurring in

seasonal bogs at high elevations in eastern Cape, Na-

tal, Transvaal and Lesotho (Fig. 7) and to the north

at high elevations in Zimbabwe, Zambia, Angola,

Guinea, Sierra Leone and the Ivory Coast.

This species differs from S. pergracilis in having a

slender, descending, subvertical rhizome with a

swollen, tuberous tip; in having a flanged-V-shaped

lamina-profile; in having hairy glumes, and in having

grey, trabeculate-verrucose achenes with the crests

of the trabeculae light to bright reddish-brown.

Vouchers: H. G. Breyer 18070; Bews 471; Killick

& Vahrmeijer 3663; Hoener 2014 (NU; UD-W);

Acocks 2196.

11. Scleria nutans Wild, ex Kunth, Enum. PI. 2:

351 (1837); Robinson in Kirkia 4: 179 (1964), Kew

Bull. 18: 502 (1966); Gordon-Gray in Ross, FI. Natal

112 (1972). Type: Venezuela, Cumana, Humboldt

s.n. (B-W 17336, holo.).

Hypoporum nutans (Willd. ex Kunth) Nees in Mart., FI. Bras.

2: 170 (1842).

Hypoporum humile Nees in Linnaea 9: 303 (1834) nomen nu-

dum.

Scleria mollis Kunth, l.c.: 352 (1837). Type: Brazil, Sellow

s.n. (K, holo.!).

S. cenchroides Kunth l.c.: 352 (1837). Type: South Africa,

between Umtentu and Umzimkulu Rivers below 500 ft, Drege

s.n. (Bt; K, lecto.l).

5. bojeri C.B.C1. in Dur. & Schinz, Consp. FI. Afr. 5: 669

(1895) nomen nudum.

S. hirtella Sw. var. f uberculata C.B.C1. in FC 8: 294 (1898);

Schonl. in Mem. bot. Surv. S. Afr. 3: 64 fig. 73 (1922). Type:

South Africa, Burke 62 (K, holo.!).

S. hirtella Sw. var. chondrocarpa Nelmes in Kew Bull. 10: 451

(1955). Type: Uganda, A. S. Thomas 95 (K, holo.!). [5. hirtella

auctores permulti, non Sw.]

Perennial. Rhizome 2-A mm thick, woody: scales

brown. Culms 0,16-0,5 m tall, glabrous or the angles

sparsely hirsute, 2-15 mm distant; bases sometimes

thickened to 7 mm. Leaves 1,5-5 mm broad, gla-

brous to villous, mouths of sheaths truncate or con-

vex, villous. Inflorescence unbranched, 25-85 mm

long. Glomerules reflexed, 3-6-(7), of 2-7 sessile or

subsessile spikelets. Bracts densely pale-, reddish- or

blackish-ciliate, awned, 4-12 mm long. Spikelets 4-5

mm long. Glumes 2-5 mm long, densely pale-, red-

dish- or blackish-ciliate, awned, pale with red

streaks. Achene ovoid to subglobose, 1,2-1 ,5 x 1

mm, glabrous, smooth or lightly tuberculate, grey or

grey-brown.

Perennial with annual aerial parts occurring in

open, permanently boggy areas in the summer rain-

fall areas of southern Africa (Fig. 8) to the north in

Zambia, Tanzania, Uganda, Zaire, Gabon and Ni-

geria, also in Madagascar and in South America, in

Brazil and Venezuela.

The features which most easily distinguish this

species from S. aterrima with which it has often been

confused, are its woody, often long, horizontal rhi-

zome; the spacing of its leaves more-or-less evenly

along the length of the culm; by the vestiture of the

glumes of some specimens being pale or reddish (not

invariably blackish) and by the absence of hypoder-

mal translucent cells in the intercostal region of the

lamina.

Vouchers: Killick & Strey 2506; C. Reid 423; Ward

5057; K. D. Huntley 705 (K; NU); Arnold 796.

12. Scleria aterrima (Ridl.) Napper in Kew Bull.

25: 145 (1971). Type: Angola, Welwitsch 7143 (BM,

holo.; K, iso.!).

S. hirtella Sw. sensu Boeck. in Linnaea 38: 439 (1874) excl.

synon., quoad Barter 1561.

S. hirtella Sw. var. aterrima Ridl. in Trans. Linn. Soc., ser. 2,

Bot. 2: 166 (1884). Type: Angola, Welwitsch 7143 (BM, holo.; K

iso.!).

S. catophylla C.B.C1. in Dur. & Schinz, Consp. FI. Afr. 5: 670

(1895) pro parte excl. synon. Amer., et in FC 7: 294 (1898), et in

FTA 8: 498 (1902); Hutch, in FWTA 2: 491 (1936); Robinson in

Kirkia 4: 182 (1964), et in Kew Bull. 18: 501 (1966); Berhaut, FI.

Sen., edn 2: 356 (1967). Type: Nigeria, Barter 1561 (K, holo.!).

Perennial with two different kinds of propagative

stems. Rhizome little more than the connective be-

tween ± contiguous culm bases; scales castaneous.

Soboles sometimes produced from culm-bases, soft,

terete, 10 mm long, with collateral vascular bundles.

Culms solitary or clustered, 0,25-0,70 m tall, villous.

Leaves 2-3 mm broad, usually villous, mostly

crowded towards base of culm; mouths of sheaths

truncate or concave, densely villous. Inflorescence

unbranched, 40-180 mm long. Glomerules 4—15, of

1-7 reflexed sessile spikelets. Bracts glumiform,

awned, densely black-hirsute, 6-7 mm long. Spike-

lets c. 6 mm long. Glumes densely blackish-hirsute,

2,5— 4,5 mm long. Achene broadly ovoid to globose,

1,2-1 ,5 mm x 1 mm, glabrous, smooth, grey or

grey-brown.

Bothalia 15, 3 & 4 (1985)

523

Perennial with annual aerial parts occurring in

open, perennially wet sandy areas. Known for FSA

area from Transkei, Natal and Transvaal (Fig. 9)

and widespread in east, central and west tropical

Africa. This species is often confused with S. nutans

from which it is distinguished by its soboles, its ab-

breviated rhizome, its basally congested leaves and

by the presence of bulliform translucent cells subja-

cent to the adaxial epidermis in the intercostal re-

gions.

Vouchers: Acocks & Naude 34; Schweickerdt

2344; Moll 4759; Nicholson 1103; Strey 10102.

FIG. 9. — Distribution map of Scleria aterrima ®; S. lacustris H;

S. transvaalensis A ; .S', aterrima and S. transvaalensis ®

13. Scleria lacustris Wright in Sauvalle in Anal.

Cienc. Habana 8: 152 (1871); Nelmes in Kew Bull.

10: 422 (1955); Robinson in Kew Bull. 18: 517, 519

(1966). Type: Cuba, Wright s.n. (K, holo.!).

5. aquatica Cherm, in Bull. Soc. Bot. Fr. 77: 279 (1930); Pierart

in Lejeunia, Mem. 13: 33, t.l, figs 20-23 (1953). Type: Gabon, Le

Testu 5845 (P, holo.!).

Annual (?), aquatic with stout basal adventitious

roots and finely branched floating roots at one-seve-

ral submerged nodes of culm. Rhizome none. Culm

solitary, up to 1,8 m tall, 7-12 mm thick, with angles

spinulose. Leaves 10-15 mm broad, scabrid on mar-

gins and major ribs; mouth of sheath produced into

an oval membranous tongue 5-10 mm long with a

narrow, blackish, hispid zone in the angles at its

base. Inflorescence branched paniculate, 0,19-1 m

long; bracts of primary axis foliaceous; higher order

bracts setaceous from auriculate base with few, stiff,

blackish hairs on auricles, 5-25 mm long. Spikelets

bisexual, female and male, 4-6 mm long. Glumes

glabrous, dark red-brown, mucronulate, 3,5 mm

long. Achene ovoid, 3-3,5 x 2-2,5 mm, smooth, gla-

brous, grey-brown: hypogynium small with three

very small triangular lobes, brown.

Aquatic annual known for FSA region only from

Okavango swamps in Botswana (Fig. 9) and else-

where in Africa from Zambia, Zaire, Sierra Leone,

Central African Republic and Gabon. Also from

Madagascar and from Cuba, French Guiana, Para-

guay and Brazil.

This species cannot be confused with any other

from the region, being distinguished by its finely-

divided floating roots, by its having androgynae-

ceous and subandrogynaeceous spikelets, and by its

very small trilobed hypogynium.

Vouchers: P. A. Smith 2718; 2796.

14. Scleria transvaalensis E. F. Franklin in Kew

Bull. 38: 35 (1983). Type: South Africa, Transvaal,

T. H. Arnold 336 (K, holo.!; PRE!).

Perennial. Rhizome little more than the connec-

tive between contiguous, woody, swollen culm-

bases; scales reddish-brown. Culms 0,5-1,25 m tall

with bases 8-10 mm thick; glabrous. Leaves 4—10

mm broad, subulate, glabrous; sheath with mouth

produced into a deltoid-rounded tongue 2-5 mm

long with a membranous extension up to 1 mm long.

Inflorescence with terminal panicle 45-90 x 20-45

mm and smaller lateral panicles 1-2 per node at 1-2

nodes exserted 20-120 mm from sheaths. Bracts foli-

aceous, exceeding their panicles, subulate. Male spi-

kelets sessile or shortly pedicillate, 4—6 mm long;

glumes reddish-brown, glabrescent. Female glumes

reddish-brown with green midrib excurrent into an

awn c. 1 mm long; 4—5 mm long. Achene oblong-sub-

globose, 2,5-3 x 2 mm, glabrous, tessellate-lacu-

nose to tuberculate-lacunose, brownish- white: hypo-

gynium obtusely trilobed, stramineous.

Perennial with annual aerial parts occurring in

seasonally damp, open or semi-shaded habitats at

temperate altitudes in Transvaal, Swaziland and Na-

tal (Fig. 9). Apparently endemic in FSA region.

This species has hitherto been confused with 5.

natalensis from which it is distinguished by its bul-

bously thickened culm-bases; its subulate leaf and

bract-apices 1-2 mm broad in the distal 10 mm; its

relatively dense inflorescences with 1-2 lateral pan-

icles per node; its bright terra-cotta red glumes and

its barrel-shaped achenes.

Vouchers: Obermeyer 3095; F. Venter 1150; R.P.

Ellis 2964; Gordon-Gray 6020 (NU) Seagrief 18

(NU).

15. Scleria natalensis C.B.Cl. in FC 7 : 296

(1898); Nelmes in Kew Bull. 11: 88 (1956), pro

parte, non C.B.Cl; Franklin in Kew Bull. 38: 35

(1983). Type: South Africa, Natal, Buchanan 352

(K, holo.!; NH!).

Perennial. Rhizome usually with very short inter-

nodes little more than the connective between al-

most-contiguous culm bases; scales reddish-brown.

Culms 0,5-0,85 m tall with bases 2-3(^4) mm thick;

glabrous. Leaves 4-11 mm broad, ensiform, gla-

brous; sheath with mouth produced into a deltoid-

rounded tongue 2-5 mm long with a membranous

extension 1 mm long. Inflorescence with terminal

panicle 25-65 x 15-20 mm and smaller lateral pan-

icles (1-) 2-4 per node at 2-3 nodes exserted 30-230

mm from sheaths. Bracts foliaceous, exceeding their

panicles, ensiform. Male spikelets sessile or pedicil-

late, 4—6 mm long; glumes stramineous with fine red-

dish striae, glabrescent. Female glumes stramineous

with reddish striae, with midrib excurrent into an

awn c. 1 mm long; 4—5 mm long. Achene subglobose,

Bothalia 15, 3 & 4 (1985)

10 — Scleria natalensis. 1,2 from Ward 4953; 3 from //ennessy 372; 4-6 from Ward 5043. 1, habit, X 0,6; 2, portion of leaf and

culm at junction of lamina and sheath, x 4,3; 3, plan of T/S half mid-lamina, x 30; 4, subandrogynaceous spikelet, x 8,5; 5,

subandrogynaeceous spikelet with lower glumes removed to reveal young achene with its hypogynium, and distal empty male

glumes, x 8,5; 6, male spikelet, x 10,2.

Bothalia 15, 3 & 4 (1985)

525

obtusely trigonous, 2,5 x 2 mm, glabrous, reticulate-

lacunose to tuberculate-lacunose, brownish- white:

hypogynium obtusely trilobed, stramineous. (Fig.

10).

Perennial with evergreen aerial parts occurring in

seasonally damp open or semi-shaded habitats at low

altitudes along coastbelt of Natal and Transkei (Fig.

11). Apparently endemic in FSA region.

This species is distinguished from S. transvaalensis

by its culm-bases not bulbously-thickened; its ensi-

form leaf and bract apices 3-4 mm broad in the distal

10 mm; its relatively less dense inflorescence with

1-4 lateral panicles per node; its dull straw-coloured

glumes with some faint red streaks near the margins

and its subglobose and obtusely trigonous achenes.

Vouchers: T. H. Arnold 435; Ward 4935; Strey

9464; Galpin 10988; Moss 5515 (K; J).

FIG. 11. — Distribution map of Selena natalensis.

16. Scleria foliosa Hochst. ex A. Rich., Tent. FI.

Abyss. 2: 509 (1851); C.B. Cl. in FTA 8: 503 (1902);

Pierart in Lejeunia, Mem 13: 40, t. 2, figs 4 & 5

(1953); Nelmes in Kew Bull. 11: 102 (1956); Robin-

son in Kirkia 2: 177 (1961) et in Kew Bull. 18: 525

(1966); Podlech in FSWA 165: 52 (1967). Syntypes:

Ethiopia, Schimper 1232 (K; BM, isosyn.!); Quartin

Dillon & Petit s.n. (K; P, isosyn.!).

S. dumicola Ridl. in Trans. Linn. Soc., ser. 2 Bot. 2: 169

(1884). Type: Angola, Welwitsch 7122 (BM, holo.!).

5. perrieri Cherm. in Bull. Soc. Bot. Fr. 70: 297 (1923). Type:

Madagascar, Perrier de la Bathie 12704 (P, holo.!).

Annual, without rhizome. Culms 0,20-0,60 m tall,

glabrous. Laminas 2-7 mm broad, glabrous, scabrid

on margins and ribs: sheaths glabrescent except near

mouths where villous; mouths truncate or with ligule

1-2 mm long. Inflorescence with terminal panicle

25-50 mm long, smaller lateral panicles single at 1-3

nodes, shortly exserted from sheaths. Bracts foli-

aceous, exceeding panicles. Male spikelets sessile or

shortly pedicillate, 4—5 mm long; glumes castaneous,

glabrous. Female glumes castaneous or dark brown

distally, paler proximally, with green midrib excur-

rent into an awn c. 1 mm long, glabrous. Achene

ovoid, 3-4 x 2-2,5 mm, alveolate lacunose proxi-

mally, smooth distally, grey: hypogynium obtusely

trilobed, stramineous.

Caespitose annual, occurring in seasonal shallow

pans in Transvaal, Swaziland, northern Namibia

(Fig. 12), widespread in east, central and west

Africa, also in Madagascar and India.

This species is distinguished from other species of

subgenus Scleria of the region by the absence of a

rhizome and by the type of surface-patterning of the

achene which is smooth distally, strongly alveolate-

lacunose proximally.

Vouchers: Junod 563; P. van Wyk 4786; M.

Muller & Giess 489; Schoenf elder 808; R. W. Haines

7043.

FIG. 12. — Distribution map of Scleria foliosa •; S. unguiculata

H; S. achtenii A and S. lagoensis *.

17. Scleria unguiculata E. A. Robinson in Kew

Bull. 18: 536 (1966). Type: Zambia, Robinson 5056

(K, holo.!).

Perennial. Rhizome usually with very short inter-

nodes little more than the connective between al-

most contiguous, slightly swollen culm bases; scales

pale brown. Culms 0,45-1,0 (-1,35) m tall, glabrous.

Leaves 1,4-5 mm broad, glabrous or sparsely pilose;

mouths of sheaths truncate or convex. Inflorescence

much elongated; terminal panicle c. 30 mm long;

shorter lateral panicles 2-4(-5) per node at 2-3

nodes on slender, unequally exserted drooping pe-

duncles. Bracts foliaceous. Male spikelets pedicillate,

4-5 mm long with pedicels l-8(-10) mm long;

glumes stramineous or castaneous with green keels,

glabrous. Female glumes similar with midrib excur-

rent into an awn c. 1 mm long; 3,5^1 mm long.

Achene ovoid or subglobose, 2,7 x 1, 7-1,9 mm.

hairy, lightly tessellate-lacunose, grey or light

brown, the hairs white or golden: hypogynium with

three unguiculate lobes clasping the achene, light

brown.

Perennial with evergreen aerial parts occurring in

perennially damp, open areas in northern Botswana

(Fig. 12), and recorded from Zambia, Tanzania, Su-

dan, Central African Republic and Togo Republic.

526

Bothalia 15, 3 & 4 (1985)

This species is easily confused with 5. nyasensis

C.B. CL, which has not been recorded from the FSA

region to date, from which it is distinguished by its

narrower leaves, by its fewer lateral panicles, by its

golden achene-hairs and by its three strongly ungui-

culate hypogynium lobes. Of the species recorded

from the FSA region it most closely resembles S.

achtenii from which it is distinguished by its nar-

rower leaves, by its more numerous lateral panicles,

by its distinctly tessellate-lacunose, more densely

hirsute, golden-haired achenes and by its three

strongly unguiculate hypogynium lobes.

Vouchers: P. A. Smith 1980; 1994; 2790; 2799.

18. Scleria achtenii De Wild, in Rev. Zool. Afr.

Suppl. Bot. 14: 16 (1926) et in PI. Bequaert. 4: 219

(1927); Pierart in Lejeunia 13: 27 (1951); Robinson

in Kew Bull. 18 : 534 (1966). Type: Congo, Achten

97 B (BR, holo.!).

S. substriatoalveolata De Wild, in Rev. Zool. Afr. Suppl. Bot.

14: 23 fig. 6 (1926) et in PI. Bequaert. 4: 240 (1927). Types:

Congo, Vanderyst 1060, 8190 and s.n. (BR. syn.!).

S. subintegriloba De Wild, in PI. Bequaert. 4: 238 (1927); as S.

achtenii var. subintegriloba (De Wild.) Pierart in Lejeunia 13: 47

(1951). Type: Congo, Vanderyst 2839 (BR, holo.!).

S. nyasensis sensu Nelmes, Kew Bull. 11: 86 (1956), pro parte,

non C.B. Cl.

Perennial. Rhizome little more than the connec-

tive between ± contiguous culm bases or elongate

with culms arising at intervals of up to 15 mm; scales

reddish-brown. Culms 0,7-1,10 m tall, glabrescent

or sparsely villous-hispidulous. Leaves 2,5-5 mm

broad, villous on abaxial ribs and sheaths, glabrous

adaxially and below mouths of sheaths; mouths con-

vex with a membranous margin 1 mm long, the mar-

gin base villous. Inflorescence with terminal panicle

up to 90 mm long, smaller lateral panicles single at

2-3 nodes exserted up to 180 mm from sheaths.

Bracts foliaceous. Male spikelets sessile or shortly

pedicillate, 7-9 mm long; glumes stramineous or

reddish-striate, glabrous. Female glumes similar,

shortly awned. Achene obovoid to subglobose, 2,5-3

x 1,8-2 mm, hairy, lightly and obscurely lacunose,

grey or brownish-grey, hairs white: hypogynium

small with 3 acuminate, acute, or bifid lobes, white.

Perennial occurring in open, perennially damp

habitats on coastbelt of Natal (Fig. 12) and in Mo-

zambique, Zambia and Zaire.

S. achtenii is distinguished from S. unguiculata by

its broader leaves, fewer lateral panicles, very faintly

reticulate-lacunose, sparsely white-hirsute achenes

and by its small trilobed hypogynium with the lobes

often 2-3-fid, not unguiculate.

Vouchers: Tinley 361; Ward 8887; E. A. Robinson

5523 (K); Ward 5508; Gordon-Gray 6215 (NU).

19. Scleria lagoensis Boeck. in Vidensk. Medd.

Dansk Naturh. Foren. Kbh. 1869: 151 (1869); Core

in Brittonia 2, 85 (1936); Robinson in Kew Bull. 18:

538 (1966). Type: Brazil. 2. iii. 1864, Warming s.n.

(C, lecto.!).

S. moritziana Boeck, in Linnaea 38: 460 (1874). Type: Vene-

zuela, Moritz 645 a (BM, iso.).

5. canaliculato-triquetra Boeck. in Flora 62: 573 (1879); Hutch,

in FWTA 2: 493 (1936); Pierart in Lejeunia, Mem. 13: 48 (1951).

Type: Sudan. Schweinfurth 2474 (K, iso.!).

S. diurensis Boeck. in Flora 62: 573 (1879). Type: Sudan,

Schweinfurth 2389 (K, iso.!; pro parte).

5. cervina Ridl. in Trans. Linn. Soc. ser. 2, Bot. 2: 171 (1884)

Type: Angola, Welwitsch 7127 (BM, holo.!).

5. mayottensis C.B. Cl. in Bull. misc. Inf. R. hot. Gdns, Addit.

Ser. Kew 8: 92 (1908). Type: Madagascar, Boivin 3043 (P,

holo.!).

S. vanderystii De Wild, in Rev. Zool. Afr. Suppl. Bot. 14: 25

(1926), et in PI. Bequaert. 4: 241 (1927). Types: Zaire, Vanderyst

3471 (BR. syn.) and 5 others.

5. canaliculato-triquetra var. clarkeana Pierart in Lejeunia,

Mem. 13: 49 (1951). Types: Zaire and Rwanda Burundi, Mullend-

ers 159, 672 (K, syn.) and 8 others.

Perennial. Rhizome little more than the connec-

tive between ± contiguous culm bases; scales light

brown. Culms 0,5-2 m tall, glabrous. Leaves 5-12

mm broad, glabrous, sometimes distally scaberu-

lous; sheaths often villous near mouths; mouths

truncate or produced into a short, membranous, gla-

brous or villous tongue. Inflorescence with terminal

panicle 30-120 mm long, smaller lateral panicles 1-3

at 2-3 nodes unequally exserted from sheaths. Bracts

foliaceous. Male spikelets sessile or subsessile, 4-6

mm long; glumes pale, reddish-brown striate, gla-

brous. Female glumes similar with midrib excurrent

into a scabrid awn c. 1 mm long; 4—5 mm long.

Achene ovoid to subglobose, 3-4 x 2-2,5 mm, hairy

proximally, glabrous distally, smooth or lightly

striate-lacunose, grey or grey-brown; hypogynium

with 3 narrowly lanceolate-acuminate lobes, creamy-

white.

Perennial occurring in open or semi-shaded, per-

ennially damp grassland. Known for FSA region

only from Swaziland (Fig. 12). Widespread in tropi-

cal Africa, in Madagascar and also in Brazil, Colom-

bia and Venezuela.

This species is distinguished from S. unguiculata

and S. achtenii, both of which have hirsute achenes,

by the absence of hairs from the top of the achene,

the absence of patterning on the achene surface or

its faintness and by the achene being ± globose in S.

lagoensis, ovoid in S. unguiculata and S. achtenii. It

is distinguished from S. adpresso-hirta (Kiik.) E.A.

Robinson by its glabrous female glumes in particu-

lar.

Voucher: Compton 29644.

20. Scleria melanomphala Kunth Enum. PI. 2:

345 (1837); C.B. Cl. in FC 7: 296 (1898) et in FTA 8:

506 (1902); Schonland, Bot. Surv. Mem. 3: 65, t .74

(1922); Cherm, in Arch. Bot. Caen 4, Mem. 7: 93

(1936); Hutch, in FWTA 2: 491, 493 (1936); Pierart

in Lejeunia Mem. 13: 26 t.l, fig. 26: 31 (1951);

Nelmes in Kew Bull. 11: 88 (1956); Robinson in Kew

Bull. 18: 546 (1966). Type: South Africa, Drege s.n.

sub C.B. Cl. 4369 (K, lecto.!; OXF).

5. macrantha Boeck. in Flora 62: 572 (1879) non S. macrantha

Boeck. (1859), nom. illegit. Type: Sudan, Schweinfurth 3746 (K,

iso.!).

S. longigluma Kiik. in Engl. Bot. Jahrb. 56, Beib. 125: 22

(1921). Types: Brazil, Vie 8066 (Bt).

S. centralis Cherm. in Arch. Bot. Caen. 4, Mem. 7: 50 (1931).

Types: Central African Republic, Le Testa 2436 (P, syn!). Tisse-

rant 1233 (P, syn.!).

Bothalia 15, 3 & 4 (1985)

527

Perennial. Rhizome elongate or contracted,

4— 5(-7) mm thick; scales reddish-brown. Culms

0,8-2 m tall, glabrous or glabrescent. Leaves

7— 15(— 20) mm broad, glabrescent or sparsely villous;

sheaths with mouths produced into a convex tongue

with pale, membranous margin. Inflorescence with

terminal ± spiciform panicle 30-90(-110) mm long,

smaller ± spiciform lateral panicles (0)-l-3 at

(0)— 1— 3 nodes, exserted up to 300 mm from sheaths.

Bracts foliaceous. Male spikelets sessile or subsessile,

10-12 mm long; glumes red-brown or blackish-red,

hispidulous or villous on midrib. Female glumes red-

brown or blackish-red with green, hispidulous or vil-

lous keels, shortly awned. Achene ovoid, 4—5,25 mm

x 2, 5-3,5 mm, glabrous or proximally sparsely

hairy, smooth, grey or brown proximally, blackish

distally; hypogynium obscurely trilobed or zoniform,

pale brown.

Perennial with evergreen aerial parts occurring in

open, perennially wet habitats in Transkei, Natal,

Transvaal, Swaziland and Botswana (Fig. 13). Wide-

spread in tropical Africa, in Madagascar and also in

Argentina, Brazil and Paraguay.

This species is distinguished from 5. greigiifolia by

the even spacing of leaves along the culm, by its very

compact, spiciform panicles, by its adaxially gla-

brous female glumes, by its beakless achene with a

black apex. Both species have brown, zoniform hy-

pogynia.

Vouchers: P. A. Smith 2789; Hemm 556; Comp-

ton 27369; Strey 8304; Ward 8837.

21. Scleria poiformis Retz., Obs. 4: 13 (1786);

Willd. Sp. PI. 4: 316 (1805); Nees in Wight, Contr.

118 (1834); Kunth, En. PI. 2: 358 (1837); Steud. Syn.

2: 179 (1855); Fischer in Kew Bull. 5: 265 (1931);

Blake in Proc. R. Soc. Queensl. 62: 89 (1952) et in J.

Arn. Arb. 35: 231 (1954); Nelmes in Kew Bull. 11:

110 (1956); Kern in Blumea 11: 178 (1961); Robin-

son in Kew Bull. 18: 547 (1966). Type: India Or.,

Koenig s.n. (LD, holo.; LZ; K, iso.!)

S. oryzoides Presl, Rel. Haenk. 1: 201 (1828); Nees in Wight,

Contr. 116 (1834); Kunth, En. PI. 2: 356 (1837); Steud. Syn. 2:

169 (1855); Miq. in FI. Ind. Bot. 3: 342 (1856); Thwait. in FI.

Austr. 7: 432 (1878); C.B.C1. in Hook. f. FI. Brit. Ind. 6: 691

(1894) et in J. Linn. Soc., Bot. 34: 101 (1898) et in FT A 8: 505

(1902) et in Philip. J. Sc. Bot. 2: 105 (1907); Ridl. in FI. Mai. Pen.

5: 177 (1925); Cherm. in Arch. Bot. Caen 7: 94 (1936); Van Stee-

nis in Bull. Jard. Bot. Btzg. 3: 399 (1948); Type: Philippines,

Haenke s.n. (PR, holo.; K, iso.!).

Perennial. Rhizome elongate, 5-17 mm thick;

scales brown. Culms 1 ,3—1 ,8(— 2) m tall, glabrous.

Leaves mostly crowded towards base of culm, 20-40

mm broad, bases sometimes corky, up to 5 mm

thick, glabrous; sheaths usually split almost to the

base from concave mouths. Inflorescence a single

terminal panicle without (very rarely with) foli-

aceous bract, 100-200 x 50-120 mm. Male spikelets

sessile, 3, 5-4, 5 mm long; glumes reddish-brown, gla-

brous or hispidulous. Female glumes similar, 3,5-5

mm long. Achene broadly ovoid to subglobose,

3-3,5 x 2, 5-2, 8 mm, glabrous, smooth, grey-brown;

hypogynium small, trilobed, lobes short, broadly tri-

angular, white or finely red-brown striate.

Perennial, stout, aquatic, with evergreen aerial

parts, occurring in open shallow lakes in coastal pans

in northern Natal (Fig. 14) and Mozambique. Also

on Zanzibar, on Mafia Island and in Madagascar

where it is said by Chermezon to be an introduction,

and in India, Malaysia, Thailand, Philippines and

northern Australia.

This species is unlikely to be confused with any

other. It is very robust, forms dense, almost pure

stands in shallow coastal pans, has leaves which are

broader and thicker than those of any other local

species and is the only southern African species

other than 5. lacustris with a very small hypogynium

with three obtusely triangular lobes. The two species

are allopatric and differ in habit.

Vouchers: Strey 8199; Ward 4024; E. F. Hennessy

373 (UD-W); R. H. Taylor 218 (NU).

22. Scleria greigiifolia (Ridl.) C.B.Cl. in FTA 8:

509 (1902); Kern in Blumea 12: 41^4, fig. 1 (1963);

Robinson in Kew Bull. 18: 546 (1966). Type: Angola

Welwitsch 6959 (BM, lecto.!).

Acriulus greigifolius Ridl. in J. Linn. Soc., Bot. 20: 336 (1883)

et in Trans. Linn. Soc. 2: 166, t.22, figs 1-5 (1884); C.B.Cl. in

FIG. 14. — Distribution map of Scleria poiformis O and S. greigii-

folia ▲ .

Dur. & Schinz, Consp. FI. Afr. 5: 675 (1895); Rendle, Cat. Afr.

PI. Welw. 2: 132 (1899). Type: Angola, Welwitsch 6959 (BM,

holo.!).

A. madagascariensis Ridl. in J. Linn. Soc., Bot. 20: 336 (1883)

et in Trans. Linn. Soc. 2: 2: 166, t.22, fig. 6-7 (1884); C.B.C1. in

Dur. & Schinz, Consp. FI. Afr. 5: 676 (1895); K. Schum, in

Engler, Pfl. Ost-Afr. C: 128 (1895); Cherm. in Mem. Acad. Mal-

gache 10: 11 & 43 (1931) et in Arch. Bot. Caen 7: 101 (1936) et in

Humbert, FI. Madag. 29e fam. 266, fig. 27, 1-2 (1937). Scleria

acriulus C.B.C1. in FTA 8: 509 (1902); De Wild., Bull. Jard. Bot.

Brux. 5: 143 (1916); De Wild., l.c. 7: 5 (1920). Types: Madagas-

car, Baron 1870 (K, syn!); Hildebrandt 3571 (K, syn.!).

A. titan C.B.C1. in Bull. misc. Inf. R. bot. Gdns., Addit. Ser.

Kew 8: 62 (1908) Type: Congo Gentil s.n. (BR, holo.!).

Scleria acriulus C.B.C1. f. leopoldiana C.B.C1. ex De Wild.,

Etude F. Bas-et Moyen Congo 1: 221 (1906). Type: Congo, Gillet

2818 (BR, holo.).

S. friesii Kiik. in Wiss. Ergebn. Schwed. Rhodesia-Kongo

Exped., 1911-12, 1: 9 (1921). Type: Zambia, Fries 743 (UPS,

holo.; K, iso.!).

Perennial. Rhizome elongate or contracted, 6-10

mm thick; scales brown or reddish-brown. Culms:

1-2 m tall, glabrous. Leaves mostly crowded towards

base of culm, 5-12 mm broad, tapering smoothly to-

wards apex or, rarely, unequally laterally prae-

morse; glabrescent or hispidulous; sheaths with con-

cave mouths, the margins minutely ciliate. Inflores-

cence with lax, copious terminal panicle c. 150 mm

long and smaller lateral panicles 4—7 per node at 2-3

nodes exserted up to 200 mm from sheaths. Bracts

foliaceous. Male spikelets sessile - shortly pedicil-

late, c. 5 mm long; glumes reddish-black proximally,

castaneous distally; midrib hispidulous, margins se-

tulose-ciliate. Female glumes similarly coloured,

shortly awned, midrib and awn hispidulous, margins

setulose-ciliate and distal half of adaxial surface with

dense indumentum of stiff, upward-pointing hairs.

Achene broadly ovoid, laterally compressed,

strongly beaked, 6 x 3,5 mm; glabrous, smooth,

pale brown or pinkish-brown sometimes with violet

blotches; hypogynium zoniform, brown.

Perennial, stout, aquatic with evergreen aerial

parts occurring in open permanent bogs or shallow

lakes. Known for the FSA area from southern Natal

coastbelt and from Eastern Shore, Lake St Lucia in

northern Natal (Fig. 14). Also in tropical Africa and

Madagascar.

This species is distinguished from S. melanom-

phala by its leaves crowded towards the base of the

culm, by its lax, copiously branched panicles, by its

adaxially densely-hirsute female glumes and by its

strongly-beaked achene which lacks a black apex.

Vouchers: P. G. Stewart 293; Moss 19166 (J); K.

D. Huntley 781 (NH; NU); Strey 7251; H. B. Nichol-

son 1598.

23. Scleria angusta Nees ex Kunth, Enum. PI. 2:

346 (1837); C.B.C1. in FC 7: 296 (1898); Cherm. in

Arch. Bot. Caen 7, Mem. 2: 97 (1936); Nelmes in

Kew Bull. 11: 73 (1956); Robinson in Kew Bull. 18:

548 (1966). Type: South Africa, Drege s.n. sub.

C.B.Cl. 4246 (Bt; K, lecto.!).

Perennial. Rhizome elongated, 4-7 mm thick;

scales brown. Culms 1, 3-2,5 m tall, glabrous below,

hirsute above. Leaves unequally laterally praemorse

distally, 6-16 mm broad, glabrous or sheaths densely

villous below mouths; mouth produced into deltoid-

rounded tongue 2-5 mm long. Inflorescence with

compact terminal panicle 35-60 mm long, lateral

panicles erect, single at 3-7 nodes, 30-40(-60) mm

long, shortly exserted from sheaths. Bracts foli-

aceous. Male spikelets sessile or subsessile, 3,5-4

mm long, glumes pale brown with reddish striae,

keels and distal margins hispidulous. Female glumes

glabrescent or hispidulous on keels and distal mar-

gins, 3-3,5 mm long, pale brown with darker reddish

streaks. Achene ovoid to ovoid-globose, 2,25-3,5 x

1,8-2, 3 mm, glabrous, smooth, violet or purple: hy-

pogynium obscurely trilobed with fimbriate margin,

brown.

Perennial, stout, aquatic with evergreen aerial

parts occurring in shade in swamp forest in coastal

Natal and Transkei (Fig. 15). Also recorded from

Mozambique and Madagascar.

This species is distinguished from all others of the

FSA region by its praemorse leaves, its violet or pur-

ple achene and its fimbriate hypogynium.

Vouchers: Tinley 254; Edwards 2559; Ward 8083;

Strey 9905; Strey 11306.

UITTREKSEL

Die 23 spesies van Scleria (Sclerieae, Caricoideae,

Cyperaceae) in suidelike Afrika word hersien. Twee

subgenera word erken, Hypoporum met een seksie,

Hypoporum, en Scleria met drie seksies, Scleria, Ac-

riulus en Schizolepis. Die tribus Sclerieae word van

die tribus Bisboeckelereae onderskei.

REFERENCES

BENTHAM, G., 1883. Cyperaceae. In G. Bentham, & J. D.

Hooker, Genera Plantarum III: 1037-1073.

BLASER, H. J., 1944. Studies in the morphology of Cyperaceae

II. The Prophyll. Am. J. Bot. 31: 53-64.

CLARKE, C. B., 1894. Scleria. In. J. D. Hooker, The Flora of

British India VI: 685-694.

CLARKE, C. B., 1908. New genera and species of Cyperaceae.

Bull. misc. Inf. R. bot. Gdns Kew, Addit. Ser. 8: 1-196.

CORE, E., 1936. The American species of Scleria. Brittonia 2,1:

1-103.

EITEN, L. T., 1976. Inflorescence units in the Cyperaceae. Ann.

Mo. bot. Gdn 63: 81-112.

FRANKLIN, E. F., 1979. A note on the hairy achenes of four

African species of Scleria Bergius (Cyperaceae). Bot. J.

Linn. Soc. 79: 333-341.

HOLTTUM, R. E., 1948. The spikelet in Cyperaceae. Bot. Rev.

14: 525-541.

HOOPER, S. S., 1971. In C. R. Metcalfe, Anatomy of the Mono-

cotyledons V — Cyperaceae; 44-46. Oxford: Clarendon

Press.

HUTCHINSON, J., 1959. The families of flowering plants II —

Monocotyledons, edn 2: 704-710. Oxford: Clarendon

Press.

KERN, J. H., 1961. Florae Malesianae Precursores 30. The genus

Scleria in Malaysia. Blumea 11: 140-218.

KERN, J. H., 1963. Acriulus once more reduced. Blumea 12:

41^14.

KERN, J. H., 1974. Scleria. In C. G. G. van Steenis, Flora Male-

siana 7,3: 722-751.

KOYAMA, T., 1961. Classification of the family Cyperaceae (1).

J. Fac. Sci. Tokyo Univ., Bot. 8: 37-148.

KOYAMA, T., 1967. Cyperaceae — Mapanioideae. Mem. N. Y.

bot. Gdn 17: 23-79.

KOYAMA, T. , 1969. Delimitation and classification of the Cype-

raceae — Mapanioideae. In J. E. Gunkel, Current topics in

plant science: 201-208. New York: Academic Press.

NEES AB ESENBECK, C. G. D., 1834. Uebersicht der Cypera-

ceen — Gattungen. Linnaea 9: 273-304.

NEES AB ESENBECK, C. G. D., 1842. Cyperaceae. In C. F. P.

de Martius, Flora Brasiliensis 2,1: 182-190.

PAX, F., 1886. Beitrage zur Morphologie und Systematik der Cy-

peraceen. Bot. Jb. 7: 287-318.

PAX, F., 1887. Cyperaceae. In A. Engler & K. Prantl, Die nattir-

lichen Pflanzenfamilien 2,2: 98-126.

RAYMOND, M., 1966. Studies in the flora of Thailand. Cypera-

ceae. Dansk bot. Ark. 23: 311-374.

SCHULTZE-MOTEL, W., 1964. Reihe Cyperales. In H. Mel-

chior, Engler’s Syllabus der Pflanzenfamilien edn 12,2:

602-607.

APPENDIX : SPECIMENS STUDIED

The specimens are listed alphabetically according

to the name of the collector. The figures in parenth-

eses refer to the number of the taxon in the text. The

herbaria in which the specimens examined are

housed, are indicated by the letter codes of Holm-

gren & Keuken, Index Herbariorum (1974), except

that of the University of Durban- Westville, UD-W,

which is as yet unlisted.

Achten 97 A,B (18) BR. Acocks 2196 (10) K, PRE; 10007 (1)

PRE, NH; 10735 (5) NH; 10738 (7) PRE, NH; 10745 (5) PRE,

NH; 10758 (5) PRE, NH; 10850 (7) PRE, NH; 11304 (1) K, PRE;

18794 (5) PRE; 20882 (5) PRE; 21929 (1) K, PRE; 22048 (5)

PRE; 22049 (2) K, PRE; 22171 (2) PRE. Acocks & Naude 34 (12)

PRE; 76 (1) PRE. Arnold 336 (14) K, PRE; 435 (15) K, PRE; 467

(8) PRE; 796 (11) PRE.

Baijnath 126 (8) PRE, NU, UD-W. Baron 1870 (22) K. Barter

1561 (12) K. Baur 311 (7) K; 759 (5) K. F. Bayer s.n. (20) NU.

Bews 471 (10) NU. Biegel & Russell 3889 (7) PRE, NU. Bingham

s.n. (5) NU. Boivin 3043 (19) P. Bolus 8274 (1) PRE, BOL. Brain

3710 (9) K; 8788 (16) K. Breyer 18070 (10) PRE; 24236 (5) PRE.

Buchanan 3 (7) K; 32 (5) K; 36 (7) K; 349 (7) K; 351 (20) K; 352

(15) K. 1272 (5) K. Burchell 2463 (7) K. Burke 62 (11) K. Burtt-

Davy 767 (5) PRE.

Chandler 1335 (22) K. Chiparawasha 349 (22) NU. Clark 334 (5)

NU. Coetzee 499 (5) PRE. Coleman 691 (5) PRE, NH. Collins 26

(5) PRE. Cooper 3365 (7) K. Compere 1534 (22) K. Compton

24985 (14) PRE; 27364 (5) PRE; 27369 (20) PRE; 27790 (5) PRE;

29644 (19) K, PRE; 30964 (2) PRE; 31805 (14) PRE; 31872 (2)

PRE; 32221 (5) PRE; 32441 (5) PRE.

Davidson & Mogg 32901 (14) PRE. Devenish 970 (7) K, PRE.

Devred 1287 (11) NU; 1901 (22) K. De Winter 3915 (4) K, PRE.

De Winter & Marais 5049 (3) K, PRE. Dieterlen 749 (10) K, PRE,

NH; 776b (1) PRE; 889 (7) PRE. Dinter s.n. (16) K. Doidge &

Bottomley s.n. (7) PRE. Downing 233 (7) PRE. Drege s.n. (3934

of C. B. Clarke) (7)K;s.n. (4246 of C.B.C1.) (23) K; s.n. (4369 of

C.B.C1.) (20) is; s.n. (4381 of C.B.C1.) (7) K. Du Plessis 880 (2)

PRE. Du Toil 257 (16) PRE.

Edwards 1127 (2) K, PRE, NU; 2405 (1) PRE, NU; 2559 (23)

PRE, NU; 2563 (8) K. Ellis 2964 (14) PRE; 3279 (5) PRE.

Feeley & Ward 7 (11) PRE, NU. Flanagan 954 (1) PRE, BOL;

988 (20) PRE, BOL; 1260 (7) K, BOL; 2035 (1) PRE; 2363 (11)

K, BOL. Forbes 803 (5) NH. Frankish 318 (1) NU. Fries 743 (22)

K.

Galpin 8581 (20) PRE, STE; 9104 (7) PRE; 10988 (15) NU.

Gentil s.n. (22) BR. Gilliland 25046 (1) PRE, J; s.n. (5) J. Gor-

don-Gray 2050 (1) NU, UD-W; 2060 (1) NU; 6020 (14) NU; 6096

(2) PRE, NU; 6182 (11) NU; 6188 (20) NU; 6210 (11) NU; 6215

(18) NU; 6217 (20) NU; 6245 (15) NU; 6251 (8) NU; s.n. (15)

NU.

Haenke s.n. (21) K. Haines 7009 (7) PRE; 7010 (7) PRE; 7031 (1)

PRE; 7043 (16) PRE. Hancock s.n. (7) J. Hancox s.n. (2) NU.

Hemm 556 (20) PRE, J; 563 (14) PRE, J. Hennessy 373 (21)

UD-W; 374 (21) UD-W; 406 (1) NU, UD-W; 407 (5) NU,

UD-W; 408 (2) NU, UD-W; 409 (1) NU, UD-W; 410 (15)

UD-W; 431 (1) UD-W; 433 (5) NH, NU, UD-W; 437 (5)

UD-W; 438 (1) UD-W;440 (5) UD-W. Hildebrandt 3571 (22) K.

Hilliard & Bum 1089 (2) NU; 13394 (5) NU; 13795 (7) NU; 13843

(2) NU. Hoener 2014 (10) NU, UD-W. Hooper & Townsend 553

(22) K. Hunt & Ramos 6067 (19) K. B. J. Huntley 696 (20) NU;

781 (22) NH, NU; 789 (20) NU; 898 (23) NU. K. D. Huntley 425

(7) K, NU; 567 (5) NU; 705 (11) K, NU; 781 (22) NH, NU; 910

(1) NU.

TO ns & Basel 129/82 (11) NU.

Jacobsen 1152 (7) PRE; 1181 (5) PRE. Jarman & Guy 404 (15)

NU. Jenkinson 8 (20) K. Johnston 17 (3) NU. Junod 563 (16)

PRE. Kafuli 88 (5) NU. Killick 249 (11) K, PRE, NU; 706 (5)

NU; 1063 (5) K, PRE, NH; 1084 (7) K, PRE, NH, NU; 1222 (1)

K, PRE, NH, NU; 1233 (2) K, PRE, NH, NU; 1254 (1) K, NH,

NU; 1579 (5) K, NU. Killick & Leistner 3218 (6) PRE. Killick &

Strey 2506 (11) PRE. Killick & Vahrmeijer 3663 (10) K, PRE;

3683 (1) K, PRE. Kluge 316 (5) PRE; 403 (1) PRE; 435 (20) PRE;

2017 (5) PRE. Koenig s.n. (21) K. Krauss 42 (20) K. Kuntze s.n.

(1) K.

Lawn 1213 (20) NH; 1665 (8) NH. Leendertz 1623 (5) PRE; 6213

(7) PRE; 16233 (5) PRE. Le Testu 2. 436 (20) P; 5845 (13) P.

Lieben 2312 (22) K; 2409 (11) NU. Lind 134 (22) K. Liskowski

10640 (22) K. Lubke 181 (7) PRE. Lukuesa 100 (22) K.

McAllister 99 (5) PRE. McCallum Webster s.n. (11) NU; s.n. (12)

NU; s.n. (20) NU; s.n. (22) NU. McClean 387 (20) PRE, NH.

Mason 30 (7) NH. Merxmuller & Giess 2081 (16) PRE. Meyer s.n.

(8) NU. Michelmore 9 (8) PRE; 10 (20) K, PRE; 44 (8) K, PRE.

Milne-Redhead & Taylor 9739 (4) K. Mitchell 23/26 (5) NU. Mogg

896 (1) PRE; 897 (5) PRE; 4318 (20) PRE, J; 5543 (5) PRE; 5637

(7) PRE; 16575 (5) PRE; 17070 (20) PRE; 36608 (14) J; A 7 (1)

K, PRE. Moll 255 (12) NH; 704 (2) PRE, NU; 1424 (7) PRE,

NU; 2188 (23) NH; 4534 (8) K, PRE, NU; 4534 A (18) NH; 4759

(12) K, PRE, NH, NU; 4778 (8) NH, NU. Moll & Strey 3907 (23)

PRE, NH. Monro 635 (5) BOL. Moss 1127 (1) J; 2358 (11) K, J;

4001 (20) K, J; 5515 (15) K, J; 7207 (5) K; 19166 (22) J; s.n. (1) K.

Moss & Rogers 452 (5) BOL. Mtombeni 32 (20) PRE. Muller 2031

(14) PRE. Muller & Giess 489 (16) K, PRE. My re & Carvalho

1147 (21) NU.

Nicholson 317 (20) NH; 1103 (12) PRE; 1141 (8) PRE; 1598 (22)

PRE; 15411 (5) PRE.

Obermeyer 3095 (14) PRE. Onderstall 388 (1) PRE.

Pegler 322 (20) BOL; 1421 (1) K; 1498 (1) K. Pellatt s.n. (1) J.

Pentz & Acocks 10277 (9) PRE, NH. Perrier de la Bdthie 12704

(16) P. Phelan 148 (5) NU. Phillips 3550 (14) K, PRE. Physick 78

(5) NU. Pole-Evans s.n. (12) PRE. Pooley 1967 (8) UD-W. Pott

15253 (5) PRE, BOL.

Quartin Dillon & Petit s.n. (5) K; s.n. (16) K.

Rehmann 5626 (3) K, BOL. Reid 423 (11) PRE. Repton 1113 (5)

PRE, NH, BOL; 3295 (7) PRE. Richards 12383 (5) NU. Roberts

3087 (1) PRE. Robinson 1102 (16) K; 1724 (11) NU; 1748 (12)

NU; 2228 (17) NU; 2267 (17) NU; 2269 (20) NU; 2270 (18) NU;

2870 (7) NU; 3027 (2) NU; 3368 (19) NU; 3405 (3) NU; 4220 (6)

K, NU; 4700 (13) NU; 5055 (12) NU; 5056 (17) K; 5523 (18) K;

5524 (8) NU. Rodin 3922 (7) K, PRE. Rogers s.n. (7) K, PRE, J.

Ruch 2034 (5) PRE. Rudatis 528 (5) STE; 736 (11) K, PRE, STE;

1083 (7) STE.

530

Bothalia 15, 3 & 4 (1985)

Sandwich s.n. (5) K. Schimper 327 (5) K; 1232 (16) K; 1557 (5) K.

Schlechter 3705 (7) K, PRE, BOL; 5532 (1) PRE; s.n. (12) Z,

PRE, BOL. Schoenfelder 808 (16) K, PRE. Schonland 4189 (20)

PRE. Schweickerdt 2189 (16) K, PRE, NU; 2344 (12) K, PRE,

NU. Schweinfurth 2193 (5) K; 2389 pro parte (19) K, P; 2474 (19)

K; 3746 (20) K. Seagrief 18 (14) NU; 23 (NU); 2286 (20) RUH;

3157 (5) RUH. Sellow s.n. (11) K& ex B. Siame 209 (22) NU; 289

(5) NU. Sim 197 (1) PRE; 2705 (1) NU. Simpson 16/60 (12) NU.

Skead 66 (7) NU. C. A. Smith 1341 (5) PRE. P. A. Smith 340 (20)

PRE; 341 (7) PRE; 1980 (17) PRE; 1994 (17) PRE; 2033 (1) PRE;

2635 (7) PRE; 2718 (13) PRE; 2789 (20) PRE; 2790 (17) PRE;

2796 (13) PRE; 2799 (17) PRE. Smook 1055 (7) PRE; 1058 (2)

PRE. Smuts & Gillet 3260 (14) PRE. Stewart 293 (22) PRE. Stir-

ton 5649 (11) PRE. Stohr 570 (20) BOL. Story 6467 (4) K, PRE.

Strey 4894 (11) K, PRE, NH; 5136 (8) PRE, NH; 5137 (21) NH;

5711 (8) K, PRE, NH, UD-W; 7035 (12) K. PRE, NH, UD-W;

7108 (23) K, PRE. NH, UD-W; 7251 (22) PRE, NH; 7705 (23)

PRE, NH; 8199 (21) PRE, NH, UD-W; 8304 (20) PRE, NH;

9464 (15) PRE, NH; 9905 (23) PRE, NH; 10102 (12) K, PRE,

NH; 10331 (15) NH; 11306(23) PRE,NH. Sutherland s.n. (20) K.

H. C. Taylor 2101 (1) NU, UD-W. R. H. Taylor 120 (23) NU;

142 (8) NU; 218 (21) NU. Tinlev 254 (23) PRE, NH. NU; 303 (20)

PRE, NU; 361 (18) PRE, NH, NU. Tisserant 1233 (20) P; 2922

(3) P. Thomas 95 (11) K. R. X. L. Thomas 1202 (5) BR. Thomp-

son 29 (11) NU. Tyson 1825 (5) PRE, BOL.

Unnamed Collector s.n. (21) PRE.

Van der Schijff 6384 (20) PRE. Vandervst 1060 (18) BR; 2839 (18)

BR; 8190 (18) BR; s.n. (18) BR. Van Wyk 4786 (16) PRE. F.

Venter 1150 (14) PRE. H. J. T. Venter 733 (15) NH; 842 (20) NH;

890 (8) NH; 4839 (11) PRE; s.n. (5) PRE. Verdick 398 (5) BR.

Vesey-Fitzgerald 230 (11) NU; 1007 (3) NU; 1258 (20) NU; 1333

(5) NU; 1353 (5) NU; 1447 (5) NU; 1507 (5) NU; 2260 (5) NU;

2345 (5) NU; 2488 (20) NU.

Wager s.n. (10) PRE. Ward 111 (23) NU; 723 (8) PRE, NU; 724

(20) NU; 1040 (20) NU; 1106 (11) NU; 1186 (23) NU; 2784 (8)

PRE, NH; 2856 (20) PRE, NH, NU; 2924 (12) NH, NU; 3648

(20) PRE, NU; 3712 (23) PRE, NU; 4024 (21) PRE, NH, NU;

4327 (20) PRE; 4716 (15) NH, UD-W; 4737 (8) NH, NU,

UD-W; 4935 (8) NU, UD-W; 4936 (20) NH, UD-W; 4937 (18)

UD-W; 4953 (15) PRE. UD-W; 5043 (15) NH, NU, UD-W;

5057 (11) PRE, NH. NU, UD-W; 5077 (20) K, PRE, NH, NU,

UD-W; 5128 (8) K, PRE, NH. NU, UD-W; 5172 (8) PRE, NH,

NU, UD-W; 5435 (8) K, PRE, NH, NU, UD-W; 5437 (18) NH,

NU, UD-W; 5470 (8) K, PRE, NH, NU, UD-W; 5508 (18) K,

PRE. NH. UD-W; 6338 (1) PRE. NH, NU, UD-W; 7158 (15)

PRE, UD-W; 7242 (8) PRE. UD-W; 7823 (12) UD-W; 7879 (18)

UD-W; 7911 (23) UD-W; 8083 (23) PRE; 8108 (8) UD-W; 8494

(1) PRE; 8738 (1) PRE; 8741 (15) PRE, UD-W; 8837 (20) PRE,

NH, NU; 8874 (20) NU, UD-W; 8886 (8) PRE, UD-W; 8887

(18) PRE, NH; 9071 (1) PRE; 9145 (8) PRE, UD-W; 9146 (18)

PRE. Warming s.n. (19) C. Weintraub s.n. (1) J. Welwitsch 6959

(22) BM; 7122 (16) BM; 7127 (19) BM; 7138 (2) BM; 7143 (12) K.

Werdmann & Oberdieck 1152 (5) K, PRE. Wild. 5467 (2) NU;

5468 (7) NU. Wilms 1586 (7) K; 1646 (5) K; 2327 (7) K. Wilson

s.n. (5) NU. Wood 1428 (12) K, NH; 1597 (20) NH, BOL; 3863

(23) K, NH, BOL; 3994 (1) K, NH, BOL; 4757 (1) K, NH; 12022

(8) NH. F. B. Wright 1946 (7) NU; 2063 (7) NU. Wright s.n. (13)

K.

Bothalia 15,3 & 4: 531-539 (1985)

Studies in the genus Riccia (Marchantiales) from southern Africa. 1.

Two new species of the section Pilifer: R. duthieae and R . alatospora

O. H. VOLK* and S. M. PEROLD**

Keywords: dorsal epithelium, Duthie, Riccia species, section Pilifer, sporangia, spores, thallus

ABSTRACT

Riccia duthieae Volk & Perold and R. alatospora Volk & Perold, two new species endemic to southern Africa,

are described. Both species belong to the section Pilifer Volk (1983), which is characterized by the dorsal epithe-

lium of the thalli consisting of loose pillars of colourless cells.

Riccia duthieae Volk & Perold, sp. nov. sectio-

nis Pilifer.

Dioica, perennis, in sicco subalbida, in tumido vi-

ridis, mediocris, gregaria vel semirosulata. From 1

ad 8 mm longa, ad 5 mm lata, 1,0 mm crassa, furcata

furcis ovato-ligulatis, emarginatis, ad apicem solum

sulcatis, in partibus adultae convexis, in sicco conca-

vis, marginibus subacutis, parvopapillata, costa sen-

sim producta plano-convexa, lateribus sub apice

recte, aliter oblique adscendentibus vel explana-

tibus; pro parte colorata. Squamae magnae, altitu-

dine frondis haud superantes, hyalinae, tenerae basi

interdum picti, cellulis parietibus rectis constructa.

Frondis sectio transversalis, superficies frondis strato

pilifero (epithelium) dense obtecta, pili conici, li-

beri, hyalini, 120-160-220 pm longi, basaliter

40-70-105 pm, apicaliter 25-40-60 pm lati, tri- (raro

quadro-) cellulati, infra cellulis auxiliaris unicellula-

ris interdum ornatis; stratum aeriferum (chloren-

chyma) 2-4/10 altitudine, lamellis suberectis, 8-10

cellulis altis, unicellulis crassis, canales aeriferi am-

plii quadro- ad octogoni formantes aedificatum. Os-

tiola mascula hyalina ad 0,5 mm longe prominentia.

Sporangia parum protuberantes, cum 400 ad 600

sporae parvae. Sporae 60-70 pm in diametro,

polarae, anguste crenato-alatae, cinnamoneae, sub-

triangulares, superficies exterior sporarum laxe re-

ticulatim ornata, 3-7 areolis grandibus, 10-75-25 pm

latae, cristae crassis, saepe reticulatis secundariis in-

completis vel ad papillae reductis, saepe ab areolae

minores circumdata; superficies interior distincte tri-

radiata, dense parvo-areolata vel lamellis ad lineas

interruptas papillisve reductis. Chromosomata n = 8

(Bornefeld).

Endemica Africae australis. Riccia parvo-areolata,

R. albovestita similis, sed diversus ab ornamentatio

sporarum et aedificationem epithelii.

TYPE. — Cape, 3224 (Graaff Reinet): Aber-

deen, next to road R57, 2 km north-east of junction

with R61, at shallow edges of vleis temporarily damp

or occasionally inundated (-AC), 1981.04.11, Volk

*Botanische Anstalten d. Univ. Wurzburg D 8700, Germany,

B.R.D.

**Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

81-273 (M, holo; PRE), associated with Marsilia

burchellii, Eragrostis homomala, Chloris virgata,

Crassula spp. and with thick layers of Cyanophy-

ceae, or on bare ground. Soil clayey loam, pH 6,9.

Dioecious, perennial, often in large, densely

crowded gregarious patches or in incomplete roset-

tes 1-4 cm across. Thallus arises from narrow rib-

bon-like base, becoming wider towards apex, simple

or furcate, segments 1-8 mm long, 0,5-5 mm broad,

about 1 mm thick (Figs 1.1-1. 5 & 2.1); mostly short

and broad, shape oblong-obcordate; emarginate at

apex, shortly sulcate; when dry (Fig. 1.2, 1.3) lateral

margins inflexed over sulcus, otherwise dorsal sur-

face plane to concave, greenish-white, felt-like,

without obvious rows of cells or openings to air ca-

nals, margins somewhat thickened; ventral surface

pale green, flanks and bases of ventral scales occa-

sionally flecked with shiny reddish black blotches

(reddish yellow in chloral hydrate), when fresh or

moistened, dorsal surface pale creamy green, vel-

vety, convex. Scales large and conspicuous, spread-

ing (Fig. 1.3) or closely adherent to flanks of thallus,

rounded, soft and hyaline, hardly reaching thallus

margin, except for apical scales, cells 5-6 sided with

thin straight walls, sizes up to 95 x 40 pm; at scale

margin cells smaller, nearly isodiametric (Fig. 1.8).

Cross section of thallus 1,5 to 3 (-4) times broader

than high, margins acute to rounded, flanks steep

near apex, otherwise sloping outwards, ventral sur-

face almost flat (Fig. 1.10); epithelium about 1/10,

assimilation tissue (chlorenchyma) 2-4/10; storage

tissue 4-5/10 the thickness of thallus; dorsal surface

densely covered by hair-like cell pillars (Table 1,

Fig. 1.9) consisting of loose, colourless cells of vary-

ing lengths 120-160-220 pm; each pillar with 3 (or

TABLE 1. — Riccia duthieae, measurements on cross section

(cultivated plants of Volk 81/273)

Breadth of thallus : 0,5-2, 4— 5 mm

Height (thickness) : 0,8-1, 1-1, 4 mm

Thickness of epithelium : 150-190-245 pm ca 2/10 thickness of

thallus

Thickness of chlorenchyma: 120-390-480 pm ca 3-4/10 thickness

of thallus

Thickness of storage tissue : 320-510-800 pm ca 4-5/10 thickness

of thallus

532

Bothalia 15, 3 & 4 (1985)

FIG 1 — Riccia duthieae ( O . H. Volk 81/273, M; PRE). Structure of the thallus. 1, fresh thalli; 2, 3, dry thalli; 4, 5, thalli in dense

colonies; 6, unicellular outgrowths at bases of pillars; 7, chromosomes; 8, ventral scale; 9, epithehum; 10, cross s^hons of

thallus at various distances from apex; 11, cell pillar tips from above; 12 air canals in horizontal sect on^ lS^etaded cross

section of thallus. (1-6, 8-12 by O. H. Volk; 7 by T. Bornefeld; 13 by R. Holcroft.) Scale bar 1, 2, 4, 5, 10 - 1 mm, 3 - 2 mm,

6, 8, 9, 11, 12 = 100 jum; 7 = 1 jmn; 13 = 500 /im.

Bothalia 15, 3 & 4 (1985)

533

very occasionally 4) thin-walled cells 25^-0-60 pm

long with shorter conical cells at blunt tip, width of

basal cells 40-70-105 pm (Table 2); at bases of pil-

lars occasional unicellular globular outgrowths

which reduce air spaces over pores (Fig. 1.6); viewed

from above pillar tips irregularly arranged (Fig.

1.11) ; assimilation tissue (chlorenchyma) consisting

of 8-10 cells in columns or plates 1-3 cells broad and

1 cell thick, enclosing 4-8-sided air canals (Fig.

1.12) , which slant slightly forwards and widen up-

wardly; storage tissue with longitudinally elongated,

thin-walled cells containing rich supplies of fatty oils

in perennating tips of thalli; epidermis of ventral sur-

face has flattened cells, bearing smooth and tubercu-

late rhizoids. Antheridia numerous with colourless

ostioles, up to 0,5 mm long; archegonia scattered in

the middle along length of thallus. Sporangia slightly

bulging, up to 0,7 mm in size, with red-brown neck

and colourless tip, 175 pm long, containing 400-600

spores each. Spores small (50-) 60-70 (-80), average

size 62 pm, colour cinnamon brown, opaque, shape

triangular-globular, polar, with wing 3-5 pm wide.

(Figs 2. 5-2. 7 & 3.1, 3.3); ornamentation on distal

(outer) face: in centre 3-5 large prominent areolae

measuring 10-15-25 pm across, these areolae often

with reduced blind-ending secondary ridges, ‘crack’

viz Arnell, and sometimes with only a central knob;

marginally surrounded by smaller areolae with

thicker, incomplete ridges; however, surrounding

ridges sometimes thin and complete (Figs 2. 5-2. 7 &

3.3, 3.4); at edge ridges often extend across narrow

wing (Fig. 2.7); wing finely and irregularly crenu-

late, slightly projecting at angles of spore, with notch

leading to pore (Fig. 3.5) in spore wall (Fig. 2.6);

facets of proximal (inner) face sharply delimited by

triradiate mark and covered with fine network (up to

5 pm) of low, toothed ridges, often only partly com-

plete or reduced to simple projections and stipplings

(Fig. 3.1, 3.2). Chromosome number n = 8 (Borne-

feld) (Fig. 1.7).

When covered in mud, or when growing in dense

colonies, thin, narrow, diverging and runner-like

branches, which ascend obliquely forwards are

formed. Branches either enlarge into broader thalli

or else become transformed into nearly globular per-

ennial bulbils as follows: proximal parts of thalli,

flanks and 2-4 cell layers of chlorenchyma under-

neath epithelium lose their contents, cell walls be-

come impregnated with red-black pigment and form

covering layer 100-150 pm thick (Fig. 1.4). On dry-

ing, thallus shrinks and scar-like pits are left behind

in soil surface.

This species is named in honour of Dr Augusta

Vera Duthie (1881-1963) who, at Stellenbosch, be-

came the first South African trained botanist to be

appointed as a university lecturer in this country

(Gunn & Codd, 1981). Her collections and publica-

tions with Garside (Duthie & Garside, 1936, 1939)

are very valuable contributions to the knowledge of

South African Riccias.

This new species belongs to the South African en-

demic section Pilifer Volk (1983) and bears a close

resemblance to R. albovestita Volk (1981) and R.

parvo-areolata Volk & Perold (1984) which, how-

ever, differ from R. duthieae by having 4- (or rarely

5-) celled epithelial pillars and different spore orna-

mentation.

TRANSVAAL. — 2629 (Bethal): 5 km W of Kriel (-AB),

March 1984, Perold 342 (PRE).

CAPE. — 3224 (Graaff-Reinet): Aberdeen (-AC), April 1981

Volk 81/272; Feb. 1984 Volk 84/646 (PRE, M). 3319 (Worcester):

Robertson (-DD), Duthie 5182 (BOL); Duthie 5193 (BOL). 3326

(Grahamstown): between Ulster & Mooi River (-BA), Dec. 1981

Smook 4036 (PRE).

In a letter dated 10 June 1929, Duthie sent ma-

terial of Riccia [AVD 5004 (BOL) = CH 1007

(PRE)] from Stellenbosch Flats, accompanied by

clear drawings, to T.R. Sim, stating that it could per-

haps be R. coronata Sim, or that it was close to it.

Sim however, thought differently and suggested in

his reply of 15 June 1929, that it was a new species

and that she should describe the plant as R. duthieae

Sim MSS (correspondence at PRE). This was not

done and a description was never published. How-

ever, in a pencilled note found with this specimen at

BOL, she named it R. alatospora’ (described be-

low). According to Duthie’s drawings and confirmed

by our investigations, AVD 5004 (BOL) = CH 1007

(PRE) differs from R. duthieae, Volk 81/273, by the

former being smaller and having inconspicuous ven-

tral scales and by its larger, very broadly winged (15

pm wide) spores, with otherwise somewhat similar

ornamentation (see below).

Arnell (1963: 21 & 22) mistakenly published the

above-mentioned drawings by Duthie under R. co-

ronata Sim (type specimen Sim 8730 had been lost),

where the thallus according to Sim’s description

(1926:9) has ‘one epidermal upper layer of similar

but hyaline cells,’ and not short 3-celled pillars as

depicted by Duthie in her notes on R. alatospora.

534

Bothalia 15. 3 & 4 (1985)

FIG. 2. — Riccia duthieae (O. H. Volk 81/273, M; PRE). Thallus, epithelium and spores. 1, thallus; 2-4,

terminal cells of the epithelium; 5, 6, distal face of spore; 7, distal face and margin of spore. (1-4 SEM

micrographs by O. H. Volk; 5, L.M. photograph by S. M. Perold; 6, 7, L.M. photographs by O. H. Volk.)

Scale bar = 50 /im.

Sim wrote further that R. coronata has ‘near the

margin of the thallus, one line of long white mam-

millate cells, erect or inflexed (which) gives a crown-

like appearance.’

Riccia alatospora Volk et Perold, sp. nov. sec-

tions Piliferae, R. duthieae Volk et Perold simile,

sed habitu frondorum pusilliorum, sporis majoris

(90-100 pm) alis sporarum conspicuis plusquam 10

pm latis (indum nomen) et interdum perforatis,

squamae inconspicuae speciebus austro-africanis dif-

fert. Endemica Africae australis.

TYPE. — Cape, 3318 (Cape Town): Stellen-

bosch, Platklip, on moist sand in hollows on granite

outcrop (-DD), June 1929 Duthie 5004 (BOL), CH

1007 (PRE).

Dioecious, perennial, in crowded gregarious

patches, plants often overlying each other. Thallus

small, single or shortly furcate, arising from narrow

base becoming wider towards apex, segments up to 5

mm long, 0,85-3 mm broad and 0,6-1, 2 mm thick,

shape obcuneate to obovate, apex obtuse and emar-

ginate (Figs 4.1 & 5.1); when dry apex and distal

sides inflexed over sulcus, which is deep and narrow

apically, but soon becomes shallow, dorsal surface

flat and at margins slightly raised, glistening green-

ish-white and felt-like; ventral surface hyaline,

flanks green to pinkish red; when fresh or moist-

ened, dorsal surface broadly grooved, pale-green

with bright green colour developing in older plants,

velvety, convex. Scales imbricate at apex and distal

margins but absent proximally, shape rounded, up-

per edge hardly reaching thallus margin, size up to

450 x 550 pm, colour pink-red with outer 2-3 rows

of cells hyaline, or entirely hyaline, margin nearly

smooth with small quadrate cells, central cells ob-

long 5-6-sided, size 100 x 25 pm, walls thin and

Bothalia 15, 3 & 4 (1985)

535

FIG. 3. — Riccia duthieae ( O . H. Volk 81/273, M; PRE, 1, 3, 4; L. Smook 4036, M; PRE, 2, 5, 6).

Spores. 1, proximal face; 2, apex; 3, 4, distal face; 5, 6, side (SEM micrographs by S. M. Perold.)

Scale bar = 50 p m.

straight (Fig. 4.11, 4.12). Cross section of thallus

1,5-2 times broader than high, margins subacute to

rounded, flanks steep to outward sloping (Fig. 4.3,

4.4, 4.7), ventral surface almost flat; epithelium

about 1-2/10, assimilation tissue (chlorenchyma)

about 5/10, storage tissue 3-4/10 the thickness of the

thallus; dorsal surface covered with short, hyaline,

free hair-like pillars 100-160 pm long (Table 3, Fig.

4.5, 4.6), 2 or 3 and occasionally 1-celled, cells 35-75

pm long, apical cell conical, mammillate or blunt

(Fig. 5.3, 5.4), basal cell up to 80 pm wide (Table 4);

assimilation tissue consisting of columns or plates 1

cell thick and about 8 cells high, enclosing 4—6-8-

sided aircanals up to 80 pm across (Fig. 4.10), these

large canals suggest a resemblance to the old subge-

nus ‘Ricciella’; storage tissue with rounded, irregu-

larly arranged thin-walled cells 55 pm across, epider-

mis of ventral surface with flattened hyaline cells

TABLE 3. — R. alatospora, measurements on cross section (cul-

tivated plants of Oliver 8058)

Breadth of thallus : 0,85-1,0-1,3 mm

Height (thickness) : 0,6-0, 9-1 ,2 mm

Thickness of epithelium : 125-145-160 /xm ca 2/10 thickness of

thallus

Thickness of chlorenchyma: 275-495-650 jum ca 4—5/10 thickness

of thallus

Thickness of storage tissue : 190-200-400 /xm ca 3-4/10 thickness

of thallus

536

Bothalia 15, 3 & 4 (1985)

100pm

FIG. 4. — Riccia alatospora [A. V. Duthie 5004 (BOL) = CH 1007 (PRE); 1—4, 10-12, Naude s.n., BOL; 9, E. G. H. Oliver 8058,

PRE, 5-8]. Structure of the thallus. 1, fresh thalli; 2, shrinking thalli; 3, 4, cross sections of thallus; 5, 6, epithelium; 7, detailed

cross section of thallus; 8, chromosomes; 9, spore with notch and perforations in wing; 10, air canals in horizontal section; 11,

12, ventral scales. (1-6, 9-12 by O. H. Volk; 7 by R. Holcroft; 8 by T. Bornefeld.) Scale bar 1,2 = 2 mm; 3, 4, 7 = 1 mm; 5, 6,

10-12 = 100 pm; 8=1 pm; 9 = 50 pm.

Bothalia 15, 3 & 4 (1985)

537

FIG. 5. — Riccia alatospora [E. G. H. Oliver 8058, PRE 1-4; A. V. Duthie 5004 (BOL) = CH 1007 (PRE), 5,

6]. Thallus, epithelium and spores. 1, 2, thallus; 3, 4, terminal cells of the epithelium; 5, proximal face of

spore; 6, distal and proximal faces of spore. (1—4 SEM micrographs; 5-6 L.M. photographs by S. M.

Perold.) Scale bar = 50 tun.

TABLE 4. — R. alatospora, size (in um) of the cells of the epithelium and its other parts on cross section (culti-

vated plants of Oliver 8058)

Average size Length: Variation in size

Length Breadth Breadth Length Breadth

538

Bothalia 15, 3 & 4 (1985)

FIG. 6. — Riccia alatospara [A. V. Duthie 5004 (BOL) = CH 1007 (PRE)]. Spores. 1, proximal face; 2,

apex; 3, 5, distal face; 4, margin; 6, side. (SEM micrographs by S. M. Perold.) Scale bar = 50,um.

which bear mostly smooth, occasionally tuberculate

rhizoids. Antheridia scattered, ostioles colourless,

150 pm long. Archegonia along centre of thallus,

neck purple, 90 pm long. Sporangia bulging, up to

0,5 mm in size, containing about 150-200 spores

each. Spores (Fig. 6) with diameter (90-) 110 (-125)

pm, colour brownish-yellow to straw-coloured,

semi-transparent, shape triangular-globular, polar,

wing broad 12,5-15 pm wide, slightly and irregularly

wavy (Fig. 6,4) margin finely crenulate to denticu-

late, sometimes partly eroded, occasionally with

round perforations (Fig. 4.9), distal (outer) face with

4 or 5 large areolae in the centre (Fig. 6.3, 6.5), up to

40 pm across, formed by thick crenulate ridges 12

pm high, partially or completely subdivided into

smaller areolae by low ridges, often radiating star-

like from the centre; marginally a ring of smaller

areolae surround large central ones; ridges occasion-

ally extend across wing (Fig. 6.5), proximal (inner)

face with apex acute and high suture ridges, trira-

diate mark distinct (Fig. 6.1, 6.2), each facet with

about 30 smallish complete or incomplete areolae,

12,5 pm across, with spinous thickenings at angles of

areolae. Chromosome number n = 8 (Bornefeld)

(Fig. 4.8).

R. alatospora grows in dense colonies in flat de-

pressions on rock outcrops, periodically submerged

or wet with winter rain and then dried out for long

periods of time. They survive by producing many

Bothalia 15, 3 & 4 (1985)

539

bulbils (60-70 per cm’): most of the thallus shrinks

and accumulates foodstuff in its tip (Fig. 4.2). Soil

very shallow, consisting of decomposed granite

mixed with fine sand and dust, dark grey in colour,

pH 5,3 ( Oliver 8058). In her letter to Sim (10 June

1929) Duthie wrote: 'I have found it growing in some

abundance in hollows on a dome-shaped boss of

granite . . . These hollows, some of which are sev-

eral feet across, are very damp during our rainy sea-

son and often fill up with water. The little

species . . . grows on damp sand.’

R. alatospora is closely related to R. duthieae. It is

distinguished from it by its smaller thalli but larger

spores with very broad and often perforated and

eroded wings. It is endemic to the Cape and since its

initial collection in 1929, is known only from Stellen-

bosch.

CAPE. — 3318 (Cape Town): Stellenbosch Platklip (-DD),

Aug. 1933 Duthie 5004 (BOL); Duthie 5324 (BOL); Oct. 1935 A.

Naude s.n. (BOL); Aug. 1983 E. G. H. Oliver 8058 (PRE); Nov.

1935 O. Pretorius s.n. (BOL).

FIG. 7. — Distribution map of • Riccia duthieae and ■+- R. alatos-

pora.

ACKNOWLEDGEMENTS

The authors wish to thank Prof E. A. Schelpe for

the loan of the specimens from BOL and Dr habill.

T. Bornefeld of the Department of Plant Physiology,

Gesamthochschule D — 3500 Kassel, West Ger-

many, for the chromosome counts and figures. Sin-

cere thanks are due to Mr E. G. H. Oliver, Stellen-

bosch, for kindly sending us a fresh specimen of Ric-

cia alatospora , which he collected nearly 50 years

after the last gathering of it had been made. Thanks

are also due to Dr H. F. Glen for his kind help with

the Latin translation.

UITTREKSEL

Twee nuwe spesies van Riccia, R. duthieae en R.

alatospora word beskryf. Albei spesies behoort tot die

seksie Pi lifer Volk (1983) wat gekenmerk word dear

die unieke samestelling van die dorsale epiteel wat be-

staan uit los kolomme selle.

REFERENCES

ARNELL, S., 1963. Hepaticae of South Africa. Stockholm: Swed.

Nat. Sc. Res. Council.

BORNEFELD, T., 1985. Chromosomentypen der Gattung Ric-

cia von S und SW-Afrika und allgemeine Bemerkungen zur

Zytogenetik der Lebermoose. Nova Hedwigia 40. im Druck.

DUTHIE/SIM CORRESPONDENCE (PRE).

DUTHIE, A. V. & GARSIDE, S., 1936. Studies in South Afri-

can Ricciaceae. I. Three annual species: R. plana Taylor, R.

cupulifera sp. nov., and R. curtisii T. P. James. Trans. R.

Soc. S. Afr. 24,2: 93-133.

DUTHIE, A. V. & GARSIDE, S., 1939. Studies in South Afri-

can Ricciaceae. II. The annual species of the section Ricciella

(concluded): R. compacta sp. nov., and R. rautanenii Steph.

Trans. R. Soc. S. Afr. 27,1: 17-28.

GUNN, M. & CODD, L. E., 1981. Botanical exploration of

Southern Africa. Cape Town: Balkema.

SIM, T. R., 1926. The bryophyta of South Africa. Trans. R. Soc.

S. Afr. 15: 9.

VOLK, O. H., 1981. Beitrage zur Kenntnis der Lebermoose (He-

paticae) aus Siidwest Africa (Namibia). II. Mitt. bot.

StSamml., Munch. 17: 245-252.

VOLK, O. H., 1983. Vorschlag fur eine Neugliederung der Gat-

tung Riccia L. Mitt. bot. StSamml., Munch. 19: 453-465.

VOLK, O. H. & PEROLD, S. M., 1984. Studies in the liverwort

genus Riccia (Marchantiales) from the south-west Cape. Bo-

thalia 15: 117-124.

Bothalia 15, 3 & 4: 541-544 (1985)

A checklist of the Pteridophytes of the ‘Flora of Southern Africa’

region

NICOLA C. ANTHONY* and E. A. C. L. E. SCHELPE*

Keywords: checklist, distribution ranges, Pteridophytes, southern Africa

ABSTRACT

The 250 species, both indigenous and naturalized, in the 75 genera of southern African Pteridophytes are listed

with their subspecies and varieties (263 taxa in all). Coded distribution ranges are given for each taxon.

INTRODUCTION

The following is a checklist of all the Pterido-

phytes of the Flora of Southern Africa area (South

West Africa/Namibia, Botswana, Lesotho, Swazi-

land, Transkei and South Africa sens, lat.) arranged

not in alphabetical order, but in the order they will

appear in the FSA. The application of the geographi-

cal symbols used is depicted in Fig. 1.

Other symbols used in the checklist are:

- distribution range extends into tropical Africa

! - taxon is endemic to the southern African region

* - introduced taxa now naturalized and spreading

in southern Africa

If the taxon is found in only one other country on

the borders of southern Africa, then that country is

given. Similarly, if the taxon does not extend into

tropical Africa, but is not endemic to southern

Africa, an indication is given of its further distribu-

tion.

FIG. 1. — Geographical symbols used in the checklist.

* Bolus Herbarium, University of Cape Town, Rondebosch

7700.

PSILOTALES

Psilotaceae

Psilotum Swartz

nudum (L.) Beauv Tk; N; L; Sw; T (N;E)

LYCOPODIALES

Lycopodiaceae

Lycopodium L.

saururus Lam C(SW;S;E); N; L; O

verticillatum L.f. C(E); Tk;N; L; O; Sw; T(N;E) *

dacrydioides Bak N; T(N;E)

gnidioides L.f. C(SW;S;E); Tk; N; Sw; T(N;E) *

ophioglossoides Lam T(N;E)

cernuum L C(SW;S;E); Tk; N; Sw;

T(N;S;C;E) "

clavatum L C(SW;S;E); Tk; N; L; Sw;

T(N;S;E) *

complanatum L.

subsp. zanclophyllum C(WS;S) Madagascar, Oceanic

(Wilce) Schelpe Islands

carolinianum L.

var. carolinianum C(SW;S;E); Tk; N; T(N;C;E)

var. grandifolium Spring.... N; Sw; T(C;E)'

SELAGINELLALES

Selaginellaceae

Selaginella Beauv.

pygmaea (Kaulf.) Alston C(SW;S;E)!

dregei (Presl) Hieron B;Tk;N;0;Sw;T(N;S;C;W;E) '

njam-njamensis Hieron B

caffrorum (Milde) Hieron. C(S;E); Tk; N; L; O;

T(N;C;E) *

imbricata (Forssk.) Spring ... SWA; N; Sw

ex Decne.

kraussiana (Kunze) A.Br. ... C(SW;S;E); Tk; N; Sw;

ex Kuhn T(N;W;E)

mittenii Bak C(E); Tk; N; L; O; Sw;

T(N;C;E)‘

ISOETALES

Isoetaceae

ISOETES L.

capensis Duthie

var. capensis C(SW) !

var. stephansenii (Duthie).. C(SW) !

Jermy & Schelpe

perrieriana Iversen SWA Madagascar

transvaalensis Jermy & N; L; 0;T(N;E) !

Schelpe

aequinoctialis Welw. ex SWA; C(N)

A.Br.

wormaldii Sim C(E) !

stellenbossiensis Duthie C(SW) !

welwitschii A.Br. ex Kuhn. ..IK "

schweinfurthii A.Br SWA; T(N) A

EQUISETALES

Equisetaceae

Equisetum L.

ramosissimum Desf. SWA; B; C(N;S;E); Tk; N; L; O;

Sw; T(N;S;C:W;E) *

542

Bothalia 15, 3 & 4 (1985)

OPHIOGLOSSALES

Ophioglossaceae

Ophioglossum L.

bergianum Schlechtd C(NW;SW) !

gomezianum Welw. ex A.Br. T(C;E)

polyphyllum A. Br SWA; B; C(NW;N;SW;S;E); Tk;

N; L; O; T(N;S;C;W;E) "

nudicaule L.f C(NW;SW;S;E) ?Zimbabwe

reticulatum L SWA; C(E);Tk;N;T(N;S;C;E) "

vulgatum L C(E); Tk; N; L; O; T(N;E) '

lancifolium Presl SWA; C(SW); N; T(C)

MARATTIALES

Marattiaceae

Marattia Swartz

fraxinea J.E.Sm. ex

J. F. Gmel.

var. salicifolia (Schrad.) C(S); Tk; N; Sw; T(N;C;E)

C.Chr.

FILICALES

Osmundaceae

OSMUNDA L.

regalis L C(SW;S;E); Tk; N; L; Sw;

T(N;S;C;E) A

Todea Willd.

barbara (L.)T. Moore C(NW;SW;S;E); Tk; N; Sw;

T(N;E) A

Gleicheniaceae

Gleichenia J. E. Sm.

polypodioides (L.)J.E.Sm. C(SW;S;E); Tk; N; L; O; Sw;

T(N;C;E) A

umbraculifera (Kunze) C(E); Tk; N; Sw; T(N;E) '

T. Moore

Dicranopteris Bernh.

linearis (Burm.f.)Underw. ..N; Sw; T(N;S;E)

Schizaeaceae

Schizaea J.E.Sm.

pectinata (L.)Swartz C(SW;S;E); Tk; N; Sw; T(N;E) A

tenella Kaulf. C(SW;S;E); Tk; N; T(N;E) ?

Zimbabwe

Anemia Swartz

dregeana Kunze C(E); Tk; N; Sw; T(N;S;E) !

simii Tardieu T(N;C;E)

Mohria Swartz

caffrorum (L.)Desv C(NW;SW;S;E): Tk; N; L; O; Sw;

T(N;S;C;E) A

hirsuta J.P. Roux N; L; O !

Lygodium Swartz

microphyllum (Cav.)R.Br. N A

kerstenii Kuhn N; Sw "

Marsileaceae

Marsilea L.

nubica A.Br SWA; B A

distorta A.Br SWA '

coromandelina Willd SWA; T(N) A

ephippiocarpa Alston SWA; B; N; Sw; T(N;E)

minuta L N; Sw A

fenestrata Launert N; Sw Mozambique

aegyptiaca Willd SWA; B; C(NW); T(S) A

farinosa Launert SWA; B; C(NW); N; T(N;C) A

macrocarpa Presl SWA; B; C(N;E); Tk; N; O;

T(N;S;C;W) A

unicornis Launert SWA Angola

schelpeana Launert C(SW;S;E) !

apposita Launert B; T(E) Zimbabwe

villifolia Bremek. & Oberm. SWA; B; C(N); T(C)!

ex Alston & Schelpe

vera Launert SWA; B Zimbabwe

capensis A.Br C(NW;N;SW;S;E): Tk; O;

T(N;S;W) A

burchellii (Kunze)A.Br SWA; B; C(N;SW;S;E); O !

Salviniaceae

Azollaceae

Azolla Lam.

pinnata R.Br SWA; B; N A

*filiculoides Lam C(N;E); O

Cyatheaceae

Cyathea J. E.Sm.

dregei Kunze C(SW;E); Tk; N; Sw; T(N;S;C;

E) A

capensis (L.f.)J.E.Sm C(SW;S;E); Tk; N; Sw; T(N;C;

E) A

Hymenophyllaceae

Trichomanes L.

reptans Swartz Tk; N; T(N;E) Madagascar, S.

America

erosum Willd.

var. aerugineum N A

(V.d. Bosch) Bonap.

rigidum Swartz Tk; N; Sw; T(N;C;E)

pyxidiferum L.

var. melanotrichum C(SW;S;E); Tk; N; O; Sw;

(Schlechtd.) Schelpe T(N;W;E) A

borbonicum V.d. Bosch N; Sw; T(E)

Hymenophyllum J. E. Sm.

marlothii Brause C(SW;S) !

capillare Desv T(E) A

peltatum (Poir.)Desv C(SW;S); N; O A

tunbridgense (L.)J.E.Sm. ... C(SW;S;E); N; Sw; T(N;E) A

capense Schrad C(SW;S;E); N; T(N;E) A

polyanthos Swartz

var. kuhnii (C.Chr.)

Schelpe T(E)A

Dennstaedtiaceae

Blotiella Try on

natalensis (Hook.)Tryon C(S); Tk; N A

glabra ( Bory ) Tryon C(S); Tk; N; Sw; T(N;E)

Histiopteris (Agardh) J.Sm.

incisa (Thunb.)J.Sm C(SW;S); N; T(E) A

Pteridium Gled. ex Scop.

aquilinum (L.)Kuhn C(SW;S;E); Tk; N; L; O; Sw;

subsp. aquilinum T(N;S;C;E)

Microlepia Presl

speluncae (L.) T. Moore SWA; B; N; T(N;C) A

Hypolepis Bernh.

sparsisora (Schrad. )Kuhn .... C(SW;S;E); Tk; N; Sw; T(N;E) A

Vittaria J.E.Sm.

isoetifolia Bory .

Vittariaceae

C(SW;S;E); Tk; N; T(N;E) A

Adiantaceae/Pteridaceae

Acrostichum L.

aureum L N

Anogramma Link

leptophylla (L.) Link C(N;SW); N

Ceratopteris Brongn.

thalictroides (L.) Brongn B

PlTYROGRAMMA Link

argentea (Willd. )Domin N; T(N;C;E)

*calomelanos (Swartz) Link

var. aureoflava Tk; N; T(E) A

(Hook.)Weath. ex Bailey

Adiantum L.

incisum Forssk SWA; B; T(N;C;E)

philippense L T(E) A

hispidulum Swartz C(SW); N; T(N)

aethiopicum L C(SW;S;E) !

capillus-veneris L SWA; C(NW;N;SW;S;E); Tk; N;

L; O; T(N;S;C;W;E) A

poiretii Wikstr SWA; C(SW;E); Tk; N; L; O; Sw;

T(N;S;C;E) A

*raddianum Presl C(SW;S;E); Tk; N; T(N)

Pteris L.

vittata L.

SWA; C(N;E); Tk; N; L; O; Sw;

T(N;S;C;W;E) A

C(SW;S;E); Tk; N; L; O; Sw;

T(N;S;C;W;E) A

Salvinia Adans.

* moles ta D. Mitch.

B; C(SW;S;E); N A

cretica L.

Bothalia 15, 3 & 4 (1985)

543

buchananii Bak. ex Sim C(SW;S;E); Tk; N; T(E)

dentata Forssk C(SW;S;E); Tk; N; L;

T(S;C;W;E) A

*tremula R.Br C(SW)

catoptera Kunze Tk; N; O; Sw; T(N;E)

friesii Hieron N; Sw; T(N;C) A

Cheilanthes Swartz

rawsonii (Pappe) Mett. ex

Kuhn SWA: C(NW) !

marlothii (Hieron.) Schelpe SWA; T(N;C;E) Angola

eckloniana (Kunze)Mett SWA; C(NW;N;SW;S;E); Tk; N;

L; O; Sw; T(N;S;C;E) Zim-

babwe

inaequalis (Kunze) Mett.

var. inaequalis SWA; N; T(N;C;E) "

var. buchananii (Bak.)

Schelpe Tk; N; T(N;E) A

induta Kunze C(NW;N;SW;S;E)!

parviloba (Swartz)Swartz .... SWA; C (SW;S;E); Tk; N; T(C) !

depauperata Bak C(SW) !

contracta (Kunze) Mett. ex

Kuhn C(NW;SW;S;E); T(E) !

hirta Swartz SWA; B; C(N;SW;S;E); Tk; N; L;

O; Sw; T(N;S;C;W;E)A

robusta (Kunze) R. Tryon ... SWA; C(NW;SW) !

hastata (L.f. ) Kunze SWA; C(NW;SW;S) !

capensis (Thunb.) Swartz ....SWA; C(NW;SW;S;E); N; L; O !

kunzei Mett SWA; C(NW) !

deltoidea Kunze SWA; C(NW;SW) !

namaquensis (Bak.) Schelpe

&N.C. Anthony SWA; C(NW;SW)!

dolomiticola (Schelpe)

& N.C. Anthony C(N); T(S;C;E) !

botswanae Schelpe &

N.C. Anthony B; T(N;C) !

involuta (Swartz)Schelpe &

N.C. Anthony

var. involuta C(SW;E); Tk; T(S;C;E) A

var. obscura (N.C.Anth- SWA; B; C(S;E); Tk; N; L; O;

ony) N.C. Anthony T(N;S;C;W;E) A

viridis (Forssk. ) Swartz

var. viridis SWA; C(SW;S;E); Tk; N; L; O;

Sw; T(N;S;C;W;E) A

var. macrophylla (Kunze)

Schelpe & N. C. Anthony ...C(S;E); Tk; N; Sw; T(E) !

var. glauca (Sim) Schelpe SWA;B;C(SW;E);Tk;N;0;Sw;

& N.C. Anthony T(N;S;C;W;E) A

quadripinnata (Forssk.) C(N;S;E); Tk; N; L; O; Sw;

Kuhn T(N;S;E) A

multifida (Swartz)Swartz

subsp. multifida SWA; C(NW;N;SW;S;E); Tk

Angola

subsp. lacerata

N.C. Anthony & Schelpe... Tk; N; L; Sw; T(N;S;C;E) A

pentagona Schelpe &

N.C. Anthony SWA; T(C;E) Zimbabwe

dinteri Brause SWA Angola

bergiana Schlechtd C(SW;S;E); Tk; N; Sw; T(N;S;

E) A

concolor (Langsd.&Fisch.) SWA; B; C(S;E); Tk; N; Sw;

R. & A. Tryon T(N;C;E) A

Pellaea Link

pteroides (L.) Prantl C(SW); ?N; ?T !

rufa A. Tryon C(SW;S) !

dura (Willd.)Hook SWA; N; Sw; T(N;S;C;E) A

boivinii Hook B; N; T(N;C;E) *

pectiniformis Bak SWA; N; T(N;C;E) A

calomelanos (Swartz)Link ...SWA; B; C(NW;N;SW;S;E); Tk;

N; L; O; Sw; T(N;S;C;W;E) A

leucomelas (Mett. ex

Kuhn)Bak C(SW;S;E) !

Actiniopteris Link

radiata ( Koenig ex

Swartz)Link SWA; B; C(N); Sw; T(N;C;E) A

dimorpha Pichi-Serm T(C;E) A

Lindsaeaceae

Lindsaea Dryand. apud

J. E. Sm.

ensifolia Swartz N A

Grammitidaceae

Grammitis Swartz

poeppigiana (Mett.) Pichi-

Serm C(SW) Circumantarctic

XlPHOPTERIS Kaulf.

flabelliformis (Poir.) Schelpe N '

Polypodiaceae

Pyrrosia Mirb.

africana (Kunze) Ballard C(E); Tk; N !

schimperiana (Mett. ex

Kuhn)Alston T(E)A

Loxogramme (Blume) Presl

lanceolata (Swartz) Presl C(E); Tk; N; Sw; T(N;E) "

Polypodium L.

vulgare L C(SW;S;E); Tk; N; L Kerguelen

I., Europe, America

polypodioides (L.)Hitchc.

subsp. ecklonii (Kunze)

Schelpe C(E); Tk; N; O; Sw; T(N;S;E) A

x Pleopodium Schelpe &

N.C. Anthony

simiana Schelpe &

N.C. Anthony C(S;E); N; T(N;E) A

Pleopeltis H. B. K. ex. Willd.

macrocarpa (Bory ex C(SW;S;E); Tk; N; L; O; Sw;

Willd.) Kaulf. T(N;S;C;E) *

schraderi (Mett.)Tardieu C(S;E); Tk; N; O; Sw; T(N;E) '

excavata (Bory ex

Willd.) Sledge T(N;C;E) A

Microgramma Presl

lycopodioides (L.)Copel N; Sw; T(E) A

Microsorium Link

punctatum (L.)Copel Tk; N A

pappei (Mett. ex Kuhn)

Tardieu N A

scolopendrium

(Burm.f.)Copel C(E); Tk; N; T(E) A

ensiforme (Thunb. )Schelpe ..C(SW;S;E); Tk !

Phlebodium J. Sm.

*aureum J.Sm N

Davalliaceae

Nephrolepis Schott

biserrata (Swartz) Schott N; T(E)A

*exaltata (L.) Schott C(SW;S); N; Sw

Arthropteris J. Sm.

monocarpa (Cordem.)

C.Chr N A

Oleandra Cav.

distenta Kunze Tk; N; Sw; T(N;C;E)

Davallia J. E.Sm.

chaerophylloides (Poir.)

Steud Tk; N; T(N;E) *

Aspleniaceae

Asplenium L.

christii Hieron N A

anisophyllum Kunze Tk; N; Sw; T(N;E)

prionitis Kunze Tk; N Madagascar

boltonii Hook, ex Schelpe ...C(E); Tk; N; T(N;C;E)

gemmiferum Schrad C(SW;S;E); Tk; N; T(N;E) A

x flexuosum Schrad C(S;E); Tk; N; T(N) !

protensum Schrad C(S;E); Tk; N; T(N;E)

friesiorum C.Chr N; T(E)

sandersonii Hook Tk; N; T(N;E) A

stoloniferum Bory C(E);Tk;N;L;0;T(E) Reunion

trichomanes L C(N;SW;S;E); N; L; O; T(E) *

platyneuron (L.) Oakes C(SW;S;E); Tk; N; L; T(E) Ja-

maica, N. America

monanthes L C(SW;S;E); Tk; N; L; O; T(N;

E) '

lunulatum Swartz C(SW;S;E); Tk; N; Sw; T(N;E) A

erectum Bory ex Willd.

var. erectum C(SW;S;E); Tk; N; T(N;E) A

var. usambarense

(Hieron. )Schelpe C(SW;S;E); N; T(N;E)

inaequilaterale Willd Tk; N; Sw; T(N;C;E)

dregeanum Kunze Tk; N A

544

Bothalia 15, 3 & 4 (1985)

preussii Hieron N; T(N)

theciferum (H.B.K.)Mett.

var. concinnum

(Schrad.)C.Chr C(S;E); Tk; N; Sw; T(N;S;E) *

rutifolium (Berg. (Kunze C(SW;S;E); Tk; N; Sw; T(N;C;

E) \

hypomelas Kuhn T(N)

lobatum Pappe & Raws C(S;E); Tk; N; Sw; T(N;E)

varians Wall, ex Hook. &

Grev.

subsp. fimbriatum

(Kunze)Schelpe C(E); N; L; T(C;E)

adiantum-nigrum L.

var. adiantum-nigrum C(NW;N;SW;S;E); N; L; O;

T(N;S;E) *

var. solidum

(Kunze)J.P.Roux C(SW;S;E); Tk !

blastophorum Hieron N A

splendens Kunze C(E); Tk; N; Sw; T(N;E) !

simii Braithw. & Schelpe C(S); Tk; N; T(N)

lividum Mett. ex Kuhn T(E) *

aethiopicum ( Burm.f '.) C(SW;S;E); Tk; N; L; O; Sw;

Becherer T(N;S;C;E)

Ceterach D. C.

cordatum (Thunb.)Desv SWA; B; C(NW;N;SW;S;E); Tk;

N; L; O; T(N;S;C;W;E) "

Thelypteridaceae

Thelypteris Schmid.

madagascariensis (Fee)

Schelpe Tk; N; Sw; T(N;E) A

interrupta (Willd.) K.Iwats. SWA; B; C(SW;S;E); Tk; N; Sw;

T(N;E) A

confluens (Thunb.)Morton SWA; B; C(SW;S;E); N; L; Sw;

T(N;S;C;E) "

pozoi (Lagasca) Morton C(SW;S;E); Tk; N; T(N;E)

pulchra (Bory ex Willd.) T(E) “

Schelpe

altissima (Holtt.)Vorster N !

dentata (Forssk.)E.St John .. C(E); Tk; N; Sw; T(N;S;E)

chaseana Schelpe SWA; T(N;E) '

gueinziana (Mett.)Schelpe .... C(S;E); Tk; N; Sw; T(N;S;C;E) '

knysnaensis N.C. Anthony

& Schelpe C(S) !

bergiana (Schlechtd.) Ching C(SW;S;E); N; Sw; T(N;E)

Macrothelypteris (H. Ito)

Ching

*torresiana (Gaud.)Ching N

Ampelopteris Kunze

prolifera (Retz.)Copel N; T(N;E) "

Athyriaceae

Athyrium Roth

schimperi Moug.ex Fee C(E); N; L; T(E) A

scandicinum (Willd. )Presl N; O; Sw; T(N;E)

Lunathyrium Koidzumi

*japonicum (Thunb.)

Kurata N; T(E)

Diplazium Swartz

zanzibaricum (Bak.)C.Chr. T(N;E) '

Dryoathyrium Ching

boryanum (Willd. (Ching T(N)

Cystopteris Bernh.

fragilis (L.)Bernh C(SW;S;E); Tk; N; L; O; T(N;

E) A

Lomariopsidaceae

Elaphoglossum Schott

aubertii (Desv.)T. Moore N; T(N;E) A

hybridum (Bory)Brack C(E); Tk; N A

spathulatum (Bory)T. Moore N; O

drakensbergense Schelpe N; L; O !

macropodium (Fee)T. Moore Tk; N; T(N;E) A

conforme (Swartz)J.Sm C(SW) A

acrostichoides (Hook. & C(SW;S;E); Tk; N; L; O; Sw;

Grev.)Schelpe T(N;C;E)

angustatum (Schrad.)Hieron. C(SW;S); N !

Bolbitis Schott

heudelotii (Bory ex

Fee) Alston T(N)

Aspidiaceae/Dryopteridaceae

Woodsia R.Br.

montevidensis (Spreng.)

Hieron.

var. burgessiana (Gerr.es :

Hook. & Bak. (Schelpe C(E); Tk; N; L; O; T(N) A

angolensis Schelpe N; T(E) Angola

Didymochlaena Desv.

truncatula (Swartz)J.Sm N; T(N) A

Dryopteris Adans.

squamiseta (Hook.)Kuntze ,.C(E); Tk; N; T(N;E) A

athamantica (Kunze)Kuntze Tk;N;L;0;Sw;T(N;S;C;W;E) A

inaequalis (Schlechtd.) C(SW;S;E); Tk; N; L; O; Sw;

Kuntze sens, lat T(N;S;C;E) A

dracomontana Schelpe &

N.C. Anthony C(S;E); N; L !

esterhuyseniae Schelpe &

N.C. Anthony N; L !

callolepis C.Chr C(SW;S) A

Cyrtomium Presl

caryotideum (Wall, ex Hook.

& Grev.) Presl

var. micropterum

(Kunze) C.Chr C(E); Tk; N; L; T(E) A

POLYSTICHUM Roth

macleai (Bak. (Diels Sw; T(N;E) !

transkeiense Jacobsen C(E); Tk; N; Sw !

pungens (Kaulf.)Presl C(SW;S;E); Tk; N; Sw; T(N;E) !

luctuosum (Kunze)T. Moore C(E); Tk; N; L; Sw; T(E)

transvaalense N.C. Anthony C(E); Tk; N; O; T(N;E)

monticola N.C. Anthony &

Schelpe C(N;SW;S;E); Tk; N; L; O ?!

alticola Schelpe &

N.C. Anthony C(SW;S;E); N; L; O; T(E) A

dracomontanum Schelpe &

N.C. Anthony N !

Arachniodes Blume

foliosa (C.Chr. (Schelpe C(E); Tk; N; T(N;E) A

Rumohra Raddi

adiantiformis (G.Forst.)

Ching C(SW,S;E); Tk; N; T(N;E) A

Hypodematium Kunze

crenatum (Forssk.)Kuhn T(S;E)

Ctenitis (C.Chr.) C.Chr. ex

Tardieu

lanuginosa (Willd. ex

Kaulf.)Copel C(S);Tk; N; T(N;C;E) A

Tectaria Cav.

gemmifera (Fee) Alston N; T(N;E)

Blechnaceae

Blechnum L.

inflexum (Kunze)Kuhn C(SW;S;E); N; O; T(E) Zim-

babwe

giganteum (Kaulf.) C(SW;S;E); Tk; N; Sw;

Schlechtd T(N;S;C;E) A

capense (L. (Schlechtd C(SW;S;E); Tk; N; Sw; T(N;E)

tabulare (Thunb. )Kuhn C(SW;S;E); Tk; N; Sw; T(N;E) A

australe L.

var. australe SWA; C(NW;N;SW;S;E); Tk; N;

L; O; Sw; T(S;C;E) A

var. aberrans

N.C. Anthony & Schelpe C(E)!

punctulatum Swartz

var. punctulatum C(SW;S;E); Tk; N; L; Sw A

var. atherstonei (Pappe &

Raws.)Sim C(SW); Tk; N; Sw; T(N;E) !

var. intermedium Sim C(E); N !

var. krebsii (Kunze)Sim ...,C(E); Tk; N !

Stenochlaena J. Sm.

tenuifolia ( Desv.)T. Moore Tk; N; T(E)

UITTREKSEL

Die 250 spesies, beide inheems en genaturaliseerd,

in die 75 genera van Pteridophyta wat in suiderlike

Afrika voorkom , word met hul subspesies en varie-

teite (263 taksa altesame) gelys. Gekode verspreidings

gebiede word vir elkeen voorsien.

Bothalia 15,3 & 4: 545-559 (1985)

Notes on African plants

VARIOUS AUTHORS

AMARYLLIDACEAE

THE IDENTITY OF NERINE FLEXUOSA

CURRENT APPLICATION OF THE NAME

NERINE FLEXUOSA

The epithet flexuosa was first applied to a nerine

by Jacquin in his Hort. Schoenbr. 1:35, t.67 in 1797

under the title Amaryllis flexuosa. Jacquin’s type

specimen is no longer extant so that the interpreta-

tion of Nerine flexuosa (Jacq.) Herb, must rest on

Plate 67 and Jacquin’s description of it. These refer

to a cultivated plant with the following characteris-

tics: obtuse-ended leaves which are short and imma-

ture at flowering and develop fully only some consid-

erable time after flowering, a large umbel (up to 15

cm diam.), open in form and 6-flowered; and per-

ianth segments about 3,8 cm long, smooth-margined

in the lower half, slightly undulate in the upper half,

clearly centrally veined with deeper pink, not

widened at the base nor narrowed above the base.

Midseason flowering. Exact locality unknown.

When Baker dealt with N. flexuosa in Flora Ca-

pensis in 1896, he cited various herbarium specimens

from the Bruintjies Hoogte area of the Somerset

East Division of the eastern Cape. Baker’s descrip-

tion, supplemented by measurements taken from the

herbarium sheets referred to by Baker and by data

from living specimens from Bruintjies Hoogte, re-

fers to plants with leaves usually well developed

(20-50 cm long) at flowering time, tapering grad-

ually to a fairly acute apex; umbels of (6)— 10— ( 16)

flowers, compact in form, averaging 9 cm in di-

ameter; and perianth segments 2, 1-3 ,4 cm long, dis-

tinctly crisped even in the lower half, mostly

widened at the base and narrowed or ‘rolled’ for a

short distance above the base, lacking a distinct cen-

tral vein of deeper pink. Late flowering.

When Miss W. F. Barker dealt with the nerine

from Pluto’s Vale, Albany Division, in Flower. PI.

S. Afr. 15: t.561 (1935), her measurements showed

that this plant was very close to the nerines found in

the Somerset East District which were used by

Baker for his Flora Capensis description. Accord-

ingly, Miss Barker followed Baker and identified the

Pluto’s Vale plants as Nerine flexuosa.

A point by point comparison between Jacquin’s

plant and the Bruintjies Hoogte-Boschberg plants

reveals clearly that the latter Eastern Province plants

are a very poor match of Nerine flexuosa (Jacq.)

Herb. However, plants collected in the Clanwilliam-

Piketberg-Laingsberg area of the South West Cape

(Pocock PRE 30286; Martin & Steytler NBG 66387,

and Logan NBG 66397) are good matches of Jac-

quin’s illustration. It seems, therefore, that a new

name is needed for the eastern species described by

Baker and by Barker, and this aspect will be dealt

with in another article.

Before turning to N. humilis, two varieties of Ne-

rine flexuosa should be mentioned. The first is Ne-

rine flexuosa var. pulchella (Herb.) Bak., illustrated

and described (by Herbert) in Curtis’s bot. Mag.

t.2407 (1823) under the title Nerine pulchella Herb.

Apparently no specimen of this plant was preserved

and the figure and description seem to be inadequate

for certain identification with material from the wild,

but it does not seem to be misplaced as a very vigor-

ous, glaucous-leafed form of N. flexuosa (Jacq.)

Herb. Secondly, in Flora Capensis (1896), Baker de-

scribed Nerine flexuosa var. sandersonii. Fortunately

the type specimen of this is at Kew and Mr J. R.

Sealy (pers. comm. 1960) stated that, from the meas-

urements taken and from the general appearance of

the type material, it was his opinion that this plant ‘is

not well placed as a variety of Nerine flexuosa and

would be better associated with N. falcata Barker

and N. laticoma (Ker-Gawler) Dur. & Schinz.’ In his

‘Review of the genus Nerine’ in Plant Life (1967), Dr

H. P. Traub came to the same conclusion as Mr

Sealy.

THE RELATIONSHIP BETWEEN NERINE FLEXUOSA

AND N. HUMILIS

The specific epithet humilis was first used for a

plant referable to Nerine by Jacquin in his Hort.

Schoenbr. 1: 36, t.69 (1797) in the name Amaryllis

humilis. Once again, Jacquin’s type specimen is no

longer available so that Nerine humilis (Jacq.) Herb,

must rest on Jacquin’s figure and description. These

refer to a small plant with obtuse-ended leaves 5-8

mm wide, produced after the flowering scape, a

small open umbel of 3 flowers about half the size of

N. flexuosa on a 15 cm peduncle with perianth seg-

ments up to about 3 cm long with a distinct central

vein.

If these details are compared with Baker’s account

of N. humilis in Flora Capensis it will be seen that his

description has diverged considerably from Jacquin’s

and seems to refer as much to Jacquin’s N. flexuosa

as to Jacquin’s N. humilis. Two of the four speci-

mens cited by Baker are also misleading, for Drege’s

collection from Table Mountain is not Nerine hu-

milis but N. sarniensis, and Zeyher’s Albany collec-

tion is usually placed under N. undulata. If Jacquin’s

descriptions and figures of his N. flexuosa and N. hu-

milis are compared, side by side, it will be seen that

there is a very striking similarity in general morpho-

logy, and that the chief difference is one of size. Ker-

Gawler recognized this when he wrote in the Bot.

Reg. 1. 172 (1817) that: ‘It is extremely difficult to de-

fine any distinctions between this species (flexuosa)

and humilis . . . which do not resolve into differ-

546

Bothalia 15, 3 & 4 (1985)

ences of size and its consequences . . . they may be

distinct species . . . but we confess ourselves unable

to elicit a single stable discriminating mark except

size.’

My own studies seem to confirm this conclusion.

Because of the very complex topographical condi-

tions and the considerable differences in microcli-

mates experienced in the South West Cape, even in

areas lying close together, one would expect a con-

siderable degree of variation within a species. Exam-

ination of the material confirms this. Now that we

have so many more specimens for study, we know

that forms intermediate in size between Jacquin’s

large N. flexuosa and his small N. humilis exist.

Moreover, N. tulbaghensis is regarded as a dwarf

form of the N. flexuosa-humilis complex. From a

visual inspection of the available material, it appears

as if there is a fairly gradual variation in size from the

smaller to the larger forms without any sharply de-

fined gap or discontinuity. Pending a detailed statis-

tical analysis of all the material, it seems advisable to

take a rather broad view of the material under dis-

cussion, consequently I propose that Jacquin’s N.

flexuosa and his N. humilis be regarded as two forms

of one widely distributed rather variable, polymor-

phic species.

What should this widely circumscribed species be

called? The epithets flexuosa and humilis were both

published in the same volume and I propose that the

name humilis should be used for the following

reasons: the name flexuosa has been so widely mis-

applied in the past that it would be wise to sink it;

and that, whereas the very large and the really dwarf

forms of the complex are comparatively rare, the

fairly small forms, which are common and are widely

distributed in the South West Cape, have always

been known to botanists as N. humilis so that the

retention of this epithet would cause the minimum of

confusion. N. flexuosa will then be relegated to syn-

onymy. This is accordingly done below.

Nerine humilis (Jacq.) Herb, in Curtis’s bot.

Mag. sub t. 2124 (1820); Baker in FI. Cap. 6, 2: 213

(1896), partly excl. Drege and Zeyher specimens;

W.F. Barker in Flower. PI. S. Afr. 15: t.564 (1935).

Type: t.69 in Jacq., Hort. Schoenbr. 1 (1797).

Amaryllis humilis Jacq., Hort. Schoenbr. 1: 36, t.69 (1797);

Ker-Gawl. in Curtis’s bot. Mag. t.726 (1804).

A. flexuosus Jacq., Hort. Schoenbr. 1: 35, t.67 (1797); Ker-

Gawl. in Bot. Reg. t . 172 (1817). Nerine flexuosa (Jacq.) Herb, in

Curtis’s bot. Mag. sub t.2124 (1820); Baker in FI. Cap. 6:211

(1896) partly, excl. var. sandersonii Bak. and certain cited speci-

mens; non W. F. Barker in Flower. PI. S. Afr. 15: t.561 (1935).

Type: t.67 in Hort. Schoenbr.

Nerine pulchella Herb, in Bot. Reg. App. 19 (1821); Curtis’s

bot. Mag. t.2407 (1823).

N. tulbaghensis W. F. Barker in Flower. PI. S. Afr. 15: t.565

(1935). Type: Ross-Frames sub BOL 20369.

Bulb ovoid to ovoid-oblong, from 1,2-3,75 cm

diam. Leaves 3-7, mostly sprouting shortly before or

during flowering and developing to maturity long

after flowering, linear, minutely and inconspicuously

punctate, flat to somewhat channelled, suberect or

spreading, green to glaucous, apex fairly blunt, ma-

ture leaves from (10)-15-25-(30) cm long and from

(3)— 5— 12— (18) mm wide. Inflorescence (l)-3-7-(9)-

flowered, umbel loose and open, 4,5 to 15 cm in di-

ameter; peduncle terete, glabrous, (10)-15-35-(45)

cm long by 2-5 mm wide, green sometimes flushed

red at base; spathe valves lanceolate, up to 5,7 cm

long; pedicels terete, glabrous, firm, 1-5,6 cm

(mostly 2,5-3 cm) long, 1,5-2 mm wide. Flowen

with a zygomorphic perianth, segments linear,

2,2-3,8-(4,5) cm long, 3-5 mm wide, usually 4 or 5

spreading-ascending and 2 or 1 descending, base not

widened, margin with basal half smooth and upper

half slightly undulate, variable in colour from very

pale pink to deep pink with a clearly defined deeper

pink median stripe on the basal half of the upper side

and, on the lower side, extending towards the apex.

Stamens cream to pink, slightly declinate, 2-4 cm

long, slightly shorter than the segments, non-appen-

diculate; pollen from pale yellow to greyish-white.

Style longer than stamens, strongly recurved when

mature; Stigma 3-lobed.

Flowering chiefly in April in the Southern Hemi-

sphere. Widely distributed through the winter- rain-

fall area of the South West Cape; stretching in an arc

from Clanwilliam in the north through Tulbagh,

Worcester, Caledon and Swellendam to Riversdale

in the east.

Specimens examined:

CAPE. — 3218 (Clanwilliam): Zebra Kop, Piketberg Mtns

(-DB), Esterhuysen 14471 ; Kapiteins Kloof Piketberg (-DC), Pil-

lans s.n. 3219 (Wuppertal): Boschkloof Cedarberg Mtns (-DC),

Pocock 30286 (PRE) and 4476; Cedarberg Mtns (-AC) Martin &

Steytler 66387 (NBG); Olifants River Mtns (-CA/DB), Esterhuy-

sen 15272. 3318 (Cape Town): Moorreesburg distr. (-BA), Herre

3942. 3319 (Worcester): Saron distr. (-AA), Andrag s.n.; Kraka-

douw Peak (-AA) Esterhuysen 15008 (BOL); Hansiesberg

(-AB), Esterhuysen 25716; Witzenberg Flats (-AC), Marloth

1705; Tulbagh Kloof (-AC), Ross-Frames 20369 (BOL); Elands-

kloof, Ceres (-BD), Barker 4477; Hex River Mtns (-BD), Botha

15426; Esterhuysen 7795; 7805; Bainskloof (-CA), Barker 2056;

4539; Du Toits Kloof (-CA), Thompson 1159; Elandskloof off Du

Toits Kloof (-CA), Van Niekerk & Esterhuysen 16452 ; Brandvlei

Lake (-CB), Van Breda 1577; April Peak, Wemmershoek Mtns

(-CC), Esterhuysen 16819; Franschhoek distr. (-CC), Ridley 3897

(NBG); Zachariashoek nr Wemmershoek Dam (-CC), Taylor

4777; near De Dooms (-CD) FI. Bolus 13199; Tweefontein nr De

Wet Stn (-DA), Van Breda 1207; Langvlei Quarry (-DB), Winter

142; Robertson distr. (-DD), Marloth 8414. 3320 (Montagu):

Hillandale, Matjiesfontein (-BA), Logan 37182 (PRE) & 66397

(NBG); Kogmanskloof (-CC), Hall 2736; Shale hills nr Bonnie-

vale (-CC), Marloth 11995; Anysberg (-DB), Hall 308 (NBG);

Wurts 1532 or 1353; Burghers Pass (-DC), Goldblatt 1684; Tra-

douw Pass (-DC), Goldblatt 1696; Montagu distr., Isaacs 461;

Salter 2114. 3321 (Ladismith): S side of Klein Swartberg

(-AC/AD), Wurts 1341; Seven Weeks Poort (-AD); Compton

13220; Swartberg Forest Reserve (-BD), Taylor 4720 or 2740.

3419 (Caledon): Happy Valley nr Greyton (-BA), Barker 2055;

Esterhuysen 5070; Rjviersonderend Mtns (-BA/BB), Lewis 3011

and 3012. 3420 (Bredasdorp): De Hoop Reserve Potberg (-AD),

Burgers 1940; Duiwelshokrivier between Vermaaklikheid and

Fort Beaufort (-BD), Lewis 5943; Strawberry Hill Heidelberg

(-DD), Barker 8953. 3421 (Riversdale): Jan Muller Bridge Gou-

rits River (-BA), Barker 9242; Zanddrift, Muir 889.

ACKNOWLEDGEMENTS

I wish to record my very grateful thanks to Miss

W. F. Barker, Dr L. E. Codd, Mrs A. A. Mauve and

Mr J. Robert Sealy (Kew) for all their assistance and

kindly advice in the preparation of this paper.

K. H. DOUGLAS*

* 6 Oatlands Road, Grahamstown 6140.

Bothalia 15, 3 & 4 (1985)

ARACEAE

A NEW RECORD OF ACORUS CALAMUS IN SOUTH AFRICA

547

In a vegetation survey, recently undertaken in the

Mooi River basin in the western Transvaal, Acorus

calamus L. was discovered growing in several colo-

nies within a 700 m-long stretch just above the Bos-

kop Dam area. This is a new record for South Africa

and the first for the Transvaal. Prior to this collec-

tion ( Ubbink 1188), apparently only one record

existed in South Africa and that was from Stellen-

bosch in the south-western Cape ( Marloth 8262),

where it has subsequently not been relocated.

The total area covered by the Mooi River colonies

is approximately 500 m2. The habitats in which A.

calamus thrives are perennial, shallow, slow-flowing

or stagnant water adjoining streambank edges and a

wet swamp area, where it occurs together with more

widespread riverine and swamp species. It grows ex-

tremely well on disturbed streambank edges and

could be regarded as a pioneer plant. This corre-

sponds with findings in Europe (Westhoff et al.,

1970).

In the Mooi River, the plant has invaded two dis-

tinct areas: (a) the area between and bounded by

Phragmites australis and Schoenoplectus corymbo-

sus, a niche usually occupied by Typha latifolia and

(b) the area at the edge of the lee side streambank,

extending into shallow water and adjoining marsh —

this, in the absence of the Phragmites, Typha and

Schoenoplectus complex.

How A. calamus came to be present in the Mooi

River is not known, but it is postulated that the plant

was brought in and planted for its presumed med-

icinal properties by some farmer living near the

river. During the colonization of the Cape of Good

Hope the plant, known as Makkalmoes in South

Africa (Smith, 1964), was introduced for its med-

icinal uses and was also sometimes planted in ponds

because of its aromatic properties (Marloth, 1917).

It is clear that A. calamus was overlooked for a con-

siderable time in the Mooi River area. At a distance

there is a striking similarity in appearance between

A. calamus and the sedge Carex cernua. This, to-

gether with the often inaccessible terrain, may be the

reason why the plant was overlooked in the past.

Acorus calamus is easily recognized by the

strongly pungent aromatic oils present in all parts of

the plant. The leaves are sword-shaped and 0,8-2 m

long; the spadix appears lateral and the spathe is

leaf-shaped and green like the spadix. No fruits have

been observed.

If it is accepted that the Mooi River plants orig-

inated directly or indirectly from Europe, we may be

dealing with a triploid strain (Wulff, 1954). This

strain does not produce fruits and multiplies by divi-

sion of the rootstock (Westhoff et al., 1970). The

question arises as to whether A. calamus should be

regarded as a potential weed or as a possibly import-

ant and welcome pioneer which would serve to vege-

tate and safeguard disturbed streambank edges in

our perennial river systems.

The origin of A. calamus, its spread over much of

the world and its economic uses as described by

Grieve (1967), make interesting reading. A. calamus

is commonly known as Sweet Flag (because of its re-

semblance to Yellow Flag, Iris pseudacorus L.),

Sweet Sedge, Calamus, Kalmoes or Kalmus and

there are other names as well. The names Kalmoes

or Kalmus are applied in South Africa to several

species belonging to the Apiaceae (Wicht, 1918).

According to Grieve (l.c.), A. calamus has become

naturalized in the northern hemisphere, but orig-

inates from eastern countries, being indigenous to

the marshes of the mountains of India. She also

states that it appears to be indigenous to the north-

ern parts of the USA.

A. calamus is well known for its medicinal and

other properties. All parts of the plant, but particu-

larly the rootstocks, contain a volatile, aromatic oil,

which gives off a pleasant fragrance. Medicinally the

plant is used for flatulence (as a carminative), dys-

pepsia, typhoid, dysentry, fever, catarrh, bronchitis

and many other ills. It is also used as a vermifuge

and insecticide in Sri Lanka and India. The volatile

oil, known as Calamus oil, is largely used in perfum-

ery. It is sometimes used as an additive to gin, beer

and wine. According to Grieve, the fragrance that

makes the leaves attractive to humans, renders them

distasteful to cattle, which avoid the plant.

REFERENCES

GRIEVE, M., 1967. A modern herbal. New York, London:

Hafner.

MARLOTH, R., 1917. The flora of South Africa. London: Wil-

liam Wesley.

SMITH, C.A., 1966. Common names of South African plants.

Mem. bot. Surv. S. Afr. 35, pp. 642.

WESTHOFF, V., BARKER, P. A., VAN LEEUWEN, C. G. &

VAN DER VOO, E. E., 1970. Wilde Planten, flora en ve-

getatie in onze natuurgebieden. 1 & 2. Nederland: Vereni-

ging tot behoud van Natuurmonumenten in Nederland.

WICHT, W. F., 1918. South African huismiddels. S. Afr. med.

Rec. 16: 306-310.

WULFF, H. D., 1954. Zur Zytologie, geographischer Verbrei-

tung und Morphologie des Kalmus. Arch. Pharm., Berl.

287: 529-541.

B. UBBINK* and G. J. BREDENKAMP*

* Department of Botany, Potchefstroom University for CHE,

Potchefstroom 2520.

548

ASPARAGACEAE

THE GENUS PROTASPARAGUS IN SOUTHERN AFRICA

Bothalia 15, 3 & 4 (1985)

Historical notes

The oldest known illustration of a species of Pro-

tasparagus is the Plate 817 in Simon van der Stel's

‘ Journey to Namaqualand, 1685’ . The plant was de-

scribed as ’Asparagi silvestris met bruin-roode bes-

sen den 6de September gevonden’. It is interesting

to note that the plant was recognized as an Aspara-

gus, for with its pungent, hooked spines it differs

markedly from the edible European species, Aspara-

gus officinalis L. The branchlets, covered by over-

lapping cladodes and the sharp, short, hooked, axil-

lary spines, identify it as Protasparagus rubicundus

(Berg.) Oberm., a Cape species, still common to-

day. Fig. 1.

FIG. 1. — Protasparagus rubicundus (Berg.) Oberm. tab. 817 in

Van der Stel’s Journey to Namaqualand 1685. (Very likely

painted by Hendrik Claudius.)

* A revision of the South African species will appear in Flora of

Southern Africa.

In his Species Plantarum (1753), Linnaeus de-

scribed three species from the Cape, viz A. aethiopi-

cus L., A. retrofractus L. and A. declinatus [now

Myrsiphyllum declination (L.) Oberm.]. He also de-

scribed A. falcatus from India, which is found as far

south as Natal. Lamarck in 1783 described A. afri-

canus and A. stipulaceus. Linnaeus fil. and Thun-

berg variously placed 8 more species (collected by

Thunberg at the Cape), under either Asparagus or

Dracaena. In 1829 Roemer & Schultes in Syst. Veg.

7, dealt with 30 species. Kunth, in 1850 in Enumera-

tio Plantarum Vol. 5, divided the genus into 3 gen-

era, viz Asparagus L., which bears unisexual, dioe-

cious flowers and is endemic to Europe with few ex-

ceptions. For the species with bisexual flowers he

chose the name Asparagopsis, which unfortunately

proved to be a homonym. This was changed to Pro-

tasparagus in Jl S. Afr. Bot. 2,3: 243 (1983). The

third genus, Myrsiphyllum Willd. (1808), was also

resuscitated (cf. Bothalia 15: 77-88, 1984).

The genus Protasparagus with 67 known species in

southern Africa and many more from the rest of

Africa, is believed to be an old genus. Not a single

hybrid was observed.

The genus is divided into two subgenera: 1, Pro-

tasparagus with 11 species and 2, Africani with 56

species.

Subgenus 1. Protasparagus

This subgenus is regarded as the oldest of the sub-

genera. It comprises shrubs usually not more than 1

metre tall, with stems and branches ending in a

spine. The roots do not form tubers (an exception is

P. suaveolens where the roots are swollen occasion-

ally). The flowers are apical or axillary along the

spineless branches but in P. flavicaulis no branchlets

are present. Here the flowers are placed halfway on

the spine (a modified branch) next to the cladode-

fascicles.

Of the 11 species in this subgenus three bear nut-

lets, which is considered to be a more primitive

character; the others, like the rest of the genus, pro-

duce berries. This subgenus is endemic to the Winter

Rainfall Region with the exception of two species,

which have spread further northwards, viz P. sua-

veolens (Burch.) Oberm., which reaches Zimbabwe

and Malawi, and P. flavicaulis Oberm. which occurs

in the Transvaal and neighbouring areas.

Subgenus 2. Africani

This subgenus, with 56 species, is more varied and

has a wider distribution than subgenus Protaspara-

gus. Here the spines are the modified lower part of

the leaf, whereas the upper part takes the form of a

soft scale, which bears the buds in its axil. The spines

point downwards or outwards and are usually very

sharp and hard. Some species, however, for instance

the well-known cultivated P. densiflorus and P.

plumosus, are almost spineless; if spines occur, they

are usually confined to the basal part of the stems.

Bothalia 15, 3 & 4 (1985)

549

The habit of members of this subgenus is more va-

ried than of subgenus Protasparagus and several

species have become scandent, attaining a length of

3 m or more. The subgenus has spread all over

Africa to Asia and some of its species have reached

Australia.

This subgenus can be divided into eight series. In

series Aethiopici and Racemosi the flowers are ar-

ranged in compound, many-flowered or simple ra-

cemes. They often flower early in spring before the

cladodes have matured. However, the flowering

period is short. The flowers in the majority of

species are arranged inside or beside a cladode-fas-

cicle. Their strong sweet scent attracts bees and

other insects.

The perianth consisting of 3 + 3 tepals, usually

shows little variation. In some species however, the

tepals fuse below to form a pericladium. Where the

flower is attached to the pedicel a distinct disk is pro-

duced. In live flowers it is often quite clear that the

attenuated lower part of the pericladium belongs to

the perianth as colour differences can be observed.

The upper part matches the perianth, whereas the

pedicel below the disk is green.

Cladodes. The short green, usually terete ‘needles’

are mostly arranged in fascicles. They have been the

subject of much discussion. Some researchers have

regarded them as leaves, whereas others have con-

sidered them to be stems. McLean & Ivimey Cook in

their textbook The Monocotyledons 1: 925 (1951),

believe them to be flower stalks. Today it is usually

accepted that the cladodes represent ‘short shoots’.

In P. mucronatus (Jessop) Oberm., P. microraphis

(Kunth) Oberm. and P. stellatus (Bak.) Oberm., the

cladodes develop a discoid base or foot, which fits

onto raised disks produced on the receptacle. This

discoid cladode-base is reminiscent of the disk found

below the pericladium at the point of attachment to

the pedicel.

Roots. Some common widespread species bear swol-

len tuberous roots (Series Racemosi) or form tubers

on side roots (Series Aethiopici). These tubers con-

tain a transparent mucilage and sustain the plants in

times of drought. When dug up after a dry winter

only the empty covering of the tubers is found.

Notes on collections from Kenya record that the tub-

ers are used medicinally.

Fruit. In three species of the subgenus Protaspara-

gus, P. recurvispinus, P. bayeri and P. glaucus the

tepals harden and become erect, enclosing the ovary

to form a nutlet. In all the other species a berry is

formed, the trilocular ovary becoming fleshy; the

withered tepals may be present below or disappear.

The red or black succulent berries, are consumed by

birds, which distribute the seeds far and wide. It is

likely that P. racemosus (and also Myrsiphyllum as-

paragoides) have reached Australia via India and In-

donesia, as a result of bird migration.

Seeds. Usually only one, or rarely two globose black

seeds develop, although each of the three locules of

the ovary produces from 4—12 ovules arranged bise-

riately. They possess a phytomelan crust (cf. Dahl-

gren & Clifford, The Monocotyledons 230 (1982).

Horticultural and other uses

The most successful cultivated species is P. densi-

florus (Kunth) Oberm. and its cultivar ‘myersii’.

They are cultivated as perennial garden or pot

plants. The cultivar 'myersii', a recent introduction,

was developed by Mr Myers from East London. The

plants become cylindrical through the suppression of

side branches.

Much used by florists as a green backing for posies

and other displays, is P. plumosus (Bak.) Oberm.,

because of its decorative, triangular, flattened, ever-

green branch systems. P. falcatus (L.) Oberm. is also

cultivated in Europe.

Vegetable. A common species in the southern and

south-eastern Cape, P. multiflorus (Bak.) Oberm.,

has been used as a vegetable in a similar way as the

European Asparagus officinalis L. The bushes are

chopped down at the end of winter. The thick new

shoots are cut off when c. 150 mm long and are pre-

pared as food by boiling. They have a slight nutty

taste, but are not considered as flavoursome as the

cultivated species. P. edulis Oberm. from the east-

ern Transvaal and eastern Orange Free State is also

eaten according to Mrs M. Jacobs of Harrismith.

ACKNOWLEDGEMENTS

When the revision of Protasparagus was under-

taken for the Flora of Southern Africa, it soon be-

came apparent that the genus had been poorly col-

lected, probably because of a short flowering period

and the presence of sharp spines.

Thanks to enthusiastic collaboration of col-

leagues, many new species were collected and com-

plete material, including roots, of many inad-

equately known species was obtained. This helped

greatly to get a better picture of this genus.

Mr M. Bruce Bayer of the Karoo Garden at Wor-

cester became an ardent collaborator of Protaspara-

gus; he discovered several new species as well as

others that were poorly represented in herbaria. His

hospitality and assistance is much appreciated.

Colleagues of the Botanical Research Institute in

Pretoria and Stellenbosch and Mrs Pauline Bohnen

of Riversdale all assisted in enriching our collection.

Special thanks are due to officers of the Forestry De-

partment at George and of the Transvaal Depart-

ment of Nature Conservation, especially Mr S. P.

Fourie, and to many others.

A. A. OBERMEYER*

* Botanical Research Institute, Department of Agriculture and

Water Supply, Private Bag X101, Pretoria 0001.

550

ERICACEAE

A NEW SPECIES OF PHILIPPIA, FROM THE DRAKENSBERG

Bothalia 15, 3 & 4 (1985)

Philippia drakensbergensis E. G. H. Oliver, sp.

nov. Philippiae evansii speciei alterae habitanti in

dracomontibus natalensis affinis, sed ab ea antheris

subbasalibus (haud basalibus), stigmate infundibuli-

formi (haud peltato), indumentoque praesertim eg-

landuloso pilis longioribus glandulosis ad ramos

ramulosque limitatis satis differt.

Frutex vel arbor ad 2,5 m altus. Rami adscen-

dentes pubescentes indumento brevi inconspicuo ad-

mixto pilis crassioribus longioribus glandulosis. Folia

3-nata, 1,5-4, 6 mm longa linearia usque anguste el-

liptica appressa vel erecta usque patentia et reflexa,

pubescentia pilis brevibus crassis, demum sparse pu-

bescentia, marginibus glandulis magnis sessilibus et

subtus glandulis paucis breviter pedicellatis vestita;

petiolo 0,6-0, 8 mm longo sparse puberulo ciliato

glandulis. Flores (2)3-6(8) nati in extremis ramulo-

rum brevium lateralium, aggregatorum sub extremis

ramorum; pedicello ad 1 mm longo glabro usque

sparse pubescenti cum vel sine glandulis paucis;

bractea totaliter recaulescenti; bracteolis deficienti-

bus. Calyx 4-lobatus, lobis fere aequalibus usque

maxime inaequalibus; lobo maximo abaxiali ad 1,7

mm longo late ovato usque elliptico, parte superiore

§ foliiformi, fere libero; lobis lateralibus asymmetri-

cis, lobo adaxiali symmetrico, omnibus tribus

0,6-0, 8 mm longis ovatis; omnibus lobis breviter pu-

bescentibus usque sparse pubescentibus, marginibus

glandulis sessilibus vestitis. Corolla 4-lobata, 1,2-1 ,6

mm longa cyathiformis glabra; lobis longitudine

corollae partes aequantibus. Stamina 8; filamentis li-

beris vel ad basin parum conjunctis 0,5 mm longis

linearibus glabris; antheris inclusis subbasaliter af-

fixis 0,7-0, 8 mm longis muticis scabris, lateribus gla-

bris exceptis; poro parvo longitudine i-s thecae par-

tes aequanti. Ovarium 4-loculare ovulis multis in

quoque loculo, late ovoideum 0,7 x 0,8 mm longis-

trorsum porcatum, pubescens pilis crassis appressis

interdum glandulosis; stylo brevi 0,2-0, 3 mm longo

glabro vel raro sparse pubescenti; stigmate 0, 3-0,4

mm lato infundibuliformi infra pubescenti, interdum

solum paucis pilis. Fructus capsularis loculicidus;

seminibus multis in quoque loculo laevigatis.

TYPE. — Natal Drakensberg, MnWeni Pinna-

cles, 7-8000 ft, July 1953, Esterhuysen 21651 (BOL,

holo.; K; LD; MO; PRE).

Shrub or single-stemmed tree up to 2,5 m tall.

Branches ascending, hairy, with short inconspicuous

indumentum intermingled with stouter longer gland-

tipped hairs. Leaves 3-nate, 1, 5-4,6 mm long, linear

to narrowly elliptic, appressed or erect to spreading

and recurved with an appressed petiole, pubescent

with short stout hairs all over when young becoming

sparsely so, with large sessile glands on the margins

and a few short-stalked glands on the abaxial sur-

face; petiole 0,6-0, 8 mm long, sparsely puberulous,

glandular ciliate. Flowers (2)3-6(8)-nate at the ends

of short lateral branchlets crowded towards the ends

of the branches; pedicel up to 1 mm long, glabrous

to sparsely pubescent, with or without a few gland-

tipped hairs; bract fully recaulescent as the abaxial

lobe of the calyx; bracteoles wanting. Calyx 4-lobed,

almost equal to highly unequal; the largest lobe

abaxial up to 1,7 mm long, broadly ovate to elliptic

with the upper § leaf- like, reaching from half the

length of the corolla tube to above the corolla inter-

stices, almost separate; lateral lobes asymmetrical

and adaxial lobe symmetrical, all three 0,6-0, 8 mm

long, ovate; all lobes shortly hairy to sparsely so and

edged with sessile glands. Corolla 4-lobed, 1,2-1 ,6

mm long, cyathiform, glabrous; lobes g-| the length

of the corolla. Stamens 8; filaments free or slightly

joined at the base, 0,5 mm long, linear, glabrous;

anthers included, sub-basally attached, 0,7-0, 8 mm

long, muticous, adhering but not properly joined,

rough with smooth sides; pore small, 5-3 the length

of the cell. Ovary 4-locular with numerous ovules

per locule, broadly ovoid, 0,7 x 0,8 mm, longitudi-

nally ridged, pubescent with short thick appressed

hairs occasionally gland-tipped around the apex;

style short, 0,2-0, 3 mm long, glabrous or rarely

FIG. 2. — Philippia drakensbergensis. 1; flower; 2, corolla; 3, va-

riation in the recaulescent bract (abaxial segment of the ca-

lyx); 4, lateral sepal; 5, anther, back, side and front views; 6,

gynoecium; all x 20; 7, leaf, X 10. All drawn from the holo-

type, Esterhuysen 21651 (BOL).

Bothalia 15, 3 & 4 (1985)

551

sparsely hairy; stigma large, 0,3-0, 4 mm broad,

deeply funnel-shaped, hairy on the undersurface

sometimes reduced to only a few distinct hairs. Fruit

a loculicidal capsule; seeds numerous per locule,

smooth. Fig. 2.

l his species has remained undescribed since it was

first collected by Dr Amy Jacot Guillarmod in Leso-

tho in 1947. Since then a number of collections have

been made in the Natal Drakensberg by Miss Elsie

Esterhuysen. It was recognized as being similar to P.

evansii N. E. Br., the only known species of Philip-

pia in the Drakensberg.

P. drakensbergensis can easily be distinguished

from P. evansii by the short pubescence on the

leaves, sepals, ovary and stigma and also on the

branches and branchlets which have in addition,

stouter longer gland-tipped hairs. P. evansii, by con-

trast, has an abundance of gland-tipped hairs also on

its leaves. The anthers of P. drakensbergensis are

subbasally rather than basally attached and the pore

is relatively smaller. The shape of the stigma in the

new species is quite different, being funnel-shaped,

whereas in P. evansii it is peltate. The flowers tend

to be more densely clustered near the ends of the

branches than they are in P. evansii. The most no-

ticeable difference in the living state would be the

stickiness of P. evansii plants.

P. drakensbergensis has a different distribution

range and perhaps also ecological preference to P.

evansii. Although the two species grow in the same

general area in the northern Drakensberg of Natal,

P. drakensbergensis appears to favour the higher

reaches of the mountains at altitudes between 2 100

and 2 500 metres. It also occurs in the drier climate

of central Lesotho. Esterhuysen notes in the type

collection ‘where cooler conditions prevail this is

commoner than Erica ebracteata or Philippia evan-

sii’. P. evansii is confined to the eastern side of the

escarpment where it grows, sometimes abundantly,

at altitudes between 1 500 and 2 000 metres. The re-

cent discovery of P. drakensbergensis on the Witte-

berg in the Cape Province extends the distribution

range considerably further southward. Fig. 3.

Dr Jacot Guillarmod notes that the species occurs

along streambanks, especially the north side, in Le-

sotho. I have only seen the species on the Witteberg

in the north-eastern Cape Province, where it was

common on south- and west-facing slopes alongside

the streams or on open dry rocky grassy slopes. In

sheltered places it formed a veritable forest of single-

stemmed trees 2-2,5 m tall, reminiscent of the

philippias in photographs of the East African ericoid

belt on the high mountains'. At an altitude of

2 100-2 500 metres the plants must be subjected to a

substantial covering of snow during winter.

NATAL. — 2828 (Bethlehem): Mont aux Sources area, Linyati

(-DD), Esterhuysen 21681 (BOL). 2829 (Harrismith): MnWeni

Pinnacles (-CC), Esterhuysen 21646 (BOL; K; LD; PRE; S; STE:

UPS); ibid. Esterhuysen 27837 (BOL; PRE); M’bundini (-CC),

Esterhuysen 27809 (BOL; E; MO; NH; NU; NY; P; S; STE).

2929 (Underberg): Cathkin Peak (-AB), Esterhuysen 17361

(BOL; NBG; NU; PRE; STE); ibid. Esterhuysen 20236 (BOL; K;

PRE; STE).

LESOTHO. — 2927 (Maseru): Makhaleng Valley near Quil-

loane Falls (-BD), Hilliard & Bunt 12043 (NU; STE); God Help

Me Pass (-BD), Killick 4226 (BM; NBG; NH; PRE). 2928 (Mara-

kabies): Little Bokong Valley (-AC), Jacot Guillarmod 309

(PRE); Mamalapi (-AC), Compton 21372 (BOL; NBG); ibid. Ja-

cot Guillarmod 1217 (PRE); Putsua Mountains, Pass from Malea-

lea to Simonkeng (-DD), Esterhuysen 13176 (BOL); Ribaneng

Stream (-DD), Esterhuysen 13203 (BOL).

CAPE. — 3027 (Lady Grey): Beddgelert in the Witteberg,

Barkly East distr. (-DA), Hilliard & Burtt 13166 (NU; STE);

ibid. Oliver 8394 (BM; GRA; MO; PRE; STE).

E. G. H. OLIVER*

* Botanical Research Unit, Department of Agriculture and

Water Supply, P.O. Box 471, Stellenbosch 7600.

FABACEAE

A NEW COMBINATION IN TEPHROSIA

In Bothalia 12: 448 (1979), Dr L. E. Codd de-

scribed a new species Mundulea pondoensis from the

Lusikisiki District in Transkei. Gillet, J. B. in

FTEA, Legum. - Papil. 1: 120 (1971) gives the dif-

ference between Mundulea and Tephrosia as fol-

lows :-

Pods coriaceous, not or very tardily dehiscent, ultimately

breaking up irregularly; lateral nerves of leaflets

curved-ascending Mundulea

Pods thinly coriaceous, dehiscent, lateral nerves of leaflets

usually closely parallel, oblique to midrib and

extended to the margin Tephrosia

552

Bothalia 15, 3 & 4 (1985)

To this may be added that leaflets of Mundulea

are generally widest below the middle, ovate or

ovate-lanceolate, whereas those of Tephrosia are

narrowed at the base, widest above the middle and

generally obovate to oblong. The pods of Tephrosia

dehisce, often explosively, the separated valves be-

coming twisted. On this basis, M. pondoensis is bet-

ter placed under Tephrosia.

A number of interesting features, however, separ-

ate it from other southern African representatives of

Tephrosia. These other members are herbs, suffru-

tices or small shrubs of relatively recent semi-arid to

moist savanna, bushveld, thornveld or grassland,

whereas this species is a shrub or often robust tree to

5 m (A.T.D. Abbott, pers. comm.) of dry evergreen

forest margins on moister slopes or drainage lines

(G. R. Nichols, pers. comm.). In addition, it is re-

stricted to the highly endemic Table Mountain Sand-

stone (T.M.S.) outcrop area of Pondoland and

southern Natal.

The habitat, arborescent habit and uncharacteris-

tic flowers which are large and orange, may indicate

a different, perhaps earlier origin than the rest of the

genus in the flora area. Whereas most species of

Tephrosia in southern Africa have deep pink to

mauve flowers, a few, T. marginella H. M. Forbes,

T. elongata E. Mey. and T. linearis (Willd.) Pers.

var. discolor (E. Mey.) Brummitt have small orange

flowers and they may provide some clue on further

study.

It has been hypothesized (A. E. van Wyk, pers.

comm.) that if this species shows sufficient ‘primi-

tive’ features when the genus is compared as a

whole, it may represent a ‘relic’ of a once much

wider and earlier distribution of a tropical flora

adapted to the infertile soils derived from sandstone

of the Msikaba formation, Natal group; closely re-

lated to the Cape supergroup.

Very much the same appears to have happened

within the present distribution of Tephrosia. South-

ern Africa has been a region of considerable radia-

tion for the genus and once again T. bachmannii

Harms has become an endemic in the same region of

Pondoland and southern Natal, although its relation

with the other species is not at all obscure.

Tephrosia pondoensis (Codd) Schrire, comb.

nov.

Mundulea pondoensis Codd in Bothalia 12: 448 (1979). Type:

Transkeii 3129 (Port St Johns): near Dindini Store (-BD), Codd

9318 (PRE. holo.!)

NATAL. — 3030 (Port Shepstone): Oribi Gorge (-CB), Ab-

bott 962, (NH), Nichols 689, (NH), Schrire 1421, (NH); Umtam-

vuna Nature Reserve (-CC), Abbott 987, (NH), Abbott 1009,

(NH).

TRANSKEI. — 3129 (Port St Johns): near Dindini Store

(-BD), Codd 9318, (PRE), Umsikaba, Ndindini (-BD), Strey

10084, (NH).

B.D. SCHRIRE *

* Botanical Research Unit, Botanical Research Institute, Depart-

ment of Agriculture and Water Supply, Botanic Gardens Road,

Durban 4001.

HEPPIACEAE (LICHENES)

CORYNECYSTIS, A NEW LICHEN GENUS FROM THE KAROO, SOUTH AFRICA

Corynecystis Brusse, gen. nov.

Thallus e haptero subfruticosus, pulvinatus, saxi-

cola, plus minusve 3 cm diametro, et 2 cm altus, ad

basim (hapteron) divisus. Lobi sublineares, inflati,

ramosi, 1-3 mm diametris, usque ad 2,5 cm longi,

cylindrici vel irregulares vel interdum complanati.

Pagina externa nigricans, laevis. Cortex (stratum go-

nidiale) paraplectenchymatus. Algae cyanescentes

(Chroococcaceae). Medulla (pagina interna) alba,

10-200 pm crassa.

TYPUS. — Corynecystis capensis Brusse.

Corynecystis capensis Brusse, sp. nov.

Thallus e haptero subfruticosus, pulvinatus, saxi-.

cola, plus minusve 3 cm diametro et 2 cm altus, ad

basim (hapteron) divisus. Lobi sublineares, inflati,

ramosi, 1-3 mm diametris, usque ad 2,5 cm longi,

cylindrici vel irregulares vel interdum complanati.

Pagina externa nigricans, laevis, hebetata. Cortex

(stratum gonidiale) paraplectenchymatus, 30-60 pm

crassus; algae cyanescentes ( Chroococcaceae ). Me-

dulla (pagina interna) alba, 10-200 pm crassa; hy-

phae 2-3 pm crassae, leviter gelatinosae. Ascomata

prima perithecioidea, hyalina, in thallis omnino im-

mersa, globosa, circa 300 pm diametris, dein 400 pm

lata et apothecioidea. Paries (excipulum) hyalinus,

plus minusve (periclinate) prosoplectenchymatus,

20-30 pm crassus, J-. Subhymenium (vel carpocen-

trum) hyalinum, 20-70 pm profundum, J + caeru-

leum. Hymenium hyalinum, circa 200 pm altum, J +

caeruleum. Paraphyses graciles, flexiles, septatae,

remote ramosae et anastomosae vel fere simplices,

leviter gelatinosae. Asci clavati vel botuliformes,

parietibus praecipue ad apicibus versus incrassatis,

J+ caeruleis, cum vaginis gelatinosis, laceratis, hya-

linis, distalibus (generis Peltulae Nyl. similes). Ascos-

porae numerosae (circa 100 vel plures), hyalinae,

simplices, ovales, 4-5 x 5,5-7 pm, J-. Pycnidia glob-

osa, in thallis omnino immersa, hyalina, 100-200 pm

diametris. Pycnidiosporae hyalinae, ellipsoideae,

circa 4 x 1, 8-2,0 pm.

TYPUS. — 3221 (Merweville): Layton, on rock

face in kloof (-BB). D.A.M.B. Shearing 589,

1984.06.25 (PRE, holo.; COLO; LD, iso.).

Thallus subfruticose from a holdfast, pulvinate,

saxicolous, about 3 cm across and 2 cm high, divided

to base (holdfast). Lobes sublinear, inflated,

branched, 1-3 mm in diameter, up to 2,5 cm long,

cylindrical to irregularly inflated, or sometimes flat-

tened. Outer surface charcoal, smooth, matt. Cortex

Bothalia 15, 3 & 4 (1985)

553

(algal layer) paraplectenchymatous, 30-60 pm thick;

algae blue-green (Chroococcaceus). Medulla (inner

surface) white, 10-200 pm thick; hyphae 2-3 pm

thick, lightly gelatinized. Ascomata first perithe-

cioid, hyaline, immersed, globose, about 300 pm in

diameter, becoming 400 pm wide and apothecioid.

Wall (exciple) hyaline, more or less (periclinally)

prosoplectenchymatous, 20-30 pm thick, I-. Subhy-

menium (or carpocentrum) hyaline, 20-70 pm deep,

1+ blue. Hymenium hyaline, about 200 pm high, 1 +

blue. Paraphyses slender, flexible, septate, remotely

branched and anastomosed or nearly simple, lightly

gelatinized. Asci clavate to botuliform; with thick-

ened walls, especially towards the apices; 1+ blue,

with a distal hyaline, lacerate gelatinous sheath (as

in the genus Peltula Nyl.) Ascospores numerous

(about a hundred or more) per ascus, hyaline,

monolocular, oval, 4-5 x 5,5-7 pm, I-. Pycnidia im-

mersed in thallus, hyaline, globose, 100-200 pm in

diameter. Pycnidiospores hyaline, ellipsoid, about 4

X 1, 8-2,0 pm.

This remarkable new lichen has the colour, size

and habit of a Thyrea, except the lobes are inflated.

Thyrea, however, has non-corticate, homoeome-

rous, dorsiventral solid lobes, with Gloeocapsa algae

as the photobiont. The asci contain fewer ascospores

(8-32 each), and lack the lacerate sheaths. Other

subfruticose or fruticose members of the Lichina-

FIG. 4. — Corynecystis capensis. D. Shearing 490. Scale in mm.

ceae (as presently circumscribed) are much smaller.

(Henssen, 1963, 1977, 1980; Poelt, 1969; Zahl-

bruckner, 1926).

Immature thalli of the new lichen (Fig. 4) resem-

ble Dactylina arctica (Hook.) Nyl. in form and size,

but the latter contains green algae, is dull ochre

coloured, and contains depsides and/or depsidones.

This new genus is, however, closest to Peltula Nyl.

both in apothecial characters, and in the structure of

the upper cortex, which is constructural with the al-

gal layer in these cases. The major difference be-

tween Corynecystis and Peltula is the large inflated

lobes of the former. This represents an order of mag-

nitude in size difference from any known fruticose

Peltula, which are also solid within. (Wetmore,

1971; Swinscow & Krog, 1979).

This new lichen is therefore best placed in the

Heppiaceae, and is presently only known from the

type locality, Layton, south east of Fraserburg in the

Cape Province.

CAPE. — 3221 (Merweville): Layton, locally common on rock

(-BB). D. Shearing 490. 1984.02.15 (PRE). Fig. 4.

REFERENCES

HENSSEN, A., 1963. Eine revision der flechtenfamilien Lichina-

ceae und Ephebaceae. Symb. bot. upsal. 18,1: 1-123.

HENSSEN, A., 1977. The genus Zahlbrucknerella. Lichenologist

9: 17^46.

HENSSEN, A., 1980. Problematik der Gattungsbegrenzung bei

den Lichinaceen. Ber. dt. bot. Ges. 92: 483-506.

POELT, J., 1969. Bestimmungsschlussel Europaischer Flechten,

pp. 757. Lehre: Cramer.

SWINSCOW, T. D. V. & KROG, H., 1979. The lichen genera

Heppia and Peltula in East Africa. Norw. J. Bot. 26:

213-224.

WETMORE, C. T. , 1971. The lichen family Heppiaceae in North

America. Ann. Mo. bot. Gdn 57: 158-209.

ZAHLBRUCKNER, A., 1926. Spezieller Teil. In A. Engler &

K. Prantl, Die Naturlichen Pflanzenfamilien, edn 2, Vol. 8:

61-270. Leipzig: Engelman.

F. BRUSSE*

* Botanical Research Institute, Department of Agriculture and

Water Supply, Private Bag X101, Pretoria 0001.

ORCHIDACEAE

A NEW NAME FOR DISA PATENS

Disa lutea Linder, nom. nov.

Ophrys patens L.f., Suppl. 404 (1781). Serapias patens (L.f.)

Thunb., Prodr. 3 (1794). Disa patens (L.f.) Thunb., FI. Cap. edn

1, 85 (1807); Linder in Contr. Bolus Herb. 10: 326 (1981), nom.

illeg. , later homonym for D. patens Sw. in K. svenska Vetensk-

Akad. Handl. 21: 214 (1800). Penthea patens (L.f.) Lindl., Gen.

& Sp. Orch. 362 (1838). Disa tenuifolia Sw. in K. svenska Vetensk-

Akad. Handl. 21: 214 (1800), nom. illeg., superfluous synonym

for Ophrys patens L.f. Type: South Africa, without precise local-

ity, Sparrman s.n. (LINN, holo.l).

H. P. LINDER*

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

554

Bothalia 15, 3 & 4 (1985)

NEW COMBINATIONS IN HERSCHELIANTHE

The discovery that the genus Herschelia is illegiti-

mate (Rauschert, 1983) has necessitated the follow-

ing additional new combinations:

Herschelianthe forficaria (H. Bol.) N. C. An-

thony, comb. nov.

Disa forficaria H. Bol., leones Orch. Austro-Afr. 1: t. 87

(1896). Type: Cape Province, Du Toit's Kloof, Drege 2211b (K,

holo.).

Herschelianthe newdigateae (L. Bol.) N. C. An-

thony, comb. nov.

Disa newdigateae L. Bol. in Flower. PI. Afr. 11: t. 415 (1931).

Type: Cape Province, Knysna, Forest Hall, Newdigate sub Bolus

6327 (BOL, holo.).

Herschelianthe schlechteriana (H. Bol.) N. C.

Anthony, comb. nov.

Disa schlechteriana H. Bol. in Trans S. Afr. phil. Soc. 16: 149

(1907). Type: Cape Province, Riversdale, Garcias Pass, Luyt sub

BOL 10571 (BOL, holo.; BM; BR; K; W).

Herschelianthe barbata (L. f.) N. C. Anthony,

comb. nov.

Orchis barbata L. f., Suppl. 399 (1781). Type: Cape of Good

Hope, Sparrman s.n. (LINN, holo.; S).

Herschelianthe spathulata (L. f.) Rauschert

subsp. tripartita (Lindl.)N. C. Anthony, comb. nov.

Disa tripartita Lindl.. Gen. Sp. Orch. 353 (1838). Type: Cape

Province, Albany, Geelhoutboom, Drege 3577a (K, holo.; P; S).

Herschelianthe lugens (H. Bol.) Rauschert var.

nigrescens (Linder) N. C. Anthony, comb. nov.

Herschelia lugens var. nigrescens Linder in Bothalia 13: 379

(1981). Type: Cape Province, Humansdorp, near Oyster Bay,

Muller s.n. (NBG, holo.).

REFERENCE

RAUSCHERT, S., 1983. Beitrag zur Nomenklatur der Orchida-

ceae. Feddes Repert. 94: 433-471.

NICOLA C. ANTHONY*

* Bolus Herbarium. University of Cape Town, Rondebosch

7700.

PTERIDOPHYTA

TWO NEW TAXA AND A NEW COMBINATION IN SOUTHERN AFRICAN PTERIDOPHYTA

Aspidiaceae

In southern Africa the name Polystichum lucidum

(Burm. f.) Becherer has traditionally been applied

to the few-scaled, forest-dwelling species of Polysti-

chum centred in Cape Province, while similarly the

name Polystichum pungens (Kaulf.) Presl has been

used to denote the high altitude rocky habitat, ferru-

gineous-scaled species. Recent examinations of the

type specimens of both taxa necessitate several

changes:

The type of Asplenium lucidum Burm. f. in Gen-

eva has been interpreted by Morton (Photographs of

Fern Specimens distributed by the U.S. National

Museum no. 3865) as Asplenium adiantum-nigrum:

‘There are two specimens in Geneva, both collected

by Burman and both labelled A. lucidum. This one

. . (a photograph of Asplenium adiantum-nigrum)

. . agrees with the description best . .

The type of Aspidium pungens in Leningrad has

been found to be the forest-dwelling species of Poly-

stichum.

The result of these discoveries is therefore a

change in the application of the epithet ‘pungens’ to

the forest species and, since there is no subsequent

epithet, a new name for the montane species: P.

monticola. Photographs of all the revelant type

specimens are lodged in the Bolus Herbarium.

Polystichum monticola N. C. Anthony &

Schelpe sp. nov.

Rhizoma repens, c. 15 mm diametro, basibus per-

sistentibus stipitum et paleis lanceolatis ferrugineis

breviter laciniatis concoloris castaneo-fasciatisve c.

10 mm longis onustum. Frondes caespitosae ad api-

cem rhizomae, arcuatae; stipes pallido-brunneae,

dense paleis brunneis vel plerumque ferrugineis latis

et angustis vestitus, denique subglaber praeter base

caespitem paleum; lamina herbacea ad tenuiter car-

noso-coriacea, ovato-truncata, c. 350 x 140 mm, bi-

pinnata ad tripinnatifida, pinnae basales aegre re-

ducta; pinnae angustissime ovatae, attenuatae; pin-

nulae primum visum lunatae, aristae non prominens,

supra glabrae, infra paleis capillaceis conspersis ves-

titae; venatio submanifesta; rhachis rhachides secun-

dariaeque paleis ferrugineis laciniatis praeditae. Sori

c. 1-1,5 mm diametro; indusium membranaceum,

erosum, c. 1 mm diametro.

TYPE. — Cape Province, Cape Peninsula,

Devil’s Peak, Dark Gorge, Esterhuysen 26685

(BOL, holo.; B; C; CHR; G; GH; K; M; MO;

NBG; NU; P; PR; PRE; S; STE).

Polystichum pungens sensu Sim, Ferns S. Afr. edn 2: 116, t. 27

(1915).

Rhizome creeping, c. 15 mm in diameter, with

persistent stipe bases and set with lanceolate ferrugi-

neous shortly laciniate concolorous or castaneous-

striped rhizome-scales c. 10 mm long. Fronds tufted

at the front of the rhizome, arching: stipe pale-

brown, thickly set with brown or more usually ferru-

gineous broad and narrow scales, becoming subgla-

brous with age except for a tuft of scales basally;

lamina herbaceous to thinly coriaceous, ovate-trun-

cate, c. 350 x 140 mm, 2-pinnate to 3-pinnatifid, the

basal pinnae only slightly reduced; pinnae very nar-

rowly ovate, attenuate; pinnules appearing lunate,

the aristae not prominent, glabrous ventrally, set

with occasional hair-like scales dorsally; venation

somewhat apparent; rhachis and secondary rhachises

Bothalia 15, 3 & 4 (1985)

555

set with numerous laciniate-based ferrugineous

scales. Sori c. 1-1,5 mm in diameter; indusium mem-

branous, erose, c. 1 mm in diameter.

P. monticola is found on rocky mountain slopes in

shaded habitats, between c. 1 000-2 000 m altitude

in the Cape Province, Transkei, Lesotho, Natal and

Orange Free State.

Vouchers

Dieterlen 695 (PRE; SAM; STE); Esterhuysen

26698 (B; BOL; C; G; GH; K; M; MO; NBG; NU;

PR- PRE; STE); Esterhuysen 35645 (B; BOL; C; G;

GH; K; M; MO; NBG; NU; P; PRE; S; STE); Hill-

iard & Burtt 11795 (NU; PRE); Schlechter 6932

(NBG; PRE).

Blechnaceae

Blechnum australe L. var. aberrans N. C. An-

thony & Schelpe var. nov.; a varietato typico Blech-

num australe L. lamina fertili non reducti et lamina

sterili simili, et soris discretis, ad angulum costae

portatis haud aegre distinguitur.

TYPE. — Cape Province, Stutterheim, Ama-

bele, Hardcastle 297 (NBG, holo.).

Easily distinguished from the typical variety of

Blechnum australe L. by the fertile lamina being sim-

ilar to the sterile lamina, and the sori being discrete

and set at an angle to the costa.

This taxon is therefore similar to Blechnum punc-

tulatum var. krebsii (Kunze) Sim which is also found

in the eastern Cape Province, and further eastwards

into the Transkei and Natal. The lack of both

pointed apices to the pinnae and minute marginal

teeth distinguish it from B. australe var. aberrans.

ISOETACEAE

Isoetes capensis Duthie var. stephansenii (Du-

thie) Schelpe & N. C. Anthony comb, et stat. nov.

Isoetes stephansenii Duthie in Trans. R. Soc. S. Afr. 17: 330

(1929).

N. C. ANTHONY* and E. A. SCHELPE*

* Bolus Herbarium, University of Cape Town, Rondebosch

7700.

x PLEOPODIUM — A PUTATIVE INTERGENERIC FERN HYBRID FROM AFRICA

Sim (1892) described a new variety of the common

epiphytic fern now known as Pleopeltis macrocarpa

(Bory ex Willd.) Kaulf. , naming it var. sinuatum due

to the departures of the frond shape from the normal

subentire form. The extreme irregularity of the lami-

nar outline in the known collections, even on a single

plant, has prompted a reinvestigation of this wide-

spread form. It now seems most probable that this

taxon is not a simple variety but an intergeneric hy-

brid.

A study was made of the available herbarium ma-

terial of Polypodium polypodioides subsp. ecklonii

(Kunze) Schelpe, Pleopeltis macrocarpa and the pu-

tative hybrid. In addition, live colonies of all three

taxa were available for study at the National Botanic

Gardens, Kirstenbosch.

The venation patterns were obtained by soaking

herbarium material in a commercial solution of so-

dium hypochlorite (3,5%) for 2 to 3 days and the

venation then drawn with the aid of a dissecting

microscope and drawing tube.

Frond shape

The gross morphologies of the putative parents

Pleopeltis macrocarpa and Polypodium polypo-

dioides subsp. ecklonii are depicted in Fig. 5A & C

respectively. A range of frond shapes of the putative

hybrid is depicted as Fig. 5B‘-B3; their most obvious

feature is the extreme variation and irregularity of

the lobing. The entire lanceolate outline of P.

macrocarpa and the deeply pinnatifid one of P. poly-

podioides subsp. ecklonii are combined very irregu-

larly, resulting in random sinuations, closely juxta-

posed deltate segments and broadly adnate, elong-

ate lobes. These are usually confined to the lower

portion of the frond, which is in turn often sterile.

Scales

A comparison of the scale morphology of both rhi-

zome-scales and lamina-scales yielded the following

results, depicted in Fig. 6:

P. macrocarpa has relatively short and broad rhi-

zome-scales, with the thickened central area dark

brown, fading towards the apex, and quite narrow

relative to the pale marginal area. The margin itself

is irregularly laciniate-erose. P. polypodioides

subsp. ecklonii has long, narrow rhizome-scales with

an almost black central stripe extending uniformly

almost to the apex, and very narrow pale margins.

The margin itself is erose and lacks the deep irregu-

lar lacerations of P. macrocarpa. To the naked eye,

the rhizome of the Polypodium is effectively darker

and relatively smooth, whereas that of the Pleopeltis

is paler and has a scurfy appearance. The rhizome-

scales of x Pleopodium simiana are intermediate in

having the elongate shape of the scales of the Poly-

podium and the margin and central thickening of the

scales of the Pleopeltis.

The lamina-scales follow the same pattern in

characteristics of the margin and central thickening

(scales of a similar shape were chosen for illustra-

tion). Those of P. polypodioides subsp. ecklonii

cover most of the lower surface of the lamina, c.

80% or more, whereas those of P. macrocarpa are

spaced one or two scale-widths apart, x Pleopodium

simiana has an intermediate covering of scales on the

lower surface, and the presence of very occasional

scales on the upper surface in P. macrocarpa (as op-

posed to the quite glabrous upper surface of P. poly-

podioides subsp. ecklonii) is repeated.

556

Bothalia 15, 3 & 4 (1985)

FIG. 5. — A, Pleopeltis macrocarpa. Natal, Lions River, ‘Braco’, Karkloof, Schelpe 5129 (BOL). B1 3 X Pleopodium simiana: B1,

Schelpe 6039 (BOL).; B2, Esterhuysen 13265 (BOL); B3, Pope 152 (BOL). B4, Pleopeltis macrocarpa mutant, Kenya, South

Kinangop, Isaac 3004 (BOL). C, Polypodium polypodioides subsp. ecklonii, Cape Province, Griqualand East, Zuurberg, Ty-

son 1778 (BOL).

FIG. 6. — Rhizome-scale and two

lamina-scales. A, Pleopeltis

macrocarpa, Natal, Lions

River, Braco’, Karkloof,

Schelpe 5129 (BOL); B, x

Pleopodium simiana, Schelpe

6039 (BOL); C, Polypodium

polypodioides subsp. ecklonii,

Orange Free State, Harri-

smith, Sterkfontein, Blom 7

(BOL).

Bothalia 15, 3 & 4 (1985)

557

Venation

In P. polypodioides subsp. ecklonii the venation is

free (see Fig 7), whereas in P. macrocarpa the veins

anastomose along the whole length of the lamina. In

x Pleopodium simiana the veins anastomose within

the lobes, but not between them, in the more deeply

sinuate portions of the lamina.

Sori

The positioning of the sori in the putative hybrid

follows the pattern of P. macrocarpa when the lob-

ing is shallow, being placed in the upper unlobed

portion of the frond. When the lobes are well-devel-

oped they often bear sori close to the margin, as in

P. polypodioides subsp. ecklonii. The sori them-

selves are intermediate in size and are usually com-

pletely surrounded by the peltate lamina-scales

when young, as in P. polypodioides subsp. ecklonii.

The young sori of P. macrocarpa are covered by

peltate paraphyses, some of which often persist to

maturity, whereas the sori of P. polypodioides

subsp. ecklonii do not have paraphyses. The sori of

the specimens of the putative hybrid examined do

not have peltate paraphyses and the spores are abor-

tive.

A note of interest here is the existence of three

collections of P. macrocarpa with irregularly-shaped

fronds from localities outside the distribution range

of P. polypodioides subsp. ecklonii. Further exami-

nation of these specimens revealed the presence of

peltate paraphyses in the sori and normal spores in

all three. It is postulated that these are mutants of

the normal form — specimens with irregularly

shaped fronds that do occasionally occur in nature,

e.g. the abnormal fronds of species of Blechnum atx-

tenuatum and B. punctulatum illustrated by Sim

(1915, plates 76 and 78). Comparable mutants of

Phyllitis scolopendrium with lobed instead of entire

fronds are well known in horticulture (see Scolopen-

drium vulgare in Lowe, 1872). Fig. 5 B4 is an illustra-

tion of a frond taken from just such a mutant speci-

men from Kenya ( Isaac 3004). The other known

sports were collected above Kirstenbosch on the

Cape Peninsula (Schelpe sub BOL 32597) and in

Uganda (Stauffer 679). The three are similar in that

the random sinuations are superimposed on the

basic P. macrocarpa frond outline and are consid-

ered to be simple enations.

The distribution ranges of the three taxa are

shown in Figs 8-10.

Wagner & Wagner (1975) describe a very similar

polypodiaceous hybrid from Jamaica, Polypodium

x leucosporum Klotz. Its putative parents are Poly-

podium lanceolatum L. (synonymous with Pleopeltis

macrocarpa ) and Polypodium thyssanolepis A. Br.

(a ‘divided-leaved’ taxon with anastomosing vena-

tion). In the case of P. x leucosporum, cytology re-

vealed the presence of both 4x and 5x forms. The

authors note that the spores are abortive and con-

clude that ‘in Jamaica, P. x leucosporum arose as a

sterile cross of tetraploid forms of the parental

species’. No cytological work was carried out on the

three southern African taxa.

x Pleopodium Schelpe & N. C. Anthony.

Pleopeltis H. B. K. ex Willd. x Polypodium L.

x Pleopodium simianum Schelpe & N. C. An-

thony hybr. nov.; filix inter Pleopeltis macrocarpa

(Bory ex Willd.) Kaulf. et Polypodium polypo-

dioides subsp. ecklonii (Kunze) Schelpe quasi inter-

media et verisimiliter ex hybratione harum taxorum

orta, ab ambobus irregulariter formo pinnatifido

frondium et sporis abortivis differt; a P. macrocarpa

paraphysum peltatorum in indusio absento et a P.

polypodioides subsp. ecklonii semper lobis subdelta-

tis et angulatioribus etiam recedit.

TYPE. — Natal, Lions River District, Ever-

glades, Moll 1263 (BOL, holo.; PRE, iso.).

Polypodium lanceolatum var. sinuatum Sim, Ferns S. Afr. edn

1: 202, t. 118 (1892), Ferns S. Afr. edn 2: 279, t. 143 (1915). Syn-

types: South Africa, Cape Province, Tsitsikamma, Atherstone s.n.

(?K), Fordyce Tree, Holland s.n. (NBG 80444!), Boschberg, Mac-

Owan s.n., above Perie Mission Station, Sim s.n., above Evelyn

Valley, Sim s.n.; Natal, Seven Mile Bush, Upper Umkomaas, on

the heights near York, Buchanan s.n.

Pleopeltis macrocarpa forma sinuata (Sim) Schelpe in Contr.

Bolus Herb. 1: 96 (1969).

A fern almost intermediate between Pleopeltis

macrocarpa (Bory ex Willd.) Kaulf. and Polypo-

dium polypodioides subsp. ecklonii (Kunze) Schelpe

and probably arising from the hybridization of these

taxa, it differs from both by the irregularly pinnatifid

FIG. 7. — Venation. A, Pleopeltis

macrocarpa, Natal, Lions

River, ‘Braco’, Karkloof,

Schelpe 5129 (BOL); B, x

Pleopodium simiana, Schelpe

6039 (BOL); C, Polypodium

polypodioides subsp. ecklonii,

Cape Province, Hogsback,

Schelpe 5014 (BOL).

558

Bothalia 15, 3 & 4 (1985)

FIG. 8. — Distribution range of

Pleopeltis macrocarpa in

Africa.

FIG. 9. — Distribution range of x

Pleopodium simiana.

Bothalia 15, 3 & 4 (1985)

559

FIG. 10. — Distribution range of

Polypodium polypodioides

subsp. ecklonii.

lamina shape and abortive spores; from P. macro-

carpa it differs also in the lack of peltate paraphyses

in the indusium, and from P. polypodioides subsp.

ecklonii by the lobes being most often subdeltate and

more steeply angled.

Rhizome creeping, c. 2-3 mm in diameter, set

with peltate, laciniate-lacerate, ovate-lanceolate,

pale rhizome-scales c. 3 x 0,7 mm with a central,

dark brown stripe. Fronds spaced 20-25 mm apart;

stipe set with peltate, rounded to ovate-lanceolate

scales, becoming subglabrous with age; lamina thinly

carnose-coriaceous, c. 140 x 20 mm, sinuate to

deeply pinnatifid, often only in the lower half, ir-

regular, the segments unequally deltate or the elon-

gate segments adnate, set at an angle to the costa,

upper half often subentire to very shallowly sinuate

around the sori, set with peltate, erose-lacerate,

rounded, dark-centred scales less than 1 mm in di-

ameter dorsally, ovate-lanceolate towards the costa,

and very occasional similar scales ventrally; veins

anastomosing in groups within the segments or

throughout the non-pinnatifid portions. Sori borne

in two rows, one on either side of the costa, in the

upper parts of the lamina, or borne in the lobes,

close to the margins of the longer lobes, oval, non-

paraphysate; spores abortive.

Known localities of the putative hybrid:

ZIMBABWE. — 1831 (Marandellas): Wedza Mountain (-DC)

Burrows 2924 (BOL). 2030 (Fort Victoria): Belingwe. Mt Buhwa

(-CB) Pope 955 (BOL; PRE; SRGH). 2031 (Bikita): Mt Horzi

(-BA) Pope 152 (BOL; PRE; SRGH).

TRANSVAAL. — 2329 (Petersburg): Houtboschberg (-DD)

Schlechter 163 (PRE), Schlechter 4452 (BOL). 2330 (Tzaneen):

Woodbush (-CC) Schelpe 6039 (BOL). 2430 (Pilgrims Rest):

Mariepskop, Bedford Footpath (-DB) V.d. Schijff 4958, 5580

(PRE); Graskop area, Erasmus Kop (-DB) Hardcastle 51 (PRE).

2530 (Lydenburg): Lydenburg, Coromandel Farm (-AB) Bur-

rows 3076 (BOL); Long Tom Pass (-CA) Balsinhas & Kersberg

2126 (PRE); Long Tom State Forest, Sabie (-CA) Burrows 3242

(BOL): Stella Mine (-DB) V. Jaarsveld311 (NBG, PRE).

NATAL. — 2830 (Dundee): Umsinga (-DA) Buchanan s.n.

2929 (Underberg): Hlatikulu Forest (-BA) Killick 1955 (PRE);

Impendhle, Seven Mile Bush (-DB) Sim; Impendhle (-DB) Ran-

dles 80 (NU). 2930 (Pietermaritzburg): Everglades (-AC) Moll

1263 (BOL; PRE); Umgeni above Midmar (-AC) Moll 1240

(BOL; NU; PRE); Zwartkop (-CB) Hill 161 (PRE), Sim sub CH

4231 (PRE), Lawson 204 (NU).

CAPE. — 3225 (Somerset East): Boschberg (-DA) MacOwan

s.n. 3226 (Fort Beaufort): Fordyce Tree (-CB) Holland s.n.

(NBG 80444); Hogsback (-DB) Esterhuysen 13265 (BOL), Giffen

511 (PRE). 3227 (Stutterheim): above Evelyn Valley (-CA) Sim;

Hogsback, Guncuka Forest (-CA) Roux 514 (NBG); Frankfort

(-CB) Sim 446 (PRE); Perie Forest (-CB) Sim 445, 447, 675

(PRE), Flanagan FH 54 (PRE); above Perie Mission Station

(-CC) Sim. 3325 (Port Elizabeth): Van Staadensberg (-CC)

MacOwan s.n. 3423 (Knysna): Knysna (-AA) Pappe s.n.; Tsitsi-

kamma (-BB) Atherstone s.n.

REFERENCES

LOWE, E. J., 1872. Ferns: British and Exotic 5. London: George

Bell.

SIM, T. R., 1892. Ferns of South Africa edn 1. Juta.

SIM, T. R., 1915. Ferns of South Africa edn 2. Cambridge Uni-

versity Press.

WAGNER, W. H., 1962. Irregular morphological development

in hybrid ferns. Phytomorphology 12: 87-100.

WAGNER, W. H. & WAGNER, F. S., 1975. A hybrid polypody

from the New World tropics. Fern Gaz. 11: 125-135.

NICOLA C. ANTHONY* and E. A. SCHELPE*

* Bolus Herbarium, University of Cape Town. Rondebosch

7700.

Bothalia 15, 3 & 4: 561-566 (1985)

Leaf anatomy of the South African Danthonieae (Poaceae). X. Pseudo -

pentameris

R. P. ELLIS*

Keywords: Danthonieae, leaf anatomy, Pseudopentameris

ABSTRACT

The leaf blade anatomy of Pseudopentameris macrantha (Schrad.) Conert and P. brachyphylla (Stapf) Conert

was studied both in transection and in surface view and is illustrated by means of photomicrographs. The leaf

anatomy is typically danthonoid. The abaxial epidermis consists of inflated, hexagonal long cells and stomata and

micro-hairs are absent. Adaxial micro-hairs were observed. The creation of a new genus to accommodate these

two species appears justified but certain species in Pentameris Beauv., Merxmuellera Conert and Pentaschistis

Stapf show similarities to Pseudopentameris and consideration should be given to their inclusion in Pseudopenta-

meris.

INTRODUCTION

Pseudopentameris Conert was described as re-

cently as 1971 to accommodate two species pre-

viously placed in Danthonia DC., namely Pseudo-

pentameris macrantha (Schrad.) Conert and P. bra-

chyphylla (Stapf) Conert (Conert, 1971). De Wet

(1956) was also of the opinion that these two species

deserved generic rank but refrained from describing

a new genus as a complete comparative anatomical,

cytological and morphological study of the genus

Danthonia was not available.

The species assigned to Pseudopentameris have 2-

flowered spikelets over 40 mm long and differ con-

siderably from all the other South African species

previously referred to the genus Danthonia. Several

authorities, including Nees, Kunth and Adamson

(Chippindall, 1955), were aware of these differences

and had placed these species in the genus Pentameris

Beauv. Conert (1971) notes that Pseudopentameris

differs from Pentameris in having many-nerved

glumes, the caryopsis structure is different and the

leaf anatomy is distinct. Chippindall (1955) mentions

that the style is greatly reduced, whereas in Penta-

meris, it is short but there is little else in the spikelet

morphology of Pseudopentameris macrantha and P.

brachyphylla to suggest affinity with Pentameris.

De Wet (1956) considers Pseudopentameris to be

closely related to both Danthonia and Pentameris

but Conert (1971) is of the opinion that the genus

occupies an isolated position with no obvious re-

lationships with other danthonoid grasses. This opin-

ion is confirmed by Renvoize (1981) who includes

Pseudopentameris in the peripheral genera of his Ar-

undineae which includes all the genera here consid-

ered as belonging to the Danthonieae. However, in

his multivariate analysis Pseudopentameris is

grouped with the ‘core’ genera of the Arundinoi-

deae.

P. macrantha and P. brachyphylla are both rather

robust, tufted perennials being easily distinguished

from each other by the pronounced rolling or curling

Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

of the lower leaves of P. brachyphylla. The genus is

confined to the south-western Cape Province from

Piketberg in the north to Riversdale in the east. P.

brachyphylla is very rare and occurs mainly in the

Hottentots Holland range where it sometimes occurs

together with P. macrantha. P. macrantha is particu-

larly common on the Cape Peninsula. Both species

appear to favour sandy habitats at the base of hill

slopes, often in seepage areas but, in the De Hoop

area, P. macrantha grows in crevices in limestone

rock.

The leaf anatomy of Pseudopentameris has re-

ceived scant attention in the past. De Wet (1954,

1960) reported that the chlorophyll is uniformly dis-

tributed throughout the mesophyll, bicellular hairs

are present, the epidermis consists of thin paren-

chyma and the costal silica cells are dumb-bell

shaped. He is of the opinion (De Wet, 1956) that the

species now included in Pseudopentameris combine

in their leaf anatomy the panicoid type of epidermis

(with micro-hairs) and the festucoid type of chloro-

phyll distribution. Renvoize (1981), on the other

hand, considers micro-hairs to be absent in Pseudo-

pentameris but does not consider the genus to belong

with the pooid grasses because ‘the subsidiary cells

are low dome-shaped, the silica bodies are saddle-

shaped or oblong and the long cells may be straight

or sinuous-walled’. These features favour the placing

of this genus in the Arundinoideae.

Conflicting reports have, therefore, been made in

the past regarding the leaf anatomy of Pseudopenta-

meris, particularly as far as the presence or absence

of micro-hairs is concerned. This paper reports on a

detailed study of a representative sample of 24 speci-

mens, most of which were freshly fixed in the field

and, as a result, should enable anatomical character-

istics to be confidently evaluated in future consider-

ations of the relationships and taxonomy of Pseudo-

pentameris. The standardized terminology of Ellis

(1976, 1979) is used in the descriptions together with

the following abbreviations:

vb/s — vascular bundle/s

Tvb/s — first order vascular bundle/s

2’vb/s — second order vascular bundle/s

3’vb/s — third order vascular bundle/s

ibs — inner bundle sheath; mestome sheath

obs — outer bundle sheath; parenchyma sheath

562

Bothalia 15, 3 & 4 (1985)

ANATOMICAL DESCRIPTION OF PSEUDOPENTAMERIS

CONERT

Leaf in transverse section

Leaf outline : broadly U-shaped (Figs 2 & 3) with

arms of lamina convex (Fig. 1) in P. macrantha :

leaves tend to be expanded and flat in P. brachy-

phylla (Figs 14 & 17). Ribs & furrows: similar adax-

ial ribs present over and furrows between all vbs;

furrows of medium depth, either cleft-like (Fig. 4) or

more open (Figs 5 & 6); ribs with rounded sides and

flat tops in both species. No abaxial rib develop-

ment. Median bundle : not structurally different from

other l’vbs; distinguishable by location only (Figs 1,

14 & 17). Vascular bundle arrangement : always 5

l’vbs in leaf section in P. brachyphylla (Figs 14 & 17)

but P. macrantha has either 5 (Fig. 1) or 7 or 9 (Figs

2 & 3). The arrangement of different orders of vbs

along the width of the blade is regular from midrib to

margin with 1 3’vb located between consecutive

l’vbs; sometimes the secondary bundle may be suffi-

ciently developed to be termed a 2’vb (Fig. 17) and

irregularities in the pattern of arrangement may ap-

pear near the margin (Fig. 4). All vbs are located in

the centre of the blade or are slightly displaced abax-

ially. Vascular bundle structure : 3’vbs circular to el-

liptical with phloem well developed (Figs 5, 6 & 16).

l’vbs elliptical with phloem adjoining the ibs; very

narrow metaxylem vessels; lysigenous cavity and

protoxylem vessels developed. Vascular bundle

sheaths: obs elliptical to almost round; entire around

some 3’vbs (Figs 6 & 15) or with abaxial interruption

(Figs 5 & 16); l’vbs with both adaxial and abaxial

interruptions; no extensions. 10-14 cells comprise

obs; cells not well differentiated from chlorenchyma

cells although obvious due to absence of chloroplasts

FIGS 1-7. — Pseudopentameris macrantha : transverse sections of the leaf blade. 1-3, outline of the leaf, x 100: 1, Ellis 2526; 2, Ellis

2512; 3, Ellis 2515; 4, leaf margin structure, Ellis 2526, x 250. 5-7, detail of vascular bundles and mesophyll: 5, compact,

isodiametric chlorenchyma, Ellis 2338, x 250; 6, isodiametric chlorenchyma, Ellis 2515, x 400; 7, unlignified sclerenchyma

girders and chloroplast arrangement, Ellis 2265, x 400.

Bothalia 15, 3 & 4 (1985)

563

and slightly thickened walls. Ibs complete around

l’vbs; cells with uniformly thickened walls (Figs 5, 7,

15 & 16). Sclerenchyma : adaxial girders associated

with all l’vbs; equidimensional in P. macrantha (Figs

5, 6 & 7) but T-shaped in P. brachyphylla (Figs 15 &

16). Abaxial girders equidimensional in both

species. Fibres with thickened walls but composed

mainly of cellulose and do not stain red with saffra-

nin. Margins with rounded sclerenchyma cap in P.

macrantha (Fig. 4) but no sclerenchyma developed

in association with the margin in P. brachyphylla

(Figs 14 & 17). Mesophyll: chlorenchyma not ra-

diate; consists of tightly packed, angular, regular,

isodiametric cells (Figs 5, 6 & 16); these chloren-

chyma cells with characteristic central vacuole with

peripheral chloroplasts. No colourless cells. Adaxial

epidermis : fan-shaped groups of bulliform cells situ-

ated at bases of furrows. No macro-hairs, prickles,

hooks or papillae seen but adaxial micro-hairs com-

monly visible in transections of P. brachyphylla (Figs

15 & 16). Abaxial epidermis : no bulliform cells; con-

sists of very conspicuous large, regular, inflated cells

with outer tangential wall slightly thickened. No ap-

pendages.

Abaxial epidermis in surface view

Intercostal long cells: usually elongated but some-

times length is only slightly greater than width (Fig.

12); side walls always angled or bowed outwards giv-

ing cells a hexagonal or inflated appearance; end

walls vertical; walls either slightly undulating (Fig.

18) or not undulate (Fig. 12) and unthickened. Cell

shape and size is noticeably consistent throughout all

intercostal zones. Pairs of short cells are present be-

tween successive long cells. No abaxial bulliform

cells. Stomata: absent on abaxial surface of both P.

macrantha and P. brachyphylla in all specimens

examined except Ellis 2515 (Fig. 8); low dome-

shaped. Intercostal short cells: silico-suberose

couples with silica cell tall and narrow with smooth

outline or kidney shaped (Figs. 12, 18 & 19); short

cell larger than silica cell and enfolding it; located

between virtually all intercostal long cells. Papillae:

absent. Prickles & hooks: not present on any prep-

arations examined. Micro-hairs: not present on

abaxial surface. Macro-hairs: absent. Costal silica

bodies: equidimensional to horizontally elongated

dumb-bell shaped (Figs 12, 13, 18 & 19); alternate

with costal long cells; costal zones narrow (3-5 files

wide).

Specimens examined:

Pseudopentameris brachyphylla

CAPE. — 3418 (Simonstown): Platberg, Kogelberg State

Forest (-BD), Ellis 2343, 2344, Boucher 357a. 3419 (Caledon):

Lebanon Forest Reserve, Grabouw (-AA), Kruger 149; Bredas-

dorp (-BD), Acocks 22476.

Pseudopentameris macrantha

CAPE. — 3218 (Clanwilliam): Versveld's Pass, Piketberg

(-DC), Ellis 1171. 3318 (Cape Town): Kirstenbosch, Table

Mountain (-CD), Ellis 2308, Sandwith 73, White 5518; Heuning-

vlei, Jonkershoek (-CD), Ellis 2265, 2266. 3418 (Simonstown);

Cape Point Nature Reserve (-AD), Ellis 2326, 2327; Cape

Hangklip (-BD), Ellis 2338, 2339; Elephant Rock near Bettys

Bay, Cleghorn 2495. 3419 (Caledon): 5 km from Kleinmond on

road to Hermanus (-AC), Ellis 2515; 10 km from Flermanus on

road to Gansbaai (-AD), Ellis 2512. 3420 (Bredasdorp): De

Hoop Nature Reserve (-AD), Ellis 1288, 2526; 5 km from Wyd-

gelegen on road to De Hoop, Ellis 1660; Breede River mouth

(-BD), Ellis 1673; 3 km from Arniston on road to Bredasdorp

(-CA), Ellis 1274. 3421 (Riversdale): Albertina (-BA), Ellis

2551.

DISCUSSION AND CONCLUSIONS

The contradictory statements in the literature re-

garding the presence or absence of micro-hairs in

Pseudopentameris deserve comment in the light of

the observations of the present study. No micro-

hairs were observed on the abaxial epidermis of any

of the 24 specimens examined in this study and this

observation is in agreement with the statement by

Renvoize (1981) that micro-hairs are absent. How-

ever, in the leaf transections of all the specimens of

P. brachyphylla, and in many of the sections of P.

macrantha, micro-hairs are clearly evident on the

sides of the furrows of the adaxial epidermis. (Figs

15 & 16). The emphasis placed on the absence of

micro-hairs by Renvoize (1981), therefore, is not

justified and Pseudopentameris can be considered as

belonging to the Arundineae together with all the

other South African genera previously placed in the

Danthonieae. Considering it as a peripheral genus,

together with the other anomalous arundinoid gen-

era, is no longer necessary in the light of the obser-

vations of the present study.

This confirmation of the presence of micro-hairs

on Pseudopentameris agrees with De Wet’s (1954,

1956, 1960) observations. However, De Wet (1956)

considers the epidermis of Pseudopentameris to be

panicoid and similar to that of genera such as Schis-

mus Beauv., Chaetobromus Nees and many species

of Pentaschistis Stapf. This is not the case, however,

as Pseudopentameris does not possess numerous

abaxial micro-hairs as do all the above genera. The

abaxial epidermis of Pseudopentameris actually re-

sembles that of Pentameris Beauv. and Merxmuel-

lera Conert. more closely. These are considered as

being festucoid by De Wet (1956) and the characteri-

zation of the epidermis of Pseudopentameris as pan-

icoid does not agree with the observations of the

present study. Of all the South African danthonoid

genera, Pseudopentameris resembles most closely

Pentameris, Merxmuellera and those Pentaschistis

species considered by De Wet (1956) as having the

festucoid type of epidermis. All these taxa lack abax-

ial micro-hairs, costal zones are often not clearly de-

fined and stomata are often absent from this surface.

Superficial morphological similarities between

Pseudopentameris and Pentameris undoubtedly exist

and Nees, Kunth and Adamson considered these

two genera to be congeneric (Chippindall, 1955).

Numerous morphological characteristics distinguish

these two genera (Chippindall, 1955; De Wet, 1956),

however, and Conert (1971) considered these differ-

ences to be of sufficient magnitude to warrant the

establishment of a new genus. He is of the opinion

that no close and obvious relationship exists between

Pseudopentameris and any other genera known to

him.

The leaf anatomy corroborates Conert’s (1971)

decision. The anatomy of Pseudopentameris differs

564

Bothalia 15, 3 & 4 (1985)

significantly from that of most Pentameris species

and anatomical indications are that Pseudopentame-

ris undoubtedly warrants separate generic status and

that it occupies a somewhat isolated phylogenetic

position nearest to Pentameris and Merxmuellera.

This contention appears to be substantiated by the

fact that there are some species in both the latter

genera possessing a large degree of anatomical simi-

larity to Pseudopentameris , indicating a somewhat

intermediate relationship for Pseudopentameris.

Pentameris dregeana Stapf, P. macrocalycina

(Steud.) Schweick., P. obtusifolia (Hochst.)

Schweick. and P. thuarii Beauv. show little anatom-

ical similarity to Pseudopentameris. However, Pen-

tameris longiglumis (Nees) Stapf and an undescribed

species represented by Ellis 2342, Taylor 3023 and

Haynes 770, bear rather close resemblances to Pseu-

dopentameris. Both these species have large, in-

flated abaxial epidermal cells which are hexagonal in

shape in surface view. Stomata and micro-hairs are

also absent and the costal zones are somewhat indis-

tinct. However, the silica bodies are not dumb-bell

shaped as in Pseudopentameris.

Both these two Pentameris species have been

found only in the Caledon District in the Kogelberg

State Forest. They grow together in the same com-

FIGS 8-13.. — Pseudopentameris macrantha : abaxial epidermis. 8-9. arrangement of costal and intercostal zones, x 100: 8, costal

and intercostal zones overlying first and third order vascular bundles, note presence of stomata, Ellis 2515; stomata absent,

Ellis 2526. 10-1 1 , regular intercostal long cells and narrow costal zones, x 160: 10. Ellis 1 28S ; 11. Ellis 1673. 12-13. detail of

epidermal cells, x 250: 12, intercostal long cells separated by silico-suberose couples and costal silica bodies irregularly dumb-

bell shaped. Ellis 2551; 13. silica bodies dumb-bell shaped and long cells slightly sinuous, Ellis 2265.

Bothalia 15, 3 & 4 (1985)

565

munity, often also in association with both Pseudo-

pentameris brachyphylla and P. macrantha. The un-

described species, in particular, appears to be inter-

mediate between Pentameris longiglumis and Pseu-

dopentameris and a hybrid origin appears most

likely. The spikelet morphology of Pentameris longi-

glumis requires detailed reassessment in order to es-

tablish its true relationships with either Pentameris

or Pseudopentameris.

In the genus Merxmuellera there is also a group of

species which differs significantly from the remain-

der of the genus in possessing epidermal long cells

with outwardly bowed walls. This group includes M.

decora (Nees) Conert, M. lupulina (Thunb.) Conert

and M. rufa (Nees) Conert and their anatomy has

been described in detail (Ellis, 1983). All these

species differ from Pseudopentameris in having

abaxial stomata and micro-hairs. However, stomata

are absent from some specimens of Merxmuellera

decora and micro-hairs are exceedingly rare in all

three species. It must also be mentioned that the in-

flated nature of the abaxial long cells is only evident

in surface view and in transection the epidermal cells

are not as large and inflated as in Pseudopentameris.

All the above species do, however, have very little

lignin present in the fibres of the sclerenchyma gir-

ders — a noticeable feature shared with Pseudopen-

tameris but not with other danthonoid grasses in gen-

eral.

Finally, Merxmuellera lupulina , in particular,

bears a strong anatomical resemblance with Pentas-

chistis involuta (Steud.) Adamson and P. viscidula

(Nees) Stapf (Ellis, 1983). These species also have

inflated epidermal cells and are exceptional in the

genus due to this fact. They, therefore, deserve con-

sideration together with Merxmuellera lupulina and

its allies.

Anatomical criteria, therefore, suggest that the

above-mentioned Merxmuellera, Pentameris and

Pentaschistis species require taxonomic re-evalua-

tion in conjunction with Pseudopentameris. This is

essential in order to establish the true relationships

of Pseudopentameris and to define the limits of the

above genera. This study of the leaf anatomy of

Pseudopentameris fully supports the creation of a

separate genus to accommodate these two very

closely related species which bear little resemblance

to the majority of the taxa previously classified in

Danthonia as well as Pentameris and Pentaschistis.

However, Pseudopentameris does share certain ana-

tomical characters with a few species from each of

FIGS 14-19. — Pseudopentameris brachyphylla : leaf anatomy and epidermal structure. 14-17, leaf in transverse section: 14, outline

of leaf, Ellis 2344, x 100; 15. detail of mesophyll and vascular bundles, note presence of adaxial micro-hairs (arrowed). Ellis

2344, x 400; 16, mesophyll, vascular bundles and micro-hair (arrowed), Ellis 2343, x 250; 17, outline. Ellis 2343, x 100. 18-19,

detail of abaxial epidermis: 18, sinuous intercostal long cells and dumb-bell shaped silica bodies, Ellis 2344. x 400, 19. intercos-

tal long and short cells, Ellis 2343, x 250.

566

Bothalia 15, 3 & 4 (1985)

these genera but the taxonomic and phylogenetic

significance of these shared characteristics is not cer-

tain at this stage and awaits cytogenetical confirma-

tion. A possible practical solution, that is suggested

by the leaf anatomy, is the incorporaton of all these

taxa with anomalous anatomy in a single genus to-

gether with Pseudopentameris. This would result in

the creation of a genus, which would be anatomically

homogeneous, and would be easily distinguishable

from all the other South African danthonoid genera

on the basis of leaf anatomy.

UITTREKSEL

Die anatomiese struktuur van die blaar van Pseu-

dopentameris macrantha (Schrad.) Conert en P. bra-

chyphylla (Stapf) Conert is bestudeer. Beskrywings

van die blaaranatomie in dwarssnee en van die abak-

siale epidermis word gegee en geillustreer deur middel

van fotomikrograwe. Die blaaranatomie is tipies van

die Danthonieae maar die abaksiale epidermis be-

staan hoofsaaklik uit geswolle, seshoekige langselle

en huidmondjies en mikrohare is afwesig. Adaksiaal

mikrohare is wel waargeneem. Die beskrywing van 'n

nuwe genus vir die twee spesies, wat huidiglik in

Pseudopentameris gehuisves word , blyk geregverdig

te wees maar sekere spesies van Pentameris Beauv. ,

Merxmuellera Conert en Pentaschistis Stapf toon

ooreenkomste met Pseudopentameris en behoort

oorweeg te word vir moontlike insluiting in hierdie

genus.

REFERENCES

CHIPPINDALL, L. K. A., 1955, In D. Meredith, The grasses

and pastures of South Africa. Johannesburg: CNA.

CONERT, H. J., 1971. The genus Danthonia in Africa. Mitt, bot,

StSamml., Munch. 10: 299-308.

DE WET, J. M. J., 1954. The genus Danthonia in grass phyto-

geny. Am. J. Bot. 41: 204-211.

DE WET, J. M. J., 1956. Leaf anatomy and phytogeny in the

tribe Danthonieae. Am. J. Bot. 43: 175-182.

DE WET, J. M. J., 1960. Leaf anatomy and morphology in South

African species of Danthonia. Bothalia 7: 303-310.

ELLIS, R. P., 1976. A procedure for standardizing comparative

leaf anatomy in the Poaceae. I. The leaf blade in transverse

section. Bothalia 12: 65-109.

ELLIS, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in

surface view. Bothalia 12: 641-672.

ELLIS, R. P., 1983. Leaf anatomy of the South African Dantho-

nieae (Poaceae). VIII. Merxmuellera decora, M. lupulina

and M. rufa. Bothalia 14: 197-203.

RENVOIZE, S. A., 1981. The subfamily Arundinoideae and its

position in relation to a general classification of the Grami-

neae. Kew Bull. 36: 85-102.

Bothalia 15, 3 & 4: 567-571 (1985)

Leaf anatomy of the South African Danthonieae (Poaceae). XI. Penta-

meris longiglumis and Pentameris sp. nov.

R. P. ELLIS*

Keywords: Danthonieae, leaf anatomy, Pentameris, Pseudopentameris

ABSTRACT

The leaf blade anatomy of Pentameris longiglumis (Nees) Stapf and that of an undescribed Pentameris species is

described and illustrated by means of photomicrographs. It is shown that the anatomical structure of the leaves of

the new Pentameris species in particular, both transverse sections and abaxial epidermal scrapes, resembles closely

that of species placed in the genus Pseudopentameris. This anatomical resemblance is closer than that with any

Pentameris species. It appears, therefore, as if transfer of this new species to Pseudopentameris is justified on the

anatomical evidence and this indication must be followed up by morphological studies. Pentameris longiglumis

shows close anatomical resemblance to P. macrocalycina (Steud.) Schweick. and P. obtusifolia (Hochst.)

Schweick. and should be classified with these taxa.

INTRODUCTION

Two little-known and rare danthonoid grass

species have been studied anatomically for the first

time. Pentameris longiglumis (Nees) Stapf is re-

presented by only four specimens in the National

Herbarium (PRE) and the other species, as yet

undescribed, by three specimens. All these speci-

mens have been included in this study and, although

they constitute only a small sample, all available ma-

terial has been examined. Fortunately both taxa

were collected and fixed in the field and conse-

quently accurate comparisons have been possible

with other danthonoid grasses included in this study.

The three unnamed specimens are considered to

represent a new species of Pentameris by the agros-

tologists of the National Herbarium (B. de Winter,

pers. comm.). One specimen was originally named

Pseudopentameris macrantha (Schrad.) Conert but

present opinion disagrees and it is felt that these

specimens show little in common with Pseudopenta-

meris and should be referred to Pentameris as a dis-

tinct, new and undescribed species.

Although this Pentameris sp. nov. is obviously

morphologically distinct, Ellis (1985) has drawn at-

tention to the strong anatomical resemblance of this

taxon to Pseudopentameris and considers this like-

ness to be taxonomically meaningful. These appar-

ently conflicting opinions require further substantia-

tion and, consequently, the leaf anatomy of this new

Pentameris species, together with that of P. longiglu-

mis, is here described in detail and copiously illus-

trated. This will enable comparisons to be made with

the anatomy of Pseudopentameris (Ellis, 1985) and

with other Pentameris species (Ellis, in press). In ad-

dition, morphological studies of the ovary and ripe

grain, in particular, are required to reliably establish

the natural relationships of these two taxa of uncer-

tain affinity. This information, together with the

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

anatomical evidence, should confirm the classifica-

tion of these two species in either Pentameris or

Pseudopentameris — two genera which can easily be

separated on the basis of caryopsis structure (Stapf,

1900; Chippindall, 1955; De Wet, 1956; Conert,

1971).

P. longiglumis and the undescribed species (which

will be referred to as Pentameris sp. nov. for con-

venience) are confined to the extreme south-western

Cape Province and appear to occur only on Table

Mountain and the Kogelberg Mountains. Both

species were collected by the author at the same

locality on the Kogelberg and it may, or may not, be

significant that Pseudopentameris brachyphylla

(Stapf) Conert was also present in the same com-

munity and in very close proximity to the two Penta-

meris species. Hybridization is, consequently, not

ruled out by spatial separation and this possibility

must be taken into account when the relationships of

these species are considered.

In the following anatomical descriptions, the ter-

minology of Ellis (1976, 1979) is employed together

with the following abbreviations;

vb/s — vascular bundle/s

l’vb/s — first order vascular bundle/s

3’vb/s — third order vascular bundle/s

ibs — inner bundle sheath; mestome sheath

obs — outer bundle sheath; parenchyma sheath

COMBINED ANATOMICAL DESCRIPTION OF PENTA-

MERIS LONGIGLUMIS AND PENTAMERIS SP. NOV.

Leaf in transverse section

Leaf outline : broadly U-shaped (Figs 4, 5 & 9) to

loosely inrolled (Figs 1 & 11). Ribs and furrows: sim-

ilar adaxial ribs present over all vbs in Pentameris sp.

nov. (Figs 10 & 12) but ribs associated with 3’vbs

smaller than those over l’vbs in P. longiglumis (Figs

2 & 3). Furrows cleft-like, of medium depth. No

abaxial rib development. Median bundle : not struc-

turally distinct from lateral l’vbs; distinguishable by

location only (Figs 1 & 11). Vascular bundle arrange-

ment: 9 or 11 l’vbs in leaf section with a single 3’vb

568

Bothalia 15, 3 & 4 (1985)

FIGS 1-8. — Leaf blade anatomy of Pentameris longiglumis. 1-6, leaf blade in transverse section. 1-3, Ellis 2341: 1, hollow, cylin-

drical outline, x 60; 2, well developed, inflated, abaxial, epidermal cells clearly seen, x 400; 3, u-shaped groups of chloren-

chyma with densely packed, isodiametric cells, X 400. 4, Taylor 7231, x 100. 5-6, Marloth 3063, x 100. 7-8, abaxial epidermis,

Ellis 2341, x 250: 7, inflated long cells without distinct costal zones; 8, long cells filled with air showing thickness of cell walls

and sinuous outer surface.

Bothalia 15, 3 & 4 (1985)

569

located between successive l’vbs. Near the margin a

pair of l’vbs may be located adjacent to one another

in P. longiglumis (Figs 4, 5 & 6). All bundles are

located in the centre of the blade. Vascular bundle

structure: 3’vbs circular to elliptical with well devel-

oped phloem; much smaller in P. longiglumis (Figs 2

6 3) than in Pentameris sp. nov. (Figs 10 & 12).

l’vbs elliptical with phloem adjoining the ibs; very

narrow metaxylem vessels. Vascular bundle sheaths:

obs elliptical; entire around some 3’vbs (Figs 3 & 12)

or with abaxial interruption (Fig. 10); l’vbs with

both abaxial and adaxial interruptions; no exten-

sions in Pentameris sp. nov. (Figs 11 & 12) but P.

longiglumis has well developed abaxial and adaxial

extensions (Figs 2 & 3). Obs cells not well differen-

tiated from chlorenchyma cells; however, they are

distinct due to absence of chloroplasts. Ibs complete

around l’vbs; cells with thicker inner tangential

walls (Figs 2 & 12). Sclerenchyma: girders associated

with all vbs; adaxial girders inversely anchor- or T-

shaped (Figs 3 & 10) with relatively long stems;

abaxial girders trapezoidal to equidimensional.

Fibres with thickened walls but, particularly in Pen-

tameris sp. nov., composed mainly of cellulose sec-

ondary walls. Mesophyll: chlorenchyma not radiate;

consists of tightly packed, angular, isodiametric cells

(Figs 3 & 10); these cells with characteristic central

vacuole and peripheral chloroplasts. No colourless

cells. Adaxial epidermis: fan-shaped groups of bulli-

form cells situated at bases of furrows (Figs 3 & 10);

bulliform cells better developed in Pentameris sp.

nov. than in P. longiglumis. In P. longiglumis epi-

dermal cells papillate and with many prickles (Fig.

2); in Pentameris sp. nov. no adaxial papillae or

prickles but micro-hairs present on sides of furrows

(Fig. 12). Abaxial epidermis: no bulliform cells; epi-

dermis consists of very large, conspicuous, regular,

somewhat inflated cells with outer-tangential wall

slightly thickened. No appendages visible.

Abaxial epidermis in surface view

Intercostal long cells: usually elongated but length

may be only slightly greater than width in Pentameris

sp. nov. (Figs 13 & 14); side walls always angled or

bowed outwards giving cells an inflated hexagonal

appearance (Figs 7 & 8, 13-16); end walls vertical;

anticlinal walls slightly undulating in Pentameris sp.

nov. (Figs 15 & 16) but less so in P. longiglumis (Figs

7 & 8). Cell shape and size is noticeably consistent

throughout all intercostal zones and even through-

out the whole abaxial epidermis. Pairs of short cells

present between successive long cells. No abaxial

bulliform cells. Stomata: absent on abaxial surface

(Figs 7 & 8, 13-16). Intercostal short cells: cork-silica

cell pairs with silica cell tall and narrow with smooth

outline (P. longiglumis) or rounded to kidney-

shaped (Pentameris sp. nov.). Associated with tall

and narrow cork cell; narrower than width of adja-

cent intercostal long cells. Papillae: absent. Prickles

and hooks: absent. Micro-hairs: none seen on abax-

ial epidermal scrapes although micro-hairs were ob-

served in the adaxial grooves of Pentameris sp. nov.

(Fig. 12). Macro-hairs: none present. Costal silica

bodies: rounded, equidimensional to slightly taller

than long (Fig. 16); costal zones narrow and not con-

spicuous.

Specimens examined:

Pentameris longiglumis

CAPE. — 3318 (Cape Town): Table Mountain (-CD), Mar-

loth 3063. 3418 (Simonstown): Platberg, Kogelberg State Forest

(-BD), Ellis 2341 , Taylor 7231 ; Kogelberg, Esterhuysen 13326.

Pentameris sp. nov.

CAPE. — 3418 (Simonstown): Platberg, Kogelberg State

Forest (-BD), Ellis 2342. 3419 (Caledon): Lebanon State Forest

(-AA), Haynes 770; Nuweberg, Caledon (-AB), Taylor 3023.

DISCUSSION AND CONCLUSIONS

A comparison of the leaf blade anatomy of Penta-

meris longiglumis and Pentameris sp. nov. with that

of the type of the genus, P. thaurii Beauv., reveals

many significant differences (Ellis, 1985a). The ana-

tomical structure of P. thuarii will be fully described

and critically evaluated in this subsequent article

(Ellis, 1985a), but here it will suffice to say that the

leaf anatomy of P. thuarii differs substantially from

that of each of the other four species presently re-

garded as belonging to the genus Pentameris. In fact

the leaf anatomy of P. thuarii closely resembles that

of several Pentaschistis species such as P. tortuosa

(Trin.) Stapf, P. silvatica Adamson and P. pallescens

(Schrad.) Stapf and it appears as if Pentameris is a

heterogeneous grouping as currently constituted.

It must be emphasized that this close likeness of P.

thuarii to several Pentaschistis species has, as yet,

not been evaluated using morphological criteria. Al-

though this similarity relates only to the anatomical

features of the leaf blade at present, the indications

are that these anatomical resemblances reflect the

natural relationships of this group of species. Of par-

ticular note are the distinctive micro-hairs, shared by

Pentameris thuarii and the Pentaschistis species men-

tioned above, in which the basal cell is very much

longer than the very short, tapering apical cell.

Micro-hair characteristics are generally considered

to be reliable indicators of taxonomic affinity (Clif-

ford & Watson, 1977) and, consequently, the group-

ing of Pentameris thuarii with these taxa sharing sim-

ilar micro-hairs, amongst other characteristics, ap-

pears justified.

On the other hand, a comparison of the leaf anat-

omy of P. longiglumis and Pentameris sp. nov. with

that of Pseudopentameris (Ellis, 1985) shows very

close agreement — particularly between Pentameris

sp. nov. and both Pseudopentameris macrantha and

P. brachyphylla. These latter three taxa have vir-

tually identical leaf anatomy in all respects and the

anatomical description given for Pseudopentameris

(Ellis, 1985) would suffice more than adequately to

describe both the transection and abaxial epidermis

of Pentameris sp. nov. Several important anatomical

diagnostic features shared by these three taxa are:

the presence of ribs of similar size and shape over all

vascular bundles; the schlerenchyma girders com-

posed of unlignified fibres; micro-hairs located in the

adaxial furrows; abaxial epidermal cells inflated to

hexagonal in shape and costal zones indistinct; ab-

sence of epidermal appendages and stomata on

abaxial epidermis. These anatomical features distin-

guish Pseudopentameris from the rest of the dantho-

noid grasses and, in combination, are shared by no

570

Bothalia 15, 3 & 4 (1985)

■ «

FIGS 9-16. — Leaf blade anatomy of Pentameris sp. nov. 9-12, leaf blade in transverse section. 9-10, Ellis 2342: 9, outline, x 100;

10, densely packed chlorenchyma cells in u-shaped groups, x 400. 11-12, Haynes 770: 11, hollow, cylindrical in outline, X 60;

12, detail of inflated abaxial epidermal cells. Note also that schlerenchyma girders are not fully lignified, X 400. 13-16, abaxial

epidermis in surface view. 13-14, Ellis 2342: 13, arrangement of costal and intercostal zones, x 160; 14, long cells with distinc-

tive shape, x 250. 15, intercostal and costal long and short cells, Haynes 770, x 250. 16, detail of costal and intercostal epider-

mal cells, Taylor 3023, x 250.

Bothalia 15, 3 & 4 (1985)

571

other genus. The almost identical anatomy of Penta-

meris sp. nov. and Pseudopentameris is surely indica-

tive of close phylogenetic relationship and this study

strongly suggests that this undescribed grass should

actually be placed in the genus Pseudopentameris

where it conforms exactly with the generic anatom-

ical circumscription.

This undescribed species is, therefore, virtually

identical to Pseudopentameris in leaf anatomy. How-

ever, Pentameris longiglumis , with which it has been

linked morphologically, and with which it grows in

the field, differs in several basic respects from the

typical Pseudopentameris- type of anatomy. These

differences will be enumerated and discussed fully

because they serve to distinguish Pseudopentameris

from the remaining group of species presently as-

signed to Pentameris. This applies particularly to P.

macrocalycina (Steud.) Schweick. and P. obtusifolia

(Hochst.) Schweick. but not to P. dregeana Stapf.

Anatomical differences between P. longiglumis and

Pentameris sp. nov. have been adequately illustrated

(Figs 1-8 and 9-16) and are briefly as follows:

a) Ribs associated with third order vascular bun-

dles smaller than those over the first order bundles.

Compare Figs 2 & 3 with Figs 10 & 12.

b) The third order bundles are much smaller in

comparison with the first order vascular bundles.

c) Abaxial and adaxial bundle sheath extensions

are present and consist of parenchyma cells which

gradually become thicker towards the epidermis

(Fig. 3). In Pseudopentameris the girder, or exten-

sion cells are more fibre-like with narrower di-

ameters (Fig. 12).

d) The cells of the bundle sheath extensions and

the fibres in contact with the epidermis are distinctly

lignified and stain red with safranin and fast green.

Fig. 2, where a green filter was used to accentuate

the red lignified tissue, should be compared with

Fig. 12 where the sclerenchyma tissue of the girders

is stained the same colour and density as the cellulo-

se-walled parenchyma of the mesophyll.

e) Adaxial bulliform cells at the bases of the fur-

rows are poorly developed in comparison to those of

Pentameris sp. nov. Compare Figs 3 & 10.

f) The adaxial epidermal cells of P. longiglumis

are distinctly papillate (Fig. 2) whereas no papillae

are present on either epidermis of Pentameris sp.

nov. (Fig. 12).

g) Adaxial micro-hairs are not distinguishable as

in Pentameris sp. nov. (Fig. 12) where they are

clearly visible on the sides of most adaxial furrows.

h) The shape of the intercostal long cells differs

slightly in surface view. The cells are more elongate,

with thicker anticlinal walls and without obvious un-

dulations in P. longiglumis, as a comparison of Figs 7

& 8 with Figs 15 & 16 will show.

i) The intercostal silica cells are tall and narrow

and often are not associated with a cork cell in P.

longiglumis, whereas in Pentameris sp. nov. they

tend to be rounded to kidney-shaped as in Pseudo-

pentameris (Ellis, 1985).

For these reasons, P. longiglumis is not consid-

ered as being as closely associated with Pseudopenta-

meris as Pentameris sp. nov. is. If all these taxa are

placed in the same genus, then leaf anatomy will be-

come so heterogeneous as to be meaningless. How-

ever, there is strong agreement in the anatomical

characteristics listed above between P. longiglumis

and P. macrocalycina and P. obtusifolia in particular

(Ellis, in press) and these three taxa again appear to

reflect a natural grouping best accorded generic

status apart from Pseudopentameris and Pentameris

thuarii. As P. thuarii is the type of the genus, a new

genus will have to be created for P. longiglumis, P.

macrocalycina and P. obtusifolia (and perhaps P.

dregeana) if the morphological evidence corrobo-

rates the anatomical evidence presented here.

ACKNOWLEDGEMENTS

The author is grateful to Mrs A. Romanowski for

photographic assistance and Mrs S. M. Thiart for

typing the manuscript.

UITTREKSEL

Die blaaranatomie van Pentameris longiglumis

(Nees) Stapf en van ’n onbeskryfde Pentameris spesie

word beskryf en geillustreer deur middel van fotomi-

krograwe. Dit word bewys dat die anatomiese struk-

tuur van die blaar van die nuwe Pentameris spesie in

besonder, beide die dwarssnitte en abaksiale epider-

male skrapings, none ooreenkomste met spesies wat

in die genus Pseudopentameris geklassifiseer word,

toon. Hierdie anatomiese ooreenkoms is groter as die

met enige van die Pentameris spesies. Gevolglik dui

anatomiese kenmerke aan dat die insluiting van die

nuwe spesie in die genus Pseudopentameris verkieslik

is bo plasing in die genus Pentameris. Hierdie aan-

duiding moet deur morfologiese studies gevestig

word. P. longiglumis, aan die ander kant, behoort

saam met P. macrocalycina (Steud.) Schweick. en P.

obtusifolia (Hochst.) Schweick. geklassifiseer te

word, weens noue ooreenkomste van die blaaranato-

mie.

REFERENCES

CHIPPINDALL, L. K. A., 1955. In D. Meredith, The grasses

and pastures of South Africa. Johannesburg: CNA.

CLIFFORD, H. T. & WATSON, L., 1977. Identifying grasses:

data, methods and illustrations. St. Lucia: University of

Queensland Press.

CONERT, H. J., 1971. The genus Danthonia in Africa. Mitt. bot.

StSamml., Munch. 10: 299-308.

DE WET, J. M. J., 1956. Leaf anatomy and phylogeny in the

tribe Danthonieae. Am. J. Bot. 43: 175-182.

ELLIS, R. P., 1976. A procedure for standardizing comparative

leaf anatomy in the Poaceae. I. The leaf blade as viewed in

transverse section. Bothalia 12: 65-109.

ELLIS, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in

surface view. Bothalia 12: 641-672.

ELLIS, R. P., 1985. Leaf anatomy of the South African Dantho-

nieae (Poaceae). X. Pseudopentameris. Bothalia 15:

561-566.

ELLIS, R. P., 1985a. Leaf anatomy of the South African Dantho-

nieae (Poaceae), XII. Pentameris thuarii. Bothalia 15:

573-578.

STAPF, O., 1900. Gramineae. In R. Thiselton-Dyer, Flora Ca-

pensis, Vol. 7. London: Reeve.

Bothalia 15, 3 & 4: 573-578 (1985)

Leaf anatomy of the South African Danthonieae (Poaceae). XII.

Pentameris thuarii

R. P. ELLIS*

Keywords: Danthonieae, leaf anatomy, Pentameris, Pentaschistis

ABSTRACT

The leaf blade anatomy of Pentameris tliuarii Beauv. is described and illustrated. It is shown that the leaf

anatomy of P. thuarii bears little resemblance to the anatomy of P. macrocalycina (Steud.) Schweick., P.

ohtusifolia (Hochst.) Schweick., P. longiglumis (Nees) Stapf or P. dregeana Stapf. The leaf anatomy of P. thuarii

was found to correspond closely with that of some specimens of Pentaschistis pallescens (Schrad.) Stapf, P. silvatica

Adamson and P. tortuosa (Trin.) Stapf. The morphology of these Pentaschistis specimens needs to be compared

with that of P. thuarii in order to establish whether the true relationships of P. thuarii lie with these taxa.

Anatomically, P. thuarii does not appear to be closely related to any species presently classified in Pentameris.

INTRODUCTION

Pentameris thuarii Beauv. is a robust perennial

with stout woody culms. It is restricted to the

mountains of the southern Cape Province from

Stellenbosch as far eastward as the Outeniqua

Mountains. This species prefers mesic habitats,

often occurring on streambanks or on seepages.

No detailed accounts of the anatomy of P. thuarii

have been published. De Wet (1956) notes that this

species has the festucoid type of epidermis and

anatomy. This implies that micro-hairs are absent

and that the outer parenchyma sheath is poorly

differentiated and that the chlorenchyma is

uniformly distributed between the bundles.

Renvoize (1981) places Pentameris in his core group

of the arundinoid grasses — a group which possesses

finger-like micro-hairs. In this respect, he differs

from De Wet (1956).

It is the purpose of this paper to describe and

illustrate the leaf anatomy of P. thuarii and to

compare and contrast this structure with that of the

other South African danthonoid grasses. In the

anatomical descriptions which follow, the

terminology of Ellis (1976, 1979) will be followed

and the following abbreviations used:

vb/s — vascular bundle/s

l’vb/s — first order vascular bundle/s

2’vb/s — second order vascular bundle/s

3’vb/s — third order vascular bundle/s

ibs — inner bundle sheath; mestome sheath

obs — outer bundle sheath; parenchyma sheath

ANATOMICAL DESCRIPTION OF PENTAMERIS

THUARII BEAUV.

Leaf in transverse section

Leaf outline: broadly V-shaped with slightly

involuted margins (Figs 1 & 2). Ribs and furrows:

rounded adaxial ribs located over all vbs; furrows of

medium depth, wide (Figs 3 & 4). No abaxial ribs.

Median bundle: no structurally distinct midrib

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

present. Vascular bundle arrangement: 5, sometimes

7, l’vbs in entire blade (Figs 1 & 2). Laterally 1 3’vb

occurs between consecutive l’vbs but this

arrangement is variable on either side of the median

vb where 2 (Fig. 2) or 3 (Fig. 1) 3’vbs may be present

between successive l’vbs; sometimes a 2’vb occurs

immediately adjacent to the median bundle (Fig. 2).

All bundles situated in the centre of the blade (Figs 3

& 4). Vascular bundle structure: 3’vbs elliptical with

xylem & phloem clearly visible, l’vbs elliptical with

phloem adjoining ibs and metaxylem vessels with

vessel width slightly greater than that of obs cells

(Figs 5 & 6). Vascular bundle sheaths: double; outer

sheath elliptical, with slight abaxial and adaxial

interruptions; extensions grade into fibres of girders

(Figs 5 & 6); obs cells inflated to elliptical, regular in

size; not very conspicuous; walls slightly thicker than

mesophyll cell walls; chloroplasts absent (Figs 5 &

6). Ibs complete; cells relatively large with inner

tangential and radial walls thickened. Sclerenchyma:

adaxial strands rounded, following shape of rib and

tapering toward junction with bundle sheath

extensions (Figs 4 & 5); abaxial strands minute in

association with 3’vbs (Figs 4 & 5) but more

conspicuous opposite l’vbs where they grade into

the bundle sheath extensions. Strands comprised of

lignified fibres; walls not markedly thickened.

Minute sclerenchyma caps present in margin (Fig. 2)

which is often very pointed. Mesophyll: indistinctly

radiate especially in adaxial ribs (Fig 3); continuous

between bundles with abaxial layer of cells almost

pallisade-like being tightly packed and regularly

arranged; remainder of chlorenchyma less regular

with many air spaces; intermediate between

isodiametric and truly irregular types. Colourless

cells: absent (Fig. 3). Adaxial epidermis: fan-shaped

groups of bulliform cells located at bases of furrows

(Fig. 3); not very restricted and up to 8 cells wide.

Epidermal cells much smaller than abaxial

epidermal cells. Prickles common on ribs (Figs 4, 5

& 6). No papillae. Micro-hairs visible near bases of

furrows (Figs 5 & 6); basal cell very elongate with

short distal cell. Abaxial epidermis: no bulliform

cells. Cells large and inflated (larger than obs cells)

with continuous cuticle (Figs 3 & 6). No macro-

hairs, prickles, papillae or micro-hairs.

574

Bothalia 15, 3 & 4 (1985)

2

FIGS 1-6. — Leaf blade anatomy of Pentameris thuarii as seen in transverse section. 1, Ellis 2221, x 100; 2, Ellis 660, X 100; 3-4,

Ellis 2220, X 250; 3, taken with red filter to enhance cellulose walls; 4, taken with a green filter to accentuate lignified tissue; 5,

Ellis 660, x 250, adaxial micro-hairs located at the bases of the furrows; 6, Ellis 660. x 400, interference contrast showing

micro-hair beneath adaxial prickles.

Abaxial epidermis in surface view

Intercostal long cells: elongated, with parallel side

walls, vertical end walls and slight undulations (Figs

7-10). Cell shape and size noticeably constant

throughout all intercostal zones. Long cells

separated by hooks, cork cells or cork-silica cell

pairs. No bulliform cells. Stomata: no abaxial

stomata. Intercostal short cells: either solitary or

paired; tall and narrow (Fig. 10). Papillae: absent.

Prickles: medium prickles with barbs sometimes

present on costal zones (Figs 7 & 8); barbs all

orientated in same direction. Hooks: intercostal

hooks common on some specimens (Figs 7 & 8);

barb orientation variable. Micro-hairs: none

observed; presumed absent. Macro-hairs: absent.

Silica bodies: very reduced and indistinct; tall and

narrow and solitary (Fig. 10) or irregularly rounded

and associated with cork cell (Figs 8 & 9). Costal

zones: very narrow; consist of only 1 or 3 files of

cells; long cells narrower than intercostal long cells.

Specimens examined:

Pentameris thuarii

CAPE. — 3318 (Cape Town): Jonkershoek State Forest,

Stellenbosch (-DD), Ellis 2220, 2221, Kruger 798. 3320

(Montagu): Tradouw’s Pass, Barrydale (-DC), Ellis 660, 1202.

3322 (Oudtshoorn): Outeniqua Pass, George (-CD), Acocks

21743, Barker 7653.

DISCUSSION AND CONCLUSIONS

Five species comprise the genus Pentameris

(Smook & Gibbs Russell, 1984). These are P.

thuarii, P. macrocalycina (Steud.) Schweick., P.

obtusifolia (Hochst.) Schweick., P. longiglumis

(Nees) Stapf and P. dregeana Stapf. The leaf

anatomy of P. dregeana differs considerably from

the other species and will be described separately in

a later paper. P. dregeana will, therefore, be

excluded from the following discussion and must not

be included in the concept of the genus Pentameris.

P. longiglumis (Ellis, 1985), P. macrocalycina and

P. obtusifolia (Ellis, 1985a) share many anatomical

features and form a recognizable, coherent

anatomical group consistent with that of most

genera. Each species has characteristic attributes of

its own, yet all have the same basic anatomical

ground plan.

A comparison of the leaf anatomy of P. thuarii

with that of these three species (Table 1) reveals

many obvious and striking differences. The,

Bothalia 15, 3 & 4 (1985)

575

FIGS 7-10. — Abaxial epidermal structure of Pentameris thuarii. 7-8, Ellis 2220; 7, costal and intercostal zones, x 160; 8,

intercostal hooks and costal prickles, x 250. 9, Ellis 2221, intercostal and costal short & silica cells; few hooks present, x 250;

10, Ellis 660, shape and structure of long cells, X 250.

illustrations of the anatomy of P. longiglumis (Ellis,

1985) and of P. macrocalycina and P. obtusifolia

(Ellis, 1985a) should be compared with Figs 1-6 and

Figs 7-10 to assess visually the magnitude of these

structural differences.

The leaf anatomy of P. thuarii differs so much

from that of the other three species that a generic

difference appears to be indicated. The heavily

lignified, thick, setaceous leaves of P.

macrocalycina, P. obtusifolia and P. longiglumis are

strongly reminiscent of members of the genus

Merxmuellera and these species show a definite

anatomical relationship with this genus. P. thuarii,

on the other hand, shares little in common with

Merxmuellera and its phylogenetic connections must

be sought elsewhere in the Danthonieae.

The leaf anatomy of P. thuarii was compared with

the anatomy of all the other 16 South African

danthonoid genera and it is only with a few species

of the genus Pentaschistis that similarities were

found. Of the 49 South African species of this genus

(Smook & Gibbs Russell, 1984), 40 were available

for comparative purposes and, of these, only three

had similar leaf anatomy to that of P. thuarii. These

were Pentaschistis tortuosa (Trin.) Stapf, P. silvatica

Adamson and P. pallescens (Schrad.) Stapf (Figs

11-19). Even in these three taxa only certain

specimens showed anatomical resemblances to

Pentameris thuarii. The following specimens of

Pentaschistis, as they are presently identified,

conform with the Pentameris thuarii type of

anatomy:

Pentaschistis pallescens

CAPE. — 3318 (Cape Town): Stellenbosch, Jonkershoek

(-DD), Ellis 2250, 2251. 3418 (Simonstown); Hottentots Holland

Mountains, Sugarloaf Peak (-BB), Ellis 2271, 2272.

Pentaschistis silvatica

CAPE. — 3318 (Cape Town): Table Mountain, Kirstenbosch

(-CD), Ellis 2309.

Pentaschistis tortuosa

CAPE. — 3418 (Simonstown): Hottentots Holland Mountains,

Sugarloaf Peak (-BB), Ellis 2276, 2277.

These Pentaschistis specimens share the following

characteristics with Pentameris thuarii but not with

other Pentameris species.

1. The leaf is thin and expanded and similar

adaxial rib and furrows are associated with all the

vascular bundles. In the Pentaschistis specimens the

ribs over the first order bundles are slightly larger

and may be flat-topped (Figs 11, 13, 14, 16 & 17).

2. The leaf margin is projected into a prominent,

narrow pointed cap (Figs 11 & 13). In P. thuarii this

projection is not always so prominent but in some

specimens (Fig. 1) there is a definite narrowing at

the pointed cap of the margin.

3. Laterally only a single third order bundle occurs

between adjacent first order bundles but in the

region of the median bundle 2-3 smaller bundles

may be present between successive first order

bundles. Compare Figs 3 & 4 with Figs 14, 16 & 17.

This condition is not found in any other Pentameris

species (Table 1).

576

Bothalia 15, 3 & 4 (1985)

« - £

<N <l> '

o »> "o •

c C Cl-

3 £ J2 C

£> — T3

E -2 5 C

JJ % £

E JS -£ -5

'5 'a o

g*

5 B

m ^ ^ cu

^ _ 8 S2 X)

U g « u

f g ,

• - w 3) 3

m &is-s

! gS 3-g E

g ° f & s £

§1?

dO^^U ^ CO

>» £

S ~

-D

> <D

<N 3

•P §

3 ^ 3

•3 x

.S S

g ts ^

2^§

C 73

(D O G G cd

+-> — ■ <D CD (J

= E £ °

CL Cj

^ CD

CD 73

JS -g

Cj CD c/5

73 X) CL

> x> ,>

~ G C3

oS «

CD ,

'j= c £ c

Cl

§ > a " p 73

■3 2 a-S I

- o ^ g x: .G

•a £ “ M o <u

CD 1— 7” , C >

•a <d <u

g 5 ^ ^

Cl c *"r' o C3

x .5 l -2 ^ £

UJ cn cn <1 ;>

O X

(D 3

c

EKe

O a a

U Z co

3 C ^ O Gx)

~ O a)

£ I s

G 73

O CD

cj c*

fJ3 3

O

cj X>

<D ■-;

hJ 0^

3 cd Cl

^ CD 3

« *o 3

> C/5 CD

g O £

g 3 13

Cj CD a) o O

73 -£ £ *c o

< £ £ cl n;

Bothalia 15, 3 & 4 (1985)

577

FIGS 11 - 19. — Anatomy of Pentaschistis species showing anatomical resemblances with Pentameris thuarii. 11-12, Pentaschistis

tortuosa, Ellis 2277: 11, transverse section showing projecting margin, x 100; 12, abaxial epidermis, x 160. 13-15, Pentaschistis

pallescens : 13 & 14, Ellis 2250, x 100, transverse sections with projecting margin and two smaller bundles between adjacent

large bundles; 15, Ellis 2272, x 160, abaxial epidermis with stomata absent. 16-19, Pentaschistis silvatica, Ellis 2309: 16-17, leaf

in transverse section; 16, x 100; 17, x 250; 18-19, abaxial epidermis, 18, X 160; 19, X 400 showing characteristic micro-hair.

4. Abaxial sclerenchyma girders are not

extensively developed and no continuous

hypodermal fibrous layer is present.

5. Well developed groups of bulliform cells occur

at the bases of the adaxial furrows (Fig. 17) as is the

case with Pentameris thuarii (Fig. 3).

6. Adaxial micro-hairs are visible in only one

specimen (Ellis 2271), but these resemble those of

Pentameris thuarii with an elongated basal cell and a

very short, tapering distal cell (Figs 5 & 6). Similar

micro-hairs were also observed on the abaxial

epidermis of some of these specimens (Fig. 19). The

hair on the adaxial surface of Fig. 13 is not a micro-

hair but a unicellular macro-hair. Similar macro-

hairs occur on both adaxial and abaxial surfaces of

all these Pentaschistis specimens but were not

observed on Pentameris thuarii.

578

Bothalia 15, 3 & 4 (1985)

7. Narrow, elongated intercostal long cells with

thin slightly undulating walls are common to

Pentameris thuarii and all these Pentaschistis

specimens (Figs 7-10 and Figs 12, 15, 18 & 19).

8. The intercostal short cells are the same width

as the long cells (Figs 18 & 19) and not much

narrower as in Pentameris macrocalycina and P.

obtusifolia.

9. Prickles sometimes occur on the costal zones

(Fig. 18) as is the case with Pentameris thuarii.

10. Costal silica bodies are inconspicuous.

These Pentaschistis specimens, therefore, bear a

very strong anatomical resemblance to Pentameris

thuarii and a morphological evaluation of this

anatomical observation is necessary to interpret

objectively the findings of this study. This is

particularly important as many other specimens of

Pentaschistis pallescens, in particular, have a totally

different anatomical structure. These specimens are

Ellis 649, 2226, 2227, 2228, 2260 and 2261. These

specimens, together with Esterhuysen 27320 (P.

tortuosa) are completely different from the

specimens considered here and show close

similarities to Pentaschistis colorata (Steud.) Stapf.

It, therefore, seems as if the identification of the

Pentaschistis specimens resembling Pentameris

thuarii is very suspect and requires checking. This is

unlikely at present as the taxonomy of Pentaschistis

is very confused and more reliable identifications are

not possible.

However, a morphological comparison, using

the Pentaschistis herbarium vouchers of specimens

resembling Pentameris thuarii in leaf anatomy,

should be undertaken. Particular attention should be

given to the structure of the ovary and of the ripe

fruit as the structure of these in P. thuarii is

diagnostic (Stapf, 1900; Chippindall, 1955; De Wet,

1956). In Pentaschistis the structure of the ovary and

fruit differs considerably from that encountered in

Pentameris (Stapf, 1900) and this comparison should

quickly establish whether this group of specimens

does show morphological relationships with P.

thuarii . As P. thuarii is the type of the genus, this

morphological comparison will enable a new

circumscription of the genus to be formulated and

compared with the condition in the other species

currently recognized as belonging to the genus. The

anatomical indications are that P. thuarii is closely

related to the Pentaschistis specimens but not to

Pentameris macrocalycina, P. obtusifolia, P.

longiglumis or P. dregeana.

ACKNOWLEDGEMENTS

The staff of the National Herbarium (PRE) are

thanked for plant identifications, Mrs A.

Romanowski for photographic assistance and Mrs S.

M. Thiart for typing the manuscript.

UITTREKSEL

Die blaaranatomie van Pentameris thuarii Beauv.

word beskryf en geillustreer. Daar word bewys dat

die anatomiese struktuur van die blaar van P. thuarii

geen ooreenkomste met die anatomie van P.

macrocalycina, (Steud.) Schweick., P. obtusifolia

(Hochst.) Schweick., P. longiglumis (Nees) Stapf of

P. dregeana Stapf toon nie. Die anatomie van P.

thuarii vergelyk baie gunstig met die van sekere

monsters van Pentaschistis pallescens (Schrad.)

Stapf, P. silvatica Adamson en P. tortuosa, (Trin.)

Stapf. Die morfologie van hierdie Pentaschistis

monsters behoort vergelyk te word met die van P.

thuarii om vas te stel of die natuurlike verwantskappe

van P. thuarii wel met hierdie groep spesies le.

Anatomiese aanduidings is egter dat P. thuarii geen

verwantskappe met enige van die spesies wat tans in

Pentameris geklassifiseer word, toon nie.

REFERENCES

CHIPPINDALL. L. K. A.. 1955. In D. Meredith, The grasses

and pastures of South Africa. Johannesburg: CNA.

DE WET, J. M. J., 1956. Leaf anatomy and phytogeny in the

tribe Danthonieae. Am. J. Bot. 43: 175-182.

ELLIS, R. P., 1976. A procedure for standardizing comparative

leaf anatomy in the Poaceae. I. The leaf blade as viewed in

transverse section. Bothalia 12: 65-109.

ELLIS, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in

surface view. Bothalia 12: 641-672.

ELLIS, R. P., 1985. Leaf anatomy of the South African

Danthonieae (Poaceae). XI. Pentameris longiglumis and

Pentameris sp. nov. Bothalia 15: 567-571.

ELLIS. R. P., 1985a. Leaf anatomy ol the South African

Danthonieae (Poaceae). XIII. Pentameris macrocalycina

and P. obtusifolia. Bothalia 15: 579-585.

RENVOIZE, S.A., 1981. The sub-family Arundinoideae and its

position in relation to a general classification of the

Gramineae. Kew Bull. 36: 85-102.

SMOOK. L. & GIBBS RUSSELL, G.E., 1984. In G. E. Gibbs

Russell et al.. List of species of southern African plants.

Mem. bot. Surv. S. Afr. 48: 1-138.

STAPF, O., 1900. Gramineae. In R. Thisleton-Dyer, Flora

Capensis, Vol. 7. London: Reeve.

Bothalia 15, 3 & 4: 579-585 (1985)

Leaf anatomy of the South African Danthonieae (Poaceae). XIII.

Pentameris macrocalycina and P. obtusifolia

R. P. ELLIS*

Keywords: Danthonieae, leaf anatomy, Pentameris, Poaceae

ABSTRACT

The leaf blade anatomy of Pentameris macrocalycina (Steud.) Schweick. and P. obtusifolia (Hochst.) Schweick.

is described and illustrated. The leaf anatomy of these two species shows many similarities suggesting a close

relationship between them. A slight problem appears to exist with the circumscription of P. obtusifolia and a minor

taxonomic adjustment may result in a classification which agrees totally with that based on leaf anatomy. This

would result in details of the leaf outline being diagnostic for these two taxa. The nomenclature of P. obtusifolia is

also very confusing and clarification is needed by reference to the relevant type specimens. P. macrocalycina and

P. obtusifolia together with P. longiglumis (Nees) Stapf, appear to form a distinct genus and do not bear close

anatomical resemblances to either P. thuarri Beauv. or P. dregeana Stapf.

INTRODUCTION

Pentameris macrocalycina (Steud.) Schweick. and

P. obtusifolia (Hochst.) Schweick. are two grass

species which are common in the mountain fynbos of

the mountain ranges of the extreme southern Cape.

Their distribution ranges from the Cedarberg in the

west to the Great Winterhoek Mountains in the east.

Both species prefer rocky habitats on Table Moun-

tain Sandstone, often occurring on very steep slopes

with a cool, south-facing aspect. They are true mon-

tane species capable of withstanding extreme cli-

matic conditions and P. obtusifolia, in particular, is

found on even the highest peaks in alpine condi-

tions. At these higher altitudes (above 1 650 m) this

species may form low, dense, cushion-like plants,

possibly in response to regular snowfalls.

These two species appear to be well adapted to

fire and most collectors note that they are conspicu-

ously common soon after fires. However, after sev-

eral seasons of regeneration of the fynbos sclero-

phyllous vegetation they may die out and be re-

placed by ericaceous and proteaceous species. On

the other hand, P. macrocalycina may remain com-

mon for many years even in the absence of burning.

This is obvious on extremely rocky substrates such as

crevices in rocks, even in mature fynbos communi-

ties.

Both species are strongly tufted perennials with

rigid and woody culms. The leaf sheaths are glabrous

except at the woolly mouth and the leaf blades are

wiry, filiform, plicate and terete. These blades are

finely pointed or pungent, very hard, glabrous and

smooth outside and densely but minutely tomentose

inside. The leaves of an individual plant may be

either flexuous and strongly curled or erect and very

straight. Plants with both leaf types often occur in

the same population and, consequently, this differ-

ence does not appear to be taxonomically or ecologi-

* Botanical Research Institute. Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

cally significant although it has a very strong visual

impact. In P. obtusifolia the leaf blades are usually

shorter, more rigid and exceptionally pungent

(Chippindall, 1955).

The ecological requirements, as well as vegetative

and spikelet morphology, of these two species are

very similar and they appear to be closely allied.

Chippindall (1955) notes a close relationship be-

tween P. obtusifolia and P. dregeana Stapf but this

was not borne out by field observations or leaf anat-

omy (Ellis, in press). In this study, the leaf anatomy

of P. macrocalycina and P. obtusifolia will be com-

pared to ascertain whether the leaf anatomy con-

firms these indications of a close relationship. The

anatomy of these species has not been published in

the literature and it will be described and illustrated

in detail using standardized terminology (Ellis, 1976,

1979). The following abbreviations will be used:

vb/s — vascular bundle/s

l’vb/s — first order vascular bundle/s

3’vb/s — third order vascular bundle/s

ibs — inner bundle sheath; mestome sheath

obs — outer bundle sheath; parenchyma sheath

ANATOMICAL DESCRIPTION OF PENTAMERIS MACRO-

CALYCINA

Leaf in transverse section

Outline of lamina : permanently infolded with

very round outline (Figs 1-6); adaxial channel a

deep, narrow cleft (Figs 5-7); 9, 11 or 13 vbs in

blade; diameter of leaf section 3,0-4,0 mm and leaf

thickness ± 2,0 mm. Ribs and furrows: adaxial ribs

and furrows between all vbs; furrows very deep and

cleft-like; rounded, massive ribs over l’vbs, smaller

ribs over 3’vbs (Figs 8-10). No abaxial ribs or fur-

rows. Median vascular bundle: structurally identical

to l’vbs or may be smaller than lateral l’vbs; no mid-

rib or keel developed. Vascular bundle arrangement:

5 or 7 l’vbs in blade; one 3’vb usually separates con-

secutive l’vbs (Figs 1 & 2) but lateral l'vbs may be

adjacent to one another (Figs 3-7). No 2’vbs. All

bundles located closer to the abaxial surface. Vascu-

lar bundle description: 3'vbs elliptical with well-de-

580

Bothalia 15, 3 & 4 (1985)

FIGS 1 - 10. — Transverse sections of the leaf blade of Pentameris macrocalycina. 1, Ellis 2500 x 100; 2, Ellis, 2557,

x 100; 3, Ellis 2540, x 100; 4, Ellis 2499, x 100; 5-6, Ellis 2508, x 160; 7, Ellis 2548, x 160; 8, Ellis 2508, x 250;

9, Ellis 2275, x 250; 10, Ellis 2499, x 250. Figs 4, 6, 8 & 10 taken with a red filter to enhance cellulose cell walls

and for the remainder a green filter was used resulting in the lignified walls being accentuated.

Bothalia 15, 3 & 4 (1985)

581

veloped phloem, l’vbs elliptical; phloem adjoins ibs;

metaxylem vessels very narrow, narrower even than

the ibs cells (Fig. 9). Vascular bundle sheaths : obs

elliptical; reduced to two lateral columns of cells on

either side of each l’vb (Figs 8-10); incomplete with

wide adaxial and abaxial interruptions; intergrades

adaxially with a conspicuous extension/girder com-

posed of thickened parenchyma (Figs 8-10). Obs

cells small, irregular in shape, with thin walls and

with no chloroplasts. Ibs entire; consists of relatively

large cells with inner tangential and radial walls

thickened. Sclerenchyma : l’vbs with adaxial, inver-

sely anchor-shaped girders of thickened parenchyma

joined to the vbs by long, thick obs extensions (Figs

9 & 10). 3’vbs may lack girders (Fig. 8). All vbs with

conspicuous, abaxial, sclerenchymatous girders

linked laterally to form a continuous, subepidermal

layer of sclerenchyma (Figs 8-10). Fibres lignified

(Figs 5 & 6) and thick-walled. No additional scleren-

chyma cap developed in the leaf margin but abaxial

hypodermal layer extends around the margin to the

apex of the first adaxial rib (Fig. 9). Mesophyll :

homogeneous chlorenchyma consisting of small,

tightly packed, isodiametric cells irregularly ar-

ranged. The cells occupy the sides and bases of fur-

rows and form either U-shaped (Fig. 10) or W-

shaped (Fig. 8) groups. No colourless cells in the

mesophyll. Adaxial epidermis : small, indistinct

groups of bulliform cells at bases of furrows (Figs 8

& 10); epidermal cells very small with individual

cuticles; prickles very common. Abaxial epidermis :

no bulliform cells; epidermis of large, uniform cells

with continuous, thick cuticle (Fig. 9). No macro-

hairs, prickles or papillae.

Abaxial epidermis in surface view

Intercostal long cells : rectangular, length about

twice width (Figs 11-18); side walls parallel, end

walls vertical; walls heavily thickened and pitted

(Figs 15-18). Cell shape and size very consistent

throughout abaxial epidermis; costal zones not easily

distinguishable. Adjacent horizontal files arranged

so that long cells and short cells are opposite one an-

other in a brick-work pattern. Long cells separated

by short cells in a single file. Intercostal short cells :

silico-suberose couples between all long cells; cork

cell crescentic, enfolding rounded silica body (Figs

15 & 16). Narrower than long cells. Stomata : no

abaxial stomata (Figs 11-18). Papillae : absent.

Prickles', absent. Micro-hairs : none seen. Macro-

hairs: absent. Costal zones : usually indistinguishable

from intercostal zones in surface view (Fig. 11);

sometimes evident due to underlying fibres and

slightly narrower long cells (Figs 12 & 14). Composi-

tion identical to intercostal zones.

Specimens examined:

CAPE. — 3219 (Wuppertal): Algeria State Forest, Cedarberg

Mountains (-AC), Ellis 2508: Sneeuberg, Taylor 5131:

Buffelshoek Pass, Koue Bokkeveld Mountains (-CA). Ellis 2499,

2500. 3318 (Cape Town): Jonkershoek, Stellenbosch (-DD), Ad-

amson 3980. 3319 (Worcester): Franschhoek Pass, Franschhoek

(-CC), Ellis 2348. 3321 (Fadismith): Garcia’s Pass, Fangeberg

(-CC), Ellis 2540. 3322 (Oudtshoorn): top of Swartberg Pass

(-AC), Ellis 2557, 2582: Robinsons Pass, Outeniqua Mountains

(-CC), Ellis 2548. 3323 (Willowmore): Potjiesrivierhoogte Pass

(-CA), Acocks 21589. 3324 (Steytlerville): Cockscomb Peak,

Great Winterhoek Mountains (-DB), Esterhuysen 28012. 3418

(Simonstown): (-AB), Ellis 2313, 2314, 2315: Sugarloaf Peak,

Hottentots Holland Mountains (-BB), Ellis 2275, 2280, 2292,

2293.

ANATOMICAF DESCRIPTION OF

PENTAMERIS OBTUSIFOLIA

Leaf in transverse section

Two different types of leaf anatomy can be recog-

nized in this species. These different anatomical

types are based entirely on differences in the leaf

outline but the remaining anatomical characteristics

of the leaf blade are very similar to those described

for P. macrocalycina. The detailed anatomical de-

scription given for P. macrocalycina can, therefore,

serve adequately to describe P. obtusifolia and here

only differences will be accentuated.

The first anatomical type of P. obtusifolia (Figs

19-22) has a tightly acicular, permanently infolded

leaf lamina which is round in section and has a nar-

row, cleft-like adaxial channel. The second type

(Figs 23-26) differs in having an inrolled leaf which

cannot be regarded as being of the permanently in-

folded type and which lacks a cleft-like adaxial chan-

nel with vertical sides.

The first type is virtually identical to P. macrocaly-

cina (Figs 1-10) in all details of leaf anatomy. The

only differences are quantitative and P. obtusifolia

has a leaf diameter of 7,5 mm as opposed to 3, 0-4,0

mm in P. macrocalycina. For comparative purposes

Figs 1-4 and Figs 19 & 21 are of identical magnifica-

tion and clearly show the larger size of the P. obtusi-

folia leaf sections.

This type of leaf anatomy was only observed in

one specimen — Ellis 2478. It may, or may not, be

significant that this was the only freshly fixed speci-

men of P. obtusifolia collected for this study and is,

therefore, not completely comparable with the other

specimens examined. It appears unlikely, however,

that the differences between this type and the other

are due to a lack of fixation and consequent de-

hydration and shrinkage of tissue. The first type of

P. obtusifolia has a maximum leaf thickness of 3,0

mm (Figs 20 & 22), whereas the second type is only

1,7 mm thick (Fig. 26). A shrinkage of almost 50% is

very difficult to imagine and it appears as if this dif-

ference does in fact represent a structural difference.

In addition, in the P. macrocalycina sections pre-

pared from herbarium material the degree of shrink-

age was insignificant, an observation which also indi-

cates that these different anatomical types in P. ob-

tusifolia may be structurally meaningful.

The second type, with inrolled leaves lacking cleft-

like abaxial channels, has a much thinner leaf blade

than either the first type of P. obtusifolia or P.

macrocalycina. Apart from this apparently signifi-

cant difference in leaf outline, however, the remain-

der of the leaf structure is identical to that described

for P. macrocalycina and the relevant descriptions

will also suffice for this type. Vascular bundle num-

ber and arrangement, mesophyll and sclerenchyma

structure and epidermal cell characteristics are all

very similar in both species and it is only the leaf out-

line of these specimens of P. obtusifolia which differs

at all from P. macrocalycina. The outline of the leaf,

582

Bothalia 15, 3 & 4 (1985)

FIGS 11 - 18. — Abaxial epidermal preparations of Pentameris macrocalycina. 11, Ellis 2293, x 160; 12 _ & 14, EW« ^82: 12, ^60,

14, x 250. Note indistinct costal and intercostal zones: 13, Ellis 2292, x 250. Epidermal cells filled with air: 15, Ellis 2280, x

400; 16, Ellis 2548, x 400; 17-18, Ellis 2499, x 400. Fig. 18 with air-filled cell lumens.

Bothalia 15, 3 & 4 (1985)

583

FIGS 19 - 26. — Leaf blade anatomy of Pentameris obtusifolia as seen in transverse section. 19-22, Ellis 2478: 19-20, green filter

used, 21-22, red filter used; 19, x 100; 20, x 250; 21, x 100; 22, x 250. 23, Esterhuysen 18210, x 100; 24, Esterhuysen 27442, x

100. 25-26, Esterhuysen 16531: 25, x 100; 26, x 250, red filter.

584

Bothalia 15, 3 & 4 (1985)

although visually very clear and distinct, may, there-

fore, represent only an insignificant anatomical dif-

ference.

Abaxial epidermis

Identical to P. macroccilycina and the reader is re-

ferred to the relevant description for structural de-

tails. No epidermal differences were noted between

specimens exhibiting the two types of leaf outline as

seen by comparing Figs 27 & 28 with Figs 29 & 30.

Specimens examined:

CAPE. — 3319 (Worcester): Leeuwfontein Peak, Gydoberg

(-AB), Ellis 2478\ Waaihoek Mountains, (-AD), Esterhuysen

18210; Buffelshoek Peak, Hex River Mountains (-BD), Esterhuy-

sen 27442; Slanghoek Mountains, Wittenberg (-CA), Esterhuysen

16531; Sneeukop, upper Wellington, Esterhuysen 26517; Fontein-

tjiesberg, Hex River Mountains (-CB), Esterhuysen 22209,

Stettynsberge (-CC), Esterhuysen 11115.

DISCUSSION AND CONCLUSIONS

The leaf anatomy of P. macrocalycina and P. ob-

tusifolia is very similar indeed and it is only in the

outline of the transections that significant anatom-

ical differences can be detected. Metcalfe (1960) is

of the opinion that in permanently infolded leaves

structural criteria such as the shape of the blade as

outlined by the abaxial surface, the number of vascu-

lar bundles present in the section, the adaxial chan-

nel and leaf width and thickness are of significant

specific diagnostic value. If this is indeed so, then the

virtually identical anatomy of all the P. macrocaly-

cina specimens and the Ellis 2478 specimen of P. ob-

tusifolia, suggests that Ellis 2478 actually should be

reclassified as P. macrocalycina. If this single speci-

men were to be transferred to P. macrocalycina,

then two distinct anatomical groups would result —

P. macrocalycina with permanently infolded acicular

leaves with cleft-like adaxial channels and P. obtusi-

folia with inrolled leaves lacking cleft-like adaxial

channels. These two taxa would then be anatom-

ically distinct and two separate, but very closely re-

lated, species could be distinguished.

The specimen of P. obtusifolia with anomalous

leaf anatomy (Ellis 2478) has been verified by the

staff of the National Herbarium (B. de Winter, pers.

comm.) as being correctly identified as P. obtusifo-

lia. Consequently the difference in leaf outline be-

tween P. macrocalycina and P. obtusifolia is not di-

agnostic and Ellis 2478 represents a clear interme-

diate. This suggests that P. macrocalycina and P. ob-

tusifolia are exceedingly closely related and only an

infraspecific separation appears justified.

It is of interest to note that the anatomical sample

of P. obtusifolia examined in this study actually com-

prises the entire collection of P. obtusifolia in the

National Herbarium. Of these six specimens, three

were originally identified as either P. macrocalycina

or P. dregeana and were only renamed at the sugges-

tion of the anatomical evidence reported here.

Freshly fixed anatomical material is required to

evaluate this apparently significant anatomical dif-

ference and until this is forthcoming it appears sen-

sible to continue recognizing these two closely re-

lated taxa at the specific level.

Leaf anatomy, therefore, indicates that these two

taxa are closely related. This observation disagrees

; s. C i i t i 5 «- 1 Si -

< r r, X 3 ;c £ ' S 'S t £ 3

i ? T t f 3; X 3 s i i

, C It X X 3 e 3 P t t

X i l i t T f C. x < 3

. c C i t t - 3 i i

■.ft £ £ e i £ i i t t-

r t c s . .i : s. 3 tic t

, - ft i t. i liK 3 3 * - * « £

' C t 5 ~i-j. i ■ * 3 ■ x 3 r . r

C 1 T ' < * 5 ; x <■ £ ? ?

i t 3 ‘ f $. i % ‘ r‘ x t 1 s !

x t t v s i i t x x s

t C x ( 3 X "x < X c e

ill t c £ t X 0 6 : 1 c

c e t tit ft ® f x X *

c £ .• 3 x J .• •• C s : J

i V 5 % i> ft ’* v a y A i •«

3. C x 1 - L 4 9 i s

i if i* 'x > " 3 . i s -

FIGS 27 - 30. — Abaxial epidermis of Pentameris obtusifolia. 27-28, Ellis 2478: 27, x 160; 28, X 250. 29, Esterhuysen 26517, x 250;

30, Esterhuysen 27442, x 250.

Bothalia 15, 3 & 4 (1985)

585

with Chippindall’s (1955) statement that P. obtusifo-

lia is allied to P. dregeana. However, due to nomen-

clatural confusion this apparent conflict may not ac-

tually exist and it appears as if the concept of P. ob-

tusifolia differs between Chippindall (1955) and that

used here. The leaf anatomy of P. dregeana differs

considerably from that of P. macrocalycina and P.

obtusifolia (Ellis, in press) and it appears unlikely

that the morphology would indicate a close relation-

ship. A more likely explanation is that P. obtusifolia,

as used today, actually represents the undescribed

species from the mountains of the Worcester District

referred to by Chippindall (1955). P. obtusifolia , as

defined here, is confined to the Worcester area in

the higher peaks of the Hex River Mountains, the

Wittenberg, the Gydoberg and the Waaihoek Moun-

tains.

Chippindall (1955) referred the name P. obtusifo-

lia to specimens collected in the Houw Hoek Moun-

tains of the Caledon District. It appears as if the

specimen actually alluded to is Burchell 8076, called

P. squarrosa Stapf by Stapf (1900). P. obtusifolia, as

it is known today, does not occur in the southern

mountains and therefore, it appears highly likely

that genuine nomenclatural confusion exists regard-

ing the name of the entity discussed here. This has

resulted in these apparently conflicting statements in

the literature.

The specimens here referred to as P. obtusifolia

form a coherent morphological, ecological and ana-

tomical entity and the type must be examined in or-

der to establish the correct name. Nevertheless,

these specimens undoubtedly constitute a recogniza-

ble and distinct species. The uniformity of this taxon

would be further increased with the removal of Ellis

2478 to P. macrocalycina. This specimen is much

more robust than the remainder of the specimens as-

signed to P. obtusifolia and also has rigid, erect seta-

ceous leaves whereas the other P. obtusifolia speci-

mens have shorter, curly leaves.

There seems to be a good case, both morphologi-

cally and anatomically, for the transfer of this speci-

men to P. macrocalycina. This would enable the re-

cognition of two distinct taxa which, together with

P. longiglumis (Nees) Stapf ( Ellis, 1985) appear to

form a natural grouping best accorded generic status

apart from P. thuarii (Ellis, 1985a) and P. dregeana

(Ellis, in press).

ACKNOWLEDGEMENTS

The author is grateful to Mrs A. Romanowski for

photographic assistance, Mrs S. Thiart for typing the

manuscript and to the staff of the National Herbar-

ium (PRE) for plant identifications.

UITTREKSEL

Die blaaranatomie van Pentameris macrocalycina

(Steud.) Schweick. en P. obtusifolia (Hochst.)

Schweick. word beskryf en geillustreer. Die anatomie

van hierdie twee spesies toon none ooreenkomste wat

verwantskappe tussen hulle aandui. Die omgrensing

van P. obtusifolia, skep sekere probleme en die takso-

nomiese regstelling hiervan behoort ’n klassifikasie te

gee wat heeltemal ooreenstem met die wat op ana-

tomie gebaseer is. In hierdie geval sal sekere ken-

merke van die blaar in deursnit vir die onderskeiding

van hierdie twee taksa diagnostiek word. Daar be-

staan ook onsekerheid oor die nomenklatuur van P.

obtusifolia wat verwarring skep. Hierdie probleem

behoort opgelos te word deur verwysing na die be-

trokke tipe monsters. P. macrocalycina en P. obtusi-

folia, saam met P. longiglumis (Nees) Stapf, vorm ’n

duidelike genus en geen anatomiese verwantskappe

met P. thuarii Beauv. en P. dregeana Stapf word aan-

gedui nie.

REFERENCES

CHIPPINDALL. L. K. A., 1955. In D. Meredith, The grasses

and pastures of South Africa. Johannesburg: CNA.

ELLIS, R. P., 1976. A procedure for standardizing comparative

leaf anatomy in the Poaceae. I. The leaf blade as viewed in

transverse section. Bothalia 12: 65-109.

ELLIS, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in

surface view. Bothalia 12: 641-672.

ELLIS, R. P., 1985a. Leaf anatomy of the South African Dantho-

nieae (Poaceae). XL Pentameris longiglumis and Pentameris

sp. nov. Bothalia 15: 567-571.

ELLIS. R. P., 1985b. Leaf anatomy of the South African Dantho-

nieae (Poaceae). XII. Pentameris thuarii. Bothalia 15:

573-578.

METCALFE, C. R., 1960. Anatomy of the Monocotyledons. I.

Gramineae. Oxford: Clarendon Press.

STAPF, O., 1900. Gramineae. In R. Thiselton-Dyer, Flora Ca-

pensis, Vol. 7. London: Reeve.

Bothalia 15, 3 & 4: 587-590 (1985)

The Kranz syndrome in the Eragrostideae (Chloridoideae, Poaceae) as

indicated by carbon isotopic ratios*

HECTOR O. PANARELLO ** and EV ANGELINA sANCHEZt

Keywords: carbon isotope ratios, Eragrostideae, Kranz syndrome

ABSTRACT

13C/12C ratios are generally regarded as being very reliable indicators of C, or C4 photosynthesis. These relative

carbon isotope ratios are expressed as a negative 8I3C and fall into two distinct groups: Kranz (or C4) plants with 6

between -9°/oo and -18°/oo and non-Kranz (C3) plants with 6 between -22°/<>o and -28°/oo. In this paper,

29 taxa, representing 12 genera, of the tribe Eragrostideae were examined by mass spectrometry for their &13C in

dried leaf tissue. All these taxa proved to be C4 plants with 613C values ranging between -13,6°/oo and

-10,9°/m. These findings confirmed published leaf anatomical observations which showed that all the studied

taxa had characteristic Kranz leaf anatomy.

INTRODUCTION

A syndrome of anatomical, cytological and phy-

siological characters, all related to aspects of the car-

bon fixation process, has been reported from several

families of Angiosperms. This syndrome has been

called the Kranz syndrome, C4 photosynthesis, and

the Hatch-Slack pathway. Kranz plants exhibit a

high degree of efficiency in the utilization of ambient

CCC, resulting in maximum photosynthetic produc-

tion at high temperature and high light intensity.

The anatomical characteristics of this syndrome have

been known for 100 years (Haberlandt, 1882) but

the physiological aspects have only been known for

nearly 20 years (Kortschak et al. , 1965 and Hatch &

Slack, 1966).

The anatomical specializations of the Kranz syn-

drome include the presence of a chlorenchymatous

bundle sheath of large, thick-walled cells containing

specialized chloroplasts distinct from those of the

mesophyll — they are a greater size, a greater num-

ber per Kranz cell are present and they accumulate

starch. In addition, the mesophyll cells adjoin the

bundle sheath and are radially arranged. Plants pos-

sessing the Kranz syndrome fix carbon initially into

four-carbon acids (oxalo-acetic acid, malic acid and

aspartic acid) in these mesophyll cells (Hatch &

Slack, 1970).

The only way to prove definitely that a plant is

Kranz or non-Kranz is to investigate both its physiol-

ogy, either directly, or indirectly by means of 13C/12C

ratios, and to examine its anatomy, particularly the

leaf anatomy. Differences in carbon isotope ratios

between Kranz and non-Kranz plants are well docu-

mented (Fritz & Fontes, 1980). 6 13C values for C4

plants range between -18%o and -9%o with an

average of about -12%o, whereas the 6 13C for C3

plants is between -38%o and -22%o with an

* INGEIS contribution No. 71

** Instituto de Geocronologia y Geologia Istotopica, Ciudad

Universitaria, Buenos Aires, Argentina.

t Museo Argentino de Ciencias Naturales, ‘Bernardino Rivada-

via’, Buenos Aires. Argentina.

average of -25%o (Fig. 1). In the Poaceae, no ra-

tios between -18%o and -22%o have been re-

ported. Plants with Crassulacean Acid Metabolism

also have high carbon isotope ratios (Fig. 1), but are

not relevant in this study since CAM has not been

demonstrated in the Poaceae (Brown, 1977).

The determination of carbon isotope ratios is a

very convenient method for classifying plants as be-

ing Kranz or non-Kranz. Only small amounts of

plant tissue are required, the data are very accurate

and objective and the plant material need not be

physiologically active. The use of dried specimens

from herbarium sheets is, therefore, possible.

MATERIALS

Leaf tissue was taken from herbarium sheets. This

plant material was obtained from many sources, and

these are all gratefully acknowledged. Sources in-

cluded: Dr Thomas R. Soderstrom, US National

Herbarium, Smithsonian Institution, Washington,

DC (US); Museo Argentino de Ciencias Naturales

‘Bernardino Rivadavia’ (BA); Herbario Gaspar

Xuarez, Catedra de Botanica, Facultad de Agrono-

mia, Universidad de Buenos Aires (BAA); Instituto

de Botanica Darwinion (SI); Instituto Miguel Lillo,

Tucuman (LIL); Instituto de Botanica Agricola,

INTA, Castelar (BAB); Instituto de Botanica, Fac-

ultad de Agronomia, Universidad Nacional del No-

reste (CTES).

The genera of the Eragrostideae examined, ar-

ranged alphabetically, are given in Table 1. All

voucher specimen sheets from which material was

taken have been annotated with labels stating the

13C/12C ratio of that specimen.

METHODS

About 8 mg of the leaf samples were dried under

vacuum at a temperature of 110°C for 8 hours or

overnight. Every sample was then mixed with 100

mg vanadium pentoxide (V205) in a quartz vial and

flame sealed under vacuum of about 10 4 mbar. The

vial, containing the sample and the vanadium pen-

toxide (V205) was allowed to react at 1000°C in an

588

Bothalia 15, 3 & 4 (1985)

%o&,3C PDB

FIG. 1. — Carbon isotopic composition of photosynthetically

fixed carbon; a, terrestrial plants; b, known C,and C4 plants;

c, known CAM plants. Reproduced from Fritz & Fontes,

1980.

electric furnace for five minutes, according to the

technique described by Panarello et al. (1983).

The resultant CO-, was introduced into the purifi-

cation line by breaking the vial in a vacuum by

means of a special ‘vial breaker’. Water and light

gases were removed and the C01 was transferred to

a sample bottle by trapping with liquid nitrogen. The

mass 45/44 ratio was then determined by measuring

the samples in a Micromass 602-D Mass spectrome-

ter. A reference and standard CO: from Carrara

marble was used (Panarello et al., 1980).

The results are expressed as 6 I3C (%o) defined

as follows:

6 13C = [ _ i | x 1000'Voo

L RPDB J

Where Rs = Isotopic ratio of the sample

RPDB= Isotopic ratio of the STANDARD

PDB (Chicago PDB Standard, Be-

lemnitella americana from the Cre-

taceous Pedee formation. South

Carolina [Craig, (1954)].

The precision of measurement is ± 0,l%o of

the ratio. If 6 13C is > 0 the sample is heavier than

the standard, if less than 0 the sample is lighter, or

depleted, in comparison with the standard. From

this definition 6 13C PDB = 0.

Carbon occurs in nature as two stable isotopes 12C

and 13C, with the following average abundance:

12C : 98,89%

13C : 1,11%

However, the 13C/12C ratio is not constant and

undergoes variations of a few parts per thousand in

this ratio due to some physical and chemical pro-

cesses. This variation in the isotopic composition is

known as ‘isotopic fractionation’. One of the most

important processes in nature which causes isotopic

fractionation of carbon is photosynthetic carbon as-

similation by green plants.

Both terrestrial and marine plants have lower

13C/12C ratios than their respective carbon sources,

atmospheric CO-,(6 — 7; -8°/no) and ocean carbon-

ates (6 ~ %o). This means that during photosyn-

thetic C02 fixation there is preferential utilization of

12C and exclusion of 13C (Craig, 1957; Smith &

Epstein, 1971).

RESULTS

The 6 13C values of the taxa examined in this study

are presented in Table 1. These carbon isotope ra-

tios are also compared with published anatomical

observations. As always, in the Poaceae, the 6 13C

values corroborate the leaf anatomy and all taxa

have C4 carbon isotope ratios (between -13,6%o

and -10,9%o) which agree with published ac-

counts of Kranz leaf, and stem, anatomy for these

taxa. This study has, therefore, confirmed the pres-

ence of the Kranz syndrome in these selected rep-

resentatives of the Eragrostideae.

ACKNOWLEDGEMENTS

The authors are indebted to Dr Enrique Linares

for facilities made available at the INGEIS and to

the staff of the Stable Isotopes Laboratory. Miss

Cristina Maetakeda (CONICET) is thanked for

technical assistance and Dr R. P. Ellis (Botanical

Research Institute, Pretoria) for reading the manu-

script.

This research was supported with CONICET

funds.

TABLE 1. — 5^C values of selected representatives of the Eragrostideae

Bothalia 15, 3

& 4 (1985)

589

oo .

. .

r* £ £

ro ^ On

^ OO <N

CQ to

< 5 5 S

M -o . -a

c <u

oo <o • a>

a: cc

2 ■ o ■

-e e .e

0 U § U

1 c 1 c

VO

rr ^ oo

( ■ — \ ro

^ w

r- _

00

^ Tf C

r~- oo

On ro

vO OO

O ON

VO vo

< <

CQ CQ

< vo

C G < ™

(yj CQ !rj

C ‘C

►h ^

CD <*3 C

M <D C

X) CQ 3

c3 cd

u n-; a:

: ov 3

S S N1

® ® | H

■S is ^ <’

■ ra CQ

Ml 2"

^ Tt

7 O

<. ro

N «

<D X

X

< d

« j

U U U (X c/3 O U 1

CQ OQ

< <

s s

r-

no oo - — - , — ,

Ooo o

't i/l ro Tt

o CA ^ r-

8 £ oo vo

^ fN CO

U.UfflN

o § < d

* &5d

o C

^ c % !Z

S-.SS

x-* al

• c 00 rV

c ^ oc

* 3 a ■

>— >’ X co CQ

_

|SSg

^ m m U

V w ° ^

00

;■ 3 «

H X 00

' a. c

CD •->

c °

O £ g £

m d ® a-

d <J ^ d

J= •“

o fO

C 00

cd on

CO

o £

^ ^ i-h al cl cd cl

. ro m . „ . „ oo „ on on on on

oooor^r-t^oor^r'-r~-r~-r-'

On ON ON .„i/) io rt ’t

r-r^-r-r^t^'^oooooo

ON ON On r-~ On ON ON ON ON

o'sS

£2^

CO

X r'~

73 ON

g ’1

CO G

.. £

I o o

! vn cS 03

i <*> <+h 73

1 a) rd

; So g

<3 o 00

VJS

. „ ON

r- r-

r- r- ^

ON ON

NNNNNNNN

.O

o o o o o

„ _

O O O

N0 G ■£

CQCQc/5cocQCQcomcoco(^i/5(y5i/5co«^CQcococo(>ocoUcoS

OOCCOOGOGGCCGCGGOCGCC

OvMroinr-^isDOOvOO

O ro h (N ri cn r-T t-h i— T co

vONOVOCOOr}-cOv>ov£)Ovor^VO<NrN

<N »-h co (N ri rT i-h ol r-H <n <N <N rn <N ri

I I I I I I I

I I I I I I I I

< C c S

rd, 2 52

0 < 4= <C ^

C c ^ b

O X

3? ’s'

^ 2

K S

U .2 u Jic

.x®5 Ti 5

-d 3 ®

: o ^ !> t?

-

/ O 3 • G V

i d o'^ “■ o ^

J X

•2 ® 5

S :a ^

5

-o -o -e

a ^

■ ■ d ^ S

S m*s

52, -c I a S

Hiss

^ « -si -

5 ^ »0 S3

•3 n. s ^

I§i!

O? g g

- frS’S'S:

s sll

Q Q Q ty

I 3

>u c y, s

s Ira I

3 <D *S q 2

■Bb a s ts -s

6h S ^ C ^

f If p.-i

^ M ‘C

' J C

s >•'

5 £

S S c3

V. o ^

Ci. Ci.^

> > > > >

>>z

VJ

►J ^ Q

n ? 33

w ^ X

•3 o'®

III

CB* 8-5

•a

s .

1 p

'C cu 0.

O Z a

k = "a

£?| S

<3 ^3 s:

! c -S o o

-G07

: 5 ,0 x ^

1 5 2 .75 R

> -s § i 1

- o

Tripogon spicatus (Nees) Ekman —11,4 Brown, 1958; Brown, 1977 Molfino s.n. (BA 39580)

590

Bothalia 15, 3 & 4 (1985)

UITTREKSEL

Die I3C/I2C verhoudings word algemeen as baie be-

troubare indikators van C? of C4 fotosintetiese roetes

beskou. Hierdie relatiewe koolstofisotoopverhou-

dings word as negatiewe 6 13C uitgedruk en val in twee

duidelike groepe: Kranz (of C4) plante met tussen

-9o/oo en -18°/oo en nie- Kranz (of C?) plante met

b tussen -22°/oo en -28°/oo. In hierdie artikel is

29 taksons, wat 12 genera van die tribus Eragrosti-

deae verteenwoordig, deur middel van massaspektro-

metrie vir hulle bI3C in gedroogde blaarweefsek be-

studeer. Dit is bewys dat al hierdie taksons C4 plante

is met b ,3C waardes wat varieer tussen -13,6°/oo en

-10, 9° loo. Hierdie bevindinge het gepubliseerde

anatomiese waarnemings van blare, wat getoon het

dat al die bestudeerde taksons kenmerkende Kranz

blaaranatomie besit, bevestig.

REFERENCES

BROWN, W. V., 1958. Leaf anatomy in grass systematics. Bot.

Gaz. 119,3: 170-178.

BROWN. W. V., 1977. The Kranz Syndrome and its subtypes in

grass systematics. Mem. Torrey bot. Club 23,3: 1-97.

CACERES, M. R.. 1950. Los caracteres anatomico-foliares de

Munroa mendocina y Blepharidachne benthamiana. Revta ar-

gent. Agron. 17: 233-240.

CACERES, M. R., 1951. La anatomia foliar de Scleropogon bre-

vifolius y sus relaciones taxonomicas. Revta argent. Agron.

23: 109-114.

CRAIG, H.. 1957. Carbon-13 in plants and the relationships be-

tween Carbon-13 and C-14 variations in nature. J. Geol. 62:

115-149.

FRITZ. P. & FONTES. J. C., 1980. A Handbook of Environ-

mental Isotope Geochemistry. Volume I. The terrestrial en-

vironment.

HABERLANDT, D. G., 1882. Vergleichende Anatomie des as-

similatorischen Gewebesystems der Pflanzen. Pringsheims

Jb. wiss. Bot. Bd. 13.

HATCH, M. D. & SLACK, C. R.. 1966. Photosynthesis by

sugar-cane leaves. Biochem, J. 101: 103-111.

HATCH. M. D. & SLACK. C. R., 1970. Photosynthetic CO; fix-

ation pathways. A. Rev. PI. Physiol. 21: 141-162.

KORTSCHAK, H.P.. HARTT C. E. & BURR. G. O., 1965.

Carbon dioxide fixation in sugar-cane leaves. PI. Physiol. 40:

209-213.

METCALFE, C. R., 1960. Anatomy of the Monocotyledons. I.

Gramineae. Oxford: Oxford University Press.

NICORA. E., 1962. Revalidacion del genero de Gramineas ‘Nee-

ragrostis’ de la flora norteamericana. Revta argent. Agron.

29: 1-11.

PANARELLO, H. O.. GARCIA. C. M.. VALENCIO, S.A. &

LINARES. E.. 1980. Determinacion de la composition

isotopica del carbono, en carbonatos, su utilization en Hidro-

geologia y Geologia. Revta Asoc. geol. argent. 35: 460—466.

PANARELLO. H. O.. ALBERO, M. C. & ANGIOLINI. F. E.

1983. Stable isotope fractionation during the benzene synthe-

sis for radiocarbon dating. Radiocarbon 25,2: 529-532.

SANCHEZ. E.. 1974. Subfamilia Eragrostoideae (Gramineae):

Tribu Eragrosteae: Anatomia foliar de las especies argenti-

nas de los generos Eleusine y Dactyloctenium. Darwiniana

18: 526-538.

SANCHEZ. E., 1975. Anatomia foliar de las especies argentinas

del genero Gouinia Fournier (Gramineae). Lilloa 34,7:

89-106.

SANCHEZ, E., 1978. Anatomia foliar de Diandrochloa glome-

rata (Walt.) Burkart (Gramineae). Lilloa 35,1: 67-72.

SANCHEZ, E., 1979a. Anatomia foliar de las especies y varie-

dades argentinas de los generos Tridens y Erioneuron (Gra-

mineae). Darwiniana 22,1-3: 159-175.

SANCHEZ, E.. 1983a. Estudios anatomicos en el genero Blepha-

ridachne Hackel (Poaceae, Eragrostoideae, Eragrosteae).

Revta. Mus. Argent. Cienc. Nat. 6,3: 73-87.

SANCHEZ, E., 1983b. Dasyochloa Willdenow (Poaceae).

genero monotipico de Norteamerica. Lilloa 36,1: 131-138.

SANCHEZ. E. , 1983c. Estudios anatomicos en el genero Munroa

Torrey (Poaceae. Chloridoideae, Eragrostideae). En prensa

revista Darwiniana 25.

SANCHEZ, E., 1984. Estudios anatomicos en el genero Munroa

(Poaceae, Chloridoideae, Eragrostideae). Darwiniana 25,

1 — 4: 43-57.

SMITH. B. N. & EPSTEIN. S., 1971. Two categories of L,C/l:C

ratios for higher plants. PI. Physiol. 47: 380-384.

SMITH, B. N. & BROWN, W. V., 1973. The Kranz syndrome in

the Gramineae as indicated by carbon isotopic ratios. Am. J.

Bot. 60,6: 505-513.

Bothalia 15, 3 & 4: 591-596 (1985)

The genus Rubus in South Africa. I. Chromosome numbers and

geographical distribution of species

J. J. SPIES* and H. DU PLESSIS*

Keywords: chromosome numbers, geographical distribution, hybridization, polyploidy, Rubus

ABSTRACT

The geographical distribution of 14 of the Rubus species in South Africa is presented. Chromosome numbers of

nine of the species were determined: six for the first time, one is confirmed and additional polyploid levels are

described for the other two species.

It is demonstrated that the South African species of the subgenus Idaeobatus contain less diploid specimens and

more polyploid specimens than their extra- African counterparts. This phenomenon could be attributed to hybridi-

zation between the subgenera Eubatus and Idaeobatus.

INTRODUCTION

The genus Rubus is spread over all continents and

is found in most climatic regions. Focke (1910-1914)

divided this genus into 12 subgenera of which only

two are present in South Africa. All the represent-

atives of the subgenus Eubatus Focke are intro-

duced, whereas the subgenus Idaeobatus Focke con-

tains both indigenous and introduced species.

Introduced Eubatus species all represent the sub-

series Suberecti of the series Moriferi and include the

species Rubus affinis Wh. & N., R. cuneifolius

Pursh., R. pascuus Bailey and R. flagellaris Willd.

The introduced Idaeobatus species, R. niveus

Thunb. and R. phoenicolasius Maxim., form part of

the series Nivei of the section Idaeanthi. The in-

digenous Idaeobatus species include the sections Ro-

sifolii (R. rosifolius Sm.), Afromontani (R. immixtus

C.E. Gust.) and Idaeanthi [various species of the se-

ries Afroidaei, including R. apetalus Poir., R. inter-

currens C.E. Gust., R. longepedicellatus (C.E.

Gust.) C.H. Stirton, R. pinnatus Willd., R. rigidus

Sm., R. transvaliensis C.E. Gust, and R. ludwigii

Eckl. & Zeyh.. with its subspecies ludwigii and spa-

tiosus C.H. Stirton]. For the purpose of this study R.

adolfi-friederici Engl., R. ecklonii Focke and R. ex-

succus Steud. are included in R. apetalus Poir. be-

cause integration between these taxa renders the

separation into distinct species impossible. Plants

collected from a hybrid swarm in eastern Transvaal

are included in R. x proteus C.H. Stirton.

The problems with Rubus taxonomy in South

Africa are aggravated by the occurrence of apo-

mixis, hybridization among indigenous species and

between indigenous and introduced species, the va-

riation produced by a breeding program and subse-

quent escape from cultivation of those plants and in-

adequate collected herbarium material. The aim of

this study is, therefore, to provide cytogenetical evi-

dence for the species delimitation in the South Afri-

can species of Rubus. To achieve this goal, the re-

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

suits of a preliminary study on chromosome numbers

and species distribution of the most important

species are presented in this paper. Other papers in

this series will include studies on meiotic chromo-

some behaviour, reproduction, hybridization and

will be concluded with a cytotaxonomic study of the

genus Rubus in South Africa.

MATERIALS AND METHODS

The plants used in this study were collected

throughout South Africa and transplanted under

quarantine in the Pretoria National Botanical Gar-

den. The following 35 plants, representing nine dif-

ferent species, were used:

Eubatus

Rubus affinis :

TRANSVAAL. — 2329 (Pietersburg): Dap Naude Dam

(-DD), Stirton 5746.

R. cuneifolius:

NATAL. — 2929 (Underberg): 14 km from Swartberg to

Underberg (-CD). Stirton 8154. 2930 (Pietermaritzburg): 3 km

from Midmar Dam to Lions River (-CB), Henderson & Gaum 93;

5 km from Pietermaritzburg to Mooi River (-CB), Liengme s.n.

3029 (Kokstad): 11 km from Harding to Weza (-DB), Stirton

8102.

R. pascuus:

TRANSVAAL. — 2430 (Pilgrim's Rest): 1 km from Graskop

to Sabie (-DD), Henderson & Gaum 18. Stirton 9800, 9861 &

9868.

R. flagellaris:

TRANSVAAL. — 2530 (Lydenburg): Kaapse Hoop (-DB),

Henderson & Gaum 2.

Idaeobatus

R. apetalus:

TRANSVAAL. — 2430 (Lydenburg): Kaapse Hoop (-DB),

Henderson & Gaum 6.

NATAL. — 2929 (Underberg): Clairmont Plantation (-DD),

G. Hemm s.n. a & b. 2930 (Pietermaritzburg): Endeni Farm

(-CC), Wells 5000.

592

Bothalia 15, 3 & 4 (1985)

R. longepedicellatus:

TRANSVAAL. — 2430 (Pilgrim's Rest): 1 km from Graskop

to Sabie (-DD), Henderson & Gaum 22, Stirton 9862. 2530 (Ly-

denburg): 5 km from Lydenburg to Sabie (-AB), Henderson &

Gaum 36; Brooklands (-BA), Henderson & Gaum 14.

R. pinnatus :

TRANSVAAL. — 2530 (Lydenburg): Brooklands (-BA),

Henderson & Gaum 15.

NATAL. — 3029 (Kokstad): 4 km from Kokstad to Weza

(-CB), Arnold 1335.

R. ludwigii :

NATAL. — 2929 (Underberg): Kamsberg Nature Reserve

(-BD), Henderson & Gaum 41.

CAPE. — 3226 (Fort Beaufort): Hogsback (-DB), Admiraal &

Drijfhout 2940.

R. x proteus:

TRANSVAAL. — 2430 (Pilgrim's Rest): Bourkes Luck

(-DB), Henderson & Gaum 27, 28, 31 & 32; Mac-Mac Falls

(-DD), Henderson & Gaum 20; 1 km from Graskop to Sabie

(-DD), Stirton 9798, 9865 & 9866.

NATAF. — 2929 (Underberg): 25 km from Himeville to

Boesmansnek (-DC): Henderson & Gaum 50 & 51. 3029

(Kokstad): Ngeli Forest (-DA), Stirton 8135.

R. transvaliensis x R. longepedicellatus :

TRANSVAAL. — 2530 (Lydenburg): Nelspruit (-BD), Hen-

derson & Gaum 10.

TABLE 1. — Chromosome numbers of some South African Rubus

species

Rubus species:

TRANSVAAL. — 2430 (Pilgrim’s Rest): 5 km from Graskop

to Sabie (-DD), Henderson & Gaum 24.

Specimens are housed in the National Herbarium,

Pretoria (PRE). Chromosome counts were made

from meiotic squashes in aceto-carmine (Darlington

& LaCour, 1976). Between 20 and 25 cells per plant

were studied.

Distribution maps were obtained by using the data

of all Rubus specimens available on PRECIS (Pre-

toria computerized information system) (Gibbs Rus-

sell & Gonsalves, in press).

RESULTS AND DISCUSSION

All chromosome numbers determined were mul-

tiples of 7 and somatic chromosome numbers of 14,

21, 28, 35, 42, 49 and 56 were observed in the inves-

tigated South African Rubus species (Table 1). Poly-

ploidy occurred in six of the nine species studied.

The different Rubus species varied in regard to their

geographical distribution.

a) The subgenus Eubatus

The most widespread of the introduced Moriferi

was Rubus affinis. Specimens representing R. affinis

were collected in the northern Transvaal, Swaziland,

Natal and southern and western Cape (Fig. 1). How-

ever, the western Cape is the only place where this

species invaded the natural vegetation and where

this species is regarded as a weed. The somatic chro-

mosome number of 28 for R. affinis (Gustafsson,

1933, 1939 & 1943; Heslop-Harrison, 1953) is con-

firmed.

R. cuneifolius is restricted to Natal, with the ma-

jority of specimens collected in western Natal (Fig.

1). In addition to the chromosome number of 2n =

14 for R. cuneifolius reported by Shoemaker & Stur-

rock (1959), polyploid forms with 21 and 28 somatic

chromosomes were observed during this study.

R. pascuus was collected in the eastern Transvaal

(Fig. 1). This species frequently hybridized with R.

longepedicellatus and the hybrids are included in R

x proteus. Since R. pascuus contains both triploid

and tetraploid specimens, it may either represent a

hybrid rather than a parental form or a diploid form

might be present in South Africa.

Two R. flagellaris specimens were collected in the

eastern Transvaal (Fig. 1). This species has a somatic

chromosome number of 28 in contrast to the pub-

lished chromosome numbers of 56 (Faasen & Na-

deau, 1976) and 63 (Einset, 1947).

The geographical distribution of R. affinis and R.

cuneifolius (Fig. 1) indicates that these species oc-

cupy different habitats. This suggests that they were

differently adapted to the South African climate.

The single specimens of R. affinis collected in Trans-

vaal, Swaziland and Natal might suggest separate in-

troductions. The occurrence of polyploidy in the

subgenus Eubatus is restricted to eastern Transvaal

and Natal.

b) The subgenus Idaeobatus

The introduced species of the subgenus Idaeoba-

tus have a limited distribution. R. niveus is restricted

to Swaziland and the adjacent Transvaal areas (Fig.

Bothalia 15, 3 & 4 (1985)

593

FIG. 1. — Geographical distribution of Rubus affinis 0. R. cunei-

folius ▲ , R. pascuus ■ and R. flagellaris 0 in South Africa.

2). R. phoenicolasius is found in the central parts of

Natal (Fig. 2).

The indigenous R. rosifolius is found in eastern

Transvaal and Natal and a single specimen was

found near Cape Town (Fig. 3). R. immixtus is

found in the eastern Cape (Fig. 2).

FIG. 2. — Geographical distribution of Rubus niveus 0, R. phoe-

nicolasius ■ and R. immixtus 0 in South Africa.

FIG. 3. — Geographical distribution of Rubus rosifolius ■ and R.

apetalus 0 in South Africa.

The series Afroidaei is indigenous and contains

species with variable distributions. R. apetalus speci-

mens were collected in eastern Transvaal, Natal and

the eastern Cape (Fig. 3). R. longepedicellatus is

found mainly in eastern Transvaal but two speci-

mens from southern Natal were also observed (Fig.

4). R. pinnatus is another widespread species and

specimens from Transvaal, Natal and eastern, south-

ern and western Cape are found in the National Her-

barium (Fig. 5). R. rigidus has the widest distribu-

tion of all Rubus species in South Africa. This

species is found in Transvaal, Natal, Orange Free

State and eastern, southern and western Cape (Fig.

6). R. ludwigii is restricted to the mountainous parts

of Transvaal, the Orange Free State, Natal and the

Cape (Fig. 7). The hybrid species, R. x proteus, was

collected in eastern Transvaal and Natal (Fig. 8).

No chromosome counts for any of the indigenous

South African Rubus species were found in the liter-

ature. Therefore, new chromosome counts for the

genus Rubus include the species R. apetalus (2n = 14

& 28), R. longepedicellatus (2n = 14, 28 & 35), R.

pinnatus (2n = 14 & 28), R. x proteus (2n = 14, 21,

28, 35, 42, 49, & 56) and R. ludwigii (2n = 14).

FIG. 4. — Geographical distribution of Rubus longepedicellatus 0

in South Africa.

FIG. 5. — Geographical distribution of Rubus pinnatus 0 in South

Africa.

594

Bothalia 15, 3 & 4 (1985)

FIG. 6. — Geographical distribution of Rubus rigidus ■ in South

Africa.

FIG. 7. — Geographical distribution of Rubus ludwigii ▲ in South

Africa.

FIG. 8. — Geographical distribution of Rubus x proieus ♦ in

South Africa.

No correlation can yet be found between chromo-

some number and geographical distribution, except

that species with a high degree of polyploidy are

present in the eastern Transvaal and Natal.

CONCLUSIONS

The genus Rubus has a basic chromosome number

of 7 and a polyploid series exists, with plants varying

from diploid to duodecaploid. The tetraploid fre-

quency of the South African Rubus species is much

higher than that of the extra-African species (the

chromosome numbers of the extra-African species

were obtained from Darlington & Wylie, 1955; Orn-

duff, 1967-1969; Fedorov, 1969; Moore, R.J.,

1970-1977; Moore, D.M., 1982), whereas the fre-

quency of diploids is much lower (Fig. 9a).

Previous chromosome studies indicated that the

majority of species in the subgenus Eubatus are

polyploids, whereas the majority of Idaeobatus

species are diploid (Longley & Darrow, 1924; Gus-

tafsson, 1933, 1934 & 1943; Brown, 1943; Einset,

1947; Heslop-Harrison, 1953; Varaama, 1954; Dar-

lington & Wylie, 1955; Jinno, 1958a & b; Aalders &

Hall, 1966; Ornduff, 1967-1969; Fedorov, 1969; Ba-

quar & Askari, 1970; Moore, R.J., 1970-1977;

Moore, D.M., 1982). The South African results indi-

cated much higher triploid and tetraploid frequen-

cies in the subgenus Eubatus (Fig. 9b). The South

African Idaeobatus species have seven times more

tetraploids than the extra-African species (Fig. 9c).

Although the sample studied was not large enough

to give a true representation of the South African

Rubus population, the comparisons were made to in-

dicate possible trends that will need further investi-

gation.

These deviations from the extra-African pattern

might possibly be attributed to hybridization. Hybri-

dization between indigenous and naturalized species

could have resulted in the formation of polyploid se-

ries especially in the subgenus Idaeobatus. The oc-

currence of diploid (2n = 14) Eubatus species in

South Africa indicates sexual reproduction. The

worldwide tendency for a very high frequency of di-

ploid Idaeobatus species could have existed in South

Africa. Since two previously allopatric sexually re-

producing taxa became sympatric in South Africa,

the elimination of geographical distribution as a re-

productive isolation mechanism could have led to

extensive hybridization and subsequent chromo-

some doubling.

The fact that polyploidy occurs mainly in the sub-

genus Idaeobatus could indicate introgression where

backcrossing to the Eubatus parent were restricted.

Polyploid forms of R. cuneifolius and R. pascuus

suggest, however, that these species might be the

only Eubatus species involved in hybridization. The

fact that these species contain different ploidy levels

and are relatively widespread, supports this theory.

However, to determine the validity of this theory,

the type of polyploidy in each species must be inves-

tigated and subsequently their modes of reproduc-

tion.

A hypothesis is consequently proposed that this

diploid Eubatus species hybridized with the diploid

Bothalia 15, 3 & 4 (1985)

595

FIG. 9. — Comparison between

frequencies in chromosome

numbers between the South

African and extra-African

Rubus species (a), Eubatus

species (b) and ldaeobatus

species (c). (Shaded areas

represent South African Ru-

bus species, whereas un-

shaded areas represent extra-

African Rubus species.

These frequencies represent

the number of species per

polyploid level and not the

number of chromosome

counts.)

14 21 28 35 42 49 56 63 70 77 84

CHROMOSOME NUMBER

indigenous ldaeobatus species. The hybridization

between two different subgenera, with possibly non-

homologous or homoeologous genomes, would only

succeed after chromosome doubling. It is, therefore,

suggested that the high polyploid frequency in the

South African ldaeobatus species is the result of in-

tersubgeneric hybridization.

ACKNOWLEDGEMENT

C. H. Stirton is thanked for his careful examina-

tion and classification of the herbarium specimens.

UITTREKSEL

Die geografiese verspreiding van 14 van die Suid-

Afrikaanse Rubus spesies word weergegee. Chromo-

soomgetalle van nege spesies is vasgestel; ses hiervan

word vir die eerste keer beskryf, een word bevestig en

addisionele poliploiede vlakke word vir die ander

twee spesies beskryf.

Daar word aangetoon dat spesies van die Suid-

Afrikaanse subgenus ldaeobatus baie meer poli-

ploiede plante bevat as wat chromosoomgetalle uit

ander werelddele sou suggereer. Hierdie verskynsel

kan moontlik toegeskryf word aan verbastering tus-

sen die subgenera Eubatus en ldaeobatus.

REFERENCES

AALDERS. L. E. & HALL, I. V., 1966. A cytotaxonomic sur-

vey of the native blackberries of Nova Scotia. Can. J. Genet.

Cytol. 8: 528-532.

BAQUAR, S. R. & ASKARI. S. H. A., 1970. Chromosome

numbers in some flowering plants of West Pakistan. Genet,

iber. 22: 41-51.

BROWN, S. W., 1943. The origin and nature of variability in the

Pacific coast blackberries. Am. J. Bot. 30: 686-698.

DARLINGTON, C. D. & LACOUR. L. F., 1976. The handling

of chromosomes. London: Allen & Unwin.

DARLINGTON, C. D. & WYLIE. A. P., 1955. Chromosome

atlas of flowering plants. London: Allen & Unwin.

EINSET, J., 1947. Chromosome studies in Rubus. Genres Herb.

7: 181-192.

FAASEN, P. & NADEAU, P., 1976. In IOPB chromosome

number reports LI. Taxon 25: 155-164.

FEDOROV, A. A., 1969. Chromosome numbers of flowering

plants. Leningrad: Academy of Science, USSR.

596

Bothalia 15, 3 & 4 (1985)

FOCKE, W. O., 1910-1914. Species Ruborum /-///. Stuttgart:

Schweizerbart.

GIBBS RUSSELL, G. E. & GONSALVES, P . in press. PRE-

CIS: a curatorial and biogcographic system. In F. A. Bisbv.

Symposium volume: computer databases in systematics. Lon-

don: Academic Press.

GUSTAFSSON, A., 1933. Chromosomenzahlen in der Gattung

Rubus. Hereditas 18: 77-80.

GUSTAFSSON. A.. 1939. Differential polyploidy within the

blackberries. Hereditas 25: 33-47.

GUSTAFSSON. A.. 1943. The genesis of the European black-

berry flora. Acta Univ. land. 39: 1-200.

HESLOP-HARRISON. Y.. 1953. Cytologies! studies in the

genus Rubus L. 1. Chromosome numbers in the British Ru-

bus flora. New Phytol. 52: 22-32.

JINNO, T. . 1958a. Cytogenetic and cvto-ccological studies on

some Japanese species of Rubus 1. Chromosomes. Bot.

Mag., Tokyo 71: 15-23.

JINNO. T.. 1958b. Cytogenetic and cyto-ecological studies on

some Japanese species of Rubus 2. Cytogenetic studies on

some F, hybrids. Jap. I. Genet. 33: 201-209.

LONGLEY. A. E. & DARROW, G. M . 1924. Cytological stu-

dies of diploid and polyploid forms of raspberries. ./. agric.

Res. 27: 737-748.

MOORE. D. VI.. 1482. Flora Etiropaea — Checklist and chromo-

some index. Cambridge: University Press.

MOORE. R.J., 1970-1977. Index to plant chromosome numbers

for 1968-1974. Regnum Veg. 68. 77, 84. 91 & 96.

MOORE. R. J.. 1973. Index to plant chromosome numbers for

1967-1971. Utrecht: Oosthoek Uitgevers.

ORNDLIFF. R.. 1967-1969. Index to plant chromosome numbers

for 1965-1967. Regnum Veg. 50. 55 & 59.

SHOEMAKER. J. S. & STURROCK.T. T.. 1959. Chromosome

relations in blackberries. Troc. Fla St. lion. Soc. 72:

327-333.

VARAAMA. A.. 1954. Chromosome numbers of some species

and hybrids of the genus Rubus. Seuran Vanamon Tiedonan-

not 8:’ 192-195.

Bothalia 15. 3 & 4: 597-606 (1985)

The genus Rubus in South Africa. II. Meiotic chromosome behaviour

J. J. SPIES . H. DU PLESSIS* and H. LIEBENBERG**

Keywords: chromosomes, hybridization, meiosis. polyploidy. Ruhus

ABSTRACT

Meiotic chromosome behaviour in the genus Rubus is relatively normal. Polyploidy occurs in both South African

subgenera, i.e. Eubatus and Idaeobatus. The subgenus Eubatus contains plants tending mostly towards autoploidy,

whereas the subgenus Idaeobatus varies from autoploid. through segmental alloploid to alloploid. It is concluded

that this apparent difference might be due to the study of a statistically insufficient number of plants and that

alloploidy originated from intcrsubgcneric hybridization.

INTRODUCTION

The genus Rubus includes the blackberries, rasp-

berries and dewberries and is regarded as one of the

world’s most complex taxonomic groups. In an at-

tempt to solve the Rubus riddle, all available taxo-

nomic methods should be exploited. Among these

methods cytogenetics could play a major role. The

cytogenetic contributions to the taxonomy of Rubus

can be separated into two distinct types of data. The

contribution may involve information regarding

either the reproductive system or it may involve the

ploidv status and genome analysis of the specimen.

The genus Rubus is subdivided into 12 subgenera,

of which two are represented in South Africa. All

species belonging to the subgenus Eubatus Focke are

exotic in South Africa, whereas the subgenus Idaeo-

batus Focke contains two exotic and nine indigenous

species (Spies & Du Plessis, 1985).

L. H. Bailey did not describe any Rubus as a hy-

brid in his classical work ‘Species Batorum’ (1941).

His criteria for hybridization included the presence

of both parents in close proximity to the hybrid and

the hybrid must be morphologically intermediate be-

tween its parents. Strict application of these criteria

resulted in the description of a multitude of different

species. Another paper in this series will deal with

the validity of these criteria. This study is an attempt

to determine, through a study of meiotic chromo-

some behaviour, w'hether the indigenous polyploid

species (Spies & Du Plessis, 1985) originated

through alio-, segmental alio- or autoploidy. The

chromosome behaviour of the indigenous species

will also be compared with that of the introduced

species.

MATERIALS AND METHODS

The plants used during this study were listed in a

previous publication (Spies & Du Plessis, 1985).

Aceto-carmine squashes of anthers (Darlington &

LaCour. 1976) were used to study chiasma frequen-

cies chromosome associations, metaphase I, ana-

* Botanical Research Institute. Department of Agriculture &

Water Supply. Private Bag X 1 0 1 . Pretoria 0001.

** Department of Genetics, Universitv of Pretoria, Pretoria

0001.

phase I and telophase II. At least 20 cells per plant

were studied.

Diakineses were used to study chiasma frequen-

cies and chromosome associations. Chiasma fre-

quencies were calculated as the average number of

chiasmata per half bivalent. This frequency was cal-

culated by dividing the total number of chiasmata

observed by the somatic chromosome number. The

expected frequencies of bivalent and multivalent for-

mation for autoploid plants were calculated by using

the method described by Spies (1984a).

RESULTS

The chiasma frequencies varied from 0,45 in a R.

x proteus specimen to 1,41 in a R. apetalus specimen

(Table 1). The low chiasma frequencies observed in

all triploid (2n = 21) and pentaploid (2n = 35)

plants, could be attributed to the high frequency of

univalents. If a correction factor of Z was added for

each univalent, values between 1,15 and 1.22 for tri-

ploid and 1,1 to 1,15 for pentaploids were obtained.

The correction factor (Z) is obtained by dividing the

sum of the number of univalents (U) and a third of

the number of trivalents (T) per cell by the somatic

chromosome number (N) {Z = [U + (T/3)]/N|. The

variation in chiasma frequency, after correction, was

decreased and varied between 1.01 and 1,41.

Chromosome associations were presented as the

percentage of chromosomes bound as uni-, bi- or

multivalents in Table 1. Differences in chromosome

behaviour between different species and even within

a species were demonstrated by the number of multi-

valents formed. The number of univalents during

metaphase I increased with an increase in ploidy le-

vel (Table 2). Anaphase I (Table 3) and telophase II

(Table 4) were normal in almost all diploid (2n = 14)

and tetraploid (2n = 28) plants, whereas abnormali-

ties occurred in triploid (2n = 21) and pentaploid (2n

= 35) plants. These abnormalities included laggards

and a maldistribution of chromosomes during ana-

phase I, as well as the occurrence of additional

micronuclei during telophase II.

The expected frequencies of bivalents and multi-

valents formed in autoploids were determined for all

the triploid and tetraploid plants. These results are

presented in Tables 5 & 6. Chromosome behaviour

in the different species was as follows:

598

Bothalia 15, 3 & 4 (1985)

TABLE 1. — Chiasma frequencies and chromosome associations in different Rubus species

* Exotic species

**Corrected values

a) The subgenus Eubatus Focke (all species are

exotic)

Rubus affinis Wh. & N.

This tetraploid (2n = 28) species had an average

chiasma frequency of 1,22 ( Stirton 5746). The aver-

age chromosome association was 0,21, 10,71 n 0,21in

1,43IV and it varied from 14u to 8n 31v. Although

early segregation of chromosomes was observed dur-

ing metaphase I (3 cells had either 1 or 2 univalents),

anaphase I was normal in all the cells studied. No

additional micronuclei were observed during telo-

phase II. The high correlation between the expected

and observed chromosome association frequencies

indicated that this species was autotetraploid.

R. cuneifolius Pursh.

This weedy polyploid species had chromosome

numbers of 2n = 14, 21 and 28 (Spies & Du Plessis,

1985). The two diploid specimens, Stirton 8102 and

8154, had respectively average chiasma frequencies

of 1,04 and 1,09. Both plants formed 7I( per cell dur-

ing diakineses. No univalents were observed during

metaphase I and a normal 7-7 anaphase I segrega-

tion was observed. Both plants exhibited normal te-

lophase II cells.

The triploid plant, Liengme s.n., had an average

chiasma frequency of 0,85 and an average chromo-

some association of 6, 3[ 6,3n 0,7in. Liengme s.n. had

an average of 5,3 univalents (varying from 3 to 7 per

cell) per cell during metaphase I. Anaphase I

showed an unequal segregation of chromosomes and

laggards. The chromosome segregation varied from

11-10 to 13-8. Telophase II appeared normal. Triva-

lents occurred and the plant was found to be an auto-

ploid.

The tetraploid specimen, Henderson & Gaum 93,

had an average chiasma frequency of 1,19 and an av-

erage chromosome association of 2,3r 7,5n l,3m

1,7IV. An average of 2,3 univalents (varied from 0 to

4) per metaphase I cell was observed. Anaphase I

occasionally had 1 to 2 laggards. One telophase II

cell contained an additional micronucleus. The ob-

served frequency of chromosome associations corre-

sponded with the expected frequency for autotetra-

ploid plants and Henderson & Gaum 93 could conse-

quently be regarded as an autotetraploid.

R. pascuus Bailey

This species escaped from cultivation and became

a weed in the eastern Transvaal. The two triploid

specimens, Henderson & Gaum 18 and Stirton 9800,

Bothalia 15, 3 & 4 (1985)

599

TABLE 2. — Number of univalents per metaphase I cell in some South African Rubus species

Percentage of cells with

_ c ■ different numbers of Average

Species Specimen no. 2n = univalents numberof

univalents

0123456789 10

Eubatus

* Exotic species

corresponded in regard to their chiasma frequencies

and chromosome associations (Table 1). They had

an average chiasma frequency of 0,83 and an aver-

age chromosome association of 6,5, 6,ln 0,8UI. The

number of univalents during metaphase I varied

from two to eight per cell with an average of 6,4.

Anaphase I was only studied in Henderson & Gaum

18 and a chromosomal maldistribution (11-10 and

12-9) was observed, as well as up to five laggards. In

spite of the abnormal anaphase I, 80% telophase II

cells appeared normal in both plants. Both plants

tended towards autoploidy.

The two tetraploid plants, Stirton 9861 and 9868,

had slightly different chiasma frequencies. These dif-

ferences are reflected in different chromosome asso-

ciations where Stirton 9861, with its lower chiasma

frequency, had more bivalents and less multivalents

than Stirton 9868, with its higher chiasma frequency.

Both plants had normal metaphase I, anaphase I and

telophase II stadia and they represented autoploids.

R. flagellaris Willd.

The tetraploid plant representing this species,

Henderson & Gaum 2, had a chiasma frequency of

1,08 and an average chromosome association of 0,1,

13,9n with a normal metaphase I, anaphase I and

telophase II. This plant was found to be an alloploid .

b) The subgenus Idaeobatus Focke (with the ex-

ception of R. phoenicolasius and R. niveus, all

species are indigenous).

R. apetalus Poir.

The diploid specimen, G. Hemm s.n., had an av-

erage of 1,11 chiasmata per half bivalent and all cells

showed only bivalents. No abnormalities were ob-

served during metaphase I, anaphase I and telo-

phase II.

The three tetraploid plants, G. Hemm s.n., Hen-

derson & Gaum 6 and Wells 5000, differed signifi-

cantly in regard to their cytogenetic behaviour. The

average chiasma frequencies varied from 1,06 in

Wells 5000 to 1,41 in G. Hemm s.n. and the average

chromosome association from 14„ to 11, ln 1,4IV in

the two plants mentioned. Metaphase I, anaphase I

and telophase II were normal in all three plants with

the exception of one anaphase I cell of Wells 5000

where one laggard was observed. The differences in

chromosome association indicate that G. Hemm s.n.

and Henderson & Gaum 6 may be considered as seg-

mental alloploids, whereas Wells 5000 may represent

an alloploid.

600

Bothalia 15, 3 & 4 (1985)

TABLE 3. — Chromosome segregation during anaphase I in some South African Rubus species

Species Specimen no. Chromosome segregation

Diploid plants 7+7 6+6+2

■"Exotic species

R. longepedicellatus (C.E. Gust.) C. H. Stirton

This species contained diploid, tetraploid and pen-

taploid specimens. The diploid specimen, Hender-

son & Gaum 22, had an average of 1,12 chiasmata

per half bivalent and an average chromosome asso-

ciation of 0,3, 6,85n. Three cells contained two uni-

valents during metaphase I and one anaphase I cell

contained two laggards. Telophase II was normal.

The two tetraploid specimens, Henderson &

Gaum 14 and Stirton 9862, had respectively an aver-

age of 1,04 and 1,11 chiasmata per half bivalent and

an average chromosome association of 0,59, 13,5n

0,1IV and 0,16, 13,92„. Metaphase I, anaphase I and

telophase II were normal in both plants. Henderson

& Gaum 14 was found to be a segmental alloploid,

whereas Stirton 9862 represented a true alloploid.

The pentaploid plant, Henderson & Gaum 36, had

an average of 0,99 chiasmata per half bivalent and an

average chromosome association of 5,2, 10,95,, 1,7„,

0,7IV. Metaphase I had an average of 5,1 (1-8) uni-

valents per cell. Additional micronuclei were ob-

served during telophase II. The absence of pentava-

lents and low frequency of multivalents suggested

that this plant was a segmental alloploid.

R. pinnatus Willd.

Henderson & Gaum 15, the diploid specimen, had

an average chiasma frequency of 1,09 and an aver-

age chromosome association of 0,1, 6,95,,. Meta-

phase I, anaphase I and telophase II were normal.

Arnold 1335, the tetraploid specimen, had an av-

erage of 1,24 chiasmata per half bivalent and an av-

erage chromosome association of 11,84„ 1,08IV. Me-

taphase I, anaphase I and telophase II were normal.

The observed frequency of multivalents was lower

than the expected and, therefore, this plant was a

segmental alloploid.

R. ludwigii Eckl. & Zeyh.

The specimens representing this diploid species,

Admiraal & Drijfhout 2940 and Henderson & Gaum

41, had average chiasma frequencies of 1,2 and 1,18

respectively and average chromosome associations

of 0,08, 6,96„ and 7,,. All studied metaphase I, ana-

phase I and telophase II cells were normal.

R. x proteus

The diploid specimen studied, Henderson &

Gaum 28, had an average chiasma frequency of 1,1

and all cells contained 7,,. No abnormalities were

seen during metaphase I, anaphase I or telophase II.

The triploid specimen, Stirton 9866, had an aver-

age chiasma frequency of 0,88 and an average chro-

mosome association of 6,2, 5,95„ 0,9,,,. There were

between three and seven univalents (average is 5)

during metaphase I. No additional micronuclei were

observed during telophase II. This plant tended to-

wards autoploidy.

Four tetraploid specimens, Stirton 9798, Hender-

son & Gaum 27, 32 and 51, were included in this

Bothalia 15, 3 & 4 (1985)

601

TABLE 4. — Number of micronuclei during telophase II in some South African Rubus species

*Exotic species

TABLE 5. — Comparison between expected and observed chromosome association frequencies in triploid Rubus specimens

E Expected frequency

O Observed frequency

species. The average chiasma frequency varied from

0,45 in Stirton 9798 to 1,18 in Henderson & Gaum

51. The low chiasma frequency in Stirton 9798 might

be attributed to asynapsis. In a number of cells no

chiasmata were formed and 28 univalents were seen.

This phenomenon resulted in an average chromo-

some association of 11, 6r 4,7n in Stirton 9798 com-

pared to an average of 0,2, 8,6„ 0,2in 0,81V in Hen-

derson & Gaum 51. The asynapsis further resulted in

an average of 6 (1-10) univalents per metaphase I

cell. The only other plant having univalents during

metaphase I was Henderson & Gaum 32 with an av-

erage of 0,4 per cell. Anaphase I and telophase II

were normal in Henderson & Gaum 27, 32 and 51,

602

Bothalia 15, 3 & 4 (1985)

TABLE 6. — Comparison between expected and observed chromosome association frequencies in tetraploid Rubus specimens

* Exotic species

** p and q were used to determine the expected values

E Expected frequency

O Observed frequency

Bothalia 15, 3 & 4 (1985)

603

whereas Stirton 9798 had either laggards or a chro-

mosomal maldistribution during anaphase I. These

abnormalities resulted in additional micronuclei dur-

ing telophase II and no normal telophase II cells

were seen. The differences in chromosome pairing

among these plants are reflected in the fact that Hen-

derson & Gaum 51 is considered as an autoploid,

Henderson & Gaum 27 as a segmental alloploid ten-

ding towards autoploidy, Henderson & Gaum 32 as a

segmental alloploid tending towards alloploidy and

Stirton 9798 as an alloploid.

The pentaploid specimens, Stirton 9865 and Hen-

derson & Gaum 20, had respectively average

chiasma frequencies of 1,05 and 0,91 and chromo-

some associations of 3,3, 14,85„ 0,25m 0,3[v and 6,3,

13, 6n 0,5m. Although both plants had an average of

3,4 univalents per cell during metaphase I, the num-

ber of univalents varied from 1 to 6 in Stirton 9865

and 0 to 10 in Henderson & Gaum 20. During telo-

phase II Stirton 9865 had two cells containing an ad-

ditional micronucleus. Stirton 9865 is regarded as a

segmental alloploid and Henderson & Gaum 20 as a

segmental alloploid tending towards alloploidy.

The hexaploid specimen studied, Henderson &

Gaum 31, had an average chiasma frequency of 1,21

and an average chromosome association of 1 ,85T

14, 4„ 2,05ni 1,3[V. All the metaphase I cells had at

least one univalent (average = 1,9). Due to the high

chromosome number anaphase I could not be stu-

died. Half of the studied telophase II cells were ab-

normal, containing 1 to 3 additional micronuclei.

The absence of higher order configurations than

quadrivalents and the relatively low frequency of tri-

and quadrivalents seemed to indicate that this speci-

men has been a segmental alloploid.

R. transvaliensis x R. longepedicellatus

This tetraploid hybrid specimen, Henderson &

Gaum 10, had an average chiasma frequency of 1,07

and 100% bivalents were formed. No abnormalities

were observed in any meiotic stage. The chromoso-

mal behaviour indicated that this specimen

represented an alloploid.

Rubus species

The herbarium personnel were unable to identify

Henderson & Gaum 24. This plant was tetraploid

and had a chiasma frequency of 1,16 and an average

chromosome association of 2,24, 7,99n 0,65m 1 ,95IV.

Anaphase I and telophase II were normal, whereas

metaphase I had an average of 2,6 univalents per

cell. According to the number of multivalents

formed, this specimen represented a segmental allo-

ploid tending towards autoploidy.

DISCUSSION

Most Rubus chromosomes have the ability to par-

ticipate in the formation of more than one chiasma.

During diplotene/early diakinesis most bivalents oc-

cur as ringbivalents (Figs la & 2a), whereas later di-

akinesis stages have a majority of chainbivalents

(Figs lb & 2b-2f). This might possibly be due to

chiasma terminalization. The formation of multiva-

lents may suppress the effect of chiasma terminaliza-

tion and thus increases the chiasma frequency.

It is sometimes difficult to determine whether me-

taphase I univalents originated from asynapsis or

early segregation (Figs lc & 2g & h). The occurrence

of univalents during diakinesis would suggest asy-

napsis. This segregation of univalents might be ran-

dom (for example see Fig. lc: only one univalent can

be seen on one side of the metaphase plate and three

on the other side of the camera lucida drawing). The

effect of univalents during metaphase I is not very

serious, because their occurrence is restricted to

either high ploidy levels or uneven polyploid levels,

where it is presumed that apomixis will occur.

Chromosome laggards are occasionally observed

in diploid and tetraploid Rubus plants. The number

of laggards increases drastically in triploid plants

where a maldistribution of chromosomes is also

found.

An exceptionally high frequency of apparently

normal telophase II cells was observed (Figs If &

2j). This may be due to the formation of microspores

FIG. 1. — Camera lucida drawings

of different meiotic stages in

R. x proteus (Stirton 9865). a,

diplotene/early diakinesis; b,

diakinesis; c, metaphase I; d,

telophase I; e, anaphase II; f,

telophase II: C. chainbiva-

lent; H, horizontal division;

L, univalents or early segre-

gating chromosomes; M, me-

taphase chromosomes; Q,

quadrivalent; R. ringbivalent;

S, secondary chromosome as-

sociation; T, trivalent; U, uni-

valent. (x 1600.)

604

Bothalia 15, 3 & 4 (1985)

with varying numbers of chromosomes rather than

excluding laggards through micronuclei. During ana-

phase II chromatid segregation might occur without

being preceded by the formation of a cell wall (Fig.

Id & 2i). Chromosome laggards from anaphase I

might thus be incorporated into the tetrad nuclei. A

study of pollen fertility might prove interesting.

The meiotic chromosome behaviour mentioned

above indicates differences in different Rubus

species complexes. These differences in chromo-

some pairing (Table 1) indicate either differences in

genome homology or the existence of genes inhibi-

ting chromosome pairing in certain species.

All species belonging to the subgenus Eubatus oc-

cur as exotics in South Africa. Both naturalized

South African polyploid Eubatus species studied, R.

affinis and R. cuneifolius, were apparently auto-

ploid. All published chromosome numbers for R. af-

finis indicated a somatic chromosome number of 28

(Gustafsson, 1933, 1939 & 1943; Heslop-Harrison,

1953; Spies & Du Plessis, 1985) and, therefore, it is

not known if the diploid form still exists. An excep-

tion from the general autoploid situation was found

in the octoploid R. cuneifolius specimen, Henderson

& Gaum 50, which was a segmental alloploid and in

R. flagellaris where the only studied specimen rep-

resented an alloploid. The fact that the South Afri-

can specimen is tetraploid, whereas extra-African

specimens were either octoploid or nonaploid (Ein-

set, 1947; Faasen & Nadeau, 1976) is an indication

that further studies of this group are necessary be-

fore any conclusions can be made.

The existence of triploid and tetraploid R. pascuus

forms might suggest that the diploid form may still

be present in South Africa. The fact that all speci-

mens represented autoploidy or tended towards au-

Bothalia 15, 3 & 4 (1985)

605

toploidy supports this suggestion. It is further sup-

ported by the fact that a diploid R. x proteus speci-

men was observed. Since R. x proteus originated as

a hybrid between R. pascuus and R. longepedicella-

tus, a diploid hybrid specimen suggests diploidy in

both parents.

The subgenus Idaeobatus contains two exotic and

nine indigenous species in South Africa. The chro-

mosomal behaviour of these species is variable and

autoploidy, segmental alloploidy and alloploidy

were observed. Different systems might be operat-

ing in some species, resulting in either autoploidy

and alloploidy (R. x proteus) or alloploidy and seg-

mental alloploidy (R. longepedicellatus & R. apeta-

lus) in the same species.

It would seem that the majority of exotic Rubus

plants in South Africa tends towards autoploidy

(77,8%), whereas the minority of indigenous Rubus

specimens tends towards autoploidy (23,5%). The

suggested ploidy types of different Rubus species are

summarized in Table 7.

The high frequency of autoploidy found in this

Eubatus population, does not support an intersubge-

neric hybridization theory (Spies & Du Plessis, 1985)

or even interspecific hybridization within the subge-

nus. However, it must be remembered that the

specific delimitation in Rubus does not conform with

the biological species concept. Therefore, ‘interspe-

cific hybridization’ within a Rubus subgenus could

lead to autoploidy. A prerequisite for this assump-

tion is that the current classification system must

represent natural relationships. Plants within a sub-

genus must, therefore, be more closely related to

one another than to any species in another subgenus.

It is further assumed that karyotipic evolution, al-

though not directly correlated with morphological

divergence, progresses along the same lines. Large

genomic differences are, therefore, more likely to be

expected in different subgenera than within a subge-

nus. Consequently, hybridization within a subgenus

is more likely to involve smaller chromosomal differ-

ences and even interspecific hybridization within the

subgenus would be more likely to produce segmental

alloploids tending towards autoploidy.

In contrast to the South African Eubatus species,

the Idaeobatus species have only 23,5% autoploids

or segmental alloploids tending towards autoploidy

and 35,3% alloploids or segmental alloploids tend-

ing towards alloploidy. These figures indicate that

hybridization might occur more frequently in the

subgenus Idaeobatus than in the subgenus Eubatus

in South Africa.

Crane & Thomas (1949) described preferential

pairing in Eubatus x Idaeobatus hybrids when each

genome is represented twice. Their finding is not

supported by this study in which tetraploid R. x pro-

teus [a hybrid between R. pascuus (Eubatus) and R.

longepedicellatus (Idaeobatus)] specimens varied

from alloploid to autoploid. Crane & Thomas (1949)

also described intergenomic pairing with little re-

striction when each genome is represented only

once. This finding is supported by the chromosome

behaviour of Henderson & Gaum 28.

The preferential pairing in Rubus is due to minor

alterations of the genetic material. The degree of

TABLE 7. — The polyploid classification of some South African Rubus species (1 represents autoploidy, 2-4 segmental alloploidy, with 2

tending towards autoploidy and 4 tending towards autoploidy and 5 represents alloploidy)

Exotic species

606

Bothalia 15, 3 & 4 (1985)

chromosomal alterations varies in different Rubus

species complexes. Due to these genomic differences

in the hybrids, the hybrids varied from sterile to fer-

tile in some interspecific crosses involving plants on

the same (Jinno, 1958; Britton & Hull, 1959; Haskell

& Tun, 1961) or at different (Crane & Darlington,

1927; Crane & Thomas, 1949; Shoemaker & Stur-

rock, 1959; Bammi, 1964) ploidy levels. The implica-

tions of these phenomena are that intersubgeneric

hybridization will apparently result in alloploidy at

tetraploid level and segmental alloploidy or alloploi-

dy at other ploidy levels. However, according to the

present study, intersubgeneric hybrids may even re-

present autoploidy. Therefore, the type of ploidy in

Rubus can make only a limited contribution to the

knowledge of hybridization among different species.

The reason for this phenomenon might be that the

current classification of this genus is artificial and

does not represent the true phylogenetic relationship

between species.

The relatively normal meioses observed in diploid

plants also seems to contradict hybridization. How-

ever, intergenomic pairing with little restriction was

described in such cases by Crane & Thomas (1949).

Consequently, seemingly normal diploid plants

might represent intersubgeneric hybrids.

Longley & Darrow (1924) described the subgenus

Idaeobatus as being diploid with almost no reproduc-

tive isolation between the different species. There-

fore, hybridization among Idaeobatus species does

not require chromosome doubling for the restora-

tion of fertility. Consequently, the high polyploidy

frequency and especially the alloploid situation in

the South African Idaeobatus species suggest hybri-

dization at intersubgeneric level. The fact that ap-

parently all hybrids represent their Idaeobatus par-

ents could be attributed to several factors. Either the

hybridization hypothesis is incorrect, or the occur-

rence of matrocliny (Markarian & Olmo, 1959) com-

bined with oneway hybridization (introgression),

suppressed the presence of hybrids representing the

subgenus Eubatus morphologically. The chances of

collecting only Fj matroclinous hybrids is extremely

small and introgression was not described in any

other intersubgeneric Rubus hybrids (Crane &

Thomas, 1949; Jinno, 1958; Britton & Hull, 1959;

Haskell & Tun, 1961; Thompson, 1961). It is, there-

fore, concluded that the assumption that hybridiza-

tion occurs only within Idaeobatus species is erro-

neous due to statistically insufficient material stu-

died arid that the original assumption of intersubge-

neric hybridization is still valid.

UITTREKSEL

Meiotiese chromosoomgedrag in die genus Rubus

is relatief normaal. Poliploi'die kom in beide Suid-

Afrikaanse subgenera, nl. Eubatus en Idaeobatus,

voor. Die subgenus Eubatus bevat plante wat meren-

deels na outoploidie neig, terwyl die subgenus Idaeo-

batus varieer van outoploied, deur segmenteel allo-

ploled tot by alloploied. Uit die data word afgelei dat

hierdie skynbare verskil toegeskryf kan word aan ’n

statisties onvoldoende aantal plante en dat alloploidie

ontstaan het na inter sub generiese verbastering.

REFERENCES

BAILEY, L. H., 1941. Species Batorum. The genus Rubus in

North America I. Gentes Herb. 5: 3-22.

BAMMI, R. B. K., 1964. Cytogenetics and natural hybridization

in Rubus procerus Muell. and R. lacinatus Willd. Ph.D. the-

sis, University of California.

BAQUAR, S. R. & ASKARI, S. H. A., 1970. Chromosome

numbers in some flowering plants of West Pakistan. Genet,

iber. 22: 41-51.

BRITTON, D. M. & HULL, J. W., 1959. A black raspberry-

blackberry hybrid. Proc. Am. Soc. hort. Sci. 73: 156-157.

CRANE, M. B. & DARLINGTON, C. D., 1927. The origin of

new forms in Rubus. Genetica 9: 241-276.

CRANE, M. B. & THOMAS, P. T., 1949. Reproductive versatil-

ity in Rubus. III. Raspberry-blackberry hybrids. Heredity 3:

99-107.

DARLINGTON, C. D. & LACOUR, L. F., 1976. The handling

of chromosomes. London: Allen & Unwin.

EINSET, J., 1947. Chromosome studies in Rubus. Gentes Herb.

7: 181-192.

FAASEN, P. & NADEAU, P., 1976. In IOPB chromosome

number reports LI. Taxon 25: 155-164.

GUSTAFSSON, A., 1933. Chromosomenzahlen in der Gattung

Rubus. Hereditas 18: 77-80.

GUSTAFSSON, A., 1939. Differential polyploidy within the

blackberries. Hereditas 25 : 33-47.

GUSTAFSSON, A., 1943. The genesis of the European black-

berry flora. Acta Univ. lund. 39: 1-200.

HASKELL, G. & TUN, N. N., 1961. A squash technique for the

small chromosomes of Rubus and Ribes. Hort. Res. 1:62-63.

HESLOP-HARRISON, Y., 1953. Cytological studies in the

genus Rubus L. 1. Chromosome numbers in the British Ru-

bus flora. New Phytol. 52: 22-32.

JINNO, T., 1958. Cytogenetic and cytoecological studies on some

Japaneses species of Rubus. 2. Cytogenetic studies on some

F, hybrids. Jap. J. Genet. 33: 201-209.

LONGLEY, A. E. & DARROW, G. M., 1924. Cytological stu-

dies of diploid and polyploid forms of raspberries. J. agric.

Res. 27: 737-748.

MARKARIAN, D. & OLMO, H. P., 1959. Cytogenetics of Ru-

bus. 1. Reproductive behaviour of R. procerus Muell. J.

Hered. 50: 131-136.

MARKS, G. E., 1951. Chromosome counts of species and var-

ieties of garden plants. A. Rep. John Innes hort. Inst. 41:

47-50.

SHOEMAKER, J. S. & STURROCK,T. T., 1959. Chromosome

relations in blackberries. Proc. Fla St. hort. Soc. 72:327-333.

SPIES, J. J., 1984a. Determination of genome homology in poly-

ploids. 5. Afr. J. Sci. 80: 44-46.

SPIES, J. J., 1984b. Genaturaliseerde plantegroei: 'n sitotaksono-

miese ondersoek van Lantana camara. Unpubl. report. Dept.

Agric. South Africa.

SPIES, J. J. & DU PLESSIS, H. , 1985. The genus Rubus in South

Africa. 1. Chromosome numbers and geographical distribu-

tion. Bothalia. In press.

THOMPSON, M. M., 1961. Cytogenetics of Rubus II. Cytologi-

cal studies of the varieties ‘Young’, ‘Boysori and related

forms. Am. J. Bot. 48: 667-673.

TISCHLER, G., 1934. Die Bedeutungen der Polyploidie fur der

Angiospermen, erlautert an den Arten Schleswig-Holsteins,

mit Ausblichten auf andere Florengebiete . Bot. Jb. 67 : 1-36.

VARAAMA, A., 1954. Chromosome numbers of some species

and hybrids of the genus Rubus. Seuran Vanomon Tiedonan-

not 8: 192-195.

Bothalia 15, 3 & 4: 607-611 (1985)

Evidence of a volatile attractant in Ficus ingens (Moraceae)*

N. P. BARKER**

Keywords: fig, gas chromatography, pollinator attraction, volatile compound, wasp pollinator

ABSTRACT

The relationship between the fig Ficus ingens (Miq.) Miq. and its wasp pollinator, Platyscapa soraria Wiebes,

was studied. A mechanism of pollinator attraction is discussed and corroboration obtained by using gas chromato-

graphy to show the presence of volatile compounds. These compounds are shown to be released at the female

phase of the fig's flowering cycle. These data are correlated to pollinator behaviour and visitation data obtained

from field observations.

INTRODUCTION

The close relationship between the fig and its wasp

pollinator has long been known, and there are nu-

merous studies that have shed much light on the

complexity of this relationship, such as Ramirez

(1970, 1974) Galil & Eisikowitch (1974), Galil

(1977), Janzen (1979a) and Wiebes (1979), to men-

tion but a few of the more modern studies.

In papers that concern pollinator behaviour sev-

eral authors such as Hill (1967), Galil (1977) and

Janzen (1979a) have stated that the method of polli-

nator attraction is olfactory, but this is assumed and

no evidence is provided, janzen (l.c. ) states in his

conclusion, as well as elsewhere in his paper, that

‘This pollination is achieved at a very low density of

flowering trees, probably by chemical attraction of

the wasps'. Ramirez (1970) cites several examples of

long-distance pollinator attraction in the genus Fi-

cus.

The complex interaction of the life cycles of the fig

and the wasp will not be discussed here in detail, as

this aspect has been well researched by workers such

as Ramirez (1970, 1974), Galil & Eisikowitch

(1968a, 1968b), Galil (1977), Wiebes (1979) and Jan-

zen (1979a).

In this study an attempt was made to discover

whether or not Ficus ingens produces an odour or

scent, and to determine if this scent production coin-

cided with the receptive female phase of the fig tree.

MATERIALS AND METHODS

Study sites

Plants of Ficus ingens were studied at a site on the

Braamfontein Spruit, in the suburb of Craighall

Park, Johannesburg. The site consisted of a popula-

tion of 23 fig trees of varying sizes (three seedlings,

12 established trees and the rest large, mature trees)

growing on the northfacing aspects of granite out-

crops. Throughout the nine months of observation,

it was noted that, at any one time, no more than one

tree was at a particular flowering stage. There was,

* Forms part of a B.Sc. Hons project undertaken in the Depart-

ment of Botany. University of the Witwatersrand.

** Botanical Research Institute, Department of Agriculture &

Water Supply. Private Bag X101, Pretoria 0001.

however, one tree at the male phase (wasp-hatching

phase) and one tree at the female phase (egg-laying

stage for the wasps) during August. This meant that

the wasps were emerging from syconia of one tree

and flying to the next, a few metres away. During

the study period evidence for flowering was found

on only seven trees in the population. Most of these

seemed to have just finished flowering. This is a little

unusual, as more of the population should have been

in flower. The serious drought and winter weather

may have had an effect on the flowering of the trees.

The site was visited every two or three days, and

every day for the duration of peak syconial activity.

Herbarium specimens of the studied trees are

housed in the Moss Herbarium, University of the

Witwatersrand (J). Specimens of the wasps are

housed in the Zoology Museum, University of the

Witwatersrand.

Scanning Electron Microscopy

Wasps emerging from syconia were captured in a

trap consisting of a wire frame and nylon stocking,

which was tied over a branch with syconia of the cor-

rect age. This age was determined by the increased

size and slight colour change in the fruit. Newly

hatched wasps could be found inside the syconium

when it was split open, indicating the correct syco-

nial state.

When a sufficient number of the wasps had

emerged, the twig was cut off, and the trap then

taken (with twig in situ ) to the laboratory. The stock-

ing was then cut away from the wire and twig, and

the wasps collected and anaesthetized. These wasps

were then stored in 100% alcohol and kept for view-

ing. They were then critical point dried, placed on a

SEM stub, coated with gold-palladium and studied.

Photographs were taken when required.

Gas chromatography

Syconia of the female phase were collected and

taken to the laboratory on ice. They were then divi-

ded up into size classes, ranging from 6 mm and less

to 9 mm and more, at intervals of 1 mm as measured

across the diameter.

These were then placed in small glass bottles with

glass stoppers. The bottles had been cleaned sequen-

tially with distilled water, acetone and chromic acid.

The atmosphere inside the bottles was then allowed

608

Bothalia 15, 3 & 4 (1985)

to equilibrate for 30 minutes, the bottles being kept

at room temperature.

The control consisted of an empty bottle that had

been subjected to the same cleaning treatment. The

atmosphere inside this bottle was analysed in the

same manner as the samples.

The method used in the chromatographic analysis

was the solvent effect, as described in Grob (1975).

This method involves the shut-down of the carrier

gas supply to gas chromatograph (GC), and the in-

jection of 25 p\ of a suitable solvent (in this case n-

Hexane). This is injected by a glass and stainless

steel syringe. Fifteen seconds after this first injec-

tion, 3 ml of the atmosphere from the bottles is in-

jected using a clean gas-tight syringe. One minute

after this second injection, the carrier gas flow is re-

sumed and the temperature program initiated. The

advantages and theory of this method are outlined in

Grob (1975).

The temperature program that was run started at a

temperature of 40°C (held for 1 minute) then increas-

ing at a rate of 1,3° per minute. This program was run

for approximately 30 minutes per run. The output of

the GC was integrated by a Hewlett Packard

HP3390A integrator, using the following par-

ameters: attenuation of 0 and an area of rejection of

10 000. The paper speed was set at 0,5 cm per mi-

nute. The column used in the GC was a 25 m-long

methyl silicone filament column, and the GC was a

Carla Erba Strumentazione Gas Chromatograph.

After the syconia had been used in the bottles,

they were removed, opened longitudinally and the

number of wasps inside noted for each size class.

Fifty syconia of Ficus ingens were examined to see

if a relationship between the presence of the exudate

and the presence of wasps existed. This was done by

collecting syconia with and without this exudate

(from one tree) and noting the presence or absence

of wasps inside the syconia. The results obtained

were statistically analysed using a two by two Chi-

squared contingency test with Yates’s correction

(Parker, 1973).

RESULTS

The results of the GC experiment show that the

developing syconia of Ficus ingens do emit a number

of gaseous compounds, and that this emission is re-

lated to the size of the syconia. The number of peaks

appearing in the different size classes also varies, but

a general pattern seems to be followed. The increase

in the number of peaks within these classes may be

due to isomerization of the compounds, decay of the

compounds, or genuine syconial products. These re-

sults are shown in Fig. la-e.

Fig. 2 is a histogram of the number of peaks in

each size class as well as the control (air from an

empty bottle plus solvent).

The results of the census of the wasps inside the

syconium are presented as histograms (Fig. 3a-c).

As may be expected, the 9 mm size class produced

the most wasps per syconium. The results of the 6

mm class have not been presented as no wasps were

found in any 6 mm syconia. On each of the bar

FIG. 1. — The integrated printouts from the GC runs of the con-

trol and the syconial size classes 6 mm to 9 mm. a, control; h.

6 mm syconia; c, 7 mm syconia; d. 8 mm syconia; e, 9 mm

syconia. Starting temperature: 40 degrees held for 1 minute,

increasing at a rate of 1,3 degrees per minute for 30 minutes.

Integrator parameters: Attenuation = 0, chart speed = 0.5

cm/s, peak width = 0.10, threshold = 4 and area of rejection

= 10 000. Scale: 1 unit = 1 minute elution time. Total run

time: 30 minutes.

charts, the mean (X) and sample size (n) has been

shown.

The results of the SEM studies showed that the

female Platyscapa soraria had thoracic and coxal

Bothalia 15, 3 & 4 (1985)

609

corbiculae and these were found to be full of pollen

in those wasps that were captured outside the syco-

nia, whereas those wasps that were removed from

the syconia just after they had hatched had empty

pockets.

The results of the presence-absence of wasps in

syconia with exudate are presented in Table 1. The

data proved to be significant at the 95% confidence

interval.

In all the collections of wasps made, no parasitic

wasps were found: only the true pollinator, Platy-

scapa soraria Wiebes was present.

DISCUSSION

The gas chromatography method requires scrupu-

lous cleanliness of equipment. This aspect was there-

fore rigidly controlled. Vitiation of results can arise

from the fact that a GC machine is not easily trans-

portable and therefore cannot be taken to the study

site. This restriction to a laboratory means that sam-

ples have to be cut from the plant and taken to the

laboratory. This immediately limits the life of the

samples, as they are now entirely dependent on nu-

tritional supplies in the twig. This in turn means that,

sooner or later, the plant material will start to decay,

and therefore there is a source of unwanted decay

products. For the purposes of this study it is assumed

that, as the samples were all analysed on the same

day as they were removed from the tree, this decay

had not yet started.

This removal may also cause the figs to respond by

emitting a different set of compounds, or to stop

emitting some and emit others, etc. As far as the

analysis by GC is concerned, the unnatural starting

temperature of the GC oven (40°C) may cause the

breakdown or modification of thermolabile com-

Fig. 2. — A histogram showing the number of peaks which were

detected by the gas chromatograph for the control (air from

an empty bottle) and each syconial size class.

pounds. For the adequate interpretation of these re-

sults, these problems will have to be kept in mind,

but assumed not to happen.

The GC results indicate that a number of volatile

compounds are emitted by the syconia, and that this

emission is related to the size (and therefore age) of

the syconia. The syconia of the 8 mm size class emit-

ted the most compounds, and the others, especially

those in the 9 mm class, emitted less (Fig. 2).

The number of peaks does not necessarily indicate

the maximum amount of activity, as the increased

number may be due to isomerization of some of the

products, or possibly, decay products. However, as

these extra peaks appear in the general region of the

main peaks, and not in totally new regions of the

chromatograph, it is possible that these peaks have

NUMBER OF WASPS PER SYCONIUM

Fig. 3. — Histograms showing the number of wasps per syconium

against frequency, expressed as a percentage of syconia sam-

pled. a, 7 mm size class; b. 8 mm size class and c, 9 mm size

class. The 6 mm size class had no wasps present in any syco-

nia sampled. The mean (x) and sample size (n) are shown.

610

Bothalia 15, 3 & 4 (1985)

been separated out l?y the column due to increased

production, so indicating peak activity.

The strong association between wasp activity and

the presence of the volatile compounds is reflected

in the visitation data. The 6 mm size class had not yet

been visited by wasps, but was starting to produce

some of the volatile compounds. This nonvisitation

may be because the wasps could not detect the com-

pounds in sufficient quantities, or else because the

ostiolar scales had not yet loosened sufficiently to

allow the entry of the wasps.

These scales are thought by Janzen (1979a) to

have assumed the function of pollinator discrimina-

tion, and Ramirez (1974) has shown a strong re-

lationship between the shape of the wasp’s head and

the ostiolar structure. The mean number of wasps

per fig in the 7 mm and 8 mm size classes (as shown

in Fig. 3a & 3b) indicates a tendency for the wasps to

search for unentered syconia. Janzen (1979c) sug-

gests that this may be a mechanism to reduce compe-

tition between the offspring of different mother

wasps. How the wasp discerns whether a syconium

has been entered or not is not known. It has, how-

ever, been reported (Janzen 1979a, 1979c) that after

the wasp enters the syconium, a fluid is secreted by

the fig which fills the syconium. This fluid is thought

to have antibiotic and antifungal properties, as de-

composition of the wasps’s bodies is seldom found.

It was observed that syconia of Ficus ingens exude

a fluid, but that the syconial cavity is not filled. This

fluid is exuded from the osteole of the syconium,

which dries to form an effective seal against further

wasp entry. The stimulus for the production of this

fluid is not clear, but may be as a result of injury to

the syconium caused by the wasp. The time span be-

tween wasp entry and fluid production was not

measured.

The results of the census of wasps in syconia with

and without exudate showed that the presence of ex-

udate was a good indication of wasp entry, at the

95% confidence limit as shown in Table 1. The time

span in which the wasps are able to enter the syco-

nium starts at the time the osteolar scales loosen,

and ends either when the production of volatiles

ceases, or when the syconia exude a fluid which

blocks up the osteole. In this time-span any number

of wasps may enter the syconium.

TABLE 1. — Presence or absence of wasps for syconia with and

without exudate at the ostiole. This table is presented in

the form of a 2 X 2 contingency table. A Chi-square test

performed on this table showed the Chi-square value to be

4,847 (with Yates’s correction). As this is greater than the

expected value of 3,84 (at the 95% confidence level), the

null hypothesis (that there is no relationship between

visitation and the presence of the exudate) must therefore

be rejected

The results of the wasp census (from the syconia

used in the GC experiment) indicate that wasps pre-

fer to enter previously unentered syconia. This is in-

dicated by the value of the mean number of wasps

per syconium in the 7 and 8 mm size classes (Fig. 3a

& 3b). This observation duplicates that of Janzen

(1979c), who observed similar behaviour in other fig

species. The mean number of wasps per syconium in

the 9 mm size class would probably be a reasonable

indication of the final number of wasps that enter a

Ficus ingens syconium (3-4 per syconium, as shown

in Fig. 3c). It was also noted that syconia in this size

class had much larger ovaries, as the wasps had

already laid their eggs, and the ovaries had then be-

gun to develop into galls, as all wasp-infested female

fig flowers do (Janzen 1979b). This resulted in the

syconium becoming much more ‘woody’ or tough, as

the galls developed. Ramirez (1970) and Janzen

(1979c) observed that if a syconium was not entered

by a wasp, it aborted and fell from the tree. No data

on this aspect were collected for Ficus ingens.

CONCLUSION

The co-incidence of the production of numerous

volatile substances and wasp visitation to the syconia

of the size and age at which the volatiles were pro-

duced indicate that Ficus ingens attracts its specific

wasp pollinator by using odours. As most of the fig

species have only one legitimate wasp pollinator

species, it is to be expected that different fig species

produce different compounds, especially if they exist

in sympatry with one or more other fig species. It

should therefore be possible to ‘fingerprint’ each Fi-

cus species by its characteristic signature of volatiles,

when analysed by using gas chromatography.

Further research in this field may include the iso-

lation and identification of these compounds, and

the behavioural and physiological analyses of the

wasps’s reactions to the components of the volatile

fraction, to determine which compounds release a

behavioural reaction. It may be found that the wasps

need more than one of these compounds before the

searching behaviour is elicited.

ACKNOWLEDGEMENTS

I thank Dr H. Baijnath and Mr S. Naicker of the

University of Durban-Westville for the initial ideas

on which this research was based. Prof. R. Crewe

and Mr M. Centner for the use of the gas chromato-

graph, Mr E. R. Robinson and Dr H. Glen for help-

ful comments on the manuscript. This research was

supported in part by a CSIR postgraduate bursary.

UITTREKSEL

Die verwantskap tussen die vy Ficus ingens (Miq.)

Miq. en sy bestuiwer, die perdeby Platyscapa soraria

Wiebes, is bestudeer. ’n Meganisme van aantrekking

van die bestuiwer word bespreek en bevestiging ver-

kry deur die gebruik van gas chromatografie om die

aanwesigheid van vlugtige verbindinge aan te toon.

Dit word bewys dat hierdie verbindinge gedurende

die vroulike stadium van die vy se blomsiklus vryge-

stel word. Flierdie data word met die gedrag van die

bestuiwer en data betreffende besoeke, verkry van

waarnemings in die veld, gekorreleer.

Bothalia 15, 3 & 4 (1985)

611

REFERENCES

GALIL, J., 1977. Fig biology. Endeavour 1: 52-56.

GALIL, J. & EIS1KOWITCH, D., 1968a. Flowering cycles and

fruit types of Ficus sycomorus in Israel. New Phytol. 67:

745-758.

GALIL, J. & EISIKOWITCH, D., 1968b. On the pollination

ecology of Ficus sycomorus in East Africa. Ecology 49:

259-269.

GALIL, J. & EISIKOWITCH, D., 1974. Further studies on polli-

nation ecology in Ficus sycomorus. II. Pocket filling and

emptying by Ceratosolen arabicus Mayr. New Phytol. 73:

515-528.

GROB, K., 1975. The glass capillary column in gas chromatogra-

phy. A tool and a technique. Chromatographia 8: 423-433.

HILL, D. S., 1967, Figs ( Ficus spp.) and fig-wasps (Chalcidoi-

dea). Jnl. Nat. Hist. 1: 413-434.

JANZEN, D. H., 1979a. How to be a fig. A. Rev. Ecol. Syst. 10:

13-51.

JANZEN, D. H., 1979b. How many babies do figs pay for

babies? Biotropica 11: 127-129.

JANZEN, D. H., 1979c. How many parents do the wasps from a

fig have? Biotropica 11: 127-129.

PARKER, R. E., 1973. Introductory statistics for biology. Lon-

don: Arnold.

RAMIREZ, W., 1970. Host specificity of fig wasps (Agaonidae).

Evolution 24: 680-691.

RAMIREZ, W., 1974. Coevolution of Ficus and Agaonidae.

Ann. Mo. bot. Gdn 61: 770-780.

WIEBES, J. T., 1979. Coevolution of figs and their insect pollina-

tors. A. Rev. Ecol. Syst. 10: 1-12.

'

■

'

Bothalia 15. 3 & 4: 613-629 (1985)

Analysis of the size and composition of the southern African flora

G. E. GIBBS RUSSELL*

Keywords: endemic species, flora, phytogeography, plant families, species diversity

ABSTRACT

The southern African flora has been surveyed for the first time at species level in the List of Species of Southern

African Plants (Gibbs Russell et at.. 1984). The numbers of taxa recorded for southern Africa are compared to the

numbers reported from other parts of Africa, and the largest families in each area are listed and compared. The

species richness of southern Africa is compared to that of other parts of the world. The numbers of genera, species

and infraspecific taxa are given for each family in the southern African flora, and compared to previous counts by

Dyer (1975. 1976) and Goldblatt (1978).

INTRODUCTION

The remarkable diversity and high level of ende-

mism in the southern African flora has been re-

ported by a number of authors, notably Adamson

(1938), Weimarck (1941), Levyns (1964) and Good

(1974), and has been discussed in great detail by

Goldblatt (1978). Until recently there was no mod-

ern inventory of the taxa on the subcontinent, but

the publication of the List of Species of Southern

African Plants (Gibbs Russell et a!., 1984) now pro-

vides for the first time a complete coverage of the

entire flora. The components of the southern Afri-

can flora can now be precisely analysed and com-

parisons can be made with floras of other parts of

Africa and with previous estimates of the southern

African flora. The relative importance of present

studies can be assessed, and future work on the

Flora of Southern Africa can be planned with a more

accurate idea of the magnitude of the task.

METHODS

The numbers of taxa reported here for southern

Africa are taken from the PRECIS list of 30 June

1984. and some modifications have been made since

the first edition of the List of Species went to press in

October 1983. The taxa in the southern African flora

were counted by computer, and have been verified

by a manual count of the List of Species. Counts for

other Floras in Africa were made by hand, as de-

scribed in Gibbs Russell (1974).

Because the Floras considered for this study differ

in the delimitation of families and in the level of re-

cognition of species and infraspecific taxa, it was

necessary to adopt a uniform treatment in order to

compare them. Families are treated sensu lato , and

the genus and species counts for the segregate fami-

lies are added to give a single count in these cases.

Notable examples of families treated in this way are

Aizoaceae (includes Mesembrvanthemaceae), As-

clepiadaceae (includes Periplocaceae), Campanula-

ceae (includes Lobeliaceae), Fabaceae (includes

Caesalpinioideae, Mimosoideae and Papilionoi-

deae), Liliaceae (includes Alliaceae, Asparagaceae,

* Botanical Research Institute. Department of Agriculture &

Water Supply. Private Bag X101. Pretoria 0001.

Asphodelaceae, Colchicaceae, Dracaenaceae,

Eriospermaceae and Hyacinthaceae) and Scrophula-

riaceae (includes Selaginaceae). The alternative to a

broad acceptance of these families would have been

to split them apart in Floras which treat them as

units. This was not done because an object of the

study is to convey an overall picture of the southern

African flora in relation to the floras of other parts

of Africa, and it was therefore more consistent to

accept these families in the broadest sense.

It was also necessary to adopt as far as possible a

uniform treatment of the lower taxa, especially those

of infraspecific rank. Because one author’s species

may be another author's subspecies or variety, the

total species number for different accounts of the

same group can differ considerably, and are not

readily comparable. For this reason, combined totals

of species plus infraspecific taxa were used on the

tables below for purposes of comparison. For ex-

ample, Crassula was revised for southern Africa by

Tolken (1977), and this revision is followed in the

List of Species. There are now 237 taxa, including

142 species with 47 subspecies, 39 varieties and 9 re-

cognized hybrids. Examination of his treatment

shows that the great majority of taxa now accepted

by him at the infraspecific level were originally re-

cognized as species. Of the 29 species given for the

Cape Peninsula by Adamson & Salter (1950), 9 are

accepted at the infraspecific level by Tolken (1977).

If these treatments were compared at the level of

species, 30% of the Crassula species in Adamson &

Salter (1950) would not be counted in the List of

Species that follows Tolken, thus giving the Adam-

son & Salter ( 1950) count a falsely high comparative

value.

Just as the different Floras considered here cannot

be directly compared because of differences in treat-

ment, so the different parts of the List of Species it-

self vary widely in their ranking of taxa because the

list is the result of taxonomic judgement by numer-

ous individuals made over at least 80 years. A re-

cently revised group such as Crassula may contain

fewer species and more infraspecific taxa, while a

group greatly in need of revision, such as the entire

family Mesembryanthemaceae, presently has a great

many species that will probably be reduced when

they are critically studied. The numbers of species

614

Bothalia 15, 3 & 4 (1985)

and infraspecific taxa is therefore used as a conserva-

tive ‘lowest common denominator’ making it pos-

sible to treat at more or less the same level the floras

of different areas studied at different times by differ-

ent individuals.

The problem of outright synonymy has been im-

possible to solve when working at a continental scale

over a time span of several decades. The ideal would

be to work through each of the Floras considered

with the aid of the most recent taxonomic revisions

and thus ensure that each taxon is recognized in the

same way. However, this would be equivalent to re-

vising the entire African flora before making the

comparisons presented here. Because this is totally

impracticable at present, each of the Floras covered

is taken as it stands and the taxa accepted by the

author of each treatment are accepted in this study,

even though some of the taxa counted have been, or

should have been, placed in synonymy. The use of

species and infraspecific taxa when making the com-

parisons eases some of the discrepancies introduced

by different levels of treatment at different times,

but cannot eliminate them.

RESULTS AND DISCUSSION

Total numbers of taxa in the southern African flora

The numbers of taxa present in southern Africa

are shown in Tables 1 & 2. Although it may appear

to be a simple matter to determine total numbers of

taxa once a basic list is prepared, the total numbers

can in fact be calculated in a number of ways de-

pending on the emphasis that is required. Infraspe-

cific, naturalized and well-known but as yet unpub-

lished taxa may be included or excluded. The two

most extreme of these different totals are used below

to make comparisons with the floras of other areas.

The total number of species and infraspecific taxa

(including naturalized taxa, unpublished taxa, sub-

species, varieties and recognized hybrids) is a ‘high’

total (Table 1), and the number of indigenous

species is a ‘low’ total (Table 2).

The total number of species and infraspecific taxa

is used to compare the Floras of different areas, and

different parts of the List of Species , for the reasons

of inconsistencies in level of taxonomic treatment

given above. The ‘high’ total is also used to make

estimates of the outstanding work for the Flora of

Southern Africa because all infraspecific and natura-

lized taxa must be dealt with. In practical terms both

of these can be time-consuming to cover for the

Flora.

The total number of indigenous species must be

used for biogeographical studies because the natura-

lized aliens, which have only recently become part of

the African vegetation, are eliminated. This ‘low’

count is used to calculate species/area ratios to com-

pare the diversity of the southern African flora to

that of other parts of the world. It is also used to

TABLE 1. — Numbers of taxa present in southern Africa

Bothalia 15, 3 & 4 (1985)

615

compare the current List of Species to the earlier es-

timates of Dyer (1975, 1976) and Goldblatt (1978),

who published only species counts.

Previous estimates and counts for the number of

indigenous species in Africa have been considerably

smaller than the totals presented in the List of

Species. For seed plants, Killick (1971) estimated

17 500 species. Dyer (1975, 1976) estimated about

18 400 species, and Goldblatt (1978) reported 18 532

species, whereas the count obtained from the List of

Species (1984) is 19 980. A hand count of the in-

digenous species of vascular plants made by Wells et

al. (1983) yielded 20 044 species, which is within 100

species of the count of 20 139 determined from the

List of Species. This is independent confirmation of

the accuracy of the computerization process. The

count was obtained by different individuals using a

different method for a different purpose, but work-

ing on the same herbarium collection at the same

time.

The largest families in the southern African flora

The 38 families that comprise more than 100

species and constitute 0,5% or more of the total

flora are shown in Table 3. All these families are

flowering plants, and account for over 82% of the

entire flora, and 87% of the flowering plants.

Twelve families have around 500 or more species

and infraspecific taxa, with a break of nearly 100

taxa between the rest, which have fewer than 400

taxa. The 12 largest families account for over 58% of

the total flora and over 60% of the flowering plants.

As currently treated, the largest family is Mesem-

bryanthemaceae, but it is believed that critical re-

vision will bring the number of accepted taxa down

to about 1 200 (Gibbs Russell & Glen, 1984). The

family would then rank third, below Asteraceae and

Fabaceae.

These 38 families are only 12% of the total num-

ber of families, and the 8 families that contain about

half the total flora are less than 3% of the total num-

ber of families. When the families are arranged

according to the number of species they contain, it is

found that over half of all families have fewer than 8

species and infraspecific taxa.

Good (1974) lists the 30 largest families in the

world. Ten of the 12 largest families in southern

TABLE 3. — The 38 families of flowering plants with more than 100 species ranked by numbers of species and subspecific taxa

a Calculated from Goldblatt (1978)

b Good (1974)

*Naturalized taxa

616

Bothalia 15, 3 & 4 (1985)

Africa are included in Good’s list, but Mesembryan-

themaceae and Iridaceae are not among the world’s

30 largest families. The 38 most important southern

African families include 18 of the largest families

listed by Good.

The species/genus ratios of the families can reflect

their phytogeographical affinities. The overall

species/genus ratio for southern African seed plants

is about 9,6. The families with a species/genus ratio

more than twice the overall ratio are those known to

have diversified extensively within southern Africa

especially in Capensis: Mesembryanthemaceae, Li-

liaceae, Iridaceae, Ericaceae, Proteaceae, Crassula-

ceae, Restionaceae, Geraniaceae, Companulaceae,

Oxalidaceae, Selaginaceae, Rhamnaceae, Thyme-

laeaceae, Sterculiaceae, Lobeliaceae and Santala-

ceae. Families with a species/genus ratio about half

the overall ratio of 9,6 are all families of worldwide

distribution or with centres of diversity in the tropics

or north temperate areas: Poaceae (worldwide), Ru-

biaceae (pantropical), and Apiaceae (north tem-

perate) (Good, 1974).

All the families with a high species/genus ratio

also have more than 88% of their species endemic to

southern Africa. However, not all families with a

high percentage of endemic species have a high

species/genus ratio. In these exceptions, either more

than half the genera in the family are endemic but

each have a moderate number of species (Rutaceae,

Apiaceae), or less than a fourth of the genera in the

family are endemic but a few genera have a large

number of endemic species (Brassicaceae — Helio-

phila , Rosaceae — Cliffortia).

In Table 3, the families are ranked by the number

of species and subspecific taxa. If the number of in-

digenous species is used for ranking instead, there is

little change in the placing of the 38 largest families.

Cyperaceae and Euphorbiaceae exchange places at

ranks 10 and 11, but there is only a difference of

three species between them. Crassulaceae falls from

15th to 20th as a result of the large number of sub-

specific taxa now accepted, and Apiaceae falls from

24th to 30th as a result of the high number of natura-

lized species. Polygalaceae and Thymelaeaceae each

gain three places in comparative ranking, mainly be-

cause they have few naturalized species or infraspe-

sific taxa, and therefore there is little difference be-

tween the total number of taxa and the total number

of indigenous species.

The 21 families of flowering plants with more than

20 genera are listed in Table 4. Only one family,

Amaranthaceae, does not also have more than 100

species. The ranking of the families by number of

genera corresponds only roughly to the ranking by

number of species and infraspecific taxa, but the 12

families with the most genera include ten of the larg-

est families ranked by species and infraspecific taxa.

Ericaceae and Cyperaceae have comparatively fewer

genera, and Rubiaceae and Apiaceae have compara-

tively more.

The families in Table 4 can be divided into three

groups on the basis of the percentage of their genera

that occur in southern Africa related to their world-

wide distribution, (a) Three families have from

nearly half to virtually all their genera represented in

southern Africa: Mesembryanthemaceae, the most

strongly ‘southern African’ of all our families; Irida-

ceae, concentrated in the southern hemisphere; and

Ericaceae, with the subfamily Ericoideae present in

Europe but concentrated in southern Africa (Good,

1974). (b) All the families with a worldwide distribu-

tion described by Dyer (1975, 1976) as simply ‘cos-

mopolitan’ have between 20 and 40% of their genera

represented in southern Africa, (c) The families with

fewer than 20% of their genera occurring in south-

ern Africa are distributed predominantly in climatic

TABLE 4. — The 21 families of flowering plants with more than 20 genera ranked by numbers of genera

3 Dyer (1975, 1976)

* Naturalized genera

Bothalia 15, 3 & 4 (1985)

617

TABLE 5. — The 35 genera of flowering plants with more than 100 species and infraspecific taxa

areas not well represented in southern Africa: Or-

chidaceae (absent from dry areas), Brassicaceae

(north temperate), Euphorbiaceae (tropical), Apia-

ceae (temperate), Acanthaceae (tropical and subtro-

pical). Amaranthaceae and Rutaceae are excep-

tions. Brassicaceae is noteworthy because over half

its genera in southern Africa are naturalized.

Two bryophyte families also have more than 20

genera, Pottiaceae (29) and Lejeuneaceae (28).

The largest genera in the southern African flora

Thirty-five genera in the southern African flora

have more than 100 species and infraspecific taxa, as

shown in Table 5. All of these genera occur in one of

the 38 largest families, and 22 occur in one of the 12

families with over 500 species and subspecific taxa.

Among the 12 largest families only two, Poaceae and

Cyperaceae, do not have a genus with more than 100

species. According to Goldblatt (1978), only five of

the large genera are endemic, three in Mesembryan-

themaceae (Ruschia, Conophytum and Lampran-

thus), one in Fabaceae (Aspalathus) and one in Ru-

taceae (Agathosma).

Not surprisingly, the 12 largest families have the

most large genera. Mesembryanthemaceae has 7

genera with over 100 taxa, Asteraceae and Fabaceae

have 3 each, and Liliaceae and Asclepiadaceae 2

each. The other families each have only a single

genus with more than 100 taxa.

Size and composition of the southern African flora

compared to the floras of other parts of Africa

Since 1950, several complete accounts of the flora

of various parts of Africa have been published, cov-

ering the areas shown in Fig. 1. The numbers of

taxa, number of vegetation types, centres of ende-

mism, and sizes of each of these areas can be com-

pared to southern Africa, in order to place our flora

into perspective with that of other parts of the conti-

nent, and with smaller areas inside southern Africa.

The treatments considered for tropical Africa were:

Flora of West Tropical Africa , 2nd edn (Keay, 1954,

1958; Hepper, 1963, 1968, 1972); The flowering

plants of the Anglo-Egyptian Sudan (Andrews 1950,

1952, 1956) and Enumeratio Plantarum Aethiopicae

Spermatophyta (Cufodontis, 1953-1970, as listed by

Meyer, 1973). For areas within southern Africa, the

following treatments were considered: Prodromus

einer Flora von Siidwestafrika (Merxmtiller,

1966-1970); Flora of Natal (Ross, 1972); Flora of

Lesotho (Jacot Guillarmod, 1971); Flora of Swazi-

land (Compton, 1966, 1976); Plants of the Cape

Flora (Goldblatt & Bond, 1984) and Flora of the

Cape Peninsula (Adamson & Salter, 1950).

618

Bothalia 15, 3 & 4 (1985)

FIG. 1. — Parts of Africa covered by floras considered in this

study. 1, Flora of West Tropical Africa; 2, the Flowering

Plants of the Sudan; 3, Enumeratio Plantarum Aethiopicae

Spermatophyta; 4, List of Species of Southern African

Plants; 5, the Flora of Swaziland; 6, Flora of Natal; 7, Flora

of Lesotho; 8, Plants of the Cape Flora; 9, Flora of the Cape

Peninsula; 10, Prodromus einer Flora von Siidwestafrika.

Tables 6, 7 & 8 compare numbers of families, gen-

era and species plus infraspecific taxa for each of the

ten floras, and Table 9 shows each of the floras

ranked by numbers of taxa, area covered, number of

vegetation types and number of centres of ende-

mism. Southern Africa has the largest number of fa-

milies, genera, and species plus infraspecific taxa.

Although it does not cover the largest area, it has the

largest number of vegetation types and centres of

endemism as mapped by White (1983). Although the

west tropical African flora covers nearly twice the

area of southern Africa, it has less than a third of the

number of species and infraspecific taxa as southern

Africa. The richness of the southern African flora is

further emphasized when areas within it are com-

pared to areas outside. The included Cape flora

ranks second in number of species, with at least

1 500 taxa more than the third-ranking west tropical

African flora, even though the area covered by the

west tropical African flora is more than 53 times the

area of the Cape flora. The included South West

African/Namibian flora is similar to the Sudan flora

in number of species and infraspecific taxa, even

though Sudan has more than twice the number of

vegetation types and nearly twice the area of South

West Africa/Namibia.

The far larger number of taxa recorded for south-

ern Africa may be ascribed to four factors:

1. southern Africa has the largest number of vege-

tation types of all the floras considered. None of the

vegetation types are shared between southern Africa

and the tropical floras north of the equator except

the habitat-limited and comparatively small aquatic,

halophytic, mangrove and afromontane vegetation

types. The southern African vegetation includes the

Cape flora, with over 8 500 species and the karroid

vegetation types, also rich in species,

2. southern Africa has four centres of endemism

(White, 1983) and a very high percentage of endemic

taxa (Goldblatt, 1978). Two of the centres of ende-

TABLE 6. — Comparison of numbers of families in African floras

Bothalia 15, 3 & 4 (1985)

619

TABLE 8. — Comparison of numbers of species and infra specific taxa in African floras

a White (1983).

b Species only. Bond & Goldblatt (1984) did not include infraspecific taxa.

mism, the Cape and the Karoo-Namib, occur only

within southern Africa (except for a small extension

of the Karoo-Namib into southern Angola), and al-

though the Zambezian centre lies mostly outside

southern Africa, it is not covered by any of the Flo-

ras considered except the List of Species of Southern

African Plants. Only the afromontane centre of

endemism is shared with other Floras considered.

Goldblatt (1978) estimates that 80% of the southern

African species are endemic, and the levels of ende-

mism in many of the largest families are considerably

higher than 80%, as shown in Table 2. Brenan

(1978) calculated an area/endemic index for various

parts of tropical Africa. This index estimated for the

whole of southern Africa is 161 (i.e. 20 000 x 0,8 =

16 000 estimated endemic seed plant species;

2 573 000 km2 -e 16 000 endemic species = 161

area/endemic). This indicates higher levels of ende-

mism for southern Africa than for anywhere in tropi-

cal Africa, where Brenan reports strongest ende-

mism in Gabon (239) and Cabinda (251). In com-

parison, he reports that Sudan, with a similar area to

southern Africa, has an area/endemic index of

50 000. At the other extreme, the area/endemic in-

dex calculated for the Cape flora using the figures of

Bond & Goldblatt (1984) is 15,

3. the counts for the southern African flora as a

whole and for the Cape flora are of more recent date

than any of the others. Many more taxa are probably

known for each of the areas now than appear in the

Floras considered, most of which are over ten years

old. For example, the numbers of seed plant taxa

reported by Jacot Guillarmod (1971) for Lesotho

were 526 genera and 1 591 species and infraspecific

taxa. A listing from PRECIS of all Lesotho speci-

mens made in 1984 showed 702 genera and 2 726

species and infraspecific taxa, an increase of 25%

and 42% respectively in 13 years. Similarly, White

(1983) estimates 8 000 species for the Guineo-Con-

620

Bothalia 15, 3 & 4 (1985)

golan centre of endemism, which forms only a part

of the area of the west tropical African flora,

4. the southern African flora is being studied pri-

marily by botanists working in the region, whereas

the floras of west tropical Africa, Sudan and Ethio-

pia have been studied primarily by botanists working

in Europe. This undoubtedly increases the propor-

tion of the existing flora that has been collected.

Even within southern Africa, mapping the numbers

of specimens per quarter degree square has shown

that collecting intensity is greatest in easily acces-

sible locations (Gibbs Russell, Retief & Smook,

1984).

The individual floras included in the entire south-

ern African area each has only a fraction of the total

number of taxa in the region, and only a small pro-

portion of the 28 vegetation types present in the

whole area. Nevertheless, only the three karroid ve-

getation types are not covered by at least one of the

included Floras considered.

The ten largest families in each of the Floras are

compared in Table 10. All the ten largest families in

the entire southern African flora are also among the

largest families in at least one other Flora consid-

ered. Each of the 24 families concerned is represen-

ted in southern Africa, and only Apocynaceae and

Annonaceae do not have over 100 species and infra-

specific taxa in southern Africa. Among the floras

considered, these two families are a major compo-

nent only of the west tropical African flora.

Only two families, Asteraceae and Fabaceae are

among the ten largest in all the floras. Asteraceae

ranks first or second in all the included southern

African floras, and if Mesembryanthemaceae were

critically revised (Gibbs Russell & Glen, 1984) As-

teraceae would be the largest family in the entire

southern African flora as well. In the tropical floras

Asteraceae is not as outstanding, ranking below Fa-

baceae and Poaceae. In contrast, Fabaceae is of first

or second rank in the tropical floras, but ranks only

from second to fourth in the southern African floras.

Poaceae is among the ten largest families in all the

floras considered except the Cape flora, and it has a

much lower ranking in southern Africa as a whole

than in the tropical floras or in the included floras of

southern Africa. Nevertheless, in absolute numbers

there are more taxa recorded for Poaceae in south-

ern Africa than for the tropical floras even though

the ranking in southern Africa is lower. This com-

paratively lower rank for southern Africa as a whole

may result from two factors. Firstly, the many wide-

spread grass species are counted separately for the

smaller included floras, but only once for the entire

southern African flora. Secondly, the Cape flora has

a very low number of grasses in comparison to Ai-

zoaceae (s.l.), Liliaceae (s.l.), Iridaceae and Erica-

ceae. The large numbers of taxa in these families in

the Cape flora give the Poaceae a lower ranking in

the southern African flora as a whole. Cyperaceae is

the only family that is a major component of all flo-

ras considered except the entire southern African

flora and the Cape flora. Its absence from the ten

largest families in these two floras is probably the

result of the same factors that give the Poaceae a

comparatively low ranking in the same floras.

Three families show differences between the ma-

jor components of the tropical and southern African

floras. Acanthaceae and Rubiaceae are among the

ten largest families both in the tropical floras north

of the equator and in the included floras of tropical

affinity within southern Africa, namely South West

Africa/Namibia, Natal and Swaziland. Asclepiada-

ceae is the only family that is a major component of

all (extra-Capensis) southern African floras but not

of the tropical floras north of the equator.

Lamiaceae shows a difference between eastern

Africa and the rest of the continent. In Africa, this

family has its greatest importance in eastern Africa,

from Sudan and Ethiopia south to Swaziland and

Natal (where it ranks 11th).

The families that are peculiarly southern African,

recognized in Table 3 by their high species/genus ra-

tios and percentage of endemism, are also clearly

shown in Table 10. Aizoaceae, Ericaceae, Iridaceae

and Restionaceae are among the ten largest families

in the Cape flora, and through it, of southern Africa

as a whole. Proteaceae and Rutaceae are among the

ten largest families only in the Cape flora. Amarylli-

daceae and Campanulaceae are major components

only of the high-altitude Lesotho flora, although

Campanulaceae ranks 11th in the Cape flora. Again,

one family exhibits an opposite pattern: Euphorbia-

ceae is one of the ten largest families in all floras ex-

cept the Cape and Lesotho.

Convolvulaceae and Malvaceae are major compo-

nents of the flora only in Sudan, although they have

more than 100 species in southern Africa.

Three families show patterns of distribution that

do not coincide with any others. Liliaceae (s.l.) is

one of the ten largest families in southern and east-

ern Africa, but not of west tropical Africa and Su-

dan. Orchidaceae, although the second largest fam-

ily in the world, is not among the major components

of the floras of South West Africa/Namibia, Sudan

or Ethiopia, perhaps because of the considerable

arid areas in these countries. In Capensis, Orchida-

ceae ranks 12th in the Cape flora as a whole and 7th

in the Cape Peninsula. Scrophulariaceae is the most

unusual in its areas of importance. It is among the

ten largest families in the Cape flora (temperate and

winter rainfall). South West Africa/Namibia and Su-

dan (both tropical and arid). Natal (of tropical affin-

ity and mesic to arid), and Lesotho (high altitude).

Species richness of the southern A frican flora com-

pared to floras of other parts of the world

The richness of the southern African flora com-

pared to floras of other large regions of the world,

both tropical and temperate, is shown in Table 11.

The species/area ratio for the whole of tropical

Africa is similar to that of its included parts, Sudan

and west tropical Africa. The ratio for southern

Africa is about five times as great, illustrating the

comparative poverty of the tropical African flora

discussed in detail by Brenan (1978). Two other

tropical areas, Brazil in tropical South America and

tropical Asia, are widely separated geographically

Bothalia 15, 3 & 4 (1985)

621

TABLE 10. — Comparison of numbers of species and infraspecific taxa in the 10 largest families in all floras considered. The top number is

the rank of the family in the flora and the bottom number is the number of species and infraspecific taxa. A dash in the top position shows

that the family is not among the ten largest in that flora

Tropical floras N of equator Included sthn African floras

Entire

10 largest families

aSpecies only. Bond & Goldblatt (1984) did not include subspecific taxa.

622

Bothalia 15, 3 & 4 (1985)

TABLE 11. — Species/area ratios for several regions of the world with areas of over

± 2 500 000 km:

"Brenan (1978).

bGood (1974).

cRide (1978) quoted in Goldblatt (1978).

dGoldblatt (1978).

Calculated from Gibbs Russell et at. (1984) and Bond & Goldblatt (1984).

but have similar species/area ratios. Although tropi-

cal areas are known to have high concentrations of

species, the species/area ratio for southern Africa is

about 1,7 times greater than either. The Australian

flora is sometimes compared to that of southern

Africa because both are southern hemisphere areas

with tropical and temperate vegetation elements and

both have high levels of endemism. However, the

species/area ratio for southern Africa is nearly 2,5

times that of Australia. The ratio for eastern North

America, in the north-temperate zone, is also far be-

low that of southern Africa. Even if the extremely

diverse Cape flora is eliminated from the determina-

tion of the species/area ratio for southern Africa, the

ratio is nevertheless higher than that for tropical

South America or Asia.

Size and composition of the southern African flora

reported in the List of Species compared to previous

recent treatments

Table 12 shows the numbers of genera and species

recorded for each family in the List of Species

(1984), Dyer (1975, 1976) and Goldblatt (1978). Fa-

milies that are treated differently in the three

sources are shown both sensu stricto and sensu lato

so that comparisons can be made. For seed plants,

the count of Goldblatt and the estimate based on

Dyer agree within 50 species. They date from the

same period, but Dyer’s work was carried out pri-

marily at PRE, whereas Goldblatt’s count was deter-

mined from several herbaria, literature, and consul-

tation with experts in various groups. The closeness

of the final count confirms both as being reasonable

determinations for that time.

The List of Species count is about 215 indigenous

genera and 1 450 species higher than Goldblatt’s

count. This discrepancy is the result of different

numbers of species recorded in a number of families.

For 63 families (30% of the number of families of

seed plants) the List of Species has a higher number

of species than either of the other counts, and 27 fa-

milies exceed Dyer’s and Goldblatt’s counts by more

than 5 species. Over a third of the species are in Me-

sembryanthemaceae, which has been previously

mentioned as having in reality far fewer species than

are presently recognized. Three other families differ

by more than 100 species, Asclepiadaceae, Fabaceae

and Liliaceae. In each of these families, certain gen-

era are under revision but others are still in need of

treatment. Goldblatt’s count has only three families

for which the highest number of species is recorded,

Orchidaceae, Restionaceae and Rubiaceae. All

three of these families have had recent changes in

species numbers as a result of taxonomic revision.

Although Dyer’s estimate has the lowest total

number of species, curiously there are 15 families for

which his species counts are the highest. Two of

these families, Asteraceae and Crassulaceae, differ

from the List of Species by more than 100 species. A

number of genera in Asteraceae and the whole of

Crassulaceae have been revised since 1975. Simi-

larly, Poaceae and Sterculiaceae are recorded with

over 50 species more by Dyer, but work done for the

FTEA (Clayton & Renvoize, 1982) and a more accu-

rate determination of naturalized species has re-

duced species number in Poaceae and revisionary

work in Hermannia has reduced species numbers for

Sterculiaceae.

It appears therefore, that differences in counts of

species in a number of families are due to revisions

completed or in progress between the time of com-

pletion of Dyer, Goldblatt and the List of Species.

However, the substantially higher counts in the List

of Species for Apiaceae, Chenopodiaceae, Cypera-

ceae, Euphorbiaceae, Fabaceae, Liliaceae and Scro-

phulariaceae cannot be solely attributed to the result

of further study in these families. Furthermore, even

though revisionary work has resulted in a lower

count of species in families such as Poaceae and

Crassulaceae, and in such genera as Helichrysum,

where 283 taxa (260 species) were recognized at

PRE in 1981, but only 260 taxa (241 species) were

recognized after Hilliard’s treatment of the genus for

the Flora of Southern Africa, revision does not

necessarily result in the recognition of a smaller

number of species. The revision of Asparagus cur-

rently being completed for the Flora now recognizes

two genera and 77 species where previously one

genus with 44 species was recognized (A. A. Ober-

meyer, pers. comm.). C. H. Stirton (pers. comm.)

predicts that there may be ultimately a total of 2 000

Bothalia 15, 3 & 4 (1985)

623

TABLE 12. — Comparison of numbers of taxa per family. Families are listed alphabetically within each group (Bryophyta, Ptcri-

dophyta, Gymnospermae, Angiospermae). The difference between counts in the List of Species (1984), Dyer (1975, 1976) and

Goldblatt (1978) is indicated when there is a difference of more than 5 species. Naturalized families and genera are indicated

by an asterisk. This table runs from p. to p.

List of Species (1984)

Family Genera Species & Indigen.

infrasp. species

taxa

Dyer (1975, 76)a

Goldblatt (1978)

Genera Species Genera Species Difference

624

Bothalia 15, 3 & 4 (1985)

625

List of Species (1984) Dyer (1975, 76)a Goldblatt (1978)

626

Bothalia 15, 3 & 4 (1985)

List of Species (1984) Dyer (1975, 76)a Goldblatt (1978)

Bothalia 15, 3 & 4 (1985)

627

a Dyer (1975, 1976) omitted species counts for a number of genera. These were estimated from PRECIS lists dating from those years.

* Naturalized genera

# Undescribed species

628

Bothalia 15, 3 & 4 (1985)

species in Fabaceae, an increase of about 400 species

in this family.

CONCLUSIONS

The number of species of seed plants estimated for

the southern African flora has increased by roughly

2 300 in the past fourteen years. The question arises

whether the present total, based on the List of

Species of Southern African Plants (1984) will be just

as quickly outdated. There are three sources of

changes in the numbers of recognized taxa: lumping

and splitting of presently recognized taxa, which can

either increase or decrease the total; ‘finding’ taxa in

the Herbarium as a result of correctly identifying

existing specimens as undescribed taxa or as records

of taxa not previously known from southern Africa,

which will increase the total; and collecting new or

newly-recorded taxa in the field, which will also in-

crease the total.

The present trend in taxonomy is generally toward

reduction in the number of species recognized (Bre-

nan, 1978). Certain families, particularly Mesem-

bryanthemaceae, are likely to have their numbers of

taxa greatly reduced when they are critically revised.

There could be a reduction of over 1 000 taxa in Me-

sembryanthemaceae alone. However, the cases of

Asparagus and Fabaceae show that although there

may be a general trend toward reduction, and great

reductions in some families, not every group will

have its number of taxa reduced when it is revised.

The numbers of unpublished species and taxa

given in Tables 1 and 2 are only those which are so

well known that they have a ‘manuscript name’

awaiting publication, and they account for only

1,4% of the total number of known taxa. Obser-

vation of the PRE herbarium shows that there are a

great many specimens in ‘spp.’ folders which await

critical work by experts, and many will probably

prove to be new or newly-recorded taxa. Doubtless

the same situation exists in other herbaria with large

holdings of southern African plants. Even if only

one genus in ten contains a new species, there could

be a further 250 species now represented by Herbar-

ium specimens.

Overall, southern Africa is reasonably well cov-

ered by plant collections. PRECIS records show that

every whole degree square has some specimens re-

corded, and since the PRECIS records represent

only about 16% of the total of Herbarium specimens

in southern Africa, the true coverage is probably

better than the sample shown by PRECIS (Gibbs

Russell, Retief & Smook, 1984). Nevertheless, the

arid central and western parts of southern Africa re-

quire much greater collecting efforts in order for

their plants to be as well represented in herbaria as

those of the more mesic south and east. Because of

the lower collecting intensity in the dry areas, it is

likely that there are more new taxa awaiting discov-

ery there than in the better-known mesic areas.

For these reasons, changes in the numbers of taxa

known for the southern African flora are bound to

occur as a result of the basic taxonomic activities of

plant collecting, herbarium curation and revision. In

the future, these changes can be easily monitored

and recorded by updating the List of Species in PRE-

CIS, so that complete or partial lists can be printed

by computer, and total numbers determined. The

total now recognized is so large that even a change of

1 000 taxa represents only 4% of the total flora. The

comparisons made here thus may change in detail,

but the overall perspective is unlikely to alter.

The comparative picture outlined here emphasizes

the richness of the southern African flora in terms of

high species/area ratios, many vegetation types and

high levels of endemism, especially in some of the

largest families and genera. This richness is unequal-

led anywhere else in the world on a subcontinental

scale, and demands that a high priority be given to

the systematic study necessary to understand it.

ACKNOWLEDGEMENTS

Mrs W. Roux has patiently checked and proofread

all taxon counts from the List of Species. Dr O. A.

Leistner, Dr P. Goldblatt and Mr E. G. H. Oliver

have made valuable criticisms of the manuscript.

UITTREKSEL

Die flora van Suider-Afrika is vir die eerste keer tot

by spesies vlak in die List of Species of Southern

African Plants (Gibbs Russell et al., 1984) onder-

soek. Die aantal taksa wat vir Suider-Afrika opgete-

ken is word vergelyk met die van taksa wat vir ander

dele van Afrika aangegee is, en die grootste families

in elke streek word gelys en vergelyk. Die rykdom

van spesies in Suider-Afrika word met die van ander

dele van die wereld vergelyk. Die aantal genera, spe-

sies en subspesies vir elke familie in die Suider-Afri-

kaanse flora word gegee net met vorige tellings deur

Dyer (1975, 1976) en Goldblatt (1978) vergelyk.

REFERENCES

ADAMSON, R. S., 1938. The vegetation of South Africa. Lon-

don: Whitefriars Press.

ADAMSON, R. S. & SALTER, T. M., 1950. The flowering

plants of the Cape Peninsula. Cape Town: Juta.

ANDREWS, F. W., 1950. The flowering plants of the Anglo-

Egyptian Sudan, Vol. 1. Arbroath: Buncle.

ANDREWS, F. W., 1952. The flowering plants of the Anglo-

Egyptian Sudan, Vol. 2. Arbroath: Buncle.

ANDREWS, F. W., 1956. The flowering plants of the Anglo-

Egyptian Sudan, Vol. 3. Arbroath: Buncle.

BRENAN, J. P. M., 1978. Some aspects of the phytogeography

of tropical Africa. Ann. Mo. bot. Gdn 65: 437^478.

CLAYTON, W. D. & RENVOIZE, S.A., 1982. Poaceae. In R.

M. Polhill, Flora of Tropical East Africa. Rotterdam: Bal-

kema.

COMPTON, R. H., 1966. An annotated check list of the flora of

Swaziland. Jl S. Afr. Bot. Suppl. 6.

COMPTON, R. H., 1976. The flora of Swaziland. Jl S. Afr. Bot.

Suppl. 11.

DYER, R. A., 1975. The genera of southern African flowering

plants, Vol. I, Dicotyledons. Pretoria: Government Printer.

DYER, R. A., 1976. The genera of southern African flowering

plants, Vol. 2, Monocotyledons. Pretoria: Government

Printer.

GIBBS RUSSELL, G. E., 1975. Comparison of the size of vari-

ous African floras. Kirkia 10: 123-130.

GIBBS RUSSELL, G. E. & THE STAFF OF THE NATIONAL

HERBARIUM, 1984. List of species of southern African

plants. Mem. bot. Surv. S. Afr. 48.

GIBBS RUSSELL, G. E. & GLEN, H. F., 1984. Register of

names and types: a comparison between Mesembryanthe-

maceae and Poaceae. Bothalia 15: 125-129.

Bothalia 15, 3 & 4 (1985)

629

GIBBS RUSSELL. G. E.. RELIEF. E. & SMOOK. L.. 1984.

Intensity of plant collecting in southern Africa. Botlialia 15:

131-138.

GOLDBLATT. P.. 1978. Analysis of the flora of southern

Africa: its characteristics, relationships and origins. Ann.

Mo. hot. Gdn 65: 369-436.

GOLDBLATT. P. & BOND. P . 1984. Plants of the Cape Flora

— a descriptive catalogue. Jl .S'. Afr. Bor. Suppl. 13.

GOOD. R.. 1974. The geography of the flowering plants, edn 4

London: Longman.

HEPPER. F. N.. (ed.). 1963. Flora of West Tropical Africa, edn

2, Vol. 2. London: Crown Agents.

HEPPER. F. N.. (ed.). 1968. Flora of West Tropical Africa, edn

2, Vol. 3,1. London: Crown Agents.

HEPPER. F. N.. (ed.). 1972. Flora of West Tropica! Africa, edn

2, Vol. 3,2. London: Crown Agents.

JACOT GUILLARMOD. A.. 1971. Flora of Lesotho. Wcin-

heim: Cramer.

KEAY. R. W. T. (ed.). 1954. Flora of West Tropical Africa, edn

2, Vol. LI. London: Crown Agents.

KILLICK. D. J. B.. 1971. Progress with the Flora of Southern

Africa. Mitt. hot. StSanunl., Miinch. 10:76-79.

LEVYNS. M. R.. 1964. Migrations and origin of the Cape Flora.

Trans. R. Soc. S. Afr. 37: 85-107.

MERXMULLER. H.. (ed.). 1966-1970. Prodromus einer Flora

con Siidwestafrika. Wcinheim: Cramer.

MEYER. F. G.. 1973. Enumeration of Ethiopian spermato-

phytes. Taxon 22: 655-657.

ROSS. J. H.. 1973. Flora of Natal. Mem. hot. Sure. .S'. Afr. 39.

WEIMARCK. H.. 1941. Phytogcographical groups, centres and

intervals within the Cape flora. Acta Unix. land. , 37: 5-143.

WELLS. M. J . ENGELBRECHT. V. M.. BALSINHAS. A A

& STIRTON. C. H.. 1983. Weed flora of South Africa 2:

power shifts in the veld. Botlialia 14: 961-965.

WHITE. F.. 1983. Vegetation map of Africa. Paris: The Uncsco

Press.

.

■

.

Bothalia 15, 3 & 4: 631-654 (1985)

Additional biographical notes on plant collectors in southern Africa

L. E. CODD* and MARY GUNN*

Keywords: biographies, plant collectors, southern Africa

ABSTRACT

Biographical notes on plant collectors, supplementary to those already published in Botanical Exploration of

Southern Africa by Mary Gunn & L. E. Codd (1981), and including some collectors not previously recorded, are

provided.

INTRODUCTION

In our publication Botanical Exploration of South-

ern Africa (Gunn & Codd, 1981) many collectors’

names are listed with little or no biographical infor-

mation. Attempts have since been made to collect

some of the missing data and have, in several cases,

met with success. The information gathered is now

presented and the opportunity is taken to make

some corrections to the original text. In addition,

several new names, not previously recorded, are

added. It is also apparent that some names, taken

over from lists published by Tolken (1971) and in the

Index Herbariorum series on collectors, are scarcely

important enough to warrant inclusion.

We especially wish to thank those readers of Veld

& Flora who came forward with information in re-

sponse to a plea which we published in that journal.

No doubt many collectors have still been over-

looked, especially modern ones, in spite of appeals

to those whose names were omitted to submit the

necessary documentation. We again appeal to all

readers to inform us of any omissions or errors that

are known to them.

ALPHABETICAL LIST OF COLLECTORS

* An asterisk indicates that the name appeared in

Gunn & Codd (1981).

Abbott, Anthony Thomas Dixon (1936- )

b. Great Britain, 22 Sept. 1936; farmer; ed. Rugby

School 1950-54 and Northampton Institute of Agri-

culture 1955-56. Came to South Africa in 1956 and

farmed near Port Edward. Encouraged by Mr H. B.

Nicholson and Mr A. E. van Wyk, he is compiling a

comprehensive check list of the Umtamvuna flora,

with special reference to the high degree of ende-

mism found on Table Mountain Sandstone forma-

tions. Fig. 1.

Commemorated in Maytenus abbottii van Wyk.

Specimens c. 1 800; in PRU, NH and in a personal

herbarium.

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

* Adams, Berenice Zoe Margaret (later Mrs Mat-

thews) (1925- )

b. Springs, Transvaal, 7 June 1925; biologist and

housewife; ed. Natal Univ., Pietermaritzburg,

1943-45, graduating B.Sc. Research chemist.

Chamber of Mines Timber Research Laboratories,

Johannesburg, 1946-49, investigating the preserva-

tion of timber and fabrics against fungal attack and

fire underground, the corrosion of steel by under-

ground water, and the culture of fungi. After her

marriage in 1949 to Dr John Quarry Matthews, she

worked for a year in the Pasteur Institute, Salisbury

(now Harare), Zimbabwe, as a clinical pathologist

concerned mainly with haematology, and has con-

tinued to assist her husband with similar laboratory

work since he settled in Richmond, Natal.

Has collected several hundred specimens, many

with accompanying illustrations, mainly in Natal,

eastern Cape and eastern parts of Zimbabwe; in

NU.

FIG. 1. A. T. D. Abbott

Constance G. Adams

* Adams, Constance (‘Daisy’) Georgina (later Mrs

Tardrew) (1883-1968)

b. Cape Town, 6 Aug, 1883; d. Johannesburg, 21

June 1968; housewife and collector. She spent the

first six years of her life on her parents’ farm 'Water-

falls’ in the Tulbagh District. Both parents took a

keen interest in wild and cultivated flowers and were

among the prime movers in the establishment of the

first wild-flower show in Tulbagh. In 1890 the family

moved to Warrenton where Connie (later called

'Daisy’) had her early schooling. At the age of 14 she

632

Bothalia 15, 3 & 4 (1985)

went to Vredenburgh High School in Cape Town

and won several prizes for botany, her favourite sub-

ject. During school holidays at home she collected

specimens for the Albany Museum Herbarium, Dr

Schonland being a great friend of her father’s.

Before leaving Cape Town, she took a short

course in teaching. In 1907 the family moved to Kim-

berley and she taught for two years. She also became

friendly with Dr Wilman, Director of the McGregor

Museum, and collected plants for the herbarium. In

1910 she married Peter Tardrew and went to live

first in Bloemhof, moving to Johannesburg in 1936,

where she became a keen member of the House-

wives’ League of S. Africa. Fig. 1.

Specimens in GRA, KMG; also some leg. Mrs

Tardrew in PRE.

Ref: Informaton supplied by her youngest daugh-

ter, Mrs C. G. Smits of Pinelands, 1983; Codd &

Gunn in Veld & Flora 70 : 67 (1984).

* Admiraal, Johannes (1916-83)

d. Pretoria, 2 July 1983.

Ames, A. H. (fl. 1853)

Assistant naturalist with the United States North

Pacific Surveying Expedition, which spent seven

weeks in Simon’s Bay in 1853.

A few specimens leg. A. H. Ames are in PH but

the major collection was made by Charles Wright

(see Gunn & Codd, 1981).

Ref.: Mears in Proc. Acad. Nat. Sci. Philadelphia

133: 155 (1981) and pers. comm. Apr. 1983.

* Anderson, Mrs E.: see Ethel West

Apstein, Dr C. {fl. 1898).

Algologist. Originally from Kiel and accompanied

the German Tiefsee-Expedition of 1898-99. Col-

lected a few phanerogams around Cape Town from

6-12 Nov. 1898 and made a collection of marine al-

gae, also at Kerguelen, Seychelles and Sumatra; in

HBG.

Ref.: pers. comm, from Prof. Kurt Walther, Ham-

burg, March 1977.

Bain, Thomas Charles John (1830-93).

b. Graaff-Reinet, C.P., 29 Sept. 1830; d. Cape

Town, 29 Sept. 1893; roads engineer and builder of

many mountain passes; trained by his father, the

famous roads engineer, Andrew Geddes Bain. At

the request of Sir Henry Barkly, governor of the

Cape (q.v. Gunn & Codd, 1981), he collected a

number of Stapelieae, several of which were de-

scribed as new by N. E. Brown. In an informative

booklet which he wrote, Knysna District in the Divi-

sion of George, Colony of the Cape of Good Hope,

London 1871, he listed 41 varieties of timber.

Commemorated in Hoodia bainii Dyer.

Ref.: White & Sloane, The Stapelieae, Pasadena

1937; Georgina Lister, Reminiscences of Georgina

Lister, Johannesburg 1960; Burman in DSAB 1: 39

(1968); Storrar, A Colossus of Roads, Cape Town

1984.

Balkwill, Kevin (1958- )

b. Cape Town, 14 Febr. 1958; student; ed. Wit-

watersrand Univ., graduating B.Sc. in 1979, H. Dip.

Ed. in 1980, B.Sc. (Hons) in 1981; continuing with

Ph.D. at Natal Univ., Pietermaritzburg.

Has collected mainly in northern Transvaal and

Natal, c. 1 000 of his own numbers and 1 400 in con-

junction with other collectors, chiefly J. C. Manning

and M. J. Cadman; NU, PRE, J.

Bamps, Paul Joseph Rodolphe (1932- )

b. Louvain, Belgium, 6 Feb. 1932; botanist; ed.

Univ. Catholique de Louvain, graduating in 1955.

Curator of the African Herbarium, Jardin Botan-

ique National de Belgique, Meise.

Collected 240 specimens in Transvaal and Natal in

January-February 1982; in BR, GENT, LG, PRE.

* Barnard, Thomas Theodore (1898-1983)

d. Furzebrook, Devonshire, 20 Aug. 1983.

Ref.: Rourke in Veld & Flora 70: 39^11 (1984).

Bean, Patricia Anne {nee Taylor) (1930- )

b. Longueuil, near Montreal, Canada, 11 July

1930; biologist and teacher; ed. Rhodes Univ.

1947-51, graduating B.Sc., and Univ. of Cape

Town, M.Sc. (1962). Teacher in charge of the Field

Biology and Nature Study School, Kirstenbosch

National Botanic Garden, 1960-64; senior lecturer

in biology at the Teachers’ College, Bulawayo,

1965-75, specializing in ethology and ecology with

emphasis on veld management and game ranching.

Returned to South Africa in 1975 and from 1977 on

the staff of the Bolus Herbarium, Cape Town, with a

special interest in Agathosma.

Specimens c. 1 500, mainly in BOL.

Behr, Cathrina Maria (1958- )

b. Johannesburg, 5 July 1958: botanical assistant;

ed. Rand Afrikaans Univ. 1977-80, graduating

B.Sc. After a short period in the botanic garden of

the Botanical Research Institute, she joined the Na-

tional Botanic Gardens of South Africa and was sta-

tioned on the Highveld Botanic Garden, Roode-

poort.

Specimens c. 700, from the Krugersdorp-Wit-

watersrand area, in NBG, PRE.

* Bell-Marley, Harold Walter (1873-1946)

b. England (probably Richmond, Surrey), 1873.

d. Durban, Natal, 27 Jan. 1946; naturalist and col-

lector. Came to southern Africa as a British soldier

and fought in Zimbabwe in 1896, in the Anglo-Boer

War for 18 months, and in the Bambatha Rebellion,

Natal, in 1906. He returned to Britain where he re-

ceived his discharge from the army and returned to

Natal shortly afterwards, obtaining a post with a firm

of shipping agents at the Point in Durban. In Aug.

1918 he joined the Natal Provincial Administration

as Principal Fisheries Officer, based in Durban, a

post he held until his retirement in 1937.

Bothalia 15, 3 & 4 (1985)

633

He paid almost annual visits of several weeks du-

ration to northern Zululand and was one of the first

to make extensive collections of birds, eggs, insects

(especially butterflies and beetles), crabs, fish and

plants in this relatively inaccessible and unhealthy

part of the country. He also collected in Zimbabwe.

He was one of the first to collect molluscs from the

stomachs of deep-sea fish, many of which were un-

described.

His paintings of fish, from fresh-caught speci-

mens, are in the Natal Museum, and some of them

were published by other authors, such as J. L. B.

Smith and Chubb. His zoological material went to

various museums in South Africa, especially Dur-

ban, and to major museums in Europe and the

United States; his collection of South African birds’

eggs is in the Pretoria Museum; his plants have been

seen only in NH.

Ref.: S.A. Museums Assoc. Bull. March 1946,

p.397; L. S. Whicher in The Entomologists’ Monthly

Magazine 85: 49 (1949) and pers. comm.; A. J.

Duke, East London, pers. comm. Aug. 1983; Direc-

tor, Natal Parks Game & Fish Preservation Board,

Pietermaritzburg, pers. comm. Sept. 1983; Director,

Durban Museum, pers. comm. Oct. 1983.

* Bigalke, Erich Heinrich (1937- )

b. Kimberley, C.P., 12 June 1937; teacher and

ethnologist; ed. Rhodes Univ. 1955-59, graduating

B.A., U.E.D. and M.A. (1969); Univ. of Edinburgh

1963-64, Dip. Social Anthropology; Ph.D. (Queen’s

Univ., Belfast) 1983. Taught in schools in S.W. Afri-

ca/Namibia, Zimbabwe and Natal 1960-65 and at

Univ. of Witwatersrand 1965. Ethnologist, East

London Museum, 1966-70, Deputy Director

1971-74, Director from 1974. Interested in social

anthropology and use of plants by native tribes

(Southern Nguni).

Specimens c. 150, from eastern Cape Province and

Transkei; in GRA, some in PRE, BOL.

FIG. 2. Gladys Blackbeard

* Bisschop, John Henry Roosegaarde (1898-1984)

d. Kokstad, 27 April 1984.

* Blackbeard, Gladys Ivy (1891-1975)

b. Grahamstown, C.P., 19 May 1891; d. Grahams-

town, 11 Sept. 1975; gardener and nature lover who

maintained a nursery for indigenous plants, espec-

ially Amaryllidaceae and succulents, on her property

Scott’s Farm on the outskirts of Grahamstown, most

of which she collected and propagated herself. She

supplied plants to the Albany Museum Herbarium

and to authorities overseas, such as Von Poellnitz,

and assisted Lotsy and Goddijn of Holland during

their visit to the eastern Cape Province in 1927. Her

collection of some 2 000 plants of Clivia was ac-

quired by Mr Gordon McNeil of Ofcolaco in 1962.

Fig 2.

Commemorated in Haworthia blackbeardiae V.

Poelln.

Some specimens in GRA.

Ref.: pers. comm, from Mrs Estelle Brink, Gra-

hamstown, Feb. 1984.

Bojer, Wenceslas(Wencelaus, Wenzel) (17957-1856)

b. Resanice, southern Bohemia (then part of the

Austrian Empire), 23 Sept. 1795 (Vaughan, l.c.) or

in Prague, Bohemia, 1797 (Stafleu & Cowan, l.c.)-,

d. Port Louis, Mauritius, 4 June 1856; horticultural-

ist and naturalist. Trained as a horticulturalist on the

estate of Count Caspar von Sternberg at Radnitz,

1810-13. Through the influence of F. W. Sieber

(q.v. Gunn & Codd, 1981), he worked in the Impe-

rial Museum at Vienna 1813-1820, where he agreed

to join Sieber’s band of collectors. Left for Mauritius

as assistant to Charles Hilsenberg (1802-24) of Er-

furt, arriving on the island in July 1821. He collected

extensively in Mauritius, Madagascar and the Como-

ros, as well as along the coast of East Africa, finally

settling in Mauritius. An active member of the local

scientific society, he was appointed Professor of Na-

tural History at the Royal College 1826-32 and Cu-

rator of the Natural History Museum from 1842. In

later years he devoted much of his attention to the

sugar industry. Fig. 3.

Commemorated in genera Bojeria DC. (1836),

Bojeria Raf. (1838) and in many species names.

Specimens in W (orig.), BM, C, G-DC, K. P etc.

(IH 2,2 : 83, 1954); according to Mr P. Bamps (pers.

comm. April 1984), there are some in BR collected

at ‘Caput b. Spei.' It was not previously realized that

he had collected at the Cape. This was no doubt dur-

ing his outward journey in 1821.

Ref. Hooker's J. Bot. Kew Gdn Misc. 8: 312

(1856); Vaughan in Proc. Roy. Soc. Art. Sci. Mauri-

tius 2: 73 (1958), with portrait; Stafleu & Cowan,

Taxonomic Literature 1: 261 (1976).

* Borchardt, E. J.

Technical assistant on Jonkershoek Forestry Sta-

tion who, encouraged by Prof. Wicht, collected

specimens at Jonkershoek.

Specimens in JF, STE.

634

Bothalia 15, 3 & 4 (1985)

* Bottomley, Averil Maud (1889-1984)

d. Johannesburg, 23 Feb. 1984.

* Bremekamp, Cornells Elisa Bertus (1888-1984)

d. Bilthoven, Holland, 21 Dec. 1984

Brown, Christopher Justin (1955- )

b. Durban, Natal, 17 June 1955, ornithologist; ed.

Natal Univ., graduating B.Sc. in 1976, B.Sc. (Hons)

in 1977 and working on Ph.D. at present. Appointed

as Ornithologist in the Nature Conservation Divi-

sion, South West Africa/Namibia, and stationed in

Windhoek.

While an undergraduate, he and B. Shapiro (pres-

ent address unknown) were employed by the Natal

Parks Board to undertake a biological survey of the

newly proclaimed Itala Nature Reserve in northern

Natal. Specimens c. 580, in PRE, NPB.

FIG. 3. W. Bojer F. A. Brusse

* Brown, M. G.

Farmer in the Vryburg District who sent occa-

sional specimens to McGregor Museum, Kimberley,

in the 1940’s from his farm Palmyra for identifica-

tion; in KMG.

Ref.: pers. comm, from Mrs Anna Ludlow, 1983.

Brusse, Franklin Andrej (1951- )

b. The Hague, Holland, 24 Feb. 1951; lichenolo-

gist; came to South Africa with his parents in 1953;

ed. Witwatersrand Univ. 1972-80 graduating B.Sc.

(1975), Hons (1976), M.Sc. (1980), for a study on

Xanthoparmelia. Appointed to National Herbarium,

BRI, Feb. 1980 as curator of the lichen collection.

Fig. 3.

Specimens c. 3 000, mainly lichens, collected

chiefly in Transvaal, Cape and S.W. Africa/Nami-

bia; in PRE.

Bruyns-Haylett, Alfonso Frederik (1891-1963?)

b. Lakeside, C.P., 30 March 1891 as A. F.

Brui jns; his parents died while he was still young and

he was adopted by the Haylett family, assuming the

name Brui jns-Haylett, subsequently changing it to

Bruyns-Haylett; d. Pietermaritzburg 1963 (?); engin-

eer; ed. South African College (later Univ. of Cape

Town) 1910-13, graduating with the Diploma in

Civil Engineering. Served in Tanzania during World

War 1, where he was taken prisoner. Joined S.A.

Railways & Harbours on his return, becoming

A.M.I.C.E. and serving as President, and also M.

Inst. T. (Member of the Institute of Transport).

Spent most of his career in Natal and retired as Chief

Civil Engineer in 1951 after which he was Chairman

of the Private Townships Board, Natal Provincial

Administraton, for several years.

Was interested in succulent plants, particularly

aloes, and collected plants, often while supervising

the construction of railway tracks, mainly in Natal.

On his death his collection, including some cycads,

was presented to the Department of Botany, Natal

University.

* Bruyns-Haylett, John Pieter ( 1927- )

b. Pietermaritzburg, 8 Aug. 1927; plant pathol-

ogist; son of A. F. Bruyns-Haylett; ed. in Faculty of

Agric., Natal Univ., 1948-51, graduating B.Sc. with

Plant Pathology as a major. Continued studies with a

bursary from the Tobacco Control Board and joined

the Tobacco Research Station, Rustenburg, as Plant

Pathologist, in Aug. 1954. Exempted from M.Sc.

examinations and obtained the Ph.D. degree in 1961

for his studies on Helminthosporium turcicum.

Joined Dow Chemical Africa (Johannesburg) in

Jan. 1961 as manager of Technical Services and De-

velopment of all agricultural products in Africa; sta-

tioned at the Dow European R & D Headquarters,

Horgen, Switzerland, 1970-73, responsible for Dow

acaracides and fungicides; from 1974 at King’s Lynn.

England, firstly as Fungicide Development Manager

and, from 1977, Herbicide Development Manager

for Europe, Middle East and Africa.

Collected succulent plants with his father and, in

1949, made a herbarium collection for the Natal

Parks Board in the Giants Castle area; specimens in

NU.

* Buitendag, Elise

Botanical Officer at the Lowveld Botanic Garden,

Nelspruit, 1971-1981 ; from Jan. 1984 Plant and Seed

Control Officer, Citrus & Subtropical Fruit Re-

search Institute, Nelspruit.

Burelli, Giovanni Guiseppe (1956- )

b. San Daniele, Italy, 7 June 1956; plant pathol-

ogist; came to South Africa in Feb. 1959; ed. Wit-

watersrand Univ. 1974—79, graduating B.Sc. (Hons).

Appointed as plant pathologist with Agricura

1979-81; field officer, S.A. Avocado Growers Asso-

ciation 1982 and Director from 1983. Fig. 4.

A keen amateur botanist and naturalist, he under-

took a survey of Bophuthatswana with two of his fel-

low students, Peeters (q.v.) and Gericke (q.v.);

specimens c. 2 000, mainly in Dept of Agric. herbar-

ium, Mafikeng, J. PRE, MO.

Burger, Gladys Hever ( nee Hoare) (1906- )

b. Pretoria, 9 Aug. 1906; housewife and conserva-

tionist; ed. Johannesburg Girls’ High School. Mar-

Bothalia 15, 3 & 4 (1985)

635

FIG. 4. G. G. Burelli Gladys H. Burger

ried Mauritz Dietz Burger (1901-81) and both be-

came interested in indigenous plants. Served on the

committee of the Tree Society of Southern Africa

and contributed to the journal Trees of South A frica

between 1966 and 1980 in the form of articles and

photographs, of which Mauritz Burger built up a col-

lection of about 1 500. Fig. 4.

Specimens c. 600, mainly from the Transvaal; in

PRE.

Cabu

Collected specimens at the Cape in August — Sep-

tember 1930 and also in the Congo ( 1936); in BR.

Ref.: 1H 2: 111 (1954) and pers. comm, from Mr

P. Bamps, April 1984.

* Campbell, Glen Kerry (1944- )

b. Pretoria, 20 March 1944; plant physiologist; ed.

Natal Univ., Pietermaritzburg, 1962-65, graduating

B.Sc., later Ph.D. Lecturer in plant physiology.

Dept of Biological Sciences, Univ. of Natal, Dur-

ban, from 1966; interested mainly in seed dormancy

and plant hormones.

Undertook a survey of trees in the Stainbank

Nature Reserve, Durban, as a postgraduate project;

specimens c. 200; in NU.

* Cassidy, T. J.

Business man in Cape Town. Made a collection of

plants above Camps Bay; in NBG (Miss Barker in

pers. comm. 1982).

* Chater, Sidney Walter (1894- )

b. London; England, 6 October, 1894; journalist

and conservationist; served in World War 1 with rank

of Captain and awarded M.C. Settled in Cape Town

in March 1925 and was editor of Organised Agricul-

ture 1932-61.

Specimens c. 200; in NBG, STE, BOL; was asso-

ciated with E.G. H. Oliver during studies on Erica in

the Western Province mountains and built up an ex-

tensive collection of colour slides of the S.W. Cape

flora.

Cheadle, Vernon Irving (1910- )

b. Salem, South Dakota, USA, 6 Feb. 1910; plant

anatomist. Collected material in S. Africa c. 1960 for

anatomical study.

* Coetzer, Lourens Abraham (1942- )

Date of birth 9 Oct. 1942, not 10 Sepi. as pre-

viously recorded (Gunn & Codd, 1981). Received

the degree of D.Sc. from Pretoria Univ. for an em-

bryological study on the South African species of Ty-

losema. Has collected c. 860 specimens, in PRU,

PRE. Fig. 5.

* Cohen, Ethel (later Mrs Gluckman)

Studied at Witwatersrand Univ. and was awarded

the Solly Scholarship at Kirstenbosch in 1938, after

which she worked for a while at Kirstenbosch on a

revision of Ehrharta, then as lecturer in botany, Wit-

watersrand Univ., for a few years until her marriage.

Settled with her husband in Israel.

Specimens in NBG (leg. E. Cohen) and in J (leg.

E. Gluckman).

* Collett, Derrick George (1917- )

b. Cradock, Cape, 24 June 1917; botanist and

farmer; ed. Rhodes Univ. 1934— 36, graduating B.Sc.

Appointed to the then Division of Plant Industry in

July 1937 and stationed in the National Herbarium,

Pretoria; Liaison Officer at Kew 1938-39; on mili-

tary service, first in the South African Artillery and

later with the Royal Artillery, in the Middle East

and Italy 1940-46. Transferred to Pasture Research

in 1948 and stationed on Towoomba Pasture Re-

search Station near Warmbaths until 1950 when he

resigned to take over the family farm, Groen Kloof,

at Fish River, Cradock District. Retired to Gra-

hamstown in 1980.

Specimens in PRE (orig.), K etc.

* Collins, Elizabeth (‘Liz’) Sophia (later Mrs Roos)

(1898-1969)

b. Pretoria, 20 July 1898; d. Pretoria, 1969;

teacher; trained at Pretoria Normal College and

taught at Sunnyside Primary School for a number of

years. Sister of 'Min’ Collins (q.v.)

Specimens less than 200, leg. L. Collins; in PRE

(ex Transvaal Museum).

Ref.: pers. comm, from her daughter Mrs Trollip,

Verwoerdburg, July 1984.

* Collins, Maria (‘Min’) Carolina (later Mrs

Holden) (7-1918)

Employed as botanical assistant to Miss Leendertz

in the herbarium of the Transvaal Museum, Pre-

toria, Sept. 1907 to Oct. 1913 when she left to get

married. After her marriage she lived in Durban,

where she died in 1918. Sister of 'Liz' Collins (q.v.).

There are 244 of her specimens now in PRE (ex

Transvaal Museum), collected in the Transvaal.

Ref.: pers. comm, from Transvaal Museum, 1984.

636

Bothalia 15, 3 & 4 (1985)

Coppejans, Eric (1948- )

b. Gent, Belgium, 6 March 1948. Marine algolo-

gist in the herbarium of the Univ. of Gent. Grad-

uated with a doctorate in botanical sciences at Univ.

of Gent and collected in Cape Province and SWA-

/Namibia during 1970; 1399 specimens in GENT

(orig.), 1 077 dupl. in BR (pers. comm, from Mr P.

Bamps, April 1984).

\

Doreen Court

Court, Grace Doreen {nee Morris) (1928- )

b. Oudtshoorn, C. P., 16 Aug., 1928; teacher and

botanist; ed. Rhodes Univ. 1946-48 and 1951, grad-

uating B.Sc., U.E.D. Awarded Solly Scholarship at

Kirstenbosch in 1949. Married Jack Gilroy Court 3

Jan. 1953. Taught in Zimbabwe 1963-78; Lecturer in

Botany, Rhodes University, from 1979. Her interest

in plants, especially succulent plants, was stimulated

by her mother, Mrs F. N. Morris (q.v.), and be-

tween 1974 and 1979 she travelled widely in southern

Africa collecting material and photographs for her

book Succulent Flora of Southern Africa , Cape

Town 1981. Fig. 5.

Collected some specimens in south-western Cape,

in NBG, leg. D. G. Morris (the labels were written

by Miss W. F. Barker who transposed the initials);

also collected succulent plants throughout southern

Africa which are chiefly maintained as a living col-

lection in Grahamstown.

* Crook, Albert Oliff (‘Paddy’) (1913-84)

d. Cape Town, 28 April 1984.

* Culverwell, James (1952- )

b. Manzini (Bremersdorp), Swaziland, 29 May

1952; naturalist and conservationist; ed. St Mark’s

School, Mbabane, 1956-67; St Martin’s School, Jo-

hannesburg, 1968-70; managed a reinforcing steel

firm in Mbabane 1973-77; worked in Sabi-Sand

Game Reserve in 1978; for the Swaziland National

Trust Commission, 1979; from 1980 as warden of the

private Mbuluzi Nature Reserve in north-east Swazi-

land lowveld.

Specimens c. 1 400, in PRE; recent collections

mainly from conservation areas in north-eastern

Swaziland and Lebombos.

Cuthbert, James Brown (fl . 1931)

M.Sc., Cape Town Univ. c. 1931 for a thesis on

the physiology of a dendritic lichen and did some

collecting in the Riviersonderend Mts.

* Dahlstrand, Karl Ake (1904-80)

d. Port Elizabeth, 2 April 1980 (pers. comm, from

Mr Bo Peterson, Gotenborg. July 1984).

Davidse, Gerrit (1942- )

b. Grijpskerke, Netherlands, 19 Dec. 1942; botan-

ist; ed. Calvin College, Grand Rapids, Michigan,

1961-65, graduating B.Sc.; Utah State Univ.

1965-68, graduating M.S.; low'a State Univ.

1968-72, graduating Ph.D. Appointed as Assistant

Curator. Missouri Botanical Garden, 1972-78, As-

sociate Curator from 1978. Mainly interested in the

tropical American flora, especially Gramineae. Fie.

6.

Commemorated in the names of several American

plants, such as Hierochloa davidsei Pohl, Ichthyo-

thera davidsei H. Robinson, Strychnos davidsei Kru-

koff & Barnaby etc.

Visited South Africa in Jan. - Feb. 1974 and col-

lected 612 specimens mainly in northern Transvaal,

northern Cape and Natal; 230 in S.W. Africa/Nami-

bia; 243 in Zimbabwe; 104 in Zambia; in MO, PRE.

FIG. 6. G. Davidse G. B. Deall

* Deacon, Hilary John (1936- )

b. Cape Town, 10 Jan. 1936; archaeologist; ed.

Univ. of Cape Town, 1953-55 and 1962, graduating

B.Sc.. B.A. (Hons), M.A., Ph.D. Professional offi-

cer and later Deputy Director, Albany Museum,

1963-71; Senior Lecturer in Dept of Archaeology,

Univ. of Stellenbosch, 1971-78, and Professor from

1979. Plant specimens were collected in connection

with a project on the prehistory of the eastern Cape

for comparison with macroscopic plant remains from

cave deposits and as a pollen reference collection.

Specimens and pollen slides in GRA.

Deall, Graham Basil (1950- )

b. Harare, Zimbabwe, 14 June 1950; ecologist;

ed. Natal Univ., Pietermaritzburg, graduating B.Sc.

in 1971, B.Sc. (Hons) in 1979. Appointed to the Bo-

tanical Survey Section of the BRI in 1980 and is

Bothalia 15, 3 & 4 (1985)

637

undertaking survey work in the eastern Transvaal.

Fig. 6.

Specimens c. 2 000, collected in Lesotho (mosses

only) and eastern Transvaal; in PRE.

* Dekenah, Albert Jacob (1907-81)

d. Riversdale, C.P. 11 July 1981.

* Denman, Mary

d. Port Elizabeth, c. 1980. Matron of Empilweni

Hospital, Port Elizabeth, for many years. She had a

Licentiate in music and painted wild flowers which

were exhibited on two occasions in Port Elizabeth.

Wife of Earle L. Denman, author of Alone to Ever-

est, and accompanied her husband to the Himalayas

where she also painted some of the wild flowers

(pers. comm, from Mrs Auriol Batten, 1982).

Specimens in GRA (Tolken, 1971).

* D’Estourgies

The collection which he made in the Transvaal

was deposited in BR in 1877 (pers. comm, from Mr

P. Bamps. Brussels, April 1984).

De Wit, Hendrik Cornelis Dirk (1909- )

b. Purmerend, Holland, 24 Oct. 1909; botanist;

ed. Amsterdam Univ. (Drs. 1937) and Pretoria

Univ. (Ph.D., Univ. of S. Africa, 1941). Employed

as Agricultural Research Officer, Dept of Agricul-

ture, S. Afr., 1938-40, when he undertook a revision

of the genus Setaria ; Botanist at Buitenzorg, Java,

1941^45; Botanist, Flora Malesiana Foundation,

Leiden, 1946-53; from 1953 Senior Lecturer in Sys-

tematic Botany, University for Agriculture, Wage-

ningen; from 1959 Professor and, from 1969 until his

retirement in 1980, Director of the Laboratory of

Plant Taxonomy and Plant Geography. Publications

deal with systematic botany, mainly of the Asian

tropics, history of botany and biology.

Specimens c. 4 000, from the Cape Province, Na-

tal, Swaziland and Transvaal; in L, WAG.

Duncan, Graham Dugald (1959- )

b. 28 Nov. 1959; horticulturalist; trained at Cape

Town Technikon obtaining the National Diploma in

Horticulture and appointed to Kirstenbosch Botanic

Garden.

Specimens c. 100 from S.W. Cape; in NBG.

* Duparquet, Rev. C. (1830-88)

Portrait reproduced from Viagens na Cimbebasia ,

Museu de Angola, Luanda, 1953. Fig. 7.

* Du Plessis, Christiaan Jacobus (1944- )

b. Johannesburg, Transvaal, 19 Nov. 1944; sci-

ence teacher; ed. Pretoria Univ. 1963-72, graduat-

ing B.Sc., later H.O.D. (1969) and M.Sc. (1972) for

a thesis entitled *’n Floristiese-ekologiese studie van

die plaas Doornkop in die distrik Middelburg,

Transvaal’. Taught from 1971-77 and from 1978

Head of the Department of Natural Science, Soweto

College of Education. Fig. 7.

FIG. 7. Rev. C. Duparquet C. J. du Plessis

Specimens c. 1 900, from Middelburg District,

Transvaal; in PRU, PRE.

* Ecklon, Christian Frederick (1795-1868)

We are grateful to Dr Alfred Hansen of the Bo-

tanical Museum, University of Copenhagen, for the

following correction: Ecklon was born in Aabenraa

(in German, Apenrade), northern Schleswig. This

part of Denmark belonged to Germany from

1864-1920 but in 1795 (and from 1920) belonged to

Denmark.

* Edwards, Elizabeth M. (fl. 1945)

Studied at Natal Univ., Pietermaritzburg, during

the early 1940’s and was appointed as biology mis-

tress at Girls’ Collegiate School in 1946. Collected

while a student, mainly in her home territory. East

Griqualand; specimens in NU (pers. comm, from Dr

K. D. Gordon-Gray, Natal Univ., 1983).

Specimens in NU, J (Tolken, 1971).

* Eschscholz, Johann Friedrich (1793-1831)

b. and d. Dorpat, or more recently Tartu, Estonia

(not Latvia as stated previously) (comm, from Dr

Alfred Hansen, Botanical Museum, Copenhagen

University, 1983).

Evrard, Charles Marie (1926- )

ed. Univ. Catholique de Louvain, graduating in

1949 and doctor of botanical sciences in 1964. Pro-

fessor at the Univ. of Louvain-la-Neuve, Belgium.

Collected 479 specimens (Nos 8810-9289) in Cape

Province and SWA/Namibia; also collected in the

Congo; in BR (IH 2,2: 188, 1957; pers. comm, from

Mr P. Bamps, April 1984).

* Fawkes, Madeleine (‘Madge’) Charlotte

(1880-1954)

b. Malta, 14 Dec. 1880; d. Black Bourton, near

Oxford, England, 15 Sept. 1954; artist and collector;

trained at the Slade School of Art. Paris, and in

Cornwall. She visited her brother, Valentine

Fawkes, who farmed near Ficksburg, in 1922-23,

1931-37 and in 1952, during which she painted wild

flowers and took an active part in the local Horticul-

tural Society. Some of her paintings, mainly of

638

Bothalia 15, 3 & 4 (1985)

flower arrangements, were exhibited by the Royal

Academy and, in 1939, the Grenfell Medal was

awarded to her by the R.H.S. for her exhibition of

paintings of Lesotho wild flowers. She also painted

landscapes and portraits. She was a keen gardener

and showed considerable skill in garden designing.

Fig. 8.

Specimens and paintings in NBG.

Ref.: letter from Mrs A. M. Tennent, Ficksburg,

Aug. 1982; Codd & Gunn in Veld & Flora 68: 93-94

(1982).

FIG. 8. Madge C. Fawkes

* Feely, James Michael (1934- )

b. Tanga, Tanzania, 17 Jan 1934; self-taught natu-

ralist; ed. Univ. of Cape Town. Game ranger at

Lake St Lucia and Umfolozi Game Reserves

1955-60; Luangwa Valley, Zambia, 1961-65 then,

after a period as game rancher, field officer for Wil-

derness Leadership School until 1982; from 1983

Senior Research Assistant in Dept of Botany, Univ.

of Transkei, Umtata. Published ‘Observations on

Acacia albida in the Luangwa Valley’ in Puku 3:

67-70 (1965), and ‘Did Iron Age man have a role in

the history of Zululand’s wilderness landscapes’ in S.

Afr. J. Sci. 76: 150-152 (1980). Currently investigat-

ing the ecological consequences and constraints re-

garding Iron Age farming settlements in Transkei.

Fig. 8.

Specimens c. 400 collected in Zululand (in NPB,

NU) and c. 200 in Zambia (in FHO).

Fellingham, A. C. (nee Bester) (1933- )

b. Vryburg, C. P., 23 Oct. 1933; botanical techni-

cian; ed. Potchefstroom Univ. c. 1954, graduating

B.Sc. After various posts in medical research, she

was appointed as Chief Technician in the Botanical

Research Unit, Stellenbosch.

Specimens c. 500, mainly from S.W. and S. Cape,

in STE.

* Ferrar, Evelyn

Assistant to Dr M. Wilman, McGregor Museum,

Kimberley, in the 1940’s who collected specimens

with Dr Wilman; in KMG.

* Forbes, Helena Madelain Lamond (1900-59)

Further references: obituary in S. Afr. J. Sci. 55:

317 (1959); Schrire in Bothalia 14: 223 (1983).

Fourie, Stephanus Petrus ( 1949- )

b. Potchefstroom, Transvaal, 3 Sept. 1949; con-

servationist; ed. Rand Afrikaans Univ., graduating

B.Sc. in 1973 and B.Sc. (Hons) in 1974. Spent his

early years in Zimbabwe (then Rhodesia) and was

awarded the shield for ‘student of the year’ in the

Rhodesian army in 1968. Appointed to the Trans-

vaal Div. of Nature Conservation in 1975; Senior

Professional Officer, 1978; Chief Professional Offi-

cer 1983 and from 1984 head of the section for Flora

and Environmental Conservation; mainly concerned

with the conservation of rare and endangered plant

species in the Transvaal.

Commemorated in Protasparagus fouriei Oberm.

Specimens c. 3 200, mainly from Transvaal; in

Nature Conservation herbarium. Lydenburg, and

PRE.

* Froembling, George Herman Walter (1859-1941)

b. London, England, Dec. 1859; d. Cape Town, 7

June 1941; pharmaceutical chemist; studied in Ger-

many, qualifying in 1889 and with Ph.D. (Munich) in

1896. Came to Cape Town in 1897 and set up a phar-

maceutical business, experimenting with drugs in his

spare time in association with the late Dr Hahn and

Dr Penther of the firm Wentzel & Schleswig, and

published papers on drugs used in S. African native

medicines. In 1909 he was President of the Cape

Pharmaceutical Society and was instrumental in the

formation of the S. African Pharmaceutical Society.

In his botanical activities he was often in contact

with Marloth, Bolus and Dummer.

Commemorated in Agathosma froemblingii

Dummer.

Specimens collected around Cape Town, mainly

from 1897-98, in B (Index Herbariorum, Collectors

2,2: 211, 1957), about 500 in NBG (Rourke in Veld

& Flora 69: 159, 1983) and some in E. Also collected

in Chile and Venezuela, 1885-86 (in M).

Ref.: Afr. World 26 July 1941, p. 61; Moore in

Veld & Flora 69: 158-60 (1983).

* Fuller, Edward Barnard (1868-?)

b. Mowbray, Cape Town, 29 June 1868; d. ?;

medical practitioner; ed. S.A. College, Cape Town,

and Edinburgh Univ. graduating M.B., C.M.

(1891), F.R.C.S.E. (1892). Started a practice in

Cape Town in 1892 and was a pioneer in many fields,

including public health, medical education and surgi-

cal urology. He lectured in clinical surgery at Cape

Town University, served on innumerable commit-

tees and was a keen sportsman.

Collected some specimens at the Cape c. 1886; in

E (Hedge & Lamond, Index of Collectors in the

Edinburgh Herbarium , Edinburgh 1970).

Ref.: South African Who's Who , Johannesburg

1937; Burrows, A History of Medicine in South

Africa , Cape Town 1958.

Bothalia 15, 3 & 4 (1985)

639

* Garabedian, Star (1895-1978)

b. 1895, according to Dr Alfred Hansen of the Bo-

tanical Museum, University of Copenhagen (pers.

comm. 1983).

* Gentry, Howard Scott (1903- )

b. Temecula, California, USA, 10 Dec. 1903; bi-

ologist; graduated B.A. at Univ. at California, Ber-

keley, in 1931 (zoology), Ph.D. at Univ. of Michi-

gan, Ann Arbor, in 1946 (botany). Farmed in Cali-

fornia 1927-31; free-lance biologist 1933^-0; fossil

hunter for Amer. Mus. Nat. Hist. New York,

1934-40; emergency war work investigating rubber

plants, 1942-45; research botanist, Allan Hancock

Foundation, Los Angeles, 1946-50; plant explora-

tion and investigating new crops with U.S. Dept

Agric., 1950-71; research botanist. Desert Botanical

Garden, Phoenix, Arizona, 1972-81; from 1982 As-

sociate Director, Amerind Agrotech Laboratory,

Sacaton, Arizona.

Collected in South Africa during 1960-61 on be-

half of the USDA New Crops Branch, about 600

gatherings (Nos 18612-19218) including many her-

barium specimens; in US, PRE, etc.

Gericke, Nigel Peter ( 1955- )

b. Johannesburg, 26 June 1955; medical student;

ed. Witwatersrand Univ., graduating B.Sc. (Hons)

in 1979 and continuing for M.B. , Ch.B. Interested in

ecology and the S. African flora, he undertook a sur-

vey of Bophuthatswana with two of his fellow stu-

dents, Peeters (q.v.) and Burelli (q.v.); specimens c.

2 000, mainly in Dept of Agric. Herbarium, Mafi-

keng, J, PRE, MO.

Germain, Rene Antoine (1914-1982)

b. Monceau-Imbrechies, Belgium, 10 March 1914;

d. Ganshoren, Brussels, 4 Feb. 1982; agronomist;

ed. Univ. Catholique de Louvain, graduating in 1934

and doctor of botanical sciences in 1950. Stationed in

the Congo c. 1940-1950 where he collected about

9 000 numbers; 84 collected in Natal and Cape Prov-

ince (Nos 1509-1592) in August-September 1942; in

BR (IH 2,2: 222, 1957; pers. comm, from Mr P.

Bamps. Brussels, April 1984).

* Gerstner, Father Jacob (1888-1948)

Father Gerstner’s collecting books are in the Dept

of Botany, University of Zululand (pers. comm,

from Prof. E.J. Moll).

* Gibbs Russell, Garland Elizabeth

Wrongly listed under Russell in Gunn & Codd,

1981.

* Gluckman, Mrs E.: see Ethel Cohen

Godfrey, Robert Kenneth (1911- )

b. Bloomsbury, New Jersey, USA, 29 Aug. 1911;

botanist. Collected in South Africa in 1952 for the

U.S. Dept of Agric. New Crops Branch, chiefly

planting material of alkaloid-producing or drug

plants, and also some herbarium specimens.

Goetghebeur, Paul (1952- )

b. Ostend, Belgium, 4 Nov. 1952. Botanist in the

Herbarium of the Univ. of Gent, where he grad-

uated in botanical science. Collected c. 230 speci-

mens in Transvaal and Natal in January — February

1982; in GENT (orig.), BR, LG, PRE (pers. comm,

from Mr P. Bamps, Brussels, April 1984).

* Gordon-Gray, John Louis (1916- )

b. in the Cape (Wynberg ?), 12 Aug. 1916; army

officer and collector; after Potchefstroom Boys’

High School he entered a cadet officer’s course in

1935. Served in Western Desert, Italy and seconded

to British Army towards end of World War II. Re-

turned to S. Africa after the war and retired in 1964,

settling at The Haven, Transkei. Brother-in-law of

Dr K. D. Gordon-Gray, who encouraged him to

make a comprehensive collection, particularly of

forest plants, within a ten mile radius of The Haven

(pers. comm, from Dr K. D. Gordon-Gray, Sept.

1982). Specimens in NU; the bryophytes were sent

to MO for Dr Magill.

* Gorter, Gerard Jacobus Marinus Anne

(1913- )

First name wrongly spelt ‘Gerhard’ in Gunn &

Codd (1981).

* Gouws, Jozef Benjamin (1909- )

b. Amersfoort, Transvaal, 13 June 1909; botanist;

ed. Pretoria Univ., graduating B.Sc. (1933), M.Sc.

(1941), D.Sc. (1947), with further study at Hohen-

heim-Stuttgart in 1962. Appointed Lecturer, De-

partment of Botany, Pretoria Univ., later Senior

Lecturer, Pretoria Bantu Normal College, then Pro-

fessor of Botany, Univ. of western Cape until his re-

tirement.

Specimens c. 1 000, from northern and north-east-

ern Transvaal, and south-western Cape; in PRU,

PRE, UWC.

Gubb, Andrew Alan (1950- )

b. Cape Town 24 April 1950; botanist; ed. Cape

Town Univ., graduating B.Sc. (Hons) in 1977. Was

a ranger at Cape of Good Hope Nature Reserve

1970-72 and, from 1973, ecological botanist and cu-

rator of the herbarium, McGregor Museum, Kim-

berley, where he has studied habitat utilization by

game and phytosociology. Also interested in remote

sensing research.

Specimens c. 16 000, mainly from N. Cape, in

KMG; some from S.W. Cape in NBG.

* Guy, Robert Douglas (1932- )

b. Pietermaritzburg, Natal, 26 Dec. 1932; farmer;

ed. Rhodes Univ., graduating B.Sc.

Specimens c. 50, collected in Zululand, in NPB,

NU.

Hargreaves, Bruce James (1942- )

b. Stockton, California USA, 18 April 1942; bot-

anist; graduated B.A., Univ. of California in 1964;

M-S-P^H. , Univ. of N. Carolina in 1970; Ph.D. New

640

Bothalia 15, 3 & 4 (1985)

York Univ. in 1974. Instructor, New York Univ.

Medical Centre 1975-76; Lecturer, Univ. of Malawi

1976-81; Senior Lecturer and head of herbarium and

botanic garden, Univ. of Lesotho from 1983. Inter-

ested in medicinal and succulent plants.

Specimens c.3000 colld in Malawi, in MAL; 800 in

Lesotho, in ROML.

* Harris, Mrs Eric (nee Zoe Anne Borlase)

(1895-1970)

b. Natal, 26 July 1895; d. Pretoria, Sept. 1970;

housewife. Married Eric Harris, who farmed near

Modder River Station, in 1923. Encouraged by Miss

Wilman she collected plants on their farm Three

Oaks, especially bulb plants, some of which were

sent on to Kirstenbosch. Retired to Pretoria.

Specimens in KMG, recorded under the name

Eric Harris (b. Kimberley, 4 May 1893).

Ref.: pers. comm, from Mr Eric Harris, Pretoria,

1982.

* Heginbotham, Marjorie Constance (later Mrs

Schirach) (1921- )

b. Izmir, Turkey, 1 Aug. 1921, and spent most of

her childhood in Greece where she was educated pri-

vately. Came to South Africa with her parents in

June 1941 and worked for several years as librarian

at Kirstenbosch until her marriage.

Collected over 300 specimens, mainly from south-

western Cape; in NBG.

* Henderson, Murray Ross (1899-1982)

b. Banchory, Aberdeenshire, 1899; d. Aberdeen,

Scotland, Oct. 1982. Studied botany at Aberdeen

Univ. after World War I. Appointed in 1921 to Mu-

seums Dept, Federated Malay States, as botanist; in

1924 Curator of the herbarium in the Singapore Bo-

tanic Garden and Director from 1949 until his retire-

ment. During World War II he spent a few years

from 1941 at Kirstenbosch where he published notes

on South African cycads, returning to Singapore in

Dec. 1945.

Ref.: R. E. Holttum in The Gardeners’ Bulletin ,

Singapore 35: 235-236 (1982).

Herman, Paul Philippus Johannes (1955- )

b. Vanderbijlpark, Transvaal, 13 July 1955; bot-

anist; ed. Pretoria Univ. 1973-76, graduating B.Sc.

in 1975, B.Sc. (Hons) in 1976, M.Sc. in 1983 for a

study entitled ‘Die stingel- en blaarmorfologie van

die Suider-Afrikaanse Pavetta- spesies.’ Appointed

to the Herbarium Section, BRI, from March 1979

and, from April 1984, with the Anatomy-Cytology

Section, specializing in wood anatomy. Fig 9.

Specimens c. 1 510, collected alone and together

with other members of BRI staff, in Transvaal, OFS,

north-east Cape and the Karoo; in PRE.

Horn, D.H.S. (fl. 1960)

On the staff of the Chemical Research Inst.,

CSIR, and collected specimens in connection with

FIG. 9. Paul Herman Nancy M. E. Horrocks

an investigation of aromatic substances in plants;

emigrated to Australia.

275 specimens in PRE.

* Horrocks, Nancy Margaret Emily (nee Gillies)

(1902- )

b. Georgetown, New Zealand, 4 April 1902; bo-

tanical assistant. Came to S. Africa in 1930 and was

Technical Assistant in Compton Herbarium 1959 -

75, during which time she mounted over 42 000

sheets. Fig. 9.

Specimens c. 200, from S.W. Cape; in NBG.

Hosten, Liliane Francine (nee Willems) (1932- )

b. Likasi, Zaire, 6 Feb. 1932; botanist and house-

wife; came to South Africa in 1950 and studied at

Stellenbosch Univ., graduating B.Sc. in 1953 and

M.Sc. (Botany) in 1957. Worked at Rijksplanten-

tuin, Brussels, during 1956 and at Compton Herbar-

ium, Kirstenbosch, 1958 - 59. After her marriage

she settled in Port Elizabeth and worked part-time at

UPE. Interested mainly in Mesembryanthemaceae

and in drift seeds.

Specimens in STE, PRE (leg. Willems) and UPE.

* Howlett, J.C.

Technical Assistant in Division of Botany and

Plant Pathology, stationed in Pretoria from 1912 -

1936 and in Durban Botanic Station 1936 until his

retirement in 1951; accompanied Dr I.B. Pole Evans

on some of his expeditions.

Ref.: Schrire in Bothalia 14: 232 (1983).

Hugo, Loretta (later Mrs Van Zyl) (1942- )

b. Moorreesburg, C.P.,8Nov. 1942; botanist; ed.

Stellenbosch Univ. 1972 - 74, graduating B.Sc. , later

M.Sc. Appointed at Botanical Research Unit, Stel-

lenbosch, in 1976 and from 1980 Curator of the Her-

barium. Married Percy van Zyl, 29 July 1982 and re-

signed from her post shortly afterwards.

Specimens c. 3 000, from S.W. Cape; in STE,

PRE.

* I’Ons, John Henry (1936- )

b. Johannesburg, 16 Jan. 1936; pasture ecologist;

ed. Natal Univ., Pietermaritzburg, 1954 - 59, grad-

uating B.Sc. (Agric.) in 1958, M.Sc. (Agric.) in

Bothalia 15, 3 & 4 (1985)

641

1961. Pasture Research Officer, Swaziland, 1959 -

68; in South African Dept Agric. as Pasture Re-

search Officer 1968 - 77 and Agricultural Counsel-

lor, Australia, 1978 - 82. Resigned in 1982 and set-

tled in Australia. Undertook an ecological survey of

Swaziland for which the specimens were identified

by Prof. Compton.

Specimens collected mainly on Malkerns Re-

search Station, Swaziland, and at Estcourt and

Dohne Research Stations; in PRE and Malkerns Re-

search Station.

Immeiman, Kathleen Leonore (1955- )

b. East London, C.P., 26 Dec. 1955; botanist; ed.

Natal Univ. , Pietermaritzburg, 1974 - 77, graduating

B.Sc., B.Sc. (Lions); Cape Town Univ. 1978 - 79,

M.Sc. for a revision of South African spp. of Holoth-

rix ; studying Justicia and allied genera for doctorate.

Appointed to Flora Research team, BRI, 1980 and

has written up Simaroubaceae and genera Acrido-

carpus and Triaspis for the Flora of Southern Africa.

Fig. 10.

Specimens c. 600, from E. Cape, northern Natal,

Transvaal and S.W. Africa/Namibia.

FIG. 10. Kathy Immeiman A. J. Joubert

* Isaac, William Edwyn (1905- )

Prof. Isaac was Professor of Botany at Nairobi

Univ. College from 1961 until his retirement in 1970

and was not subsequently at Cardiff Univ. as pre-

viously stated (Gunn & Codd, 1981).

* Jacobs, Marius (1929-1983)

d. Leiden, Holland, 28 April 1983; taxonomist,

conservationist and biohistorian.

Ref.: Van Steenis in Flora Malesiana Bull. 36:

3869 - 3871 (1983); Kalkman in Blumea 29: 1 - 12

(1983), with portr. and list of publications.

* Jaeot Guillarmod, Amy Frances May Gordon

(1911- )

Dr Amy Jaeot Guillarmod graduated with the de-

gree of D.Sc. from St Andrews Univ., Scotland, not

Ph.D. as previously stated (Gunn & Codd, 1981).

* Jenkins, Thomas J.

Appointed in June 1907 as a clerk at the Transvaal

Museum, Pretoria, and also did collecting and cata-

loguing while at the Museum, concentrating largely

on plant collecting. In 1913 he was transferred on

promotion to the Dept of Lands and shortly after-

wards to the Dept of Inland Revenue, where he

spent most of his career in the Public Service.

Over 1 200 specimens in PRE (ex Transvaal Mu-

seum), collected mainly in the Transvaal, also

around Durban and in Mozambique. His collecting

activities ceased after leaving the Transvaal Museum

in 1913.

Ref. : pers. comm, from Transvaal Museum, 1984.

* Johnstone, Douglas (‘Das') Ian ( fl . 1950)

Studied at Natal Univ., Pietermaritzburg, 1945 -

49, during which time he collected quite extensively

on Noodsberg. Employed at Hlobane where he con-

tinued to collect for some time (pers. comm, from

K.D. Gordon-Gray, Sept. 1982).

Specimens in NU, NH.

* Joubert, Adriaan Jacobus (1901- )

b. on farm Derdeheuwel, Montagu district, C.P.,

27 May 1901; science and biology teacher; ed. Stel-

lenbosch Univ. graduating B.Sc. in 1923, H.S.O.D.

in 1927, and B.Ed. in 1933. Spent his teaching career

at Hoerskool Ladismith; appointed as science and

mathematics teacher in 1924, later Vice Principal

and as Principal from 1953 until his retirement in

1966. With the introduction of biology to the sylla-

bus in 1926, he undertook the teaching of the sub-

ject, although without previous training in this direc-

tion. One of his first undertakings was to build a

rock garden at the school and he later maintained a

collection of succulents at his home. His collection

came to the notice of Prof. G.C. Nel of Stellenbosch

Univ. with the result that plants were sent to Stellen-

bosch and to Kirstenbosch. He specialized at various

times in the Stapelieae, Gibbaeum and Lithops.

Prof. Nel acknowledges his assistance in the Gib-

baeum Handbook , London 1953, and White &

Sloane presented him with copies of their books on

the Stapelieae and Euphorbieae in recognition of

material of living plants supplied. After retiring in

1966 his interest in Lithops induced him to under-

take temporary teaching posts at inland centres such

as Petrus Steyn and Merweville. He contributed arti-

cles, with colour photographs, to the Landbou-

weekblad on Lithops (31 Oct. 1981) and Gibbaeum

and Muiria (12 Nov. 1982). Fig. 10.

Commemorated in Hereroa joubertii L. Bob,

Conophytum joubertii Lavis, Sceletium joubertii L.

Bol.

Specimens in BOL.

Karis, Per-Ola (1955- )

b. Stockholm, Sweden, 28 Aug. 1955; botanist;

ed. Stockholm Univ. 1976-80, graduating B.Sc. At

Kirstenbosch Sept. 1980 to Feb. 1981 studying Met-

alasia and Lachnospermum.

Specimens c. 200; in S, NBG.

* Keit, Julius Wilhelm (1841-1916)

His first name was inadvertently omitted in Gunn

& Codd (1981).

642

Bothalia 15, 3 & 4 (1985)

* Kent, Charles C.

Plant Inspector, Div. of Entomology, Dept of

Agric., from 1921; stationed in Durban until 1954

when he was transferred to East London.

Ref.: Schrire in Bothalia 14: 223 (1983).

* Konig (Koenig) Johann Gerhard (1728-85)

b. Courland, Latvia, which was not a province of

Denmark as stated (Gunn & Codd, 1981).

Krantz, P.A. ( fl . 1894)

Technical Assistant and taxidermist at Transvaal

Museum who collected material for the Museum, in-

cluding some plant specimens (e.g. Lekkerkraal and

Matlabas River, near Thabazimbi, Jan. 1894); in

PRE.

Ref.: Staats Almanak, Z.A. Republiek, 1898,

1899.

* Kresfelder, Louis J. (1898- )

b. Pretoria, 1 June 1898 and attended school in

Pretoria. Joined the government service in March

1912 as clerical assistant and transferred to the Divi-

sion of Botany and Plant Pathology in October 1914.

In addition to clerical work, he was required to pay

regular visits to the Groenkloof Experiment Station

and, from 1916, was also involved in the citrus can-

cer campaign and in nursery inspection work (see

Pole Evans, I.B.). Later became a laboratory assist-

ant under E.M. Doidge (q.v.) and her successor,

J.E. Vanderplank (q.v.), finally retiring on 1 June

1958 after which he settled in Villiers, OFS.

Specimens in PREM (Tolken, 1971).

* Lamb, James Dugal Cameron (1853-1937)

(As Lamb, J.A.C., in Tolken, 1971; I.H. 2,3; 406,

1972; Gunn & Codd, 1981). b. probably in

Berkshire, England, 1 June 1853, of Scottish de-

scent; d. Cape Town 1937; horticulturalist. After a

period of training at Kew, he settled in Rondebosch

and was the gardener and estate manager at Bel-

mont Park House, the home of Alfred Ebden, until

1918 when the property was sold. He then moved to

Oxford Road, Observatory, where he worked pri-

vately on garden lay-outs and maintained a nursery.

He was a member of the Cape Town Mountain club

from 1898 - 1923 and collected plants, geological

specimens and other natural history material during

his walks.

His collecting numbers exceed 4 000, collected

mainly around Cape Town, but also at places such as

Worcester, the Karoo, Benoni, Mafikeng, Botswana

and Zimbabwe. They were kept as a private collec-

tion, the best of which (over 1 000) were donated to

SAM in 1932, where they were remounted with the

original labels in his handwriting. Unfortunately

only the month and not the year of collecting was

recorded. Fig. 1 1 .

Ref.: Annual Rep. S. Afr. Mas. for 1932; pers.

comm, from his granddaughter, Mrs Molly Dowson

of Pinelands, Cape, his great granddaughter, Mrs

Cynthia Kemp of Johannesburg and Dr J.P.

Rourke, Compton Herbarium, Cape Town; Codd &

Gunn in Veld & Flora 70: 61 (1984).

FIG. 11. J. D. C. Lamb D. J. McDonald

* Lang, Herbert (7-1957)

A German by birth, he emigrated to the United

States where he was associated with the Field Mu-

seum of Natural History, Chicago, for some years.

Lived in Cape Town 1926 - 27 and in Pretoria from

1927 until his death in 1957. Married the owner of

Eaton Hall, a boarding house in Skinner Street near

the Transvaal Museum, and accompanied several of

their expeditions, during which he collected and

took excellent photographs. Organized a major ex-

pedition to the Kalahari sponsored by Mr Arthur S.

Vernay on behalf of the Chicago Museum, known as

the Vernay-Lang Kalahari Expedition, March-Sep-

tember 1930. In 1931 he was the first to make a sub-

stantial herbarium collection in the northern Kruger

National Park. In 1957, shortly before his death, the

Transvaal Museum acquired his extensive collection

of zoological and botanical photographs.

Albums of his plant photographs and c. 600 speci-

mens (ex Transvaal Museum) in PRE.

Ref.: FitzSimons & Brain in Tvaal Mas. Bull. No.

13 (1972).

Latrobe, Peter (1795-1863)

Associated with the Moravian ministry in Ful-

neck, England. There are some specimens in PH

(Mears in Proc. Acad. Nat. Sci. Philadelphia 133:

155, 1981, and pers. comm. Apr. 1983) attributed to

‘Latrobe, Peter, CBS' ex Herb. Schweinitz, but La-

trobe may not himself have visited the Cape. One

wonders if he was perhaps the son of C.I. Latrobe,

also of the Moravian ministry, who made an exten-

sive tour of mission stations in South Africa and pub-

lished Journal of a Visit to South Africa in 1815 and

1816 , London 1818.

* Lavoipierre, Michael {fl. 1944)

Studied at Natal Univ., Pietermaritzburg, during

World War II and collected some plants while a stu-

dent, mainly together with E. A.C.L.E. Schelpe dur-

ing an excursion to Port St Johns. Specimens in NU

(pers. comm, from Dr K.D. Gordon-Gray, Sept.

1982).

* Lawson, Archdeacon George Mervyn( 1865-1945)

b. on one of the Channel Islands (probably Jersey)

in 1865; d. Kimberley, 17 Aug. 1945; grew up in

London. After obtaining a B.A. degree, he pre-

Bothalia 15, 3 & 4 (1985)

643

pared himself for the ministry and, after taking Or-

ders, he emigrated to South Africa to do mission

work. He made his headquarters at Papkuil in Gri-

qualand West, attending to various missions from

Schmidtsdrift to west of Upington.

He was interested in succulent plants and took

specimens to Dr Wilman of the McGregor Museum

for identification, some of which were transmitted to

Bolus Herbarium. Encouraged by Archdeacon F.A.

Rogers, he also made herbarium specimens.

Commemorated in Ruschia lawsonii (L. Bol.)

Schw.

Specimens in KMG.

Ref.: Information supplied by Mr W.M. van Zyl

of Paarl and Mr A. Gubb, McGregor Museum, Kim-

berley; Codd & Gunn in Veld & Floral 0 : 68(1984).

* Leighton, Frances Margaret (later Mrs W.E.

Isaac) (1909- )

Daughter of Francis Leighton, a brother of James

Leighton (not daughter of James Leighton as stated,

Gunn & Codd, 1981).

* Lemaire, C.H.

The information included in Index Herbariorum,

Part II, Collectors, 3 (I - L): 429 (1972), appears to

be incorrect. Mr P. Bamps of BR informs us (April

1984) that Capt. Charles Lemaire (1863 - 1926)

gathered some pictures of plants, without speci-

mens, during his ‘Mission scientifique du Katanga’ in

1898 - 1900. H. Lemaire (no dates), agronomist, col-

lected some 550 specimens in the ‘Equateur’ District

of Congo in 1913 - 14. There is no evidence in BR

that either of them, or ‘Lemaire et al.’ collected in

the Cape Province, so this name should be excluded

from our list.

* Lennox, Fay (later Mrs Boik) (1924- )

b. Harding, Natal, 4 Sept. 1924; farmer and

nature lover; ed. privately and at Durban Girls’ Col-

lege until 1942. Interested in wild life and indigenous

plants and, although she has not collected exten-

sively herself, has assisted others, such as Mr L.E.

Taylor, Prof. Olive Hilliard and Mr B.L. Burtt, in

the Weza/Ingeli area. Married Reginald Bruce Boik,

farmer and business man, in Sept. 1977.

Specimens c. 200; in NU, E.

* Levring, Carl Tore Christian (1913-82)

b. Lund, Sweden, 21 Febr. 1913; d. Gallstad, Swe-

den, 30 Jan. 1982; phycologist; ed. Lund Univ., Fil.

Dr Docent 1940; Asst. Prof., Goteborg, 1942; head

of the Marine Botanical Institute, Goteborg, 1948;

Professor 1969.

Collected marine algae and visited S. Africa in

1947; specimens in GB etc. (IH 2,3: 437, 1972).

Commemorated in genera Levringia Kylin, Lev-

ringiella Kylin.

Ref.: pers. comm, from Mr Bo Peterson, Gote-

borg, July 1984.

* Liebenberg, Louis Christiaan Cronje (1900-1985)

d. Pretoria, 26 Feb. 1985.

Lindner, Otto (1852-1945)

b. Berlin, Germany, 10 Aug. 1852; d. Ypres, Bel-

gium, 16 Feb. 1945; technician. Adopted Belgian

nationality in 1888. He was sent by the ‘Etat

independant du Congo’ to Damaraland from Feb. to

July 1886 in order to recruit man-power and during

his stay he collected plants which he gave to the then

Director of the Jardin botanique de l‘Etat (BR).

Ref.: pers. comm, from Mr P. Bamps, Brussels,

Aug. 1984.

* Linley, John Stanley (1909- )

b. Caledon, C.P., 9 March 1909; pharmacist and

conservationist; ed. Cape Technical College, qual-

ifying as a pharmacist in 1931. Was part-time lec-

turer in the Dept of Pharmacy 1932-48, specializing

in pharmacognosy; Chief Pharmacist, Cape Town

City Council, 1935-65.

Collected several hundred specimens in the south-

western Cape Province mainly for (and often with)

Dr Joyce Lewis, hence largely Iridaceae but also Or-

chidaceae, Liliaceae, Amaryllidaceae and Erica-

ceae; accompanied Dr John Hutchinson for a day

during his visit to Cape Town, and represented the

Tercentenary Foundation in certain nature conser-

vation projects. Specimens in SAM.

* Louw, Hester E.

d. Kimberley c. 1977. Taught at the Preparatory

School, Beaconsfield, in the 1970’s.

Collected in the Cape Province, especially in the

Kimberley area; specimens in KMG (pers. comm,

from Mr A. Gubb, KMG).

Lowrey, Timothy Kemper (1953- )

b. Woodland, California, USA, 12 Aug. 1953;

botanist; ed. Utah State Univ. 1971 - 76, graduating

B.Sc, in 1974 and M.Sc. in 1976; and Univ. of Cali-

fornia, Berkeley, 1976 - 81, graduating Ph.D. Post-

doctoral Research Fellow, Ohio State Univ., 1981 -

82. Lecturer in Botany and Curator of the Moss Her-

barium, Witwatersrand Univ. from Oct. 1982.

Mainly interested in biosystematics of Compositae

and plant reproductive biology.

Specimens c. 300, from Cape Province, Natal,

Witwatersrand area and Savuti (northern Bo-

tswana); in J.

* McCIean, Alan Percy Douglas (1902- )

Dr McCIean was transferred to the Durban Bo-

tanic Station in 1926 but was placed in charge in 1928

when the then Officer in Charge, Dr H.H. Storey,

left for the Amani Research Station in Tanganyika.

Ref.: Schrire in Bothalia 14: 232 (1983).

McDonald, David Jury (1956- )

b. Florida, Transvaal, 7 Aug. 1956; ecologist; ed.

Natal Univ., Pietermaritzburg, graduating B.Sc. in

1977, B.Sc. (Hons) in 1978 and M.Sc. (Univ. of

Cape Town) in 1983. Appointed in 1981 to Botanical

Research Unit, BRI, Stellenbosch, and studying

phytosociology of mountain fynbos vegetation. Fig.

11.

644

Bothalia 15, 3 & 4 (1985)

Specimens c. 900, from northern Natal, Richters-

veld and south-western Cape; mainly in STE.

* Maguire, Brian (1922-83)

d. Johannesburg, 17 Aug. 1983.

Ref.: Davidson in Forum Botanicum 21: 88

(1983).

Marais, Elizabeth Maria (1945- )

b. Moorreesburg, C.P., 16 Oct. 1945; botanist;

ed. Stellenbosch Univ. 1964-67, 1976-81, graduat-

ing M.Sc. Lecturer in Botany at Akademie vir Ter-

siere Onderwys, Windhoek, 1981 - 83 and from 1983

Lecturer at Univ. of Stellenbosch.

Specimens c. 100, from south-west Cape, in

STE-U.

* Markotter, Erika Irene (1906-83)

b. Bethel, E. Cape, 21 Feb. 1906; d. Kuilsrivier,

C.P., 22 Sept. 1983 (pers. comm, from Mr R.O.

Moffett, Bellville, 1984).

* Meebold, Alfred (1863-1952)

b. Heidenheim, Wiirttemberg, Germany, 24 Sept.

1863; d. Havelock North, New Zealand, 6 Jan. 1952;

traveller and collector; travelled to India, Ceylon

and Burma from 1904 - 1912, collecting 18 000 num-

bers; to United States, Pacific Islands, Australia and

New Zealand in 1928 - 30 and 1932 - 33, visiting Na-

tal, Cape Province, Transvaal and Rhodesia in 1933,

during which period a further 10 000 numbers were

added. He maintained a private botanic garden on

his property in Heidenheim.

Commemorated in genus Meeboldia Pax &

Hoffm. (India).

Specimens in M, dupl. in HBG, NY etc.; some in

PRE.

Ref.: Pax & Hoffman in Pflanzenfam. 17b: 187

(1936); Suessenguth in Boissiera 7: 21 (1943); IH 2,

4: 521 (1976); Mr. R. O. Moffett, pers. comm. 1983.

Meyer, Helmut Ernst (1908- )

b. Kommagas, C.P., 8 Nov. 1908; horticultural-

ist; son of Rev. L. G. Meyer (q.v.). Appointed as

assistant to A. G. S. Herre (q.v.) 1930, under whom

he received 3 years training as horticulturalist. Ad-

ditional training, with emphasis on care of succulent

plants, in Germany at Tubingen (1933-1935), Ho-

henheim at Stuttgart (1935-1936), and Miinchen

(1937). Returned to Stellenbosch in 1938 where he

worked as horticulturalist in the botanic garden of

the University until his retirement in 1973. Under-

took collecting trips to Namaqualand together with

Herre, under whose name the specimens were cata-

logued. Most of the specimens thus collected are no

longer in existence. In his spare time grew Disa uni-

flora in his backyard greenhouse, where he pro-

duced numerous spectacular selections and hybrids.

Fig. 12.

Commemorated in Conophytum helmudi Lavis,

Lithops helmutii L. Bol.

Ref.: pers. comm, from Dr P. Vorster, Stellen-

bosch, July 1984.

FIG. 12. Back: H. E. Meyer, Luisa Meyer jr.

Seated: Rev. L. G. Meyer, Mrs Luisa Meyer

Front: Werner Meyer

Meyer, Louis Gottlieb (1867-1958)

b. Near Kleine Marpe, Dettmold, Germany, 16

Oct. 1867; d. Stellenbosch, Sept. 1958; missionary;

initially trained as agriculturalist; arrived in South

Africa as missionary on 7 Nov. 1894 to practise in

Namaqualand, firstly at Kommagas and later at

Steinkopf, which also included the Richtersveld. In-

terested in natural history, especially plants, since

childhood, his mission station became headquarters

to Marloth on the latter’s visits to Namaqualand,

and they sometimes collected together. Meyer also

sent plants collected on his own to Marloth and to

Herre (q.v.), and insects to Dr H. K. C. Andreae

(q.v.) which are at present housed in the South Afri-

can Museum. Interned at Pietermaritzburg for a

time during World War I. After his retirement he

settled in Stellenbosch, where he died and was bur-

ied. Fig. 12.

Commemorated in the genus Meyerophytum

Schwant. and in Aloe meyeri Van Jaarsveld, Ana-

campseros meyeri v. Poelln., Cheiridopsis meyeri

N.E. Br., Conophytum meyeri N.E. Br., Euphorbia

meyeri Nel, Herreanthus meyeri Schwant., Lithops

meyeri L. Bol., Meyerophytum meyeri (Schwant.)

Schwant., Nelia meyeri Schwant., Ruschia meyeri

Schwant., Stomatium meyeri L. Bol.

Specimens from Namaqualand in STE (Tolken,

1971); BOL, PRE.

Ref.: pers. comm, from Dr P. Vorster, Stellen-

bosch, July 1984.

Bothalia 15, 3 & 4 (1985)

645

Meyer, Luise ( nee Olpp) (1873-1956)

b. Gibeon, SWA, 1 June 1873; d. Stellenbosch,

1956; daughter of Johannes Olpp (b. 5 July 1837)

who was one of the pioneer missionaries who landed

at Angra Pequina (Liideritz) in 1868; wife of Rev. L.

G. Meyer (q.v.) whom she often accompanied on

collecting trips. Fig. 12.

Commemorated in Conophytum meyerae

Schwant., Ruschia meyerae Schwant. Both type

plants were attributed to Rev. G. Meyer.

Ref.: pers. comm, from Dr P. Vorster, Stellen-

bosch, July 1984.

Meyer, Luise (1905- )

b. Kommagas, C.P., 1905; daughter of Rev. L.

G. Meyer (q.v.) and Luise Meyer (q.v.); she some-

times accompanied her parents on collecting trips.

Currently (1984) living in Stellenbosch. Fig. 12.

Commemorated in Conophytum luisae Schwant.

Ref.: pers. comm, from Dr P. Vorster, Stellen-

bosch, July 1984.

* Moffett, Rodney Oliver (1937- )

Is revising Rhus for the Flora of Southern Africa.

Specimens c. 3 600, mainly Rhus from all over South

Africa. Fig. 13.

* Morris, D.G.: see Mrs Grace Doreen Court.

FIG. 13. R- O. Moffett Florence N. Morris

Morris, Florence Nightingale (nee Seller)

(1897- )

b. Cradock, C.P., 25 May 1897; housewife and

plant lover; daughter of Rudolf Seller who was park

curator in Cradock; mother of Mrs Grace Doreen

Court (q.v.). Talented in music, she obtained Licen-

tiates (Trinity College, London) in both piano and

singing by the age of 18. After her marriage to Gil-

bert Morris they settled in Oudtshoorn until 1942

and became keenly interested in succulent plants, es-

pecially Haworthias. She carried on an active corres-

pondence and exchange with Grace Britten in Gra-

hamstown, F. R. Long in Port Elizabeth and W.

Triebner in Windhoek. Haworthia morrisiae V

Poelln. is named after her. Fig. 13.

Miicke, M. (fl. 1910)

Collected about 100 specimens between Walvis

Bay and Windhoek in 1910 and then collected more

extensively in East Africa.

Ref.: Strey, MS (1961).

Netshiungani, Eric N. (1952- )

b. Mauluma, Dzanani Distr., Venda, 1952; for-

ester; ed. Fort Cox College, Ciskei, obtaining the

diploma in Forestry. Appointed as a forester in

Venda Dept of Agric. and Forestry in 1977 and

transferred to Venda Herbarium, Tate Vondo, in

June 1978 as assistant to G. Hemm; later curator of

the herbarium. Specimens in Tate Vondo herbar-

ium, PRE, PRU.

Nicholas, Ashley (1954- )

b. Chingola, Zambia, 15 Aug. 1954; botanist; ed.

Natal Univ. Pietermaritzburg, graduating B.Sc. in

1976; B.Sc. (Hons) in 1979; M.Sc. in 1982. Ap-

pointed in 1981 as Curator of the Donald Killick

Herbarium, Dept of Forestry, Natal (CPF). Inter-

ested in Droseraceae and Asclepiadaceae in addition

to forest flora.

Collected c. 1 900 specimens in South Africa,

mainly Natal, in CPF, NH, NU, PRE, K, MO; c.

500 in Zimbabwe, in SRGH.

Nichols, Geoffrey Richard (1953- )

b. Nairobi, Kenya, 22 Oct. 1953; horticulturalist;

obtained Diploma in Agriculture, Cedara, 1974;

National Diploma in Horticulture, Durban, 1977;

National Diploma in Park and Recreation Admin.

1980. Horticultural scholar at Kirstenbosch, 1979.

Since 1982 with the Durban Parks Dept. Fig. 14.

Specimens c. 650, collected mainly in Natal; in

NH, NBG.

FIG. 14. G. R. Nichols C. G. Osbeck

Nilsson, Eva Birgitta Maria (1955- )

b. Boras, Sweden, 3 Sept. 1955; botanist; ed.

Stockholm Univ. 1977-80, graduating B.Sc. At

Kirstenbosch Sept. 1980 to Feb. 1981 studying Wa-

chendorfia.

Specimens c. 200; in S, NBG.

646

Bothalia 15, 3 & 4 (1985)

* Oates, Lawrence Grace (1943- )

h. Cape Town, 13 March 1943; nature conserva-

tionist; great-grand nephew of the naturalist-collec-

tor Frank Oates (q.v. Gunn & Codd 1981 ) and grand

nephew of Capt. Oates of Scott’s Antarctic Expedi-

tion; ed. Stellenbosch Univ. 1962-66, graduating

B.Sc. (Agric.), and Pretoria Univ. 1967, B.Sc.

(Hons) (Wildlife Management). Agricultural Exten-

sion Officer, Dundee, 1968-69; joined Division of

Nature Conservation, Transvaal, in 1969 and en-

gaged on an ecological study of the Hans Merensky

Nature Reserve, 1969-71; officer in charge of S.A.

Lombard Nature Reserve, 1971-73; Regional Re-

presentative (Management) from 1973, concerned

mainly with the balanced management of nature re-

serves in western Transvaal.

Specimens c. 1 000, collected mainly on the Hans

Merensky Nature Reserve, Letaba District. Trans-

vaal; in PRE and in a local herbarium in the Re-

serve.

Ornduff, Robert (1932- )

b. Portland. Oregon, USA, 13 June 1932; botan-

ist; ed. Reed College, Univ. of Washington, and

Univ. of California (Berkeley), 1949-61 graduating

A.B., M.Sc., Ph.D. On faculties of Reed College,

Duke University and, from 1963, Univ. of Califor-

nia, Berkeley. Director of the University Herbarium

and Jepson Herbarium, 1968-83, University Botan-

ical Garden from 1973, and Professor from 1969. In-

terested in reproductive biology, phytogeography

and chemosystematics, and specialist in reproductive

systems of heterostylous Cape plants. Publications

on heterostyly in South African plants (review); cy-

totaxonomy of Villarsia , Lachenalia, Cyanella, Hae-

modoraceae; reproductive systems of Wachendorfia,

Cyanella, Nivenia; taxonomy of Oxalis.

Visited South Africa in 1970-71, 1974 and 1981

and collected c. 800 specimens as well as living plants

for the University of California Botanical Garden,

Berkeley, from the south-western Cape; in UC.

* Ortendahl (Oertendahl), Ivar Anders (1870-1935)

Further ref.: Norlindh in Veld & Flora 70: 46

( 1984), with portrait.

* Osbeck, Carl Gustaf (1766-1841)

Portrait kindly provided by Mr Bo Peterson,

Goteborg and published with acknowledgement to

the Nordiska Museet, Stockholm. Fig. 14.

Ottley, Alice Mario (1882-1971)

b. Seneca Castle, New York, 20 Nov. 1882; d.

Chula Vista, California, 22 July 1971 ; botanist; A.B.

(Cornell Univ.) 1904; M.A. (Wellesley College)

1906; Ph.D. (Univ. of California) 1921. Assistant

and instructor in the Botany Department, Wellesley

College, 1907-19; Associate Professor 1922-34 and

full Professor from 1934 until her resignation in

1939. In 1925 she was Exchange Professor at Wit-

watersrand Univ. and collected about 600 specimens

in Transvaal and Mozambique, often together with

Margaret Moss, who was a contemporary of hers at

Wellesley College, and with R. G. N. Young.

Specimens in WELC. Also collected in the USA.

Ref.: IH 2,5: 630 (1983); comm, from Wellesley

College, Feb. 1984.

* Otzen, Max (7-C.1947)

b. in Schleswig-Holstein and came to South

Africa in the 1880’s. Moved to South West Africa/

Namibia when the diamond boom was on and was

later appointed to the board of directors of Consoli-

dated Diamond Mines. Retired to Cape Town where

he maintained an extensive collection of succulent

plants. See also Gunn & Codd (1981).

Ref.: Krueger in Veld & Flora 68: 87 ( 1982), who

took some of the photographs of Euphorbias pub-

lished in White, Dyer & Sloane, The Succulent Eu-

phorbieae, Pasadena 1941.

* Owen-Smith, Norman (1942- )

b. Durban, Natal, 10 Jan. 1942; ecologist; ed. Na-

tal Univ. and Univ. of Wisconsin, graduating M.Sc.

and later Ph.D. for a study on behavioural ecology

of the white rhinoceros. Lecturer, University of

Rhodesia (now Zimbabwe), 1975-78; research eco-

logist in Centre for Resource Ecology, Witwaters-

rand Univ. from 1979, currently studying interrela-

tions between vegetation and large herbivores.

Specimens c. 200, in NPB (Tolken. 1971).

Palmer, Antony Riordan (1954- )

b. Springs, Transvaal, 8 Aug. 1954; ecologist and

computer botanist; ed. Univ. of Natal, Durban,

1973-76, graduating B.Sc. (Hons); Rhodes Univ.

1979-81, M.Sc. with a thesis dealing with the vegeta-

tion of the Andries Vosloo Kudu Reserve. Ap-

pointed to Cape Dept of Nature and Environmental

Conservation from 1979. Interested in floristics of

conservation areas in semi-arid parts of the Cape

and the interpretation of satellite imagery for map-

ping and monitoring vegetation. Fig. 15.

Specimens c. 1 200, mainly from Great Fish River

Valley and the mountains near Graaff-Reinet; in

GRA, PRE.

* Papenfuss, George Frederik (1903-81)

d. Berkeley, California, 8 Dec. 1981.

Ref.: Forum Botanicum 20: 33 (1982); Taxon

31: 613-615 (1982).

FIG. 15. A. R. Palmer C. P. Peeters

Bothalia 15, 3 & 4 (1985)

647

Paterson, M. (//. 1926)

Specimens in MO (orig.), BR (IH 2,5: 651. 1983).

According to Dr Peter Goldblatt, Curator of African

Botany, MO (pers. comm.), this is a relatively small

collection of c.70 specimens acquired with the her-

barium of Adele Grant (q.v. Gunn & Codd. 1981).

The specimens were collected mainly in the Herma-

nns area in 1926 and 1927.

Peeters, Christian Paul (1956- )

b. Brussels. Belgium, 30 April 1956; took up resi-

dence in Johannesburg in 1971; entomologist; ed.

Witwatersrand Univ. graduating B.Sc. (Botany and

Zoology) in 1978, B.Sc. (Hons)^ (Zoology) in 1979.

At present part-time lecturer engaged on post-grad-

uate study in entomology at Witwatersrand Univ.

Fig. 15.

Undertook a survey of Bophuthatswana with two

of his fellow-students, Gerieke (q.v.) and Burelli

(q.v.), and collected c. 900 specimens; in J, BOL,

PRE and in a collection maintained in Mafikeng.

* Penfold, Oliver Crace (c. 1875-1954)

A dentist by profession who grew up in England

and qualified L.D.S. and R.C.S. in London in 1898.

Came to South Africa in 1900 and registered with

the S.A. Medical and Dental Council on 1 Oct.

1900. Practised in Cape Town until his retirement in

1940 and died on 16 September 1954. At first he col-

lected moths but soon turned his attention to plants

and was friendly with Dr G. J. Lewis, Prof. R. H.

Compton, Miss W. F. Barker and Capt. T. M.

Salter. He was a member of the Botanical Society of

S. Africa from 1919. His wife ran the tea-room at

Kirstenbosch for a number of vears until into the

1940’s.

Specimens in NBG, SAM.

Ref.; pers. comm, from Mr C. D. C. Dickson,

Cape Town; from his stepdaughter, Mrs D. E.

Clark. Durban; and from the Registrar, S.A. Medi-

cal and Dental Council. Pretoria; Codd & Gunn in

Veld & Flora 70: 67 (1984).

Perold, Sarie Magdalena (nee Lombard) ( 1928- )

b. Johannesburg, 19 May 1928; research techni-

cian; ed. Witwatersrand Univ. 1946-49, graduating

B.Sc. in anatomy and histology; Pretoria Univ.

1977-78 (2nd year courses in botany and zoology).

Worked in S.A. Inst. Medical Res. as technician,

1950-52; with a private firm of pathologists,

1952-56; scientific assistant (chemical pathology) in

Witwatersrand Univ. Medical Faculty, 1956-67;

from 1979 research technician. BR1. responsible for

electron microscopy and studies in Hepaticae (genus

Riccia). Married to J. M. Perold in 1953. Fig. 16.

460 specimens (Riccia), mainly from Transvaal, in

PRE.

Peter, Gustav Albert (1853-1937)

b. Gumbinnen, east Prussia (now in USSR), 21

Aug. 1853; died Gottingen. Germany, 4 Oct. 1937;

botanist; ed. Konigsberg Univ. 1870-74. After grad-

uating with a doctorate, he was appointed at

FIG. 16. Mrs S. M. Perold G. A. Peter

Miinchen Univ. where he and Prof. Nageli worked

on the genus Hieracium. From 1888 to T923 he was

Professor of Botany at Gottingen Univ. and. al-

though he did not publish a great deal, he built up

the botanic garden and was interested in compara-

tive anatomy and cryptogams. In 1891-93 he con-

tributed the Convolvulaceae, Polemoniaceae and

Hydrophyllaceae for Engler & Prantl’s Natiirl.

Pflanzenfam. In 1903 he travelled extensively in

Europe and in 1907 embarked on a two year world-

wide tour.

In July 1913 he left on an extended collecting tour

in Africa which lasted until Feb. 1919. In S.W. Afri-

ca/Namibia from 20 Aug. — 17 Oct. 1913, he col-

lected 540 numbers, travelling from Swakopmund -

Okahandja - Tsumeb - Waterberg - Windhoek -

Keetmanshoop and Fish River - Luderitz. In South

Africa from 24 Oct. — 29 Nov. 1913, he collected

about 1 000 numbers, in Cape Town (where he met

Dr Marloth), the Peninsula - George - Knysna -

Avontuur - Port Elizabeth - Graaff-Reinet - Pre-

toria. He then proceeded to Bulawayo - Victoria

Falls - Harare - Umtali - Beira and the Mozambi-

que coast. The rest of the time was spent in Tanzania

but unfortunately much of his material remained be-

hind in Dar-es-Salaam when he left and had to be

destroyed because of insect infestation. A second

journey to Africa was made in 1925-26 during which

he collected 75 numbers in S.W. Africa/Namibia

(Swakopmund and Walvis Bay) from 7 — 11 Aug.

1925. and 310 numbers in South Africa (Cape Town

and coastal ports) from 14 — 24 Aug., followed by

further extensive collecting in Tanzania. His studies

were presented in his ‘Flora von Deutsch-Ostafrika'

in Feddes Reprium Beih. 40. Part 1 (1929 — 38) and

40, Part 2 (1932-38). Fig. 16.

Apparently the specimens were numbered after

his return to Germany but he died before the num-

bering was complete. His herbarium of about 50 000

numbers was acquired by Berlin-Dahlem in 1936.

The numbering was continued in B, starting with

No. 50001, but not in chronological order because

the material had been sorted into families in the

meantime. Duplicates are in BR. GOET. K. MO,

WAG and possibly W.

Ref.: NotizbI. Bot. Gart. Mas. Berl. 13: 166

(1936); Schmucker in Ber. dt. bot. Ges. 56: (203)

(1938), with portr. reproduced herewith; Exell &

648

Bothalia 15, 3 & 4 (1985)

Hayes in Kirkia 6: 85 (1967); personal communica-

tions in 1983 from Prof. P. Hiepko, Berlin-Dahlem,

and Mr R. O. Moffett of Bellville, C.P.

* Pfeil, Joachim Friedrich Graf von (1857-1924)

Full title as given in Index Herbariorum (Collec-

tors) 2,5: 677 (1983).

Pienaar, Barendina (“Barnie”) Jacoba (nee De Vil-

liers) (1926- )

b. Somerset East, C.P., 26 Sept. 1926; botanical

technician; ed. Natal Univ., Pietermaritzburg,

1944-47, graduating B.Sc., U.E.D.; Pretoria Univ.

1982-83, B.Sc. (Hons.); presently studying South

African Vigna spp. for M.Sc. Taught geography at

Newcastle school 1948-52; temp, lecturer (geogra-

phy), Pretoria Onderwyskollege 1956 and at Univ.

College of Zululand 1960; Herbarium technician,

Univ. College of Zululand, 1963-66; Natal Herbar-

ium, Durban, 1979-81; National Herbarium, BRI,

from 1981. Married Dr Benjamin Pienaar (1923-81),

educationist, in 1949.

Specimens c. 50 leg. De Villiers in NU; c. 670 leg.

Pienaar in ZULU (50), NH, PRE, inch about 360

cultivated plants collected in Durban.

Radloff, Ellen Miranda (7-1982)

b. Kimberley, date ?; d. Kimberley, 6 July 1982;

physiologist; ed. Huguenot Univ. College, Welling-

ton, obtaining the Diploma in Domestic Science and

B.Sc. Awarded the Abe Bailey Travelling Scholar-

ship and took courses in nutrition, pathology and

public health at Cambridge Univ. and Bedford Col-

lege, London, obtaining the M.Sc. degree. With the

award of a Stirling Research Scholarship she at-

tended Yale Univ. for three years doing research in

respiration and circulation for which she received

the Ph.D. degree. Returning to S. Africa she lec-

tured in the Physiology Dept of the Medical School,

Witwatersrand Univ. 1936-43. At the request of

Fort Hare Univ. she established the Physiology Dept

and served as Professor until 1946 when she returned

to her home, Sekretarius Farm near Kimberley.

She collected specimens near Kimberley; in

KMG.

Ref.: pers. comm, from Mr A. Gubb, KMG.

Raitt, Lincoln Miles (1948- )

b. Somerset West, C.P., 7 June 1948; plant eco-

physicist; ed. Helderberg College, B.Sc. (S.A.) in

1970, B.Sc. (Hons) (Stell.) in 1971, M.Sc. (Stell.) in

1974. Technical Officer in the Botany Dept, Univ. of

Stellenbosch, 1971-80; from 1980 Lecturer, Univ. of

Western Cape. Research interest in ecophysiology

and reproduction.

Specimens c. 800, mainly S.W. Cape; in STE,

uwc.

* Rehmann, Anton (1840-1917)

Further ref.: Codd & Gunn in Bothalia 14: 1-14

(1982), with portrait. Dr Christian Puff kindly

searched through the collection in W for types of

species described by Szyszylowicz, but found none

except Greyia radlkoferi Szyszyl.

Prof. Dr Kornas of the Jagiellonian Univ.

Krakow, Poland, informs us (March 1984) that Reh-

mann wrote two books, in Polish, on his travels in

South Africa in 1875 - 77 and 1879 - 80, in which

information on his itineraries is recorded.

Reid, Clare (1955- )

b. Johannesburg, 20 Aug. 1955; botanist; ed. Na-

tal Univ., Pietermaritzburg, 1973-77, graduating

B.Sc. (1976), Hons (1977). Appointed to National

Herbarium, BRI, in 1979, and concerned mainly

with Cyperaceae and petaloid monocots, especially

Cyrtanthus. Fig. 17.

Specimens c. 800, together with several hundred

collected jointly with other staff members, in various

parts of South Africa and S.W. Africa/Namibia; in

PRE.

* Renny, Alexander Timbrell (1903-83)

d. Tzaneen, Transvaal, 28 June 1983.

Riley, Herbert Parkes (1904- )

b. Brooklyn, New York, 28 June 1904; cytogene-

ticist; ed. Princeton Univ. graduating A.B. (1925),

A.M. (1929), Ph.D. (1931); Hon. D.Sc., Univ. of

Kentucky, 1976. National Research Council Fellow,

Harvard Univ. 1932-34; Asst Professor, Tulane

Univ. 1934—38; Univ. of Washington (Seattle)

1938-42; at Univ. of Kentucky as Professor,

1942-56, Distinguished Professor, 1956-74 (Head of

Department, 1942-65), and, since his retirement in

1974, Distinguished Professor Emeritus. As Full-

bright Lecturer at Pretoria Univ. Aug. - Dec. 1955

and at Cape Town Univ. Dec. 1955 - May 1956.

Publications include Families of Flowering Plants of

Southern Africa , Univ. of Kentucky Press, 1963, and

The Aloineae, A Biosystematic Survey (with S. K.

Majumdar), Univ. of Kentucky Press, 1979.

Specimens collected in S. Africa in PRU, MEM;

also collected in USA.

Ref.: IH 2,5: 762 (1983).

Ringgold & Rodgers

Commanders in the United States North Pacific

Surveying Expedition (1853-56), comprising five

well-equipped ships. Leaving Norfolk, Viginia, in

June 1853, they spent seven weeks in Simon’s Bay

undergoing repairs and refitting. A substantial col-

FIG. 17. Claire Reid Max Schlechter

Bothalia 15, 3 & 4 (1985)

649

lection was made by the naturalist, Charles Wright

(q.v.) and to lesser extent, assistant naturalist, A.H.

Ames (q.v.). Some specimens are in PH, leg. Ring-

gold & Rodgers.

Ref.: Mears in Proc. Acad. nat. Sci. Philadelphia

133: 155 (1981) and pers. comm. April 1983.

* Robertson, Colin Charles (7-1946)

A Yale graduate, appointed in 1913 as head of the

Research Section, Dept of Forestry, Pretoria.

Specimens in A, PRF.

Ref.: Immelman, Wicht & Ackerman, Our green

heritage, Cape Town 1973; IH 2,5: 768 (1983).

Roux, Cecilia (1947- )

b. Pretoria, 31 Dec. 1947; mycologist; ed. Pre-

toria Univ., graduating B.Sc. in 1967 and B.Sc.

(Hons) in 1974, and Rand Afrikaans Univ., M.Sc. in

1978. Appointed as taxonomic mycologist in Plant

Protection Research Institute, Pretoria, from 1975.

Specimens c. 120, chiefly mycotoxic samples col-

lected in the Karoo, also some collections in S.W.

Africa and Mozambique; in PREM.

* Roux, C.H.D. {fl. 1935)

Lived in the Victoria West-Carnarvon area and

collected plants, especially mesems, for Mrs L. Bo-

lus.

Commemorated in Pleiospilos rouxii L. Bol., Rhi-

nephyllum rouxii L. Bob, Stomatium rouxii L. Bob

Specimens in BOL.

Ref.: pers. comm, from Mrs O’Connor-Fenton,

Bolus Herbarium.

Royle, John Forbes (1798-1858)

b. Cawnpore, India, 20 May 1798; d. Acton. Eng-

land, 2 Jan. 1858.

Surgeon in East India Co. A keen naturalist, he

was Curator of the Saharunpur Botanic Garden,

Punjab, 1823-31, in addition to his medical duties.

Professor of Materia Medica, King’s College, Lon-

don, 1837-56 and Secretary of Horticultural Society

of London. 1837-58.

Collected some plants at the Cape, no doubt dur-

ing voyages to India; also collected in India, Chile

and elsewhere; specimens in LIV. K, PH and else-

where (see also IH 2,5: 794. 1983).

Ref.: Taxon 5: 11 (1956); Desmond, Dictionary of

British and Irish Botanists and Horticulturalists, Lon-

don 1977; Mears in Proc. Acad. Nat. Sci. Philadel-

phia 133: 155 (1981) and pers. comm.

* Russell, G.E. (neeGibbs):seeGibbs Russell, G.E.

* Schirach, M.C.: see Heginbotham, M.C.

* Schlechter, Max (1874-1960)

Photograph provided by Dr P. Vorster and pub-

lished with acknowledgement to Mr H. E. Meyer,

Stellenbosch. Fig. 17.

* Schmitz, Marthe Odile (nee Ruch) (1935-82)

b. Strasbourg, France, 27 Nov. 1935; d. Maseru,

Lesotho, 1 April 1982; teacher and missionary; ed.

Strasbourg Univ., graduating Lie. Sc. Nat. in 1958.

Came to Lesotho in 1958 as Lecturer in Botany at

the University College of Lesotho, Roma, where she

gave courses in plant systematics, anatomy, ecology

and genetics, and was in charge of the herbarium.

She was acting head of the Department in 1960 and

1965. In 1962 she married Dr Gerhard Schmitz who

became Professor of Geography at Roma. In 1965

she gave up her post as lecturer but continued with

her studies, and had just completed a guide to the

flowers of Lesotho at the time of her death due to

injuries sustained in a motor accident.

Ref.: Talukdar in Forum Botanicum 20: 33-35

(1982).

Schrire, Brian David (1953- )

b. Johannesburg, 20 Oct. 1953; botanist; ed.

Univ. of Witwatersrand 1972-1977, graduating

B.Sc. (Hons); studying Desmodieae for M.Sc. at

Univ. of Durban Westville. Appointed to National

Herbarium, Pretoria, 1978 and from Aug. 1979 in

charge of the Durban Unit and Curator of the Natal

Herbarium. Fig. 18.

Specimens c. 1 500. mainly from eastern Trans-

vaal and Natal; in NH (mainly), PRE.

* Sevdel, Richard Heinrich Wilhelm (1885-1972)

d. Usakos, S.W. Africa/Namibia, 10 June 1972.

Specimens c. 2 000 in B, purchased between 1954

and 1967; dupl. in A, C, K, LD, MO, NY, PRE,

SRGH, US, WAG etc.

Ref.; pers. comm, from Prof. Hiepko, Berlin,

May 1983.

Skarpe, Christina (1946- )

b. Sweden, 23 March 1946; ecologist; ed. Uppsala

Univ. from 1966, graduating M.Sc. Appointed as re-

searcher in Institute of Ecological Botany, Uppsala,

studying interactions between large herbivores and

plants, with particular reference to the Kalahari and

Zimbabwe.

Specimens c. 450, from the Kalahari, in UPS, K,

PRE etc.

* Skead, David Michael (1937-1983)

b. Grahamstown, Cape, 24 Oct. 1937; d. near

Viljoenskroon in northern OLS in an airplane crash

while on survey work, 16 Oct. 1983; ornithologist;

ed. Natal Univ., Pietermaritzburg, graduating M.Sc.

in 1977 and D.Sc. (Potchefstroom University) in

1980. Worked as Ranger with the Natal Parks,

Game and Fish Preservation Board, mainly at

Giants Castle and Hluhluwe 1956-61; with the Divi-

sional Council of the Cape, at Rondevlei Bird Sanc-

tuary and Cape of Good Hope Nature Reserve

1962-64; with Transvaal Nature Conservation Divi-

sion since July 1964 and officer in charge of the Bar-

berspan Ornithological Research Station since 197 1 .

650

Assisted C. J. Ward with maintenance of the her-

barium at Hluhluwe in 1958 and has collected in the

Giants Castle and Barberspan areas, mainly in con-

nection with ornithological studies, with special re-

ference to the Yellow-billed Duck.

Specimens c. 700; in NU, PRE.

Ref.: Oatley in African Wildlife 37: 180 (1984).

FIG. 18. B. D. Schrire Deidre Snijman

Snijman, Deidre Anne (1949- )

b. Brakpan, Transvaal, 27 June 1949; botanist; ed.

Natal Univ., Pietermaritzburg, 1967-72, graduating

M.Sc. with a thesis dealing with aspects of growth

and differentiation of Nicotiana tabacum callus, and

U.E.D. Taught for a year in Natal and, from 1974,

Professional Officer in the Compton Herbarium,

Kirstenbosch; mainly interested in the autumn-flow-

ering bulbs of the winter rainfall region, especially

Amaryllidaceae, and has published a revision of the

genus Haemanthus. Fig. 18.

Specimens c. 750, mainly S.W. Cape; in NBG.

Stadler, Anne-Marie (1960- )

b. Pretoria, Transvaal, 14 July 1960; botanical

technician; ed. Pretoria Univ., graduating B.Sc. in

1982. Appointed to Botanical Survey Section, BRI,

Jan. 1983.

Specimens c. 150 from northern coastal Natal; in

PRE.

Stapleton, Clarence Cowper (1892-1974)

b. Ladismith, Cape, 19 March 1892; d. 28 Oct.

1974; forester; ed. at Diocesan College, Cape Town,

and after serving in World War I with the rank of

Lieut, in the S.A. Artillery, attended Edinburgh

Univ. 1920-22, obtaining the degree of B.Sc. (For-

estry). Appointed to the Dept of Forestry Sept. 1922

and was stationed at various centres: Transkei

1922-23, Knysna to 1929, Pietersburg to 1934,

Bloemfontein to 1939, Port Elizabeth to 1944, retir-

ing as Conservator of Forests, King William’s Town,

Sept. 1948.

Wrote Common Transvaal Trees (Dept Agr. &

For. Bull. 164, 1937) and Trees of the Knysna —

George Natural Forests (Knysna Publicity Assoc.

1949). Specimens in PRF.

Bothalia 15, 3 & 4 (1985)

* Stohr (originally Stohr or Stoehr), Frederick Otto

(1871-1946)

b. Brighton, England, 1 Dec. 1871; d. Cape

Town, 7 April 1946; medical practitioner; graduated

M.A. (classics) at Oxford Univ. and in medicine at

Guy’s Hospital, London. Came to S. Africa in 1900

with the R. A.M.C. during the Anglo-Boer War and

started collecting birds for the S.A. Museum. Took

part, as medical officer, in the expedition which sur-

veyed the border between the Congo and N. Rhode-

sia in the early years of the century and did research

on the tsetse fly for the Belgian government. About

this time he acquired a farm at Munshiwemba, in

central Zambia, to which he returned periodically,

though practising as a psychiatrist in Johannesburg.

Married the well-known pianist, Elsie Hall, in Lon-

don on 22 Nov. 1913.

Always a keen ornithologist, he also collected

some plants at Munshiwemba, Zambia, and a few in

the Transvaal in 1942. Specimens in PRE (Exell &

Haines in Kirkia 6: 102, 1967) and SAM (Tolken,

1971).

Ref.: pers. comm, from his son. Dr P.S. Stohr of

Cape Town, June 1981.

Swart, Jan Philippus Jakobus (1950- )

b. Bredasdorp, C.P., 24 May 1950; wood anato-

mist; ed. Stellenbosch Univ. 1971-74, graduating

B.Sc. (For.) and B.Sc. (Hons) in 1976. Appointed to

the Wood Anatomy Section of the Faculty of For-

estry, Stellenbosch Univ. Fig. 19.

Specimens collected in S.W. Cape; in PRE.

* Theron, Johannes Jacobus (1905-1980)

d. Pretoria, 27 Aug. 1980 (pers. comm, from Dr

L.A. Coetzer, Pretoria Univ., Aug. 1984).

* Thienel, Sister O. (fl. 1944)

Roman Catholic Sister from Mariannhill who stu-

died at Natal Univ., Pietermaritzburg, during World

War II and made a collection while a student. Speci-

mens in NU (pers. comm, from Dr K.D. Gordon-

Gray, Sept. 1982).

Thompson, Louis Clifford (1920- )

b. Haenertsburg, Transvaal, 29 Feb. 1920; farmer,

near Magoebaskloof. Was Flight Sergeant in the

R.A.F. and piloted torpedo bombers during World

War II. A keen naturalist, he has explored many

little known desert areas and mountain regions of

southern Africa and occasionally collects unusual

specimens or plants of horticultural interest for his

sister, Sheila (q.v.).

Collaborated in an interesting collection of over

100 specimens in the Baynes Mts, S.W. Africa/Na-

mibia, in July 1959, leg. Davies, Thompson &

Miller, in PRE.

* Thompson, Mary Fraser (later Mrs Rand)

(1941-82)

d. Cape Town 10 Dec. 1982 (Oliver in Forum Bo-

tanicum 21: 1, 1983; and in Veld & Flora 69: 30,

1983.)

Bothalia 15, 3 & 4 (1985)

651

* Thompson, Sheila Clifford (1917- )

b. Haenertsburg, Transvaal, 23 April 1917; nur-

seryman-farmer; daughter of Dr Thompson of Ma-

goebaskloof; ed. Witwatersrand Univ., graduating

B.Sc. (Hons). She maintains a nursery near Magoe-

baskloof in which she grows many indigenous plants,

including unusual species which she has collected

herself in eastern Transvaal, Zimbabwe and the

Cape.

Some specimens in J, PRE.

* Tosh, Patricia A. (later Mrs Mulder) (fl. 1950)

Studied at Natal Univ., Pietermaritzburg, in the

mid 1950’s and collected plants near her home at Ni-

bela Store and on the Nibela Peninsula (northern

end of Lake St Lucia).

Specimens in NU (pers. comm, from Dr K.D.

Gordon-Gray, Sept. 1982).

Tredgold, Edith Kathleen (later Mrs Macintosh)

(1900- )

b. Cape Town, 22 Feb. 1900; housewife, social

worker and museum technician; ed. Cape Town

University 1918-23, graduating M.A. (Botany).

Demonstrator in Botany Dept 1922-23; in 1927 mar-

ried J.C. Macintosh, Advocate, and settled in Gra-

hamstown until her husband’s death in 1943. Social

Worker 1944-53 and Technician in Geology Dept,

South African Museum, 1954-65. Published ‘A

guide to the rocks, minerals and gemstones of South-

ern Africa,’ 1976.

Made a small collection of plant specimens in

south-western Cape during an expedition from Ash-

ton to Hex River in 1922; in SAM.

Ref.: George & Dorothy Randall, We All Lived

Here, Queenstown 1977.

Tsuane, William {fl. 1928-68)

Assistant at Albany Museum Herbarium, Gra-

hamstown, where he was appointed in 1928 as mes-

senger and gardener. Retired Nov. 1976. His duties

included maintaining the display of wild flowers in

the entrance to the Museum.

Made some herbarium specimens, e.g. the rare

Iboza barberae N.E. Br., in GRA.

Turner, Beverley Joy (1960- )

b. Pretoria, 2 Nov. 1960; ecologist; ed. Natal

Univ., Pietermaritzburg, graduating B.Sc. in 1981,

B.Sc. (Hons) in 1982. Appointed to the Botanical

Survey Section of the BRI and engaged on ecologi-

cal studies in eastern Transvaal.

C. 200 specimens; in PRE.

* Van Dam, Gerhardus Petrus Frederick (7-1927)

Appointed as Lay Assistant in the Lower Verte-

brates Department, Transvaal Museum, Pretoria,

May 1908, and died in a motor accident in Pretoria

in 1927.

Made collections of various kinds for the Museum

while oh field excursions, including 645 plant speci-

mens, now in PRE (ex Transvaal Museum), many

recorded jointly with more senior staff members,

e.g. V. FitzSimons.

* Van der Merwe, C.P. {fl . 1908-38)

Entomologist in Dept of Agriculture from 1908.

Was stationed in Durban from about 1914-27 and

then in Pretoria until about 1938.

Specimens in PREM (Tolken, 1971).

Ref.: Schrire in Bothalia 14: 223 (1983).

Van Heerde, Pieter (1893-1979)

b. Cape Town, 30 April 1893; d. Springbok, Na-

maqualand, 15 May 1979; teacher and collector; ed.

Univ. of Cape Town, graduating M.A. (chemistry)

and with a teaching diploma. After lecturing at the

University and at the General Louis Botha Training

Ship, he taught at Wittedrif and Bredasdorp before

accepting the post of headmaster of Springbok

school in 1926, a post he held with distinction until

his retirement in 1952.

He was interested in natural history and particu-

larly in the succulent plants of Namaqualand, which

he propagated in his private garden and later on the

Hester Malan Nature Reserve, where he was em-

ployed from 1966 until his health failed in 1967. He

was always helpful to visiting botanists and passed

on his discoveries to specialists, especially to Dr

Louisa Bolus and Hans Herre. The localities re-

corded were sometimes deliberately misleading in

order to protect species in their natural habitat.

Commemorated in the genus Vanheerdea L. Bol.,

Namaquanthus vanheerdei L. Bol., Conophyllum

vanheerdei L. Bob and Astridia vanheerdei L. Bob

Specimens in BOL.

Ref.: pers. comm, from Mr E.J. van Jaarsveld,

Sept. 1983; Dr H. Glen, Dec. 1983; Mr C.D.C.

Dickson, Cape Town, March 1984.

1

FIG. 19. J. P. J. Swart

* Van Niekerk, Grace (7-1983)

d. Somerset West, C.P., July 1983; teacher and

botanical assistant. Gave up teaching to join the staff

of the Bolus Herbarium during the 1950’s and early

1960’s and did some work on the genus Aspalathus.

Returned to teaching for some years and again

J. van Rooy

652

Bothalia 15, 3 & 4 (1985)

worked at BOL for about two years in the early

1970’s.

Specimens in BOL.

Ref.: pers. comm, from Mrs O’Connor-Fenton,

Bolus Herbarium, Feb. 1984.

Van Rooy, Jacques (1953- )

b. Bloemfontein, OFS, 11 Oct. 1953; bryologist;

ed. UNISA, graduating B.Sc. in 1982; Pretoria

Univ., B.Sc. (Hons) in 1983. Appointed to the BRI

as Technical Assistant, 1978; Asst Agr. Researcher

from 1982, studying Archidiaceae, Bryaceae (Bryo-

phyta). Fig. 19.

Specimens c. 2 000, collected mainly in eastern

South Africa; in PRE (orig.), MO, NY, S, H.

Van Rooyen, Noel (1950- )

b. Nongoma, Natal, 25 Dec. 1950; ecologist; ed.

Pretoria Univ., graduating B.Sc. (Agric.) in 1973,

B.Sc. (Hons) in 1975 and M.Sc. in 1978. Lecturer in

Botany Dept, Pretoria Univ., from 1978, mainly in-

terested in classification of vegetation. Fig. 20.

Specimens c. 4 000, from Transvaal; in PRU,

PRE.

Van Wyk, Christina Maria (1955- )

b. Stellenbosch, C.P., 20 April 1955; botanist; ed.

Stellenbosch Univ. 1973-1980, graduating M.Sc.

Appointed to the Herbarium, Botanical Research

Unit, Stellenbosch, in 1980.

Specimens c. 1 000, from S. and S.W. Cape and

Namaqualand; in STE.

* Von Below, Irma Ina (later Mrs Booysen)

(1920-84)

d. Cape St Francis, C.P., 21 Jan. 1984.

Ref.: Oliver in Veld & Flora 70: 59-60 (1984).

* Vorster, Pieter Johannes (1945- )

D.Sc. awarded by Pretoria Univ. in 1979, not

Ph.D. as previously recorded (Gunn & Codd, 1981).

* Vorster, Thomas Butler (1948- )

Date of birth should be 31 March 1948, not 1931

as previously recorded (Gunn & Codd, 1981).

* Wager, Vincent Athelstan (1904- )

Correction: Dr Wager was transferred to the Na-

tal Region in 1946 and was stationed at the Natal

Herbarium until his retirement in 1966.

Ref.: Schrire in Bothalia 14: 223 (1983).

* Walters, Ian Basil (1917-1983)

b. Worcester, C.P., 23 Oct. 1917; d. Worcester,

11 Jan. 1983; dental surgeon; ed. Witwatersrand

Univ. 1935, Stellenbosch Univ. 1936 and Guy’s Hos-

pital, London, 1937^10, graduating L.D.S., R.C.S.

Served as a bomber pilot and later as an instructor in

the R. A.F. with the rank of Flight Lieutenant during

World War II and then returned to practise in Wor-

cester, C.P. Started collecting as a hobby in 1962 and

devoted most of his time to this occupation after be-

ing obliged to give up his practice in 1977 due to ill-

ness.

Collected mainly in western Cape, especially the

upper Breede River Valley; specimens c. 2 500, in

NBG, BOL, PRE.

Commemorated in Lebeckia waltersii Stirton.

Ref.: Rourke in Veld & Flora 69: 148-49 (1983).

Wanntorp, Hans-Erik (1940- )

b. Stockholm, Sweden, 8 May 1940; botanist; ed.

Stockholm Univ. where he is at present Lecturer in

Botany. Visited South West Africa/Namibia with his

wife in 1980 and collected 1 159 numbers; in S, PRE.

Wanntorp, Henni (nee Palson) (1943- )

b. Stockholm, Sweden, 15 April 1943; botanist;

ed. Stockholm Univ.; Curator of the Herbarium,

Bergius Botanic Garden, Stockholm. Collected in

South West Africa/Namibia with her husband, H. E.

Wanntorp (q.v.) in 1980.

FIG. 20. N. van Rooyen Michael Webb

Webb, Mabel Margaret (later Mrs Gerber)

(1918- )

b. Pietermaritzburg, 19 May 1918; teacher and

housewife; ed. Natal Univ., Pietermaritzburg, grad-

uating B.Sc. (1938), H.E.D. (1939). Taught in Natal

and Transvaal until her marriage in 1942, and finally

settled in Cape Town where she undertook part-time

work in the office of the Kirstenbosch Botanic Gar-

den.

Made a small herbarium collection while a stu-

dent; specimens in NU.

* Webb, Michael (1924- )

b. Durban, Natal, 18 March 1924; zoologist and

marine biologist; ed. Natal Univ., Pietermaritzburg,

1942-44, graduating B.Sc., later M.Sc. (1947) and

D.Sc. (1957). Appointed to Dept of Zoology, Fort

Hare, as Lecturer (1948-51), Senior Lecturer

(1952-56), Professor (1957-59); Lecturer in Zool-

ogy, Natal Univ. (Medical School), 1960-63; Univ.

of Durban Westville, Senior Lecturer 1964—74, As-

sociate Professor 1975; from 1976 Professor of Zool-

Bothalia 15, 3 & 4 (1985)

653

ogy, Univ. of Stellenbosch. Studies mainly in the

field of marine zoology. Also interested in horticul-

tural pursuits and in plants in general. Fig. 20.

Collected while a student at Natal University,

mainly in Natal and Transkei; specimens c. 500, in

NU.

* Weeks, G.T. (fl. 1915)

Started a collection of the Witwatersrand flora in

1915 and this formed the nucleus of the Moss Her-

barium, Univ. of Witwatersrand; specimens in J.

* Weimark, August Henning (1903-80)

b. Hammar, Narke Province, Sweden, 22 May

1903; d. Lund, Sweden, 12 June 1980. His first ap-

pointment was as assistant at the Botanical Museum,

University of Lund, in 1928.

Ref.: Almborn in Taxon 31: 616-618 (1982).

Weitz, Frans Manuel (1954- )

b. Bredasdorp, C.P., 1 Aug. 1954; botanist; ed.

Univ. of Western Cape, graduating B.Sc. in 1974,

B.Sc. (Hons) in 1976. Appointed to Univ. of West-

ern Cape as Senior Laboratory Assistant, 1975-79,

and as Lecturer from 1980. Undertaking a revision

of the genus Corymbium.

Specimens c. 400, from S.W. Cape; in UWC.

* West, Ethel (later Mrs Anderson) (c. 1870-1939)

b. London, England, c. 1870; d. Grahamstown, c.

1939; housewife and keen naturalist. Came out with

her mother and two brothers to Port Elizabeth,

where her elder stepbrother had a business, and

later married Thomas Anderson, who was manager

of the Riet River Syndicate near Douglas. While in

Port Elizabeth she was a member of the Eastern

Province Naturalists Society and E.P. Photographic

Society.

Collected over 500 specimens: (a) leg. E. West,

mainly in the eastern Cape and some around the

Peninsula, in GRA, BOL, K, NH, PRE; and (b) leg.

E. Anderson, mainly in Griqualand West, in BOL,

KMG, SAM. Her collecting register is in GRA.

Commemorated in Leonotis westiae Skan.

Ref.: pers. comm, from her nephew, Dr Oliver

West; Codd & Gunn in Veld & Flora 69: 145 (1983).

Westendorp, Gerard-Daniel (1813-68)

b. The Hague, Netherlands, 1813; d. Dender-

monde, Belgium, 1868; medical doctor in the Bel-

gian army. Collected some specimens around Cape

Town; in BR (pers. comm, from Mr P. Bamps,

Brussels, April 1984).

Whellan, James Arden (1915- )

b. Liverpool, England, 1915; entomologist; ed.

Liverpool Univ. 1934-40, graduating B.Sc. Served

in World War II with rank of Lieut. R.E.M.E. Gov-

ernment entomologist in Rhodesia (now Zimbabwe)

1947-67; Malawi 1967-74; Malaysia 1974-76. Settled

in Australia after retirement.

Specimens c. 5 000, collected mainly in Zimbabwe

and Malawi, and some while on holiday in South

Africa; in SRGH, PRE.

Commemorated in Euphorbia whellanii Leach.

* Wild, Hiram (1917-1982)

d. Johannesburg, 28 April 1982.

Ref.: Ernst in Vegetatio 51: 125-128 (1983), with

portrait and list of publications.

Willems, Liliane Francine: see Hosten, Liliane Fran-

cine

Williams, Benjamin Murray Burdon (1956- )

b. Cape Town, 20 Nov. 1956; horticulturalist.

After qualifying as a photogravure engraver, he

joined the staff of the National Botanic Gardens,

Kirstenbosch.

Specimens c. 300, from western Cape, in NBG.

*Williams, Charles R. Farmer (7-1980)

Was Chief conservator in the Indian Forest Ser-

vice and, after serving in World War II, he joined

the Faculty of Forestry, Stellenbosch Univ. as lec-

turer in forest botany 1951-59, and was largely re-

sponsible for building up their herbarium. After his

retirement he lived in Somerset West where he died

in 1980 (pers. comm, from H.C. Taylor and Prof.

D.G.M. Donald of Stellenbosch, 1982). Fig. 21.

Specimens in FFS, which Herbarium bears his

name.

Ref.: Codd & Gunn in Veld & Flora 70: 67 (1984).

FIG. 21. C. R. F. Williams

Williamson, Graham (1932- )

b. Harare (Salisbury), Zimbabwe, 7 March 1932;

dental surgeon; ed. Natal Univ. 1951; Witwaters-

rand Univ. 1952-57, graduating B.D.S. and, in 1982,

M.Sc. (Wits) for a dissertation on the orchid flora of

Zambia. From 1958-77 he practised in Zambia,

making extensive collections in this territory and in

Malawi, with special reference to Orchidaceae. In

1977 he joined the Anglo-American Corporation of

S. Africa as senior Dental Officer and was posted at

G. Williamson

654

Bothalia 15, 3 & 4 (1985)

Oranjemund, S.W. Africa/Namibia. Wrote Orchids

of S. Central Africa (1977), completed Orchidaceae

for the Flora of Witwatersrand (in preparation) and

is presently working on Orchidaceae for the FI.

Zambesiaca. In 1981 he was awarded the Harry Bo-

lus Medal by the Botanical Society of S. Africa. Fig.

21.

Commemorated in Habenaria williamsonii Cribb,

Stolzia williamsonii Cribb, Cardiochilos williamsonii

Cribb, Eulophia williamsonii Cribb, Euphorbia wil-

liamsonii Leach.

Collected 2 760 numbers in Central Africa (in K,

SRGH) in addition to many collections jointly from

time to time with L.C. Leach, R.B. Drummond and

B.K. Simon of SRGH; 550 specimens collected in

Diamond Area No. 1 and in Richtersveld (BOL, li-

chens in BM, PRE).

* Winkler, Dorothea Gudrun (later Mrs Stielau)

(1932- )

b. Masama Mission Station, Moshi, Tanzania, 6

June 1932; came to South Africa in May 1947; ed.

Stellenbosch Univ. 1951-55, graduating B.Sc.,

M.Sc. and later H.E.D. (Natal Univ.) 1974. Botan-

ical Assistant, Compton Herbarium, 1955-56; tem-

porary lecturer in botany. Natal Univ. 1957; teacher

of mathematics in Pietermaritzburg from 1974. Mar-

ried to Werner Johannes Stielau, Professor of Ani-

mal Science, Faculty of Agriculture, Natal Univ., 9

July 1958.

Specimens c. 200, Cape Province, in NBG.

* Wood, John Medley (1827-1915)

Further ref.: Schrire in Bothalia 14: 223 (1983).

Woods, Denis Hamilton (1911-1977)

b. Salisbury (now Harare), Zimbabwe, 1911; d.

Cape Town, 1977; businessman, mountaineer and

conservationist. On military service during World

War II. During his rambles he studied the indigen-

ous flora and fauna and wrote many articles in the

Journal of the Botanical Society of S.A., the Journal

of the Mountain Club of S.A. and African Wild Life,

championing the cause of conservation.

Specimens from S.W. Africa/Namibia, S.W. Cape

and Namaqualand; in NBG.

Ref.: Veld & Flora 64: 13 (1978), with portrait.

UITTREKSEL

Biografiese notas oor plantversamelaars, bykom-

stig tot die wat reeds in Botanical Exploration of

Southern Africa deur Mary Gunn en L. E. Codd

(1981) gepubliseer is, met inbegrip van 'n aantal ver-

samelaars nie voorheen vermeld nie, word verskaf.

REFERENCES

GUNN, M. & CODD, L. E., 1981. Botanical exploration of

Southern Africa. Cape Town: Balkema.

INDEX HERBARIORUM PART II, COLLECTORS. 1954-83.

Published in the Regnum Vegetabile series by the I.A.P.T.

(1) A-D (1954) by Lanjouw & Stafleu; (2) E-H (1957) by

Lanjouw & Stafleu; (3) I-L (1972) by Chaudhri, Vegter &

De Wal; (4) M ( 1976) by Vegter; (5) N-R ( 1983) by Vegter.

TOLKEN, H. R.. 1971. Index Herbariorum Austro- Africanorum.

Distributed in cyclostyled form by the S. African Associa-

tion of Botanists.

Bothalia 15, 3 & 4: 655-688 (1985)

The plant ecology of the farm Groothoek, Thabazimbi District. II.

Classification

R. H. WESTFALL*, N. VAN ROOYEN** and G. K. THERON**

Keywords: floristic classification, phytosociology, savanna, structural classificaton, vegetation

ABSTRACT

The vegetation of the farm Groothoek, Thabazimbi District, situated in the Sour Bushveld of the Waterberg,

Transvaal, is classified according to the Braun-Blanquet method, using the PHYTOTAB-program package. Five

major vegetation types with 18 communities are described with reference to the main environmental factors influ-

encing vegetation composition and structure.

INTRODUCTION

The plant ecology of the farm Groothoek, Thaba-

zimbi District, was studied in order to supply data on

the Sour Bushveld (Acocks, 1975) for the natural re-

source classification of the Department of Agricul-

ture & Water Supply. This veld type is found mainly

in the Waterberg of the Transvaal and little is known

about the vegetation. The study included classifica-

tion of the vegetation in terms of both floristics and

structure, as well as correlation of the vegetation

classification with the environment to facilitate later

study of the entire Waterberg area.

STUDY AREA

The study area consists of the farm Groothoek 278

KQ situated between 24°28' and 24°31' south lati-

tude and 27°32' and 27°39' east longitude. The

farm covers approximately 4 000 ha over an altitudi-

nal range of 1 200 m to 2 100 m. The summit of

Kransberg forms the northern boundary with the

greatest farm area consisting of two plateaux at ap-

proximately 1 500 m and 1 200 m altitude to the

south. Bakker’s Pass lies to the west allowing vehicu-

lar access to the upper plateau.

The Kransberg Massif consists of sandstone from

the Kransberg Series of the Waterberg System. Out-

crops of sandstone, shale and conglomerate from the

Nylstroom Series of the Waterberg System, are

found on the upper plateau while sandstone out-

crops are found on the lower plateau. A diabase

dyke, of post-Waterberg age, is exposed in Bakker’s

Pass. The soils arising from these parent materials

are mainly of the Mispah Form, Mispah Series. The

topography and a north-south profile of the study

area are shown in Figs 1 & 2.

The study area is representative of Acock’s (1975)

Sour Bushveld in a relatively undisturbed condition

with sufficient variation in vegetation and environ-

ment for observations to be relevant to other areas

in the Waterberg where this veld type occurs.

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

** Department of Botany, University of Pretoria, Pretoria 0002.

CLIMATE

The Waterberg is classified as Cwa according to

Koppen’s classification (Schulze, 1947) although this

is not supported by direct evidence as no full clima-

tological station has ever been maintained in the

area. Rainfall statistics indicate the area as falling

under the category Cw, warm temperate with sum-

mer rainfall, while the high probability that the

January mean exceeds 22°C indicates the Cwa classi-

fication (Schulze, 1947). Rainfall records for the ten-

year period July 1963 to June 1973 on the upper pla-

teau suggest a mean annual rainfall of 680 mm. The

consensus of local opinion, however, is that this

period was unusually dry.

METHODS

The Braun-Blanquet method of sampling and syn-

thesis followed in the study is described by Westhoff

& Van der Maarel (1973), Mueller-Dombois & El-

lenberg (1974) and Werger (1974).

The study area was stratified into physiognomic-

physiographic units determined by interpretation of

1: 33 000 aerial photographs. Sampling sites were lo-

cated by means of random co-ordinates within each

physiognomic-physiographic unit. This procedure is

accommodated within the flexible Braun-Blanquet

method (Coetzee, 1975). 170 quadrats, 10 x 20 m,

were sited in homogeneous stands representing the

different physiognomic-physiographic units. The 10

m x 20 m quadrat size used is considered adequate

for bushveld vegetation (Coetzee et al ., 1976; Van

der Meulen, 1979).

Data recorded at each sampling site include the

species present in each quadrat with estimated can-

opy cover-abundance according to the Domin-Kra-

jina scale (Mueller-Dombois & Ellenberg, 1974), ve-

getation formation, altitude, slope, aspect, soil form

and series, soil depth, water-table depth, chemical

soil characteristics of the A-horizon, lithology and

estimated percentage rock cover.

The vegetation was classified according to the

Braun-Blanquet method using the PHYTOTAB

program package (Westfall et al., 1982a). The main

environmental factors influencing the communities

were ordinated on floristic data by detrended corre-

27 32 27 34' A 27 36 27 38

656

Bothalia 15, 3 & 4 (1985)

E

-SC

<N — i

b

"3

XI

P

-C

C3

H

O

x:

o

O

6

a

<D

-X

o

<0

I 1 S

ra Q. >-

.EE 05

5) ro

C «

c

(0

a> 0)

>»

■D

3

to

o O

>. 2*

<5 <2

■o

c

_ C T3 —

3 3 ™ O

O O O C

CD m DC Q.

j

i

I

Bothalia 15, 3 & 4 (1985)

657

LU

o

111

i $

^ X /n t>=

«3 «3 i i

CO < <

CO

o

ui r O cd

CD _ UJ 5

O) CD O 5

Z O N

< Z < <

CD < w D

at (/) <

E

J£

CM

(S|BAJ9JU! iu 09)

aanimv

658

Bothalia 15, 3 & 4 (1985)

TABLE 1. — A classification of the plant communities of the farm Groothoek,

Thabazimbi District

A. Kloof Forest communities on moderately deep soil in moist sheltered habitats.

1 Erythrina lysistemon-Celtis africana Kloof Forest

2 Osyris lanceolata-Celtis africana Kloof Forest

3 Asplenium splendens-Celtis africana Kloof Forest

B. Woodland Phase of Acocks’s ( 1975) Sour Bushveld, on moderately deep to

deep soils, in moderately exposed habitats.

4 Panicum maximum- Combretum molle Closed Woodland

5 Euclea crispa- Combretum molle Closed Woodland

6 Setaria megaphylla- Combretum molle Closed Woodland

7 Terminalia sericea-Combretum molle Closed Woodland

8 Aristida diffusa— Combretum molle Open Woodland

8. 1 Strychnos madagascariensis- Aristida diffusa- Combretum molle Variation

8.2 Vitex rehmannii- Aristida diffusa -Combretum molle Variation

9 Landolphia capensis- Combretum molle Closed Woodland

9.1 Burkea africana- Landolphia capensis- Combretum molle Variation

9.2 Tapiphyllum parvifolium- Landolphia capensis-Combretum molle Variation

10 Coleochloa setifera— Combretum molle Open Woodland

1 1 Heteropogon contortus-Combretum molle Closed and Open Woodlands

11.1 Rhus dentata-Heteropogon contortus-Combretum molle Closed Woodland

Variation

1 1.2 Chaetacanthus costatus- Heteropogon contortus-Combretum molle Open

Woodland Variation

12 Themeda triandra- Combretum molle Open Woodland

13 Argyrolobium transvaalense-Combretum molle Open Woodland

14 Pachy carpus schinzianus-Combretum molle Open Woodland

15 Protea caffra-Combretum molle Open Woodland

C. Grassland Phase of Acocks’s (1975) Sour Bushveld on moderately deep soils

in exposed dry habitats.

1 6 Eragrostis pollens- Andropogon appendiculatus Grassland

D. Woodland Phase of Acocks’s (1975) North-Eastern Mountain Sourveld on

moderately shallow soils in moderately exposed habitats.

1 7 Helichrysum nudifolium-Protea roupelliae Sparse Woodland

E. Grassland Phase of Acocks’s (1975) North-Eastern Mountain Sourveld on

shallow rocky soils in exposed habitats.

1 8 Eragrostis racemosa- Trachypogon spicatus Grassland

spondence analysis (DCA) (Hill & Gauch, 1980) us-

ing the DECORANA program (Hill, 1979; Westfall

et al. , 1983). Syntaxonomic nomenclature is accord-

ing to the preliminary rules and recommendations of

the South African Syntaxonomic Nomenclature

Committee.

Total canopy cover was estimated for each qua-

drat within the eight height classes used by Van der

Meulen & Westfall (1980) for bushveld vegetation.

The height classes are independent of vegetation

height. Structure is illustrated by means of layer dia-

grams (Ito, 1979) giving mean percentage cover

within each height class. The structural features are

summarized according to Ito (1979) by grouping the

eight height classes into three strata, namely a

herb/grass stratum (0-1 m), a shrub stratum (>l-5

m) and a tree stratum (>5 m), and determining the

highest mean percentage cover in each stratum. The

following symbols, determined by the highest mean

cover in a stratum, are used to classify each layer

diagram type:

Layer diagram type Cover of strata

L - type herb > shrub > tree

rL - type herb < shrub < tree

D - type herb < shrub > tree

C - type herb > shrub < tree

I - type herb = shrub = tree

RESULTS

Classification of the vegetation, according to the

Braun-Blanquet method, revealed 21 (possibly 22)

plant communities within five major vegetation

types (Table 1). The North-Eastern Mountain Sour-

veld Communities (Acocks, 1975) are outliers of this

veld type and are found above 1 500 m altitude on

the Kransberg massif. The phytosociological classifi-

cation of communities is shown in Tables 2 & 3 with

Table 2 showing the diagnostic species and Table 3

showing the general and infrequent species. Symbols

denoting environmental parameters are shown

above the matrix in Table 2 and explained in Appen-

dix A. Species with the highest mean percentage

cover per community are shown above the matrix in

Table 3. Percentage constancy and mean percentage

cover together with the densities of trees greater

than 2 m tall per releve and per hectare, are shown

to the right of the matrix in both tables. The taxa on

the left of the matrix, in both tables, are grouped

into simplified life-form classes to facilitate the veld

condition assessment of the vegetation (Westfall et

al ., 1983).

The spatial relationships of the communities are

shown in the form of a vegetation map (Fig. 3).

. Thabazimbi District, with diagnostic species

■gusli

I till

i'll-

,|l!i |

I If

-Ilf

k ft; ^

? Q

* v

(C) sedge; (F) forb; (G) grass; (GEO) geophyte.

658

spc

ing

et i

ing

the

Co

dn

Me

The uei£

height. St

grams (Ii

within eai

summariz

eight heij

herb/grass

m) and a t

highest me

following ;

cover in a

diagram ty

Layer d,

L

rL

I

no

TABLE 3. — Phytosocio logical classification of the farm Groothoek, Thabazimbi District, w ith dominant and non-diagnostic species

%9‘8

smvoids uoSodXifDoug

%6‘frZ snuoiuoo uoSodojajafj

« k.

10 f) iflO

«o -~i

o' O*

— <N

o' o'

<N —

V

o o

V V

m (N — — — < _

V V V v

o o o o o o

V V V V V v

V V

o' o' o' o' <N o" o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o' o* o'

v v V V VVVVV V VVVVV V V V V V V V V V V V V V V V V V V V V V V V V V

'O t- — "l o' o' o' o' o' o' o' o' o' o'

c o

V V

o o o o o o o

V V V V V V V

o' o' o' o' o'

V V V

VO VD OJ ^ —

ri ^ x mg Tt n -

O oc oc oc oo r"

NOOir)Tj-'tfnmriMrirN(N

<N ^ rsi

r- <n

r- Tt

VO Tf

r- —

vo —

vO co

VO

vo

r- u-i

VO vo

r- m

r- O

vo vo

VO O

vO 0C

04 *—

oi oi

oj — +

oi ~

04 -<t

— < r-

— < o

04

— ON

0 ov

— < vo

— • VO

*— • ^t

04 — '

04 04

04 vo

(N vo

01 co

04 oi

oi ^

04 —

— ' 04

CM »-H »—

— < 00

04 o

Tj* fO

%0‘S

VSOW90DJ sijsojSdj j

%e>

VSOW3DVU SUSOjSVjq

Tt

— « vo

— « Ov

CO — <

oi O'

— < vo

*— 04

%8‘S

VSOUI99VJ SliSOjSVUTJ

%0‘£I

DUpUDJJl DpdlUdllX

%8‘6 snuoiuoo uoSodouoiopf

o/0?‘hI snuoiuoo uoZodojoiofj

04 ro

cm 04

— 00

04

<0 04

m ro

04 OO

04 VO

co O

04 VO

C\l ON

fO 04 CO

VO Tt

VO VO

ON *“<

On Tt

On 04

On CO

On vo

04 On

04 00

04 r-

CO 04

VO oo

CO O

04 04

VO 04 04

CO ’<st*

CO lO 04

— 04

Tt On

rt O

VO VO

VO O

vo On

vo O'

vo O

vo 04

04 04

Tf 04

^ CO

%9‘ei

vuvoup voo/ung

vo O

rj- On

vo On

04

O'

Tt vo

CO ON

Tt fO

vo ^

Tt 04

CO 04

04 04

CO 04 CO

CO

%0‘Z

3]jow tuniduqiuoj

vo 00

vo CO

vo 04

VO Tt

vo O-

VO — «

VO vo

vo VO

CO 04

04 04

CO 04

04 04

%S‘fr

3]]oui mnidJiquioj

Tf vo

CO Tf

co CO

CO — H

Tf 04

CO 00

Tt CO

Tt

co r-

CO VO

co VO

vo O

04

— I CO

%6‘sn

vuvoufv voyjng

^ o

ON O'

rj* vo

Tf —

CO 04

co On

Tf- Tf

Tf oo

%e£e

3]]oiu uimsuqiuoj

O Tf

ON VO

oo r-

On O

00 On

OO VO

CO CO

04 04

CO 04

%0‘S

vsnjjip vpijsuy

% T ‘83

do ottos mpmuudg

%rei

vn^ndvSdiu muviog

OO — I

OO 00

00 co

00 Tf-

00 vo

00 04

r- O'

00 O

r- on

04 04

ON ON

O 04

o —

O CO

%s‘s

d\iow uinjduqiuoj

%I *L

diiotu mn}djiqmo{ j

%0tS

vuvoujv Slipj

CO ’’st"

vo ^

VO

rf —

CO 00

Tf Tt

Tt 04

CO to

vo vo

CO O'

co co

CO \Q

T t 00

— *

co 04

CO 04

co CO

O O'

^ CO

O 00

O VO

O' vo

O VO

— O' CO

%8‘K

vjvjodouv] sui(SQ

r-' O

\D —

VO O'

VO 04

vo oc

vo ON

vo

vo vo

CO

04

co vo

co 04

co VO CO

CO VO 04

CO

VO 04

%£‘Zg snijo/poi snduvoopoj

o o

O' oc

c c

c3 3

JD k,

au o

S a

* 3

z

I

Q S3

c ^3

§ 5

Co Co

r3 W

Co Co

1 1

II

.S3

| S3

,11

•« P

^ V v

cq cq ^

sj So

Sj _

3 § S3

S3. C'.S

Q "O jV*

'C S ^

5 0“

•a ss -s

g 8 §

CD co

§•&!

•' -2 .3

V r- '2

•g •§ ■§ S

Sj.asi

* « 3

S) 3 ^

S

3 .0

^ Sj

S3

3

.SJ

O

Is

?o;S5 «

X a, O Q si.

Sk c

§ §

'C |

5 X!

sS S

0 >5

QC §

3

2. ,, Si

■s gl

1 15

'ei -c a ai ^ s

p

C § sa;

-5 v; vi fC to

ss -3 a

<3 Q

N Ct,

C Sog

S 8 c

l.s§

K "sj

5^3

Cq < O

‘C

8 5

d§ ?

S« 5 S3

5 2 -a S S

s 3 ^ '

K

Q

C

•a -«

s a

B.-S

c t;

a. o

O '

■a ^ a, ss

^ § 9- &

i co

•H ^

a 5 'C

35^

ist

SJ, S';

« O N

g Si s

5 Co 2

o c

Si R Si

-O "5 -o

s Ks ■

r° ^ r°

Cj OC Cj

>s <:

s? c

5 A

B ss o, .2

a H 3 Si

t? SS SU Si

S3 « ^ °

■S -S

oo >

g I

5 o

1 i-S

r, S S3

P rS3 Si

S S 3

lls

3 ~

5-3

S-2

b§

Its

ae b <S,

<u § 3? -~

S &5 C

Si ss q a

2

5 5

3

S3 02

3 Si

3 so sj

5 5 S

OSS

S3. -3 O

P; cs.-p

3J Si ^r

&

^ Si

o

S3

•§ O S3

^ t; -3

•2

2 5

5^5

X: cj,x»

S3 3

cq X cq

•2 -2

Si 13

8.*

CV 0

_ 3

C Co Si V.

2 ss o si Si

P -2

Si I S)

< v 5

C/O

<

d

O

CD

S3 5

^5

•a s

co

y

3 ? 3

O 5 CJ

s-

0^0

S S.S

^•3

o 5 S

fc | 8

S3 3 -j-

3 0 sfc,

S3- be S3

<*> .8 GO

S3 &C

S3 Si

3

cti .S

sc 2

?5 U

2-2^3

3 .ss o -a ~~

£ g •§ fc -S

Jk Cl. ^ Si

o *4

§••2

•Q 3

K *3

5 ,S»C

D'q

■s ^

3 a -2

?c -a a

£ t:

o

s ^

3 ■§

•2 §

11

IS® §

Q Cj'

X! S3

S3 Si

3 c a.

3 X, -2

-3 .Sc

o c

1 S

.Co CD

C CL

o

o o

3. 3.

O S'O S3

_ S: | i

it -g i

3 3 b C

k_- a x 5

si QQ "X <0

0 „ 3

§ « s .c

■jS -a 2 ^ -s

s^«8 !•&?

5

S3 g ,

cj v; ^ Q

^ C tt -5 ^

S 5 ’C s rg

5 -S *? •

ca

O ^

3

.so .*> .2 S

g ? g. 3

a. 3 o a.

o

.3

■J- N =0

c 3 .3 -3 3;

-S Si "3 3- Si

5 ^ N P S3

X Si 8 ^ ~

-3

a.

2 5

X

H

o

Si 3

o Si

02

so .O

>5 §

O 03

V. VS 5 2 ^

o'^'a.'C .3

|| l|8

o?

o c

S '

o

.CO

L> ^

^ rc:

cl co

rr- r>»

S O -2 O

^ -5:

.0

2 -2 -5:

c a 5

Sj 3 8 3.

3 so 2

8-2-fe

5 S .8

-3 Si

^ ^ -3 -2

$ C cs: =2

2X2

« 1 3

g* -o

•2 s s

3 -a

O SI

5-g S

£ £ N

0 a a

1 ^ 2 g

C S3 .S3 Sso

S Si N 3

- v Ss Si

2 3-3

5 ro p

U D 'S

o 5 2

C 3 -3

Si S3 S3

S cv ^

S .s

3 !S

3 8

. 5 fc

boo

OJ 01 oq

§

ROgs,

3 Sj 3 a

?s 5 -2

~ ’§••&

jr 65 3

•3 q Si

■a §^,5

| jo ■§

R -3 -3

S 3

S3

lii-sl

3 S3 i ■?

P ss X! ^ t:

5 3. a 5

? Si S3

85

3^8

O X!

A 5

3

0

3 -3

8 & S 3 £

•sj Cj 05 CO ^

•C: ^ s: c:

« g >f Si -3

3

X) S

.0 .3

5 S

s i" &

s|.s

•S .2 3

^£5 5

•3Sg:

X. -C. pN

3 g-

■§5 5

05 S- *S

1 |

l|

t.g

.5 C

^ .s

^ Sr5 -S. ^

s o

I ^

5 ^

-V ^

•2 •- S

£ ^

so Go

•2 -a ^ si

3 Si

•e -s

3

.55 5

I

g 8 |

o5 g

Sl|

3 2 §

CD ^

>: vj ^

subsp. barbertonensis

658

sp

in

et

in

tli

C

di

M

Tne m

height,

grams

within i

summa

eight h

herb/gr;

m) and

highest

followin

cover ii

diagram

Layi

663

Bothalia 15, 3 & 4 (1985)

The structural classification together with each

layer diagram type is shown in Fig. 4.

DESCRIPTION OF THE PLANT COMMUNITIES

In the community descriptions woody and herba-

ceous species are both listed in order of constancy

followed by mean percentage cover, with the re-

spective values indicated next to each species.

Species characteristics of each community are

omitted from the descriptions because they are di-

rectly apparent from Table 2.

1. Erythrina lysistemon - Celtis africana Kloof

Forest, found below 1 101 m altitude on sand-

stone,

2. Osyris lanceolata-Celtis africana Kloof Forest,

found at altitudes of 1 300 to 1 400 m on dia-

base,

3. Asplenium splendens - Celtis africana Kloof

Forest, found at altitudes of 1 600 to 1 850 m

on sandstone.

A. Kloof forest communities on moderately deep

soils in moist sheltered habitats

The kloof forest communities are found in kloofs

in the east (Communities 1 & 3) and in the west

(Community 2) of the study area (Fig. 3) at altitudes

of 1 050 m to 1 850 m (Table 2). The vegetation is

represented by three communities, namely Com-

munities 1 to 3 (Tables 2 & 3).

Habitat

The soils are mainly of the Mispah Form, Mispah

Series, and Shortlands Form, Bokuil Series, derived

from sandstone and diabase respectively. The soil

depth varies from 140 mm to 470 mm. The kloofs are

the least exposed of the landform classes (Appendix

A) found in the study area and as a result probably

have the narrowest temperature range and highest

humidity of the communities in the study area.

Streamflow in the kloof communities is season?'

Floristics

Although the kloof vegetation is physiognomically

homogeneous, the communities have few species in

common, viz the species of the Diospyros whyteana

species-group (Table 2F). The Olea europaea

species-group (Table 2 C) is common to the first and

second communities (Communities 1 & 2) and the

Myrsine africana species-group (Table 2 E) is com-

mon to the second and third communities (Com-

munities 2 & 3) (Table 2)

The Cyperus albostriatus species-group (Table 2

H) is the only species-group the kloof communities

: (Communities 2 & 3) have in common with the other

main vegetation types in the study area. This

species-group contains only two large woody plants

viz Secamone alpinii , a shrubby climber, and Ptero-

celastrus rostratus whose ecological amplitudes sug-

gest that they occur on forest margins. Celtis africana

is diagnostic and physiognomically conspicuous in

the kloof communities. Podocarpus latifolius is non-

diagnostic (Table 3) but physiognomically conspicu-

ous in all the kloof communities and also in portions

of the main vegetation type B and occasionally in

types D and E. An example of kloof forest vegeta-

tion in the west of the study area (Community 2) is

shown in Fig. 5.

In the phytosociological classification, the kloof

forest communities are classified as follows (Tables

1, 2 & 3):

1. Erythrina lysistemon - Celtis africana Kloof For-

est

This forest is found below 1 101 m in a deep kloof

in the south-east of the study area (Fig. 3). It is rep-

resented by releves 98 and 100 with 17 and 21 species

per releve respectively. This forest community (Ed-

wards, 1983) has an rL structure (Ito, 1979; Fig. 4a)

with the greatest average cover of 65% in the upper

height classes higher than eight metres. The kloof is

subjected to cold air drainage (Coetzee, 1975) and

being situated lower than the other two kloof forest

communities, is not only cooler at night but can

probably reach higher day temperatures resulting in

a greater temperature range than the other kloof for-

est communities in the study area.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Alma Graywacke

Formation. The average soil depth varies from 190

mm to 450 mm and the surface rock cover averages

60%. The kloof slopes from 3° to 6° in a south to

south-west direction. The electrical resistance of the

soil is the highest in relation to the other kloof com-

munities and the T-value is the lowest (Table 2), in-

dicating nutrient-poor soils. This may be attributed

to the low altitude with consequent greater stream-

flow than the high-lying kloofs. The nutrient-poor

soils may also be attributed to periodic flooding as

the quadrats were placed below the observable

floodline when sampling. The soils are strongly acid

(MacVicar et al., 1977) with a pH of 4,8 when satu-

rated with water.

Floristics

The community is diagnosed by the Plectranthus

verticillatus species-group (Table 2A). The species

diversity per unit area is low for the study area with

an average of 4 species/m: for this community.

Trees and shrubs

Conspicuous woody species with more than 5%

mean cover and occurring in more than 50% of the

releves representing this community are:

664

Bothalia 15, 3 & 4 (1985)

SSVd S.NIXNVd 01

} se/v\ o9 uojjeuiioap oi}au6eiu 36ejaA\/ ^

flotation map of the farm Groothoek, Thabuzimbi District.

diabase: post-Waterberg Group. Watcrberg Group: S3, Sandriviersbcrg I ormation; S2 & shale, Aasvoelkop Formation; SI & Conglomerate, Alma Gray wacke Formation.

666

Bothalia 15, 3 & 4 (1985)

Herbs

Herb species occurring in more than 50% of the

releves representing the community are:

Plectranthus verticillatus (forb) 100% 0,50%

Glycine wightii (forb) 100% 0,28%

Mohria caffrorum (fern) 100% 0,05%

General

Communities 1 & 2 are related to each other

through the common presence of the Olea europaea

species-group (Table 2C) and Communities 1, 2 & 3

are related to each other through the shared pre-

sence of the Diospyros whyteana species-group

(Table 2F). Community 1 has no species-group in

common with the other main vegetation types found

in the study area, which may be attributed to the low

altitude of Community 1 and the surrounding vege-

tation not having been sampled. It is, therefore, sug-

gested that this community could have a higher oc-

currence of species in common with the low-lying

valley vegetation.

2. Osyris lanceolata - Celtis africana Kloof Forest

This forest is found at altitudes of 1 300 m to 1 400

m in Bakker’s Pass, on the escarpment in the west of

the study area (Fig. 3). It is represented by eight re-

leves (Table 2) with 13 to 23 species per releve. This

forest community (Edwards, 1983) has an rL struc-

ture (Ito, 1979; Fig. 4b) with the greatest average

cover of 50% in the > 5-8 m height class. There are

several non-perennial streams in the community

which, in area, is the largest of the other kloof com-

munities (Fig. 3). The community is sheltered and is

situated at an altitude between that of the other

kloof communities (Table 2), probably resulting in

the narrowest temperature range and highest humid-

ity of the kloof forest communities.

Habitat

The soils are of the Shortlands Form, Bokuil Se-

ries, derived from diabase of the post-Waterberg

Group. The average soil depth varies from 140 mm

to 470 mm and the surface rock cover varies from 15

to 40% . The kloof slopes from 3° to 17° in a southerly

to westerly direction. The electrical resistance of the

soil is the lowest recorded for all the communities in

the study area, whereas the T-value is amongst the

highest (Table 2), indicating nutrient-rich soils. The

soils are moderately acid (MacVicar et al., 1977)

with a pH of 6,5 when saturated with water.

Floristics

This community is distinguished by the Osyris lan-

ceolata species-group (Table 2 B). The species diver-

sity per unit area is the lowest for the kloof com-

munities with an average of 2,6 species/nr for the<

eight releves.

Trees and shrubs

Conspicuous woody species with more than 5%

mean cover and occurring in more than 50% of the

releves representing the community are:

Herbs

The only herb species occurring in more than 50%

of the releves representing the community is:

Cyperus albostriatus (sedge) 63% 0,8%

The presence of a sedge as the most conspicuous

species in the herb stratum is indicative of the mesic

conditions in this community.

General

Communities 1 & 2 are related to each other

through the common presence of the Olea europaea

species-group (Table 2C) and Communities 2 & 3

are related to each other through the shared pres-

ence of the Myrsine africana species-group (Table

2E), whereas the Diospyros whyteana species-group

(Table 2F) is common to Communities 1, 2 & 3. The

Cyperus albostriatus species-group (Table 2H) shows

affinities with the main vegetation type B, but as the

affinity is limited to Communities 4 & 5 this species-

group may be regarded as representing forest-mar-

gin species.

The community is accessible to grazing in places

and the area represented by these releves contain

grass patches as forest incursors.

3. Asplenium splendens - Celtis africana Kloof For-

est

This forest is found at altitudes of 1 600 m to 1 850

m in two kloofs in the north and north-east of the

study area (Fig. 3) and is represented by six releves

(Table 2) with 13 to 24 species per releve. It (Ed-

wards, 1983) has an rL structure (Ito, 1979; Fig. 4c)

with the greatest average cover of over 50% in the

higher than twelve metres height class. It is found at

the highest altitude for the kloof forest communities

in the study area and is consequently influenced by

mist which occurs frequently on the Kransberg

massif, but because the kloofs are generally shallow

the community is more exposed than the other kloof

communities and should, therefore, experience a

greater amplitude in temperature.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Sandriviersberg For-

mation. The average soil depth varies from 170 mm

667-668

a)

Community 1 Type rL

b)

Community 2 Type rL

Community 3 Type rL

d)

>8m-i2m

> 5m- 8m

>3m — 5m

>2m-3m

> 1m- 2m

>0,5m— i m

>0m-0,5m

II

CLOSED

WOODLAND

El

25

50

75 100

Community 4 Type D

e)

>8m-l2m

>5m” 8m

>3m~ 5m

>2m - 3m

>im- 2m

>0.5m- 1 n

>0m-0,5m

■*-| CLOSED

IT WOODLAND

b

0

25

50

75 100

Community 5 Type L

CLOSED

WOODLAND

50

75 100

Community 6 Type D

g)

CLOSED

WOODLAND

50

75

100

Community 7 Type rL

h)

>3m-5m

>2m-3m

>1m-2m

>0,5m- 1m

>0m-0,5m

L

OPEN

WOODLAND

25

50

75 100

Community 8.1 Type L

OPEN

WOODLAND

0 25 50 75 100

Community 8.2 Type L

j)

> 5m-8m

>3m-5m

> 2m -3m

- 1 m-2m

>0,5m-1 m

>0m-0,5m

CLOSED

WOODLAND

0 25 50 75 100

Community 9.1 Type 0

k)

> 5m-8m

> 3m- 5m

> 2m -3m

> im-2m

> 0,5m- 1m

>0m>0,5m

I

CLOSED

WOODLAND

25

50

75 100

Community 9.2 Type D

OPEN

WOODLAND

25

50 75 100

Community 10 Type L

m)

> 5m -8m

> 3m-5m

>2m-3m

> 1 m-2m

>0,5m- im

>0m-0,5m

?

a

CLOSED

WOODLAND

25

50

75

100

Community 11.1 Type D

n)

OPEN

WOODLAND

25

75

100

Community 11.2 Type L

o)

OPEN

WOODLAND

25

50

75

100

Community 12 Type L

P)

>5m - 8m

> 3m- 5m

> 2 m- 3m

> 1 m- 2m

>0,5m-l m

>0m-05m

OPEN

WOODLAND

h

25

50

75 100

Community 13 Type L

q)

> 5m -8m[l QpEN

>3m -5m J _ .

WOODLAND

>2m -3m

> 1 m -2

>0,5m - im

>0 m-Q5m

Q

25

50

75 100

Community 14 Type L

r)

»-5m ]

> 3m

> 2m-3m

> 1 m-2m

>0,5m -im

>0m-0,5m

OPEN

WOODLAND

25

50

75 100

Community 15 Type L

s)

t)

u)

V)

> Im- 2m I

0,5m- i m _[

>0

>0m-05m

CLOSED

GRASSLAND

25

50

75 100

Community 16 Type L

SPARSE

WOODLAND

50

75

100

Community 17.1 Type L

SPARSE

WOODLAND

50

75 100

Community 17.2 Type L

CLOSED

>2m-3mfl GRASSLAND

> o»5m-1iri J

> 0 m -0. 5m |___L

25

50

75 100

FIG. 4. - Average community structure of the vegetation of the farm Groothoek, Thabazimbi District, showing forma, ton

Community 18 Type L

classes, height classes, strata, mean percentage cover and layer-diagram type.

Bothalia 15, 3 & 4 (1985)

669

FIG. 5. — An example of kloof

forest vegetation in the west

of the study area with Faurea

saligna - Combretum mode

Open Woodland in the fore-

ground.

to 390 mm and the surface rock cover varies from 40

to 95%.

The kloofs slope from 17° to 30° in a southerly to

south-easterly direction. The electrical resistance of

the soil is low at 2 200 ohms with a high T-value of

22,0 me/100 g indicating a richer nutrient status than

for Community 1. This may be attributed to the pro-

tection afforded to the soil, found mostly in pockets

between boulders, by the large rocks and boulders

found in this commu^’ty. The effect of streamflow

on the leaching of nutrients from the soil should

therefore be minimized. Furthermore, the stream-

flow in these two kloofs is considerably less than in

the other kloofs because these two kloofs are situ-

ated at a high altitude and do not have feeder

streams. It is also suggested that this community is

not subjected to flooding as severe as in the other

kloof communities. The soils are strongly acid (Mac-

Vicar et al., 1977) with a pH of 4,1 when saturated

with water.

Floristics

The community is differentiated by the Asplenium

splendens species-group (Table 2D). The species di-

versity per unit area averages 2,8 species/nv for the

six releves.

Trees and shrubs

Conspicuous woody species, with more than 5%

mean cover and occurring in more than 50% of the

releves representing this community are:

Celtis africana (tree) 100% 52%

Podocarpus latifolius (tree) 83% 47%

Myrsine africana and Diospyros whyteana both oc-

cur with 100% constancy but with 2,5% and 2,2%

mean cover respectively. Widdringtonia nodi flora

occurs with 33% constancy and 1,3% cover but also

occurs in the grassland Community 18 as a forest in-

itial (Edwards, 1967).

The smaller kloof represented by releve number

113 has a high cover-abundance (ll%-25%) of the

tree fern Cyathea dregei (Table 2). This smaller

kloof is relatively inaccessible, while the larger kloof

is more accessible, being the most used route to the

top of Kransberg. There are, however, isolated

stumps of Cyathea dregei in the larger kloof, suggest-

ing that the habitat is suitable for growth of the tree

fern. It is suggested, therefore, that were it not for

the removal of Cyathea dregei this species would be a

character species for Community 3 with a high

cover-abundance.

Herbs

Herb species occurring in more than 50% of the

releves representing the community are:

General

Communities 2 & 3 are related to each other

through the common presence of the Myrsine afri-

cana species-group (Table 2E) and the Diospyros

whyteana species-group (Table 2F) is common to the

three kloof forest communities. The Cyperus albos-

triatus species-group (Table 2H) shows affinities

with the main vegetation type B but only as regards

forest-margin species. Community 3 is similar to

Community 2 in respect of common species-groups,

however, unlike Community 2, Community 3 does

not have the Olea europaea species-group (Table

2C) in common with Community 1 . This difference

may be attributed to the higher altitude in which

Community 3 occurs indicating that the Olea eu-

670

Bothalia 15, 3 & 4 (1985)

ropaea species-group (Table 2C) does not occur

above 1 400 m altitude.

B. Woodland phase of Acocks’s (1975) Sour Bush-

veld, on moderately deep to deep soils in moderately

exposed habitats

The woodland communities representing Acocks’s

(1975) Sour Bushveld are found south of the Krans-

berg massif in the study area (Fig. 3) below 1 600 m

altitude.

Habitat

The soils are mainly of the Mispah Form (Mispah

Series) with the Hutton Form (Middelburg Series),

Shortlands Form (Bokuil Series) and the Westleigh

Form (Sibasa Series) also occurring. Soil depth va-

ries from 40 mm to more than 1 000 mm with 57% of

the soil depths recorded being greater than 130 mm.

The communities are more exposed than the kloof

communities, but more sheltered than the communi-

ties on the upper slopes of the Kransberg massif and

on the upper summit (Fig. 2), being sheltered by the

Kransberg massif.

Floristics

The vegetation is structurally heterogeneous, va-

rying from closed woodland to open woodland and

occasionally sparse woodland (Table 2). The Com-

bretum molle species-group (Table 2 AC) is differen-

tiating for the woodland communities representing

Acocks’s (1975) Sour Bushveld (main vegetation

type B). The grasslands of Acocks's (1975) Sour

Bushveld (main vegetation type C) are related to the

main vegetation type B through the common pres-

ence of the Schizachyrium sanguineum species-

group (Table 2W). The Aristida aequiglumis species-

group (Table 2Q), which contains species such as

Burkea africana and Ochna pulchra, has a narrower

ecological amplitude than the Combretum molle

species-group (Table 2AC) and is common to Com-

munities 5 to 11. An example of the vegetation is

shown in Fig. 6.

In the phytosociological classification, the com-

munities of the main vegetation type B are classified

as follows (Tables 1, 2 & 3):

4. Panicum maximum - Combretum molle

Closed Woodland

5. Euclea crispa - Combretum molle Closed

Woodland

6. Setaria megaphylla - Combretum molle

Closed Woodland

7. Terminalia sericea - Combretum molle

Closed Woodland

8. Aristida diffusa - Combretum molle Open

Woodland

8. 1 Strychnos madagascariensis - Aristida diffusa

- Combretum molle Variation

8.2 Vitex rehmannii - Aristida diffusa - Combre-

tum molle Variation

9. Landolphia capensis - Combretum molle

Closed Woodland

9.1 Burkea africana - Landolphia capensis -

Combretum molle Variation

9.2 Tapiphyllum parvifolium - Landolphia ca-

pensis - Combretum molle Variation

10. Coleochloa setifera - Combretum molle

Open Woodland

11. Heteropogon contortus - Combretum molle

Closed and Open Woodlands

11.1 Rhus dentata - Heteropogon contortus -

Combretum molle Closed Woodland Varia-

tion

11.2 Chaetacanthus costatus - Heteropogon con-

tortus - Combretum molle Open Woodland

Variation

12. Themeda triandra - Combretum molle Open

Woodland

13. Argyrolobium transvaalense - Combretum

molle Open Woodland

14. Pachycarpus schinzianus - Combretum molle

Open Woodland

15. Protea caffra - Combretum molle Open

Woodland.

4. Panicum maximum - Combretum molle Closed

Woodland

This woodland is found at altitudes of

1 450 m-1 600 m on the lower slopes of the Krans-

berg massif in the north of the study area (Fig. 3). It

is represented by ten releves with 21 to 33 species

FIG. 6. — Woodland representa-

tive of Acocks’s (1975) Sour

Bushveld with the Panicum

maximum - Combretum

molle Closed Woodland and

Euclea crispa - Combretum

molle Closed Woodland in

the foreground and the Land-

olphia capensis - Combretum

molle Closed Woodland in

the background.

Bothalia 15, 3 & 4 (1985)

671

per releve. This closed-woodland community (Ed-

wards, 1983) has a D structure (Ito, 1979; Fig. 4d)

with the greatest average cover of about 27% in the

> 3 m-5 m height class. Because the slopes on which

this community is found are south facing, the tem-

perature is likely to be lower than that of the north-

facing slopes (Theron, 1973) and if night tempera-

tures are equal, should have a smaller temperature

range than that of the north-facing slopes.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Aasvoelkop Forma-

tion. The average soil depth varies from 150 mm to

500 mm and the surface rock cover varies from 1%

to 40%. The terrain slopes from 3° to 25° in a south-

erly direction. The electrical resistance of the soil is

4 700 ohms and the T-value is 8,5 me/100 g which is

moderate for the study area indicating a moderate

nutrient status. As the community is situated on the

lower slopes, nutrient accumulation from run-off is

greater than for the upper slopes (Russell, 1961),

where greater leaching can be expected. The soils

are strongly acid (MacVicar et al., 1977) with a pH of

4,8 when saturated with water.

Floristics

This community is differentiated by the Achy-

ranthes aspera species-group (Table 2G). The

species diversity per unit area averages 5,8

species/nr for the Community. The character species

for the species-group are mainly pioneer forbs,

which could have a wider presence than that indi-

cated on Table 2. However, at the time of sampling

they were characteristic of Community 4 in the study

area and are, therefore, classified accordingly.

Trees and shrubs

Conspicuous woody species with more than 5%

mean cover and occurring in more than 50% of the

releves representing this community are:

Combretum molle (tree/shrub) 90% 7,1%

Rhus leptodictya (tree/shrub) 80% 5,1%

Faurea saligna (tree/shrub) 60% 6,2%

Herbs

Herb species occurring in more than 50% of the

releves representing the Community are:

General

Communities 2, 3, 4 & 5 are related to each other

through the shared presence of the Cyperus albo-

striatus species-group (Table 2H). Communities 4 &

5 are structurally similar (Figs 4d & 4e) and, because

they occur next to each other in the study area, can-

not be separated on aerial photographs. They are,

therefore, mapped as a single unit in Fig. 3. These

two communities represent a transitional vegetation

zone between the main vegetation types A and B be-

cause they do not have perennial character species

which are at present apparent (Table 2). If the Achy-

ranthes aspera species-group (Table 2G) were pos-

sibly eliminated by reduced grazing pressure, then

this community would be floristically more similar to

Community 5 than is indicated on Table 2.

5. Euclea crispa - Combretum molle Closed Wood-

land

This woodland is found at altitudes of 1 500 m to

1 550 m on the lower slopes of the Kransberg massif

in the north of the study area (Fig. 3). It is represent-

ed by releves 145, 153 and 134 with 25, 30 and 34

species per releve respectively. This closed-wood-

land community (Edwards, 1983) has a D structure

(Ito, 1979; Fig. 4e) with the greatest average cover

of 23% in the > 3 m-5 m height class. Because the

community is adjacent to Community 4, their tem-

perature and moisture regimes should be similar.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Aasvoelkop Forma-

tion. The average soil depth varies from 150 mm to

240 mm and the surface rock cover varies from 20%

to 60%. The terrain slopes from 8° to 16° in a south-

erly to south-easterly direction. The electrical resist-

ance of the soil is 4 700 ohms and the T-value is 8,5

me/100 g which is similar to that of Community 4.

The soils are strongly acid (MacVicar et al., 1977)

with a pH of 4,8 when saturated with water.

Floristics

The community is differentiated by the absence of

character species, as well as the absence of the Achy-

ranthes aspera species-group (Table 2G), and the

presence of the Aristida aequiglumis species-group

(Table 2Q) and the Cyperus albostriatus species-

group (Table 2H). The species diversity per unit

area averages 4,7 species/nr for the Community

which is less than for Community 4, possibly as a re-

sult of the high surface rock cover in Community 5

with1 consequently less soil area for vegetation as

well as fewer annual and pioneer species.

Trees and shrubs

Conspicuous woody species with more than 5%

mean cover and occurring in more than 50% of the

releves representing the Community are:

672

Bothalia 15, 3 & 4 (1985)

General

Communities 2, 3, 4 & 5 are related to each other

through the shared presence of the Cyperus albos-

triatus species-group (Table 2H) and Communities 5

to 11 are related to each other through the shared

presence of the Aristida aequiglumis species-group

(Table 2Q). Community 5 together with Community

4 forms a transitional vegetation zone between the

main vegetation types A and B. The pioneer forbs

present in Community 4 (Table 2G) are absent in

Community 5. Selaginella dregei is present in two of

the three releves representing Community 5 indicat-

ing the high surface rock cover found in this com-

munity. Although this community is accessible to

cattle, the high surface rock cover may possibly re-

sult in a grazing preference for Community 4.

6. Setaria megaphylla — Combretum molle Closed

Woodland

This woodland is found at altitudes of 1 075 m to

1 375 m in an open kloof in the east of the study area

(Fig. 3). It is represented by releves 103, 101, 102

and 99 with 27, 26, 27 and 21 species per releve re-

spectively. This closed woodland community (Ed-

wards, 1983) has a D structure (Ito, 1979; Fig. 4f)

with the greatest average cover of about 22% in the

> 3 m-5 m height class. Although this community is

situated in a kloof, the kloof is very broad and shal-

low so that the vegetation is more exposed with a

consequently greater temperature range and is less

moist than the kloof forest communities. The vege-

tation is, therefore, a closed woodland and not a for-

est as could be expected from the landform class H

(Table 2).

Habitat

The soils are of the Mispah form, Mispah Series

derived from conglomerate of the Alma Graywacke

Formation in the case of releves 103, 101 and 102

and sandstone of the Alma Graywacke Formation in

the case of releve 99. The average soil depth varies

from 150 mm to 210 mm and the surface rock cover

varies from 30% to 70%. The kloof slopes from 3° to

5° in a southerly to south-westerly direction. The

electrical resistance of the soil is 4 300 ohms with a

T-value of 9,3 me/100 g indicating soils of a moder-

ate nutrient status for the study area. The soils are

strongly acid (MacVicar et al. , 1977) with a pH of 4,8

when saturated with water.

Floristics

This community is differentiated by the Setaria

megaphylla species-group (Table 21). Setaria mega-

phylla is also found in Community 1 with 100% con-

stancy but with a lower mean percentage cover

(Table 2). The species diversity per unit area aver-

ages 5,3 species/nr for the community.

Trees and Shrubs

The only conspicuous woody species with more

than 5% mean cover and occurring in more than

50% of the releves representing the community is:

Elephantorrhiza burkei (shrub) 75% 7%

Herbs

Herb species occurring in more than 50% of the

releves representing the community are:

General

Communities 5 to 11 are related to each other

through the shared presence of the Aristida aequi-

glumis species-group (Table 2Q). The vegetation

represented by releve numbers 103 and 101 is on the

same side of the stream as that used by cattle for

drinking and the grazing pressure on Setaria mega-

phylla, a palatable grass, is high, hence the low

cover-abundance values of this species in releve

numbers 103 and 101.

7. Terminalia sericea — Combretum molle Closed

Woodland

This woodland is found at altitudes of 1 415 m to

1 425 m in the eastern central part of the study area

on the upper plateau (Fig. 3). It is represented by

releves 79, 80 and 77 with 30, 30 and 22 species per

releve respectively. This closed-woodland com-

munity (Edwards, 1983) has an rL structure (Ito,

1979; Fig. 4g) with the greatest average cover of

about 37% in the upper height class of > 5 m - 8 m.

The community is probably subjected to tempera-

ture inversion at night because it is situated near a

stream in a depression. The temperature range

should, therefore, be wide for the study area.

Habitat

The soils are of the Hutton Form, Middelburg Se-

ries, derived from conglomerate of the Alma Gray-

wacke Formation. The average soil depth varies

from 150 mm to 650 mm and the surface rock cover

varies from 1% to 2%. The terrain slopes from 1° to

2° in a northerly direction. The electrical resistance

of the soil is 6 400 ohms and the T-value is 5,9

me/100 g indicating a poor nutrient status for the

soils, which may be attributed to leaching. The soils

are strongly acid (MacVicar et al. , 1977) with a pH of

4,1 when saturated with water.

Floristics

The community is distinguished by the Terminalia

sericea species-group (Table 2J). The species diver-

sity per unit area averages 7 species/nr for the com-

munity which is relatively high for the study area.

Trees and shrubs

Conspicuous woody species with more than 5%

mean cover and occurring in more than 50% of the

releves representing the community are:

Terminalia sericea (tree) 100% 28,0%

Burkea africana (tree/shrub) 67% 8,5%

Bothalia 15, 3 & 4 (1985)

673

Herbs

Herb species occurring in more than 50% of the

releves representing the community are:

General

Communities 5 to 11 are related to each other

through the common presence of the Aristida ae-

quiglumis species-group (Table 2Q). This com-

munity is restricted to the moderately deep soils of

the upper plateau which are well drained, but as

these soils are limited in extent, the community is

consequently small in area. On the northern side of

the stream, opposite Community 7, there is much

cultivated land on moderately deep soils, which

could support the same vegetation as that of Com-

munity 7.

8. Aristida diffusa — Combretum molle Open

Woodland

This woodland is found at altitudes of 1 250 m to

1 400 m on the lower slopes of the lower plateau in

the south of the study area with an outlier of the

community above the eastern kloof on the upper

plateau (Fig. 3). This community is differentiated by

the Aristida diffusa species-group (Table 2K) which

only has one character species, namely Aristida dif-

fusa. The community is separated into the following

two variations, based on floristics:

8.1 Strychnos madagascariensis - Aristida diffusa

- Combretum molle Variation found at, or be-

low 1 325 m altitude,

8.2 Vitex rehmannii - Aristida diffusa - Combre-

tum molle Variation found at or above 1 325

m altitude.

8.1 Strychnos madagascariensis - Aristida diffusa

- Combretum molle Variation

This variation is found at altitudes of 1 250 m to

1 325 m, being the lowest altitude of the communi-

ties on the lower plateau in the south of the study

area (Fig. 3). It is represented by six releves with 23

to 38 species per releve. This open- woodland varia-

tion (Edwards, 1983) has an L structure (Ito, 1979;

Fig. 4h) with the greatest average cover of about 20

per cent in the lowest height class of 0,0 m - 0,5 m.

The lower plateau is exposed and a wide tempera-

ture range could be expected, while the open vegeta-

tion indicates a low moisture status for this variation.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Alma Graywacke

Formation. The average soil depth varies from 120

mm to 170 mm and the surface rock cover varies

from 50% to 70%. The terrain slopes from 1° to 9° in

a southerly to south-westerly direction. The electri-

cal resistance of the soil is 3 500 ohms and the T-val-

ue is 7,1 me/100 g indicating a moderate nutrient

status. The soils are strongly acid (MacVicar et al.,

1977) with a pH of 5,4 when saturated with water.

Floristics

This variation is distinguished by the Schrebera

alata species-group (Table 2L). The species diversity

per unit area averages 5,5 species/m2 for the Varia-

tion.

Trees and shrubs

Conspicuous woody species with a 2,5% or more

mean cover and occurring in more than 50% of the

releves representing this variation are:

Pseudolachnostylis maprouneifolia (tree/shrub) 100% 2,6%

Diplorhynchus condylocarpon (tree/shrub) 100% 2,5%

The value of 2,5 % or more mean cover for the

conspicuous woody plants has been reduced from

the arbitrarily selected 5% used for the previous

communities because Community 8 is the first com-

munity with an open formation class. The 5% mean

cover value is too high for open formation and no

woody species would be detected.

Herbs

Herb species occurring in more than 50% of the

releves representing the variation are:

Aristida diffusa (grass) 100%

Schizachyrium sanguineum

(grass) 100%

Commelina africana (forb) 100%

Diheteropogon amplectens

(grass) 83%

Aristida aequiglumis (grass) 83%

Heteropogon contortus (grass) 67%

Brachiaria nigropedata (grass) 67%

Elionurus muticus (grass) 67%

Rhynchelytrum nerviglume

(grass) 67%

Loudetia simplex (grass) 67%

Vernonia staehelinoides (forb) 67%

Tephrosia longipes (forb) 67%

5,0%

2,2%

0,2%

2,9%

1,8%

3,0%

1,0%

0,7%

0,5%

0,3%

0,03%

General

Communities 8 & 9 are related to each other

through the common presence of the Diplorhynchus

condylocarpon species-group (Table 2N) and Com-

munities 5 to 11 are related to each other through

the shared presence of the Aristida aequiglumis

species-group (Table 2Q), while Communities 8 to

16 are related to each other through the shared pres-

ence of the Schizachyrium sanguineum species-

group (Table 2W).

8.2 Vitex rehmannii - Aristida diffusa - Combre-

tum molle Variation

This variation is found at altitudes of 1 325 m to

1 400 m upslope of variation 8.1 on the lower pla-

teau in the south of the study area and on the slopes

leading to a kloof on the upper plateau in the east of

the study area (Fig. 3). It is represented by six re-

leves with 20 to 30 species per releve. This open-

woodland Variation (Edwards, 1983) has an L struc-

ture (Ito, 1979; Fig. 4i) with the greatest average

cover of 25% in the lowest height class of 0,0 m - 0,5

m, which is greater than that for Variation 8.1. The

674

Bothalia 15, 3 & 4 (1985)

temperature and moisture regimes should be similar

to that of Variation 8.1.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone in the case of releves 86, 89,

90, 87 and 96 and conglomerate in the case of releve

104, both of the Alma Graywacke Formation. The

average soil depth varies from 100 mm to 190 mm

and the surface rock cover varies from 60% to 70%.

The terrain slopes up to 28° in a southerly to south-

easterly direction in the case of releves 86, 89, 90, 87

and 96 and in a northerly direction in the case of re-

leve 104. The electrical resistance varies from 3 500

ohms to 6 700 ohms and the T-value varies from 7,1

me/100 g to 10,5 me/100 g, indicating soils of a mod-

erate nutrient status. The soils are strongly acid

(MacVicar et al., 1977) with a pH range of 4,5 to 5,4

when saturated with water.

The difference in the soil factors found in this va-

riation may be attributed to the difference in parent

materials in the case of releve 104 and possibly to the

leaching effect of a seasonal stream which causes

periodic flooding in the vicinity of releve 96, with

increased runoff and consequent leaching.

Floristics

This variation is differentiated by the absence of

character species and notably the absence of the

species of the Schrebera alata species-group (Table

2L) which differentiates this variation from the

Strychnos madagascariensis - Aristida diffusa - Com-

bretum molle Variation. The species diversity per

unit area averages 4,8 species/nr for this variation

which is lower than for Variation 8.1.

Trees and shrubs

Conspicuous woody species with 2,5% or more

mean cover and occurring in more than 50% of the

releves representing the variation are:

Combretum molle (tree/shrub) 100% 2,9%

Diplorhynchus condylocarpon (tree/shrub) 83% 4,6%

Vitex rehmannii is present with 83% constancy

and 1,4% mean cover.

Herbs

Herb species occurring in more than 50% of the

releves representing the variation are:

General

Variation 8.2 has the same affinities with other

communities as that described for Variation 8.1

9. Combretum molle - Landolphia capensis

Closed Woodland

This woodland is found at altitudes of 1 425 m to

1 550 m south of the upper plateau in the southern

half of the study area (Fig. 3). The community is di-

agnosed by the Landolphia capensis species-group

(Table 2M).

This community is differentiated into the follow-

ing two variations, based on floristics:

9.1 Burkea africana - Landolphia capensis -

Combretum molle Variation found on soils

with a moderate surface rock cover,

9.2 Tapiphyllum parvifolium - Landolphia capen-

sis - Combretum molle Variation found on

soils with a moderate to high surface rock

cover.

9. 1 Burkea africana Landolphia capensis - Com-

bretum molle Variation

This variation is found at altitudes of 1 425 m to

1 550 m south of the upper plateau in the southern

half of the study area (Fig. 3). It is represented by

eight releves with 19 to 31 species per releve. This

closed-woodland variation (Edwards, 1983) has a D

structure (Ito, 1979; Fig. 4j) with the greatest aver-

age cover of 15% in the > 3 m - 5 m height class.

The variation occurs on north-facing slopes (Theron,

1973) and should therefore have a greater tempera-

ture range than the south-facing slopes, if night tem-

peratures are the same.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from conglomerate of the Alma Graywacke

Formation. The average soil depth varies from 40

mm to 250 mm and the surface rock cover varies

from 5% to 25%. The terrain slopes up to 6° in a

northerly to easterly direction. The electrical resist-

ance is 6 700 ohms and the T-value is 10,5 me/100 g

indicating soils of a moderate nutrient status for the

study area. The soils are strongly acid (MacVicar et

al., 1977) with a pH of 4,5 when saturated with

water.

Floristics

The Variation is differentiated by the absence of

character species as well as the absence of the Selagi-

nella dregei species-group (Table 2P) which differen-

tiates this variation from the Tapiphyllum parvifo-

lium - Landolphia capensis - Combretum molle Va-

riation. The species diversity per unit area averages

5 species/nr for the variation.

Trees and Shrubs

Conspicuous woody species with more than 5%

mean cover and occurring in more than 50% of the

releves representing the variation are:

Burkea africana (tree/shrub) 100% 15,9%

Combretum molle (tree/shrub) 100% 6,0%

Ochna pulchra (tree/shrub) 88% 7,6%

Burkea africana has the highest constancy value in

the study area in Variation 9.1.

Herbs

Herb species occurring in more than 50 per cent of

the releves representing the variation are:

Bothalia 15, 3 & 4 (1985)

675

General

Communities 8 & 9 are related to each other

through the joint occurrence of the Diplorhynchus

condylocarpon species-group (Table 2N) and Com-

munities 5 to 11 are related to each other through

the shared presence of the Aristida aequiglumis

species-group (Table 20) while Communities 8 to 16

are related to each other through having the Schiza-

chyrium sanguineum species-group (Table 2W) in

common.

9.2 Tapiphyllum parvifolium - Landolphia capen-

sis - Combretum molle Variation

This variation is found at altitudes 1 425 m to

1 525 m south of the upper plateau in the southern

half of the study area (Fig. 3). It is represented by

twelve releves with 21 to 34 species per releve. This

closed-woodland variation (Edwards, 1983) has a D

structure (Ito, 1979; Fig. 4k) with the greatest aver-

age cover of 10% in the > 3 m - 5 m height class.

This variation can, therefore, be considered to be

more open than Variation 9.1 which could indicate a

drier moisture regime than for Variation 9.1. The

temperature regime should be similar to that of Va-

riation 9.1.

to each other through the shared presence of the Di-

plorhynchus condylocarpon species-group (Table

2N), the Aristida aequiglumis species-group (Table

2Q) and the Schizachyrium sanguineum species-

group (Table 2W) respectively which corresponds to

that for the Burkea africana - Landolphia capensis -

Combretum molle Variation. The Selaginella dregei

species-group (Table 2P) is common to Community

10 and Variation 9.2, but is absent in Variation 9.1.

Selaginella dregei has a 75% constancy which is high

for the study area and is indicative of the high sur-

face rock cover for Variation 9.2 because Selaginella

dregei only occurs where there are sheet outcrops.

The high surface rock cover is a possible reason for

the structural differences between the two variations

of Community 9 and can also contribute to the flor-

istic differences between the two variations.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from conglomerate of the Alma Graywacke

Formation. The average soil depth varies from 50

mm to 230 mm and the surface rock cover varies

from 5% to 80% . The terrain slopes from 1° to 8° in a

north-easterly direction. The electrical resistance of

the soil is 6 700 ohms and the T-value is 10,5

me/100 g indicating a moderate nutrient status for

the study area. The soils are strongly acid (MacVicar

et al . , 1977) with a pH of 4,5 when saturated with

water.

Floristics

This variation is differentiated by the absence of

character species and is differentiated from the Bur-

kea africana - Landolphia capensis - Combretum

molle Variation by the presence of the Selaginella

dregei species-group (Table 2P). The species diver-

sity per unit area averages 4,4 species/m: for this va-

riation, which is lower than for Variation 9.1.

Trees and shrubs

The only conspicuous woody species with more

than 5% mean cover and occurring in more than

50% of the releves representing the variation is:

Burkea africana (tree/shrub) 67% 7,35%

Combretum molle has a 100% constancy in Varia-

tion 9.2. but only 4,4% mean cover, whereas Ochna

pulchra has 83% constancy and only 4,1% mean

cover indicating the lesser cover of these two species

in this variation compared with Variation 9.1.

Herbs

Herb species occurring in more than 50% of the

releves representing this variation are:

10. Coleochloa setifera - Combretum molle Open

Woodland

This woodland is found at altitudes of 1 525 m to

1 550 m south of the upper plateau in the southern

half of the study area (Fig. 3). It is represented by

eight releves with 26 to 33 species per releve. This

open-woodland community (Edwards, 1983) has an

L structure (Ito, 1979; Fig. 41) with the greatest ave-

rage cover of 5% in the lowest height class of

0,0 m-0,5 m. Because the vegetation occurs on the

lower summit, this community is more exposed than

the adjacent communities (Communities 8 & 9), and

a greater temperature range and a drier moisture

regime than in Communities 8 & 9 could be ex-

pected.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from conglomerate of the Alma Graywacke

Formation. The average soil depth varies from 40

mm to 80 mm and the surface rock cover varies from

60% to 80%. The terrain slopes from 18° to 25° in a

southerly direction. The electrical resistance is 6 700

ohms and the T-value is 10,5 me/100 g indicating

soils of a moderate nutrient status for the study area.

The soils are strongly acid (MacVicar et al . , 1977)

with a pH of 4,5 when saturated with water.

Floristics

The community is distinguished by the Coleochloa

setifera species-group (Table 20). The species diver-

sity per unit area averages 5,3 species/nr for the

community.

Trees and shrubs

The only conspicuous woody species with 2,5% or

more mean cover and occurring in more than 50% of

the releves representing the community is:

Heteropyxis natalensis (shrub) 100% 2,5%

676

Bothalia 15, 3 & 4 (1985)

Combretum- molle, Vitex rehmannii, Bequaertio-

dendron magalismontanum and Brachylaena rotun-

data occur in all the releves representing the com-

munity, but have a mean cover of less than 2,5%.

Herbs

Herb species occurring in 50% of the releves re-

presenting the community are:

General

Variation 9.2 and Community 10 are related to

each other through the common presence of the Sel-

aginella dregei species-group (Table 2P) while Com-

munities 5 to 11 and 8 to 16 are related to each other

through the shared presence of the Aristida aequi-

glumis species-group (Table 2Q) and the Schizachy-

rium sanguineum species-group (Table 2W) respect-

ively. The condition of the grass layer is poor with

respect to cover (Fig. 41). The high surface rock

cover, caused by conglomerate sheet outcrop, is re-

flected in the relatively high proportion of Selagi-

nella dregei in Table 2.

11. Heteropogon contortus - Combretum molle

Closed and Open Woodlands

These woodlands are found at altitudes of 1 400 m

to 1 600 m in the north, south-east and south-west of

the study area (Fig. 3). This community is differen-

tiated by the absence of character species as well as

the absence of the Cyperus albostriatus species-

group (Table 2H), the Setaria megaphylla species-

group (Table 21), the Terminalia sericea species-

group (Table 2J), the Diplorhynchus condylocarpon

species-group (Table 2N), the Selaginella dregei

species-group (Table 2P), the Stoebe vulgaris species

group (Table 2U) and the Senecio erubescens

species-group (Table 2V) in the main vegetation

type B, as well as the presence of the Aristida ae-

quiglumis species-group (Table 2Q). This com-

munity is separated into the following two varia-

tions, based on floristics:

11.1 Rhus dentata - Heteropogon contortus -

Combretum molle Closed Woodland Varia-

tion on predominantly nutrient-rich soils,

11.2 Chaetacanthus costatus - Heteropogon con-

tortus - Combretum molle Open Woodland

Variation on predominantly nutrient-poor

soils.

11.1 Rhus dentata - Heteropogon contortus -

Combretum molle Closed Woodland Variation

This variation is found at altitudes of 1 400 m to

1 575 m on the lower slopes of the Kransberg massif

in the north as well as in the south-west and south-

east of the study area (Fig. 3). It is represented by

ten releves with 19 to 32 species per releve. This

closed-woodland variation (Edwards, 1983) has a D

structure (Ito, 1979; Fig. 4m) with the greatest aver-

age cover of 20% in the > 3 m - 5 m height class.

Because the variation has a closed-woodland struc-

ture, the temperature range is likely to have a nar-

rower amplitude and the moisture regime is likely to

be higher than for Variation 11.2.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Aasvoelkop Forma-

tion, in the case of releves 154, 151, 143, 139, 146,

147 and 24, and conglomerate of the Alma Gray-

wacke Formation in the case of releves 59, 49 and

50. The average soil depth varies from 40 mm to 230

mm and the surface rock cover varies from 20% to

60%. The terrain slopes up to 25° in a southerly di-

rection. The electrical resistance of the soils varies

from 2 100 ohms to 6 700 ohms. The T-value varies

from 7,4 me/100 g to 15,8 me/100 g indicating soils of

a predominantly high nutrient status. The soils are

strongly acid (MacVicar et al. , 1977) with a pH of 4,1

to 4,5 except for releve 24 which has a pH of 6,0

when saturated with water and is, therefore, moder-

ately acid. The habitat factors for this variation have

a wide range which may be attributed to the distribu-

tion of this variation which occurs as small areas,

widely dispersed through the study area (Fig. 3).

Floristics

This variation is differentiated by the absence of

character species as well as the absence of the

Helichrysum sp. species-group (Table 2 AB), which

distinguishes it from the Chaetacanthus costatus

- Heteropogon contortus - Combretum molle Varia-

tion. The species diversity per unit area averages 4,8

species/nr for this variation.

Trees and shrubs

The only conspicuous woody species with more

than 5% mean cover and occurring in more than

50% of the releves representing this variation is:

Burkea africana (tree/shrub) 70% 13%

Rhus dentata (shrub) occurs in 70% of the releves

and has a mean cover of 1,4%.

Herbs

Herb species occurring in more than 50% of the

releves representing this variation are:

Bothalia 15, 3 & 4 (1985)

677

General

Communities 5 to 11 are related to each other

through the shared presence of the Aristida diffusa

species-group (Table 2Q) and Communities 8 to 16

are related to each other through the shared pre-

sence of the Schizachyrium sanguineum species-

group (Table 2W).

11.2 Chaetacanthus costatus - Heteropogon con-

tortus - Combretum molle Open Woodland Varia-

tion

This variation is found at altitudes of 1 500 m to

1 600 m on the lower slopes of the Kransberg massif

in the northern half of the study area (Fig. 3). It is

represented by eight releves with 20 to 27 species per

releve. This open-woodland variation (Edwards,

1983) has an L structure (Ito, 1979; Fig. 4n) with the

greatest cover of 25% in the 0,0 m-0,5 m height

class. Because the Variation is generally situated up-

slope of Variation 11.1, it is probably more exposed

with a consequently greater temperature amplitude

and a drier moisture regime than Variation 11.1.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Aasvoelkop Forma-

tion. The average soil depth varies from 110 mm to

210 mm and the surface rock cover varies from 15%

to 60%. The terrain slopes from 3° to 21° in a south-

erly direction. The electrical resistance is 4 400 ohms

to 5 100 ohms for the soils and the T-value varies

from 6,3 me/100 g to 8,5 me/100 g which is generally

lower than that for Variation 11.1 indicating soils of

a lower nutrient status. This may be as a result of

Variation 11.2 being generally upslope of Variation

11.1 with nutrients lost by runoff from Variation

1 1 .2 to 1 1 . 1 . The soils are strongly acid (MacVicar et

al., 1977) with a pH range of 4,3 to 4,8 when satu-

rated with water.

Floristics

This variation is marked by the absence of charac-

ter species and is distinguished from Variation 11.1

in Community 11 by the presence of the Helichry-

sum sp. (Westfall 921) species-group (Table 2AB).

The species diversity per unit area averages 5,4

species/m2 for this variation which is higher than that

for Variation 11.1.

Trees and shrubs

Conspicuous woody species with 2,5% or more

mean cover and occurring in more than 50% of the

releves representing the variation are:

Combretum molle (tree/shrub) 88% 2,9%

Burkea africana (tree/shrub) 75% 7,1%

Protea caffra (tree/shrub) 63% 2,8%

Bequaertiodendron magalismontanum has 100%

constancy in Variation 11.2 but only 2,1% mean

cover.

Herbs

Herbaceous species occurring in more than 50%

of the releves representing this variation are:

General

Communities 5 to 11 and 8 to 16 are related to

each other through the shared presence of the Aris-

tida aequiglumis species-group (Table 2Q) and the

Schizachyrium sanguineum species-group (Table

2W) respectively. Variation 11.2 and Communities

12, 17 and 18 are related to each other through the

shared presence of the Helichrysum species-group

(Table 2AB).

12. Themeda triandra - Combretum molle Open

Woodland

This woodland is found at altitudes of 1 400 m to

1 600 m in the south, north and west of the study

area (Fig. 3). It is represented by thirteen releves

with 21 to 29 species per releve. This open-woodland

community (Edwards, 1983) has an L structure (Ito,

1979; Fig. 4o) with the greatest average cover of

38% in the 0,0 m-0,5 m height class. The tempera-

ture range should be wide and the moisture regime

should be dry, because of the relatively exposed situ-

ation of this community.

Habitat

The soils are mainly of the Mispah Form, Mispah

Series, derived from sandstone of Aasvoelkop and

Alma Graywacke Formations and shale of the Aas-

voelkop Formation. Soils of Shortlands Form, Bo

kuil Series, derived from diabase of the post-Water-

berg Group, are also found in this community (re-

leves 165 and 164). The class limits set for the soil

classification used in this study (MacVicar et al.,

1977) are not necessarily the same as the limits influ-

encing the vegetation (D. Edwards, pers. comm.),

hence the variation of soil within this community.

The Shortlands Form recorded for this community

can, furthermore, be regarded as atypical because of

the shallow soil depth of 80 mm recorded (Table 2).

The average soil depth varies from 80 mm to 300 mm

and the surface rock cover varies from 5% to 60%.

The terrain slopes from 3° to 24° in a south-easterly to

westerly direction. The electrical resistance of the

soils varies from 3 200 ohms to 6 000 ohms and the

T-value varies from 4,4 me/100 g to 7,4 me/100 g,

indicating soils with a moderate nutrient status. The

soils are strongly acid (MacVicar et al., 1977) with a

pH range of 4,5 to 5,0 when saturated with water.

Floristics

The Community is diagnosed by the absence of

character species as well as the absence of the Aris-

678

Bothalia 15, 3 & 4 (1985)

tida aequiglumis species-group (Table 2Q) and the

Senecio erubescens species-group (Table 2V) and the

presence of the Helichrysum sp. species-group

(Table 2AB) in the main vegetation type B, distin-

guished by the Combretum molle species-group

(Table 2AC). The species diversity per unit area av-

erages 5,8 species/m: for this community.

Trees and shrubs

Conspicuous woody species with 2,5% or more

mean cover and occurring in 50% of the releves rep-

resenting this community are:

Protea caffra (tree/shrub) 69% 2,5%

Faurea saligna (tree/shrub) 62% 3,4%

Faurea saligna occurs in eleven of the twelve com-

munities of the main vegetation type B represented

by species-group AC (Table 2), but has the highest

mean percentage cover of its range in the study area

in Community 12.

Herbs

Herb species occurring in more than 50% of the

releves representing this community are:

General

Communities 8 to 16 are related to each other

through the shared presence of the Schizachyrium

sanguineum species-group (Table 2W) and Variation

11.2 and Communities 12, 17 and 18 are related to

each other through the shared presence of the

Helichrysum sp. species-group (Table 2AB).

13. Argyrolobium transvaalense - Combretum

molle Open Woodland

This woodland is found at altitudes of 1 435 m to

1 575 m on the lower slopes of the Kransberg massif

in the north-west of the study area (Fig. 3). It is rep-

resented by five releves with 22 to 31 species per re-

leve. This open-woodland community (Edwards,

1983) has an L structure (Ito, 1979; Fig. 4p) with the

greatest average cover of 20% in the 0,0 m-0,5 m

height class. This community is more exposed, as a

result of its situation on prominent ridges, than the

other communities of the lower slopes and large

temperature fluctuations and a dry moisture regime

can consequently be expected.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Aasvoelkop Forma-

tion. The average soil depth varies from 120 mm to

150 mm and the surface rock cover varies from 7%

to 15%. The terrain slopes up to 27° in a south-east-

erly to south-westerly direction. The electrical resist-

ance of the soils is 2 100 ohms indicating a moderate

nutrient status. The soils are moderately acid (Mac-

Vicar et al. , 1977) with a pH of 6,0 when saturated

with water.

Floristics

This community is distinguished by the Vernonia

oligocephala species-group (Table 2R). The species

diversity per unit area averages 7,8 species/m: for

this community which is relatively high for the study

area.

Trees and shrubs

There are no woody species with a mean cover of

2,5% or more. Woody species occurring in more

than 50% of the releves representing this community

are:

Argyrolobium transvaalense

Herbs

Herb species occurring in more than 50% of the

releves representing this community, are:

16 are related to each other through the shared pres-

ence of the Schizachyrium sanguineum species-

group (Table 2W).

14. Pachycarpus schinzianus - Combretum molle

Open Woodland

This woodland is found at altitudes 1 425 m to

1 465 m on the upper plateau and adjacent south-

facing lower slopes in the centre of the study area

Bothalia 15, 3 & 4 (1985)

679

(Fig. 3). It is represented by six releves with 24 to 27

species per releve. This open-woodland variation

(Edwards, 1983) has an L structure (Ito, 1979; Fig.

4q) with the greatest average cover of 10% in the 0,0

m-0,5 m height class. Being situated adjacent to the

upper plateau, probably subjected to marked tem-

perature inversion at night, this community can be

expected to have a wide temperature range and rela-

tively dry moisture status.

Habitat

The soils are mainly of the Mispah Form, Mispah

Series, derived from sandstone of the Aasvoelkop

Formation but releves 22 and 23 represent soils of

the Westleigh Form, Sibasa Series, also derived

from sandstone of the Aasvoelkop Formation. The

soils of the Westleigh Form appear in isolated pock-

ets and are not large in extent, resulting in no appar-

ent vegetation change at the scale of this study. The

average soil depth varies from 80 mm to deeper than

1 000 mm and the surface rock cover is 4% or less.

The terrain slopes up to 5° in a southerly direction

with one releve (17) representing a 2° slope in a

northerly direction. The electrical resistance of the

soils is 2 100 ohms to 2 800 ohms and the T-value

varies from 3,4 me/100 g to 9,3 me/100 g indicating a

moderate nutrient status for the Mispah Form soils

and a low nutrient status for the Westleigh Form

soils. The soils are moderately acid (MacVicar et a!.,

1977) with a pH of 6,0 when saturated with water.

Floristics

This community is distinguished by the Pachycar-

pus schinzianus species-group (Table 2S). The

species diversity per unit area averages 7,3 species/

nrfor the Community.

Trees and shrubs

The only conspicuous woody species with 2,5% or

more mean cover and occurring in 50% of the re-

leves representing this community is:

Combretum molle (tree/shrub) 67% 2,5%

Herbs

Herb species occurring in more than 50% of the

releves representing the community are:

General

Communities 13 to 16 are related to each other

through the shared presence of the Senecio erubes-

cens species-group (Table 2V) and Communities 8 to

16 are related to each other through the shared pres-

ence of the Schizachyrium sanguineum species-

group (Table 2W).

15. Protea caffra - Combretum molle Open

Woodland

This woodland is found at altitudes of 1 425 m to

1 575 m on the upper plateau and the adjacent

south-facing lower slopes in the northern half of the

study area (Fig. 3). It is represented by eight releves

with 18 to 30 species per releve. This open-woodland

community (Edwards, 1983) has an L structure (Ito,

1979; Fig. 4r) with the greatest cover of 20% in the

0,0m-0,5 m height class. Being situated adjacent to

the upper plateau, this community can be expected

to have a wide temperature range and dry moisture

status.

Habitat

The soils are mainly of the Mispah Form, Mispah

Series, derived mainly from sandstone of the Aas-

voelkop Formation but also conglomerate, in the

case of releve 78, of the Alma Graywacke Forma-

tion. Releve 14 represents soils of the Westleigh

Form, Sibasa Series, derived from sandstone of the

Aasvoelkop Formation. The average soil depth va-

ries from 80 mm to deeper than 1 000 mm and the

surface rock cover varies from 1% to 15%. The ter-

rain slopes from 1° to 19° in a south to south-easterly

direction with releve 78 on the upper plateau rep-

resenting a slope of 2° in a northerly direction. The

electrical resistance of the soils varies from 2 100

ohms to 4 700 ohms and the T-value varies from 3,4

me/lOOg to 9,3 me/100 g, indicating soils with a low

to moderate nutrient status. The soils are moder-

ately to strongly acid (MacVicar et al. ,1977) with a

pH range of 5,0 to 6,0, when saturated with water.

Floristics

This community is distinguished by the absence of

character species and the absence of the Vernonia

oligocephala species-group (Table 2R), the Pachy-

carpus schinzianus species-group (Table 2S) and the

Cyperus denudatus species-group (Table 2T) to-

gether with the presence of the Senecio erubescens

species-group (Table 2V). The species diversity per

unit area averages 6,5 species/m: for this com-

munity.

Trees and shrubs

All woody species have less than 2,5% mean

cover. Woody species occurring in more than 50% of

the releves representing this community are:

Protea caffra (tree/shrub) 75% 1,9%

Rhus dentata (shrub) 63% 0,5%

Herbs

Herb species occurring in more than 50% of the

releves representing this community are:

General

Communities 13 to 16 are related to each other

through the shared presence of the Senecio erubes-

680

Bothalia 15, 3 & 4 (1985)

cens species-group (Table 2V) and Communities 8 to

16 are related to each other through the shared pres-

ence of the Schizachyrium sanguineum species-

group (Table 2W). The Stoebe vulgaris species-

group (Table 2U) indicates a partial affinity between

Communities 15 & 16.

C. Grassland phase of Acocks’s (1975) Sour Bushveld

on moderately deep soils, in exposed dry habitats

The grassland phase of Acocks’s (1975) Sour

Bushveld is represented by one community in the

study area, viz the Eragrostis pallens - Andropogon

appendiculatus Grassland.

16. Eragrostis pallens - Andropogon appendicula-

tus Grassland

This grassland is found at altitudes of 1 400 m to

1 465 m on the upper plateau (Fig. 7) in the central

part of the study area (Fig. 3). It is represented by

twelve releves with 12 to 28 species per releve. This

grassland community (Edwards, 1983) has an L

structure (Ito, 1979; Fig. 4s) with the greatest aver-

age cover of 23% in the 0,0 m-0,5 m height class.

However, releve 1 represents vegetation classified as

dwarf shrubland (Edwards, 1983) because of the

presence, of Stoebe vulgaris which has a mean cover

of 42% in this releve (Fig. 8). The upper plateau is

possibly subjected to temperature inversion at night

because of the slopes to the north and south, result-

ing in wide temperature fluctuations. Except for the

rainy season, when the water table is high or at the

soil surface for large areas of the upper plateau, this

community is very dry, possibly because of the ex-

posed nature of the habitat and lack of high vegeta-

tion cover.

Habitat

The soils are mainly of the Mispah Form, Mispah

Series, derived from shale of the Aasvoelkop For-

mation and the Kroonstad Form, Sibasa Series, de-

rived from shale of the Aasvoelkop Formation and

conglomerate of the Alma Graywacke Formation.

FIG. 7. — The Eragrostis pallens -

Andropogon appendiculatus

Grassland in the foreground

with the north-facing slopes in

the background.

FIG. 8. — Stoebe vulgaris Closed

Dwarf Shrubland in the fore-

ground, represented by releve

1.

Bothalia 15, 3 & 4 (1985)

681

The average soil depth varies from 150 mm to

deeper than 1 000 mm but the impermeable shale

causes a high water table, varying from 150 mm to

deeper than 1 000 mm at the time of sampling, that

could have a limiting effect on maximum root depth.

No surface rocks were recorded for this community.

The terrain slopes up to 3° mainly in a south-easterly

to south-westerly direction but also northerly in the

case of releve 8. The electrical resistance varies from

3 100 ohms to 4 800 ohms and the T-value varies

from 5,9 me/100 g 1 1 ,8 me/100 g indicating soils of a

predominantly high nutrient status, possibly owing

to cation adsorption by the clay fraction derived

from the shale. The soils are strongly acid (MacVicar

et al. , 1977) with a pH of 4,5 to 4,7 when saturated

with water.

Floristics

The community is diagnosed by the Cyperus denu-

datus species-group (Table 2T). The species diver-

sity per unit area averages 6,1 species/m2 for the

Community.

Trees and shrubs

Only an isolated Protect caffra tree was recorded

for this community. The dwarf shrub Stoebe vulgaris

with a constancy of 53% has a mean cover of 4% for

the releves representing this community. However,

in releve 1 , Stoebe vulgaris has a mean cover of 42% ,

resulting in this releve being classified as closed

dwarf shrubland. If releve 1 is not taken into ac-

count, the mean cover of Stoebe vulgaris is less than

1% for the community as a whole, hence the com-

munity is classified as grassland. The floristic compo-

sition of releve 1 is such that it forms a part of Com-

munity 16 (Table 2). Stoebe vulgaris can be regarded

as an invader species, without which releve 1 would

have potentially closer floristic affinities with Com-

munity 16. The presence of Stoebe vulgaris causes

much concern amongst the local 'farmers because of

its encroachment upon grassland used for grazing.

Herbs

Herb species occurring in more than 50% of the

releves representing this community are:

through the shared presence of the Senecio erubes-

cens species-group (Table 2V) and Communities 8 to

16 are related to each other through the shared pres-

ence of the Schizachyrium sanguineum species-

group (Table 2W). The Stoebe vulgaris species-

group (Table 2U) indicates a partial affinity between

Communities 15 & 16. Roux (1969) suggests that se-

lective grazing of grass results in grass tufts often

being left ungrazed, which provides essential shade

for Stoebe vulgaris seed germination. Where grass is

grazed short and species composition is not affected,

i.e. non-selective grazing (Acocks, 1966), Stoebe

vulgaris will be unable to germinate for lack of

shade.

D. Woodland phase of Acocks’s (1975) North-East-

ern Mountain Sourveld on moderately shallow soils,

in moderately exposed habitats

The woodland phase of Acocks’s (1975) North-

Eastern Mountain Sourveld is represented by one

community in the study area, viz the Helichrysum

nudifolium - Protea roupelliae Sparse Woodland.

17. Helichrysum nudifolium - Protea roupelliae

Sparse Woodland

This woodland (Fig. 9) is found at altitudes of

1 600 m to 1 900 m in the northern part of the study

area (Fig. 3). It is represented by 17 releves with 17

to 32 species per releve. Releves 120 and 118, al-

though floristically different from the rest of the

community, are not regarded as a separate variation

because of the lack of character species and because

only two releves are different. In order to ascertain

whether structural differences exist, the two differ-

ent floristic units were illustrated separately as Fig.

4t, for releves 120 and 118, and Fig. 4u for the rest of

the community. This sparse-woodland community

(Edwards, 1983) has an L structure (Ito, 1979; Figs

4t and 4u) with the greatest average cover of 36% for

releves 120 and 118, and 30% for the other releves in

the community, in the 0,0 m to 0,5 m height class.

This community is exposed, being situated at a high

altitude and considerable temperature fluctuation

could be expected. The moisture regime, however,

should not be the driest in the study area although

this community is much exposed, because of the fre-

quent occurrence of mist on the upper reaches of the

Kransberg massif, as observed during the course of

fieldwork.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Aasvoelkop and

Sandriviersberg Formations. The average soil depth

varies from 70 mm to 180 mm and the surface rock

cover varies from 50% to 80%. The terrain slopes up

to 38° in a south-easterly to westerly direction. Re-

leve 122 representing the 0° slope is situated on a flat

area at 1 750 m altitude. The electrical resistance of

the soils varies from 4 300 ohms to 8 300 ohms and

the T-value varies from 6,0 me/100 g to 11,8 me/100 g

indicating a low to high nutrient status. It is sug-

gested that the high nutrient status occurs in bottom-

lands where runoff from the interfluves causes de-

pletion of nutrients on the interfluves and accumula-

tion of nutrients in the bottomlands. The soils are

strongly acid (MacVicar et al., 1977) with pH from

4,4 to 4,9 when saturated with water.

Floristics

This community is differentiated by the Helichry-

sum nudifolium species-group (Table 2X) which

682

Bothalia 15, 3 & 4 (1985)

FIG. 9. — The Helichrysum nudi-

folium - Protea roupelliae

Sparse Woodland on the up-

per slopes of the Kransberg

massif.

does not occur in releves 120 and 118. The species

diversity per unit area averages 6,2 species/nr for

this community. It appears from aerial photographs

of the study area that the vegetation represented by

releves 120 and 118 falls into an apparent permanent

shadow caused by the Kransberg cliffs and with ad-

ditional sampling could prove to be a variation of

Community 17.

Trees and shrubs

The only conspicuous woody species occurring in

50 per cent of the releves representing this com-

munity is:

Protea caffra (tree/shrub) 71% 2,8%

The mean percentage cover for Protea caffra is

higher than indicated for a sparse woodland (Ed-

wards, 1983). However, the high cover can be attri-

buted to individual plants, under 2 m tall, as is also

the case with Protea roupelliae which has a 47% con-

stancy and 1,3% mean cover.

Herbs

Herb species occurring in more than 50% of the

releves representing this community are:

General

Communities 17 & 18 are related to each other

through the common presence of the Protea roupel-

liae species-group (Table 2Z) and the Helichrysum

kraussii species-group (Table 2 A A), but the former

species-group is not represented in releves 120 and

118, whereas the latter is represented in these two

releves. Variation 11.2 and Communities 12, 17 and

18 are related to each other through the shared pres-

ence of the Helichrysum sp. Westfall 921 species

group (Table 2AB).

The high altitude and steep slope on which this

community is found, together with the high surface

rock cover make accessibility and hence grazing by

cattle difficult. The grass is not moribund, however,

probably owing to periodic accidental fires.

E. Grassland phase of Acocks’s (1975) North-Eastern

Mountain Sourveld on shallow rocky soils in exposed

habitats.

The grassland phase of Acocks’s (1975) North-

Eastern Mountain Sourveld is represented by one

community in the study area, viz the Eragrostis race-

mosa - Trachypogon spicatus Grassland.

18. Eragrostis racemosa - Trachypogon spicatus

Grassland.

This grassland (Fig. 10) is found at altitudes of

1 900 m to 2 080 m in the extreme north of the study

area (Fig. 3). It is represented by 15 releves with 19

to 35 species per releve. This grassland community

(Edwards, 1983) has an L structure (Ito, 1979; Fig.

4V) with the greatest average cover of 18% in the

0,0 to 0,5 m height class. The community is exposed,

being situated on the upper summit, and a wide tem-

perature range could be expected. As in the case of

Community 17, the moisture regime should not be

the lowest in the study area, although the com-

munity is exposed, because of frequent mists, as ob-

served during the course of fieldwork.

Habitat

The soils are of the Mispah Form, Mispah Series,

derived from sandstone of the Sandriviersberg For-

mation. The average soil depth varies from 40 mm to

Bothalia 15, 3 & 4 (1985)

683

FIG. 10. — The Eragrostis race-

mosa - Trachypogon spicatus

Grassland with an isolated

Podocarpus latifolius in the

centre.

100 mm and the surface rock cover varies from 60%

to 90%. The terrain slopes up to 14° mainly in a

south-westerly through northerly to south-easterly

direction. The absence of a southerly aspect may be

attributed to the cliff face to the south of the upper

summit. The electrical resistance of the soils is 4 200

ohms and the T-value is 11,0 me/100 g indicating

soils of a high nutrient status which may be attri-

buted to the predominantly flat nature of the upper

summit which inhibits runoff and consequent loss of

nutrients. The soils are strongly acid (MacVicar et

al., 1977) with a pH of 5,2 when saturated with

water.

Floristics

This community is distinguished by the Trachypo-

gon spicatus species-group (Table 2Y). The species

diversity per unit area averages 6,0 species/m: for

this community.

Trees and shrubs

Isolated Protea roupelliae trees and shrubs occur

in 60% of the releves representing this community

and the shrublet Fadogia monticola occurs with 60%

constancy. Isolated Podocarpus latifolius trees are

found in boulder clumps and the shrub Widdringto-

nia nodiflora occurs in 33% of the releves. The can-

opy cover of the woody species is, however, less than

0,1 per cent, resulting in this community being classi-

fied as a grassland (Edwards, 1983).

Herbs

Herb species occurring in more than 50% of the

releves representing this Community are:

General

Communities 17 & 18 are related to each other

through the common presence of the Protea roupel-

liae species-group (Table 2Z) and the Helichrysum

kraussii species-group (Table 2AA). Variation 11.2

and Communities 12, 17 and 18 are related to each

other through the shared presence of the Helichry-

sum sp. species-group (Table 2AB).

DISCUSSION AND CONCLUSIONS

The present study has resulted in a classification of

the vegetation and a correlation of the main environ-

mental factors influencing the vegetation (Westfall et

al., 1983). Although the vegetation is predominantly

open woodland, the formation classes found in the

study area range from forests to grasslands, with a

diversity of communities, along a temperature/mois-

ture gradient. Soil depth also appears to play an im-

portant role in community differentiation. The kloof

forest communities are diagnosed by the species Cel-

tis africana and Diospyros whyteana, whereas the

woodland communities representing Acocks’s

(1975) Sour Bushveld are distinguished by species

such as Combretum molle, Faurea saligna, Ozoroa

paniculosa and Heteropyxis natalensis. Species such

as Burkea africana, Ochna pulchra and Strychnos

pungens represent the Aristida aequiglumis species-

group (Table 2Q) and have a more restricted range

than the species that distinguish the woodland phase

of Acocks’s (1975) Sour Bushveld. The ordination of

communities (Westfall et al., 1983) shows that the

communities characterized by the Aristida aequiglu-

mis species-group (Table 2Q) occupy a central posi-

tion on the temperature/moisture gradient. It can,

684

Bothalia 15, 3 & 4 (1985)

therefore, be inferred that the extremes of the tem-

perature/moisture gradient are limiting factors for

the species of the Aristida aequiglumis species-group

(Table 2Q). The grassland phase of Acocks’s (1975)

Sour Bushveld, in the study area, appears to be re-

lated to the reduced effective soil depth resulting

from seasonally high water tables. If the drainage

were to improve, it is possible that an open-wood-

land formation could become established. Tinley

(1977 & 1982) also attributes the occurrence of

grasslands in dambos to seasonal waterlogging.

The communities representing Acocks’s (1975)

North-Eastern Mountain Sourveld are found above

1 501 m altitude in the study area. Acocks (1975) de-

scribes North-Eastern Mountain Sourveld as having

had a high-forest climax. Isolated Podocarpus latifo-

lius trees are found in Community 17, amongst boul-

der clumps (Fig. 11). The boulders probably afford

protection from fire. Edwards (1967) describes the

occurrence of Podocarpus latifolius amongst boulder

clumps in the Protea savanna in Natal as developing

forest clumps. It is likely that Podocarpus latifolius

could have a higher mean percentage cover, than at

present, if protected from fire in Community 17,

with the possibility of forest as a climax.

A comparison of communities described by The-

ron (1973), Coetzee (1975), Coetzee et al., (1976)

and Van der Meulen (1979), who all worked in

Transvaal Bushveld, shows that none of the com-

munities are similar to communities described in this

study, in terms of character species for the communi-

ties. A complete floristic comparison of the com-

munities described by the aforementioned authors

and those in this study would entail joint synthesis of

all the communities involved.

It would appear from the community descriptions

that in an open bushveld situation, woody species

encroachment could become a problem for grazing

management, when the canopy cover of the woody

species is less than two crown diameters apart or

more than 9% canopy cover forming a closed wood-

land. Although the floristic composition of the grass

stratum appears different in Acocks’s (1975) Sour

Bushveld as represented in the study area, when the

canopy cover of the tree stratum is greater than 20%

for any height class (Fig. 4), it is likely that floristic

change could take place before a 20% canopy cover

is achieved.

The removal of the tree ferns, Cyathea dregei,

from the larger kloof in which Community 3 is

found, emphasizes the need for protection of the

smaller kloof in which Cyathea dregei still occurs, if

the species is to be conserved in its natural habitat in

the study area. Although there is an abundance of

streams in the study area, none are perennial. The

experience gained at the Thabamhlope Research

Station, near Estcourt in Natal (Westfall et al.,

1982b) indicates that protection of kloofs from graz-

ing and fire should prolong the periods in which

streams flow, which would be of benefit to the farms

south of Kransberg.

The agricultural potential of the study area ap-

pears to be limited because of the limited extent of

deep soils, generally high surface-rock cover and

lack of perennial free water. Furthermore, the grass

cover is not suited to year-round grazing because of

its sour nature (Booysen, 1967). That potential for

agriculture appears limited, is supported by the

many part-time farmers in the study area. Paddock

fences generally do not follow natural community

boundaries. This results in portions of camps being

selected for grazing far more than others. Fences

should be erected so that the enclosed vegetation is

as homogeneous as possible to encourage uniform

grazing. Where communities, such as Community

16, have local differences in veld condition with re-

spect to grazing, these could also be separated by

fences to ensure uniform grazing.

Recommendations based on this study may be

summarized as follows:

1. Species requiring conservation in their natural

habitats include Cyathea dregei as well as Encepha-

lartos eugene-maraisii which has been reported in the

FIG. 11. — An isolated Podocar-

pus latifolius in a boulder

clump in the Hclichrysum nu-

difolium - Protea roupelliae

Sparse Woodland.

Bothalia 15, 3 & 4 (1985)

685

Waterberg (Coates Palgrave, 1977) and on the upper

north-facing slopes of the Kransberg massif by local

farmers, but not observed during the course of this

study.

2. Protection of the catchment areas, from grazing

by means of fences and from fire by means of fire-

breaks, to improve the water supply.

3. Improvement of species composition of the

veld to enhance its grazing value and to overcome

problems caused by invader species such as Stoebe

vulgaris by ensuring that paddock fences coincide

with community boundaries where feasible and ad-

justing grazing pressure as the proportion of In-

creaser species and Decreaser species dictates.

4. Provision of facilities for, and control of activi-

ties of sightseers, so that the landscape is not littered

and species such as Cyathea dregei are not removed .

Conservation of the entire study area could

achieve the abovementioned recommendations and

might be feasible because of the apparently low agri-

cultural potential of the study area. Furthermore,

conservation would ensure that Acocks’s Sour Bush-

veld, which is poorly conserved (Edwards, 1974),

could be represented in the Waterberg where the

greatest area of this veld type is found (Acocks,

1975). The study area includes the Kransberg massif

which is frequently visited by members of mountain

clubs, as witnessed during the course of fieldwork,

and is considered by mountain-club members to be

probably the best rock-climbing facility in the Trans-

vaal. The study area is approximately 200 km from

Pretoria which is an accessible range for visitors

from Pretoria, Johannesburg and vicinity. Introduc-

tion of grazing game into the area would not im-

prove the grass species composition without a graz-

ing policy being implemented. Several private game

reserves in the Waterberg area have a woody species

canopy cover, including Dichrostachys cinerea thick-

ets which, from cursory observation, appear to be

impenetrable. A grazing policy in which the species

composition is frequently monitored and grazing

pressure adjusted accordingly is, therefore, essen-

tial. A further advantage of conservation of the

study area is that the area could provide emergency

grazing for farms in the Mixed and Sourish Mixed

Bushveld (Acocks, 1975), in times of drought, but

not on a permanent basis.

In conclusion, the classification of the farm Groot-

hoek, Thabazimbi District, has revealed a diversity

of plant communities and habitats requiring con-

tinual surveillance by farmers and conservation

authorities if the full potential of the area is to be

realized.

ACKNOWLEDGEMENTS

The authors thank Dr J. C. Scheepers for com-

ments and suggestions, the staff at the National Her-

barium in Pretoria for the identification of plant

specimens and Mrs J. Schaap for drawing the fig-

ures.

UITTREKSEL

Die plantegroei van die plaas Groothoek, Thaba-

zimbi-distrik wat in die Suurbosveld van die Trans-

vaalse Waterberg gelee is, is volgens die Braun-Blan-

quet metode, met gebruik van die FITOTAB pro-

gramme, geklassifiseer. Vyf hoof plantegroei-tipes

met agtien gemeenskappe is beskryf met verwysing na

die hoof omgewingsfaktore wat die plantegroei bei'n-

vloed, asook na plantegroeistruktuur.

REFERENCES

ACOCKS, J. P. EL, 1966. Non-selective grazing as a means of

veld reclamation. Proc. Grassld Soc. sth. Afr. 1: 33-39.

ACOCKS, J. P. H., 1975. Veld types of South Africa. 2nd edn.

Mem. bot. Sur\!. S. Afr. 40: 1-128.

BOOYSEN, P. de V., 1967. Grazing and grazing management

terminology in southern Africa. Proc. Grassld Soc. sth. Afr.

2: 45-57.

COATES PALGRAVE, K., 1977. Trees of southern Africa.

Cape Town: Struik.

COETZEE, B. J., 1975. A phytosociological classification of the

Rustenburg Nature Reserve. Bothalia 11: 561-580.

COETZEE, B.J., VAN DER MEULEN, F., ZWANZIGER, S.,

GONSALVES. P. & WEISSER, P. J., 1976. A phytosocio-

logical classification of the Nylsvley Nature Reserve. Botha-

lia 12: 137-160.

EDWARDS, D. , 1967. A plant ecology survey of the Tugela Ba-

sin. Mem. bot. Surv. S. Afr. 36: 1-285.

EDWARDS, D., 1974. A survey to determine the adequacy of

existing conserved areas in relation to vegetation types: a

preliminary report. Koedoe 17: 2-37.

EDWARDS, D., 1983. A broad-scale structural classification of

vegetation for practical puposes. Bothalia 14: 705-712.

HILL, M. O., 1979. DECORAN A — A FORTRAN program for

detrended correspondence analysis and reciprocal averaging.

Unpublished report. Ecology and Systematics, Cornell Uni-

versity, Ithaca, New York.

HILL, M. O. & GAUCH, H. G., 1980. Detrended correspon-

dence analysis: an improved ordination technique. Vegetatio

42: 47-58.

ITO, K., 1979. A tentative study of stratification diagrams. In A.

Miyawaki & S. Okuda, Vegetation und Landschaft Japans.

Yokohama: Yokohama Phytosociological Society.

MACVICAR, C. N., DE VILLIERS, J. M. LOXTON, R. F.,

VERSTER, E.. LAMBRECHTS, J. J. N., MERRY-

WEATHER, F. R.. LE ROUX, J., VAN ROOYEN, T. H.

& HARMSE, H. J. VON M., 1977. Soil classification: a bi-

nomial system for South Africa. Pretoria: Department of

Agricultural Technical Services.

MUELLER-DOMBOIS, D. & ELLENBERG, H., 1974. Aims

and methods of vegetation ecology. New York: Wiley.

ROUX, E., 1969. Grass: a story of Frankenw aid. Cape Town:

Oxford University Press.

RUSSELL, E. J., 1961 Soil conditions and plant growth. London:

Longman.

SCHULZE. B. R., 1947. The climates of South Africa according

to the classification of Koppen and Thornthwaite. S. Afr.

Geogr. J. 29: 32—42.

SOUTH AFRICAN COMMITTEE FOR STRATIGRAPHY

(SACS), 1980. Stratigraphy of South Africa. Part 1. Litho-

stratigraphy of the Republic of South Africa, South West

Africa/Namibia and the Republics of Bophuthatswana,

Transkei and Venda. Handb. geol. Surv. S. Afr. 8: 1-690.

THERON, G.K., 1973. ’n Ekologiese studie van die plantegroei

van die Loskopdamnatuurreservaat. D.Sc. thesis. University

of Pretoria.

TINLEY, K. L., 1977. Framework of the Gorongosa ecosystem.

D.Sc. thesis. University of Pretoria.

TINLEY, K.L., 1982. The influence of soil moisture balance on

ecosystem patterns in southern Africa. In B. J. Huntley & B.

H. Walker, Ecology of tropical savannas. Berlin: Springer.

VAN DER MEULEN. F.. 1979. Plant sociology of the western

Transvaal Bushveld, South Africa: syntaxonomic and syn-

ecological study. Dissertationes Botanicae 49. Vaduz:

Cramer.

VAN DER MEULEN. F. & WESTFALL. R. H., 1980. Structu-

ral analysis of bushveld vegetation in Transvaal, South

Africa, j. Biogeogr. 7: 337-348.

WERGER, M. J. A., 1974. On concepts and techniques applied

in the Zurich-Montpellier method of vegetation survey. Bo-

thalia 11: 309-323.

686

Bothalia 15, 3 & 4 (1985)

WESTFALL, R. H.. DEDNAM, G., VAN ROOYEN, N. &

THERON G. K., 1982. PHYTOTAB — A program pack-

age for Braun-Blanquet tables. Vegetatio 49: 35-37.

WESTFALL, R. H., EVERSON, C. S. & EVERSON, T. M.,

1982. The vegetation of the protected plots at Thabamhlope

Research Station. S. Afr. J. Bot. 2: 15 — 25.

WESTFALL, R. H., VAN ROOYEN, N. & THERON, G. K.,

1983. Veld condition assessments in Sour Bushveld. Proc.

Grassld Soc. sth. Afr. 18: 73-76.

WESTHOFF, V. & VAN DER MAAREL, E., 1973. The Braun-

Blanquet approach. In R. H. Whittaker, Handbook of vege-

tation science. The Hague: Junk.

APPENDIX: EXPLANATION OF CLASS SYMBOLS USED IN TABLE 2

Vegetation Formation

Feature

Canopy cover

(classes, according to

Edwards, 1983)

Formation

(classes, according to

Edwards, 1983)

Habitat Data

Landform

Symbol

C

O

S

F

W

S

D

G

CF

D

E

H

Class

Closed (0-2 canopy diameters apart)

Open (> 2-8 canopy diameters apart)

Sparse (> 8-27 canopy diameters apart)

Forest (< 0 canopy diameters apart)

Woodland

Shrubland

Dwarf shrubland

Grassland

Summit: upper

lower

Plateau: upper

lower

Cliff face

Upper slope

Lower slope

Steep bank/kloof

Bothalia 15, 3 & 4 (1985)

687

Chemical analysis of the A-Horizon

688

Bothalia 15, 3 & 4 (1985)

Bothalia 15, 3 & 4: 689-703 (1985)

The development from kinetic coefficients of a predictive model for the

growth of Eichhornia eras sipes in the field. I. Generating kinetic

coefficients for the model in greenhouse culture

C.F. MUSIL* and C.M. BREEN**

Keywords : Eichhornia crassipes , growth kinetics, model development

ABSTRACT

The kinetics of N- and P- limited growth of Eichhornia crassipes (Mart.) Solms were investigated in greenhouse

culture with the object of developing a model for predicting population sizes, yields, growth rates and frequencies

and amounts of harvest, under varying conditions of nutrient loading and climate, to control both nutrient inputs

and excessive growth in eutrophied aquatic systems. The kinetic coefficients, maximum specific growth rate

(Umax), half saturation coefficient (Ks) and yield coefficient (Yc) were measured under N and P limitation in

replicated batch culture experiments. Umax values and Ks concentrations derived under N limitation ranged from

5,37 to 8,86% d'1 and from 400 to 1 506 pig N tl respectively. Those derived under P limitation ranged from 4,51

to 10,89% d1 and from 41 to 162 pig P C respectively. Yc values (fresh mass basis) determined ranged from 1 660

to 1 981 (87 to 98 dry mass basis) for N and from 16 431 to 18 671 (867 to 980 dry mass basis) for P. The reciprocals

of Yc values (dry mass basis), expressed as percentages, adequately estimated the minimum limiting concentra-

tions of N and P (% dry mass) in the plant tissues. Kinetic coefficients determined are compared with those re-

ported for algae. The experimental method used and results obtained are critically assessed.

INTRODUCTION

Eutrophication, the enrichment of aquatic systems

with inorganic nutrients (Stewart & Rohlich, 1967),

is a world-wide water quality problem (Stumm,

1974) . Eichhornia crassipes (Mart.) Solms (water

hyacinth), a free-floating, aquatic plant (Penfound

& Earle, 1948; Bock, 1966), which has a high growth

rate (Penfound, 1956; Yount & Crossman, 1970;

Boyd, 1976) and produces a large standing crop per

unit area (Knipling et al., 1970; Boyd & Scarsbrook,

1975) , is the most promising floating, vascular aqua-

tic plant species for removing nutrients from eutro-

phied aquatic systems (Boyd, 1970). This species ab-

sorbs large quantities of N and P, the nutrients gen-

erally associated with eutrophication (Mackenthun,

1964; 1965), from sewage effluents (Clock, 1968;

Miner et al., 1971; Cornwell et al., 1977). In addi-

tion, it removes heavy metal and other chemical pol-

lutants from secondary waste-water effluents (Wol-

verton, 1975; Wolverton & McDonald, 1975a;

1975b; 1976; Wolverton & McKown, 1976) and re-

duces levels of suspended solids, biochemical oxygen

demand substances and other chemical factors in

such effluents to levels below the standards set by

some pollution control agencies (Wolverton & Mc-

Donald, 1975c; 1975d). Its cultivation and removal

may, therefore, constitute an effective means of

withdrawing nutrients from effluents prior to their

release into natural waters (Yount & Crossman,

1970). Similarly, the removal of water hyacinths

growing in eutrophied aquatic systems may also

assist in controlling excessive growth of plants by re-

ducing nutrient levels.

To achieve maximum nutrient removal efficiency

by E. crassipes in a nutrient removal scheme, it is

*Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag, X101. Pretoria 0001.

**Botany Department, University of Natal, P.O. Box 375, Pieter-

maritzburg 3200.

necessary to establish how much and how frequently

to harvest the population. Clearly, if the population

is continually over-harvested, the size of the popula-

tion and its effectiveness in removing nutrients will

be progressively reduced. Alternatively, if the popu-

lation is under-harvested, nutrient removal may be

ineffective and other adverse effects may arise.

Maintenance of a high growth rate and nutrient

removal capacity by E. crassipes is facilitated if the

size of the population required to maintain desirable

nutrient concentrations in the water, under varying

conditions of nutrient loading and climate, can be

predicted. Since harvesting is required to control the

population size, amounts and frequencies of harvest

must also be predicted.

From the kinetic standpoint, it is theoretically

feasible to construct a mathematical model for E.

crassipes from which population sizes, yields, growth

rates and frequencies and amounts of harvest, under

varying conditions of nutrient loading and climate,

can be predicted to control both nutrient inputs and

excessive growth in eutrophied aquatic systems

(Toerien, 1972; Musil & Breen, 1977). The follow-

ing relationships, however, require mathematical

formulation:

(i) The relationship between the yield of E. cras-

sipes, i.e. the mass of plant material produced and

the mass of a specific *limiting nutrient absorbed.

The following mathematical expression describes

this relationship:

So - St

where Yc = yield coefficient; Xo = initial biomass;

Xt = final biomass; So = initial concentration of

*Nutrient present at concentrations below that required for maxi-

mum plant growth and hence restricting the growth rate.

690

Bothalia 15, 3 & 4 (1985)

limiting nutrient; St = final concentration of limiting

nutrient.

(ii) The relationship between the specific growth

rate of E. crassipes, i.e. the increase in mass of

plants, per unit mass of plant material, per unit time

(Malek & Fencl, 1966; Radford, 1967) and the con-

centration of a specific limiting nutrient. Various

models have been used to quantify this relationship

in algae and bacteria (Shelef et al ., 1968; Toerien et

al., 1971; Goldman, 1972). The most important are

Blackman’s first order-zero order model, Teisser’s

exponential model and Monod’s rectangular hyper-

bola model, the last defined as:

where U = specific growth rate; Umax = maximum

specific growth rate; S = concentration of limiting

nutrient; Ks = half saturation coefficient = S when

U = 0,5 Umax.

(iii) The relationship between the maximum speci-

fic growth rate of E. crassipes and temperature.

Under a constant light intensity, the maximum speci-

fic growth rate (Umax) may be described solely as a

function of temperature, as shown by Goldman

(1972) and Goldman & Carpenter (1974) for various

species of marine and fresh water algae, by an Ar-

rhenius equation, defined as:

Umax = AeE/RT

where A = constant day-1; E = activation energy

cal. mole-1; R = universal gas constant cal. mole-1

°K-1; T = temperature on Kelvin scale °K.

Incorporating the Arrhenius equation into the

Monod model, the following mathematical expres-

sion is obtained in which the specific growth rate (U)

is related to both temperature and the limiting nutri-

ent concentration:

U = Ae-E/RT x —

Ks + S

The predictive abilities of such models have been

demonstrated in algae, for example, by Toerien &

Huang (1973) where the P-limited growth rate of

Selenastrum capricornutum in batch cultures was ac-

curately predicted from its kinetic coefficients and

by Bhagat et al. ( 1972) where the algal concentration

of a Vancouver Lake was adequately predicted by a

water quality simulation model also using kinetic co-

efficients.

A number of restrictions to the general use of the

above equations, however, do exist. Firstly, for each

plant species the Arrhenius equation is applicable

only over a defined temperature range as shown by

Sorokin (1960) for various algal species. Secondly,

there is evidence of a strong interaction between

light intensity and temperature. Sorokin (1960;

1971) found that for a given temperature the acti-

vation energy decreases with increasing light energy

and Shelef (1968) has shown that the saturation light

intensity is highly temperature dependent. Thirdly,

the half saturation coefficient for nutrient uptake is

also sensitive to changes in temperature (Shelef et

al., 1970). A further potential complication is the

possible temperature dependency of the yield coeffi-

cient, since minor variations in the yield coefficient

have been found with high and low temperature

strains of Chlorella grown under N03-N limitation

in continuous cultures (Shelef et al. , 1970) and in the

bacterium Aerobacter aerogenes (Topiwala & Sin-

clair, 1971).

No attempts have, as yet, been made to model the

effects of temperature on the half saturation (Ks)

and yield coefficients (Yc), although in Aerobacter

aerogenes and Escherichia coli, Topiwala & Sinclair

(1971) and Sawada et al. (1978) demonstrated that

Ks changes with temperature and that an Arrhenius

plot of the change is linear. The difficult task of de-

termining temperature dependent kinetic coeffi-

cients such as Ks and Yc in natural systems may re-

strict their application to well defined laboratory

conditions. On a seasonal basis, however, it should

be possible to assess the significance of these kinetic

coefficients in modelling.

Numerous references exist in the literature on the

nutrient uptake and growth characteristics of E.

crassipes. Despite this, the necessary mathematical

relationships required for the proper evaluation and

potential design of a predictive model have not been

adequately formulated. In a preliminary study, Mu-

sil & Breen (1977) measured the kinetic coefficients,

Umax, Ks and Yc for E. crassipes in one N03-N-

limited batch culture experiment. They illustrated

how these coefficients could be used in a predictive

model, although its validity was not tested under

field conditions. Since both N and P are the nutrients

most frequently limiting for E. crassipes under natu-

ral conditions (Wahlquist, 1972), this investigation

was designed to generate kinetic coefficients for E.

crassipes growing under N and P limitation with the

objective of developing and validating a predictive

model.

MATERIALS AND METHODS

The batch culture method or non steady-state ap-

proach (Toerien et al., 1971) was used to measure

kinetic coefficients for E. crassipes growing under

specific nutrient limitation. Batch culture experi-

ments were repeated, five times under N and three

times under P limitation.

In each experiment, ca 120, vegetatively-propa-

gated offsets (daughter plants) of uniform size (pos-

sessing two pseudolaminae with bulbous petioles

and having a fresh mass ranging from ca 4 to 10 g)

were sampled from a loosely crowded population in

a sewage maturation pond. Plants were rinsed

through three changes of deionised-distilled water,

shaken to dislodge adhering water and their fresh

masses recorded on an electric, top-loading balance.

They were placed into 5€ capacity, inert polyethy-

lene vessels (buckets) each containing 5€ of culture

solution deficient in either N or P. One or two plants

were used as an inoculum in each vessel (Table 1).

A modified culture solution based on that of Ham-

ner et al. (1942) was used (Table 2) in which the con-

centrations of either of the anions, NO3 or

PO47 could be varied independently with minimum

influence on the concentrations of cations and other

anions. Reduced cation concentrations, resulting

Bothalia 15, 3 & 4 (1985)

691

TABLE 1. — Treatment differences between experiments designed to measure kinetic coefficients

for E. crassipes growing in N-and P-limited cultures

from the lowering in concentration of an anion in the

culture solution, were restored by supplementing it

with the appropriate additional cations. These were

added predominantly as chlorides. The total salinity

of the culture solution was 0,31%o. This is well

below the salinity of 16,6%o reported by Haller et

al. (1974) to inhibit E. crassipes growth rate in cul-

ture. Ions were supplied to the culture solution in

the inorganic form and in sufficient quantities not to

be limiting for E. crassipes (Musil, 1982). Culture

solutions were changed and adjusted to pH 7,0

weekly using 5% H2S04 and 10% NaOH. Evapora-

tion loss from cultures was replaced daily with deio-

nised-distilled water.

Experiments were conducted in an air-condi-

tioned greenhouse during summer when light inten-

sities (radiant flux densities) and air temperatures

were high. Maximum daytime air temperatures in

the greenhouse were maintained at ca 30°C required

for maximum growth of plants (Knipling et al.,

1970). Diurnal air temperature and relative humidity

fluctuations in the greenhouse, recorded on a ther-

mohydrograph, did not exceed the ranges 6 to 11°C

TABLE 2. — Chemical composition and ionic concentration of culture solution

used for growing E. crassipes

Ionic concentration

692

Bothalia 15, 3 & 4 (1985)

and 50 to 90% respectively, recommended by Rob-

bins (1946) for culturing higher plants.

Every two to four days, plants were removed from

culture, allowed to drain for two minutes above the

culture vessels, shaken to dislodge adhering water,

their fresh masses recorded and returned to culture.

Plants were grown in either N- or P-deficient cul-

tures until they showed a reduced growth rate, evi-

dent as a deviation from linearity in a plot of their

fresh mass against time, indicating a N or P defi-

ciency. They were then harvested from culture and

necrotic or damaged leaves and roots removed. Cul-

ture solutions were changed, fresh masses of plants

redetermined and plants returned to culture.

In each experiment, N- or P-deficient cultures

were spiked, at this stage, with six different levels of

N or P to obtain six treatments (16 to 20 replicates

per treatment) in which N concentrations in N-limi-

ted cultures ranged from 0 to 11,29 x 103 ) ig N U1

and P concentrations in P-limited cultures ranged

from 0 to 2,09 x 103 pg P €_1 (Table 1). A rando-

mized block design was adopted (Rayner, 1967).

After spiking, mass recordings, which included

both fresh as well as dead mass of plants arising

through necrosis of plant material during growth,

continued every two to four days for all plants until

no significant increase was recorded in the total fresh

mass (fresh and dead mass) of all plants grown at

each level of N or P supplied.

Culture solutions were not changed again. How-

ever, they were topped-up daily with deionised-dis-

tilled water and adjusted to pH 7,0 weekly, since the

Ks may be influenced by pH (Goldman, 1972) and

maximum growth of E. crassipes occurs at this pH in

culture (Chadwick & Obeid, 1966). Concentrates of

the culture solution deficient in either N or P were

added to the cultures at two weekly intervals to en-

sure an adequate supply of nutrients, other than the

specific limiting nutrient, to the plants. In P-limited

culture experiments, additional N at a concentration

of 9,03 x 103 pg N €-1 was also added to cultures in

the intervening weeks to ensure that N concentra-

tions remained above those limiting for E. crassipes

(Musil, 1982). The total nutrient additions after

spiking, however, did not increase the salinity of cul-

tures above l,6%o, i.e. 10% of the inhibitory sali-

nity value of 16,6%o for E. crassipes (Haller et

al., 1974).

When mass recordings were terminated, plants,

including their offsets, were harvested from culture

allowing the culture solution retained by plants to

drain back into each vessel. Plants were shaken to

dislodge adhering water and reweighed. They were

then dried in a forced draft oven at 60°C to a constant

weight and their dry masses determined. The dry

plant tissues were ground in a mill, redried at 60°C in

a forced draft oven to a constant weight, and stored

in sealed glass bottles for later chemical analysis.

After plants had been harvested, the culture solu-

tions in three vessels taken at random from each

treatment in each experiment were topped-up to the

5€ mark with deionised-distilled water and analysed

for remaining N or P using published methods (En-

vironmental Protection Agency, 1974; American

Public Health Ass.: Standard Methods, 1975). Loss

of the specific limiting nutrient (N or P) from cul-

tures, resulting from shaking of plants at each weigh-

ing interval, could not be accounted for, but was

considered to be small.

The minimum concentration or subsistence quota

(Rhee & Gotham, 1981) of the specific limiting nu-

trient in harvested plants was analysed in three

batches of dry, ground, harvested plant tissues cho-

sen at random from each treatment in each experi-

ment using published methods (Association of Offi-

cial Agricultural Chemists, 1975).

In each experiment, specific growth rates were

calculated according to Malek & Fencl (1966) and

Radford (1967) for each plant between each weigh-

ing interval for a period of ca 21 days after spiking.

The highest specific growth rate attained by each

plant in each treatment during this period was taken

as its specific growth rate at that particular N or P

concentration. The Umax value and Ks concentra-

tion were extrapolated for E. crassipes in each ex-

periment from the intercepts of a reciprocal plot of

specific growth rates of plants against limiting nutri-

ent concentrations (Lineweaver & Burk, 1934; Cur-

rie, 1982). The Yc value was derived in each experi-

ment from the slope of the line relating total fresh

mass yields of plants to quantities of limiting nutrient

absorbed. Simple linear regressions were used to ob-

tain the best straight lines through all points (Ray-

ner, 1967). All linear regressions were subjected to

an analysis of variance (Rayner, 1967).

RESULTS AND DISCUSSIONS

Growth in deficient culture

Plants with two pseudolaminae introduced into N-

(Experiments 1 to 5) and P-(Experiments 6 to 8) de-

ficient cultures showed an initial lag phase in growth

lasting ca two to four days (Figs 1 & 2). Growth of

plants in N- and P-deficient cultures then proceeded

more or less linearly until they showed a reduced

growth rate, at which stage the growth rate of plants

was assumed to be N- or P-limited. No significant

differences (Ps? 0,05) existed at this stage between

the mean fresh masses of groups of plants that were

to comprise each treatment in each experiment (Mu-

sil, 1982). In each experiment, a different growth

period in deficient cultures was required to induce in

plants a N- or P-limited state. This was attributed

partly to different quantities of N and P stored in

plants collected on different occasions from the field

for each experiment. No correlation was evident be-

tween the duration of plant growth in deficient cul-

tures, required to induce N or P limitation, and en-

vironmental conditions recorded in the greenhouse

(Musil, 1982).

Growth after spiking

In each experiment, the addition of the limiting

nutrient caused an increase in growth rate with a

short (three to four day) period of maximum growth

rate which was proportional to the level of N or P

supplied. The periods of mean maximum growth

rate of each group of plants for each treatment were

MEAN FRESH MASS g MEAN FRESH MASS g MEAN FRESH MASS

Bothalia 15, 3 & 4 (1985)

693

FIG. 1. — Experiment 1. Change in fresh mass (means of 20 plants/treatment) of E. crassipes grown under varying conditions of N

supply. All treatments were grown under N-deficient conditions for 18 days before N was added. A, no N; B 11 290, pg N

(2 260 pg N C, 22 580 pg N (4 520 pg N C1); D, 33 870 pg N (6 770 pg N € '); E, 45 160 pg N (9 030 pg N € '); F,

56 450 pg N (11 290 pg N tl). Standard deviations of means are shown by bars: a = lag phase of growth; b = period of

maximum growth rate; c = termination of fresh mass recordings; d = projected growth in the absence of N.

694

Bothalia 15, 3 & 4 (1985)

DAYS

FIG. 2. — Experiment 8. Change in fresh mass (means of 18 plants/treatment) of E. crassipes grown under varying conditions of P

supply. All treatments were grown under P-deficient conditions for 38 days before P was added. A, no P; B, 1 300 gg P (260 /ug

P C, 3 260 gg P (650 gg P € '); D, 5 220 gg P (1 040 gg P € >); E, 7 830 gg P (1 570 gg P ( '); F, 10 440 gg P (2 090

gg P P1). Standard deviations of means are shown by bars: a = lag phase of growth; b = period of maximum growth rate; c =

termination of fresh mass recordings; d = projected growth in the absence of P.

Bothalia 15, 3 & 4 (1985)

695

evident from the maximum slopes of curves relating

growth (fresh mass) and time (Figs 1 & 2B, C, D, E,

F). Thereafter, the growth rates of plants decreased,

progressively until there was no measurable increase

in the total fresh mass (fresh including dead mass

produced during growth) of plants. This required ca

75 to 95 days after the addition of N and ca 50 to 65

days after the addition of P, in those treatments

where these limiting nutrients were supplied at the

highest levels to cultures. In Experiments 4 and 5,

mass recordings were terminated prior to cessation

of plant growth, i.e. about 21 days after spiking.

Nitrogen- and P-limited plants responded differ-

ently to the different levels of limiting nutrient sup-

plied to cultures. In Treatments 3 to 6,. where the

limiting nutrients were supplied at levels above 2 260

pg N €-i and 260 pg P £-i (Figs 1 & 2C, D, E, F),

plants generally attained a maximum growth rate

much later after the addition of N and P than in

Treatment 2 (Figs 1 & 2B), where the limiting nutri-

ents were supplied at lower levels. This could not be

reasonably explained by a restricted uptake of N or

P in E. crassipes due to a limited nitrate reductase or

alkaline phosphate activity in plants resulting from

their growth in deficient cultures (Schwoerbel &

Tillmans, 1974). Oaks et al. (1972) in a study of the

induction kinetics in the roots of Zea mays seedlings

have shown that the induction of nitrate reductase is

very rapid with maximum levels of nitrate reductase

being achieved four to six hours after transference of

seedlings from a N03-N-deficient medium to one

containing N03-N. Fitzgerald & Nelson (1966) and

Fitzgerald (1969), on the other hand, have reported

that alkaline phosphatase activity increases in algal

cells and higher aquatic plants such as Ceratophyl-

lum demersum L. with increasing P deficiency. It

would appear, therefore, that in those treatments

where plants were exposed to high levels of N03-N

and P04-P, these nutrients may have been accumu-

TABLE 3. — Statistical analysis of regressions of 1/U against

1/limiting nutrient concentration for E. crassipes grown

in N-(Experiments 1 to 5) and P-(Experiments 6 to 8)

limited cultures

lated in a pool and then reduced and incorporated

into metabolism at a later stage, i.e. the assimilation

of N and P by plants and their incorporation into

new growth did not keep pace with their uptake in

culture. Further research on the depletion of N03-N

and PO4-P in the culture solution, levels and loca-

tion of nitrate reductase and alkaline phosphatase in

the plant, however, will be required before any

meaningful conclusions can be drawn from this prob-

lem.

Maximum specific growth rate (Umax)

Lineweaver-Burk plots of the reciprocals of speci-

fic growth rates (1/U), i.e. the highest specific

growth rate attained by each plant after the addition

of N or P, against the reciprocals of limiting nutrient

concentrations (Figs 3 & 4) showed that the relation-

ship between 1/U and 1/N or 1/P was linear in each

experiment with a high degree of correlation, signifi-

cant at P=s0,01 (Table 3). An analysis of variance of

the regressions showed that the slopes and intercepts

T3

FIG. 3. — Experiment 1. A Lineweaver-Burk plot of specific growth rates of E. crassipes (means of 20 plants/treatment) against

levels of N supplied in culture. Broken lines show 95% confidence limits on either side of the regression line. Standard devia-

tions of means are shown by bars. U = Umax.

696

Bothalia 15, 3 & 4 (1985)

FIG. 4. — Experiment 8. A Lineweaver-Burk plot of specific growth rates of E. crassipes (means of 18 plants/

treatment) against levels of P supplied in culture. Broken lines show 95% confidence limits on either side of

the regression line. Standard deviations of means are shown by bars. U = Umax.

were significant at P^0,05. The Umax value was ex-

trapolated for E. crassipes in each experiment from

the intercept of the regression line on the y axis, cal-

culated from the regression equation.

The Umax values determined ranged from 0,0537

to 0,0886 g fresh mass g-1 d1 (5,37 to 8,86% d-1) in

N-limited experiments and from 0,0451 to 0,1089 g

fresh mass g-1 d-i (4,51 to 10,89% d^1) in P-limited

experiments (Table 4). An exponential relationship

was not evident between the Umax values derived

under N and P limitation and the reciprocals of mean

daily air temperatures (expressed as °K) recorded in

the greenhouse (Arrhenius plot). In addition, no

correlation was evident between the Umax values

determined and mean daily relative humidities re-

corded in the greenhouse. Significantly lower Umax

values, however, were obtained in Experiments 2, 3,

7 and 8 where longer growth periods in deficient cul-

tures were required to induce N or P limitation

(Table 4).

The latter observation could not be explained in

terms of non-competitive inhibition, i.e. by a re-

duced uptake rate of the limiting nutrient by plants

resulting from their longer growth periods in defi-

cient cultures. Investigations of the uptake kinetics

of higher plants have shown that growth of plants in

starvation (deficient) media causes a subsequent in-

crease in their nutrient uptake rate with a corres-

ponding reduction in the half saturation coefficient

(Km) for uptake. Glass (1978), for example, has

shown that the uptake characteristics for K+ of bar-

ley plants grown initially with or without K+are very

different, the Km for K* uptake being reduced in the

starved plant from 0,1 to 0,03 mM. The same occurs

for other ions, as for NO3 (Smith, 1973) and PO;j“

(Cartwright, 1972), and for other species. Doddema

et al. (1979), for example, have shown a reduction in

Km from 111 to 40 mM NO3 brought about by N

starvation in Arabidopsis thaliana. It is suggested,

therefore, that the different Umax values derived for

TABLE 4. — Maximum specific growth rates (Umax) and half saturation coefficients (Ks) derived

for E. crassipes in N-(Experiments 1 to 5) and P-(Experiments 6 to 8) limited cultures

Bothalia 15, 3 & 4 (1985)

697

E. crassipes under N and P limitation possibly re-

flect:

(i) the different physiological state of plants grown

for different spans in deficient cultures and collected

on different occasions from the field for each experi-

ment;

(ii) differences in the ratio of plant mass at spiking

to levels of limiting nutrient supplied to culture,

since a larger plant mass resulted at spiking in those

experiments where longer growth periods in defi-

cient cultures were required to induce N or P limita-

tion;

(iii) variations in light intensity in the greenhouse

between experiments.

Half saturation coefficient (Ks)

The Ks was extrapolated for E. crassipes in each

experiment from the intercept of the regression line

of 1/U against 1/N or 1/P on the x axis, calculated

from the regression equation (Figs 3 & 4). The Ks

concentrations determined ranged from 399,8 to

1 505,6 pg N €_1 in N-limited experiments and from

41,1 to 161,8 (tig P £~l in P-limited experiments

(Table 4.) They showed no correlation with mean

daily air temperatures and relative humidities re-

corded in the greenhouse or with the duration of

plant growth in deficient cultures required to induce

N or P limitation. The same reasons given for the

different Umax values determined may also partly

explain the different Ks concentrations measured for

E. crassipes under N and P limitation.

The Ks concentrations derived for E. crassipes

under N limitation are in the range of those reported

for various species of algae, whereas those derived

under P limitation are much higher (Table 5). This

indicates that E. crassipes has a potential similar to

algae to produce a high growth rate in N-limited

waters, but a potential lower than algae to produce a

high growth rate in P-limited waters. Since P is the

nutrient most frequently limiting algal growth rate in

relatively oligotrophic waters (Toerien et al.. 1975),

it would appear that in such waters P may also be the

nutrient limiting for E. crassipes.

The mean Ksn concentration of 976 pg N U1 de-

termined for E. crassipes from the five N-limited ex-

periments falls in the range 500 to 1 000 pg N €_1,

interpreted by Center & Spencer (1981) from the

N/P uptake rates of E. crassipes of 5 to 10 €_1

( Boyd, 1970; 1976; Dunigan et al. , 1975) as being the

critical limiting N concentrations in the water for E.

crassipes in the field, i.e. below which the growth

rate of this plant is significantly influenced by the N

concentration in the water. The mean Ksp concen-

tration of 94,1 pg P U1 determined for E. crassipes

from the three P-limited experiments compares fa-

vourably with 100 pg P €_1 reported by Haller et al.

(1970) and Knipling et al. ( 1970) as being the critical

limiting P concentration in the water for E. crassipes

in the field.

The ratio of the mean Ksn/Ksp concentrations,

derived for E. crassipes under N and P limitation,

suggest an optimal N/P ratio in the water for E. cras-

sipes of ca 10, i.e. below which N and above which P

concentrations in the water become growth rate limi-

ting for this plant. This value is well below the opti-

mal N/P ratio of 30 (cell and medium) reported by

Rhee (1974, 1978) for algae. It should, however, be

pointed out that, although the limiting nutrient can

often be indicated from the N/P ratio in the water, in

many instances the growth rate of phytoplankton is

controlled by P even when the N/P ratio in the water

is relatively low (Welch et al., 1978).

Yield coefficient (Yc)

With the exception of Experiments 4 and 5, where

mass recordings were terminated prior to cessation

of plant growth, the quantities of limiting nutrient

remaining in three culture solution samples taken at

random from each treatment, after plants had been

harvested, were below 0,1% of that initially added

(Musil, 1982). It was assumed, therefore, that in all

culture solutions, with the exception of Experiments

4 and 5, the N or P added had been absorbed by

plants and incorporated into growth.

Plots of the total fresh mass yields of plants (fresh

including dead mass produced during growth)

TABLE 5. — Half saturation coefficients (Ks) reported for various species of algae

compared with those determined for E. crassipes

* High temperature strain

** Emersion strain

698

Bothalia 15, 3 & 4 (1985)

TABLE 6. — Statistical analysis of regressions relating total fresh

mass yields to quantities of limiting nutrient supplied for

E. crassipes grown in N-(Experiments 1 to 3) and P-(Ex-

periments 6 to 8) limited cultures

against the quantities of limiting nutrient added

(Figs 5 & 6) showed that the relationship between

these two factors, in each of the first three N- and P-

limited experiments, was linear with a high degree of

correlation, significant at 0,001 (Table 6). An

analysis of variance of the regressions showed that

the slopes and intercepts were significant at P=?

0,001. The Yc value (fresh mass basis) was derived

for E. crassipes in each experiment from the slope of

the regression line given by the regression equation.

The Yc values (fresh mass basis) determined ranged

from 1 659,6 to 1 981,1 in N-limited experiments

and from 16 431,2 to 18 670,6 in P-limited experi-

ments (Table 7).

The mean water contents of plants, harvested

from each of the first three N- and P-limited experi-

ments, are given in Table 7. Water contents ranged

from 94,72 to 95,05% and showed no significant dif-

ferences (P=5 0,05) between experiments. They com-

pare favourably with the average water content of

94,75% derived from values reported by Penfound

& Earle (1948), Westlake (1963) and Bock (1969).

From the mean water contents of plants, the Yc va-

FIG. 5. — Experiment 1. The relationship between total fresh

mass yields of E. crassipes (means of 20 plants/treatment)

and quantities of N supplied in culture. Broken lines show

95% confidence limits on either side of the regression line.

Standard deviations of means are shown by bars.

lues (fresh mass basis) were converted to a dry mass

basis.

The Yc values (dry mass basis) determined ranged

from 86,9 to 98,1 in N-limited experiments and from

867,1 to 980,2 in P-limited experiments (Table 7).

Slightly higher Yc values (both fresh and dry mass

basis) were obtained in Experiments 1 and 6 where

plants were grown for the shortest spans in deficient

cultures to induce N or P limitation.

In all experiments, some growth (yield in plant

material) was produced by E. crassipes grown in the

absence of N or P (Figs 5 & 6). This indicated that,

although limiting N and P concentrations were exist-

ent in the plants, sufficient quantities were present

to allow some growth. In fact, higher yields were

produced by E. crassipes grown in the absence of the

limiting nutrient in Experiments 1 and 6, where

TABLE 1 -Yield coefficients, Yc, (g of fresh mass yield of plant material per g

of limiting nutrient absorbed by plants) derived for E. crassipes in N-(Ex-

periments 1 to 3) and P- (Experiments 6 to 8) limited cultures. Yield coeffi-

cients (dry mass basis) are estimated from the mean water contents of plants

determined in each experiment

Analysis of variance

Variance ratio (F value) 2,99

Degrees of freedom (n-1) 615

Significance level NS

% (P=0,05)

NS = not significant

Bothalia 15, 3 & 4 (1985)

699

TOTAL N ADDED X10"3g

FIG. 6. — Experiment 8. The relationship between total fresh

mass yields of E. crassipes (means of 18 plants/treatment)

and quantities of P supplied in culture. Broken lines show

95% confidence limits on either side of the regression line.

Standard deviations of means are shown by bars.

plants were grown for the shortest periods in defi-

cent cultures to induce N or P limitation, than in

other experiments (Musil, 1982). This suggests that

the limiting nutrients (N or P) were present at higher

concentrations in plants at spiking in these two ex-

periments than in other experiments. In principle,

however, higher limiting concentrations of N and P

present in plants at spiking in Experiments 1 and 6

respectively should not have had an influence on the

Yc values determined, since these were derived from

the slopes of regression lines relating total fresh mass

yields of plants to quantities of limiting nutrient (N

or P) supplied in culture. Consequently, the slightly

higher Yc values measured for N and P in Experi-

ments 1 and 6 respectively could not be readily ex-

plained.

The Yc values (dry mass basis) derived for E. cras-

sipes under P limitation are in the upper range of

those reported for various species of diatoms and

other algae, whereas those derived under N limita-

tion are much higher (Table 8). This indicates that

E. crassipes has the potential to produce a similar

biomass per unit quantity of P absorbed, but a much

larger biomass per unit quantity of N absorbed, than

diatoms and other algae. Furthermore, the Yc val-

ues suggest that E. crassipes has the potential to re-

move similar quantities of P, but smaller quantities

of N, per unit amount of plant mass than diatoms

apd other algae. Maximum specific growth rates re-

ported for algae (Shelef et al., 1968; Zabat et al.,

1970; Toerien et al., 1971; Goldman, 1972), how-

ever, are considerably higher than those determined

for E. crassipes. In eutrophic waters, therefore, in

which the limiting nutrient concentrations are high

and specific growth rates of both algae and E. cras-

sipes approach their Umax values, E. crassipes

would need to be present with a proportionately

larger biomass than algae to compensate for its

lower growth rate to ensure a potential similar to al-

gae for removing nutrients. E. crassipes lower poten-

tial than algae to produce a high growth rate in

waters where P is limiting, as evident from a com-

parison of its Ks concentrations for P with those of

algae (Table 5), suggests that in relatively oligotro-

phic waters E. crassipes would also be less efficient

in removing nutrients than algae, at least where both

plants are present with the same biomass.

The minimum limiting concentrations of N and P

(% dry mass) in plants harvested from each treat-

ment (means of 3 batches), in each of the first three

N- and P-limited experiments, are shown in Table 9.

The minimum limiting concentrations of N and P

(subsistence quotas) in the dry plant tissues ranged

from 0,94 to 1,28% N in N-limited experiments and

from 0,09 to 0,14% P in P-limited experiments. They

TABLE 8. — Yield coefficients (Yc) reported for various species of diatoms and

other algae compared with those determined for E. crassipes

700

Bothalia 15, 3 & 4 (1985)

I A15L.I 9. Minimum limiting concentrations of N and P

(means of 3 batches) in E. crassipes harvested from

culture

NS — not significant

showed no significant differences (Pi? 0,05) between

experiments (Table 9).

Toerien et al. (1971) and Coetzer et al. (1977)

pointed out that the yield coefficient (dry mass basis)

for a specific limiting nutrient, when expressed as a

reciprocal and a percentage, should estimate the mi-

nimum concentration of the limiting nutrient in the

dry plant tissue. The Yc values (dry mass basis) de-

rived for E. crassipes under N and P limitation, when

expressed as reciprocals and percentages (1/Yc x

100), adequately estimated the minimum limiting

concentrations of N and P in plants harvested from

culture (Table 10). This suggests that the Yc values

determined for E. crassipes are fairly reliable. The

average minimum limiting concentrations of 1,10%

N and 0,11% P, estimated in E. crassipes plant tis-

sues from the mean Yc values for N and P respec-

tively, also compare favourably with the minimum

concentrations (% dry mass) of 1,33% N and 0,14%

P reported by Boyd & Vickers ( 1971) in E. crassipes

growing in the field, and with the minimum concen-

tration (% dry mass) of 0,098% P reported by Halier

& Sutton (1973) in E. crassipes growing in the ab-

sence of P in culture.

Droop (1968) and Rhee (1973) showed that the

minimum concentration of a specific limiting nutri-

ent in algal cells is equal to, or not significantly dif-

ferent from, the intracellular half saturation coeffi-

cient (Kq) for the limiting nutrient. Consequently, if

it is assumed that a similar situation exists in E. cras-

sipes, then the ratio of the average minimum N/mini-

mum P concentrations in E. crassipes , derived from

the mean Yc values for these nutrients, give an opti-

mal N/P ratio in E. crassipes of ca 10. This value

compares favourably with the optimal N/P ratio in

the water for E. crassipes of ca 10, estimated from

the ratio of the mean Ksn/Ksp concentrations de-

rived under N and P limitation in culture.

CONCLUSIONS

Maximum specific growth rates (Umax) and half

saturation coefficients (Ks) were not adequately de-

termined for E. crassipes growing in N- or P-limited

batch cultures. In contrast, yield coefficients (Yc)

were determined with sufficient accuracy. With bet-

ter facilities, it is possible that the batch culture

method used for measuring kinetic coefficients for

E. crassipes growing under specific nutrient limita-

tion in this investigation could be improved. For ex-

ample, if plants for culture were collected from

populations grown under controlled environmental

conditions in a standardized culture medium, it is

possible that a uniform growth period required to

induce in plants a N- or P-limited state could be ob-

tained. This might decrease the variability in Umax

values and Ks concentrations determined. It is sug-

gested, however, that precise measurements of

Umax and Ks may only be obtained for E. crassipes ,

under specific nutrient limitation in culture, by

growing plants under constant environmental condi-

tions in some type of continuous flow culture system

in which the limiting nutrient concentrations could

be maintained at constant levels. In such a system,

therefore, it would not be necessary to grow plants

initially in deficient cultures to induce N or P limita-

tion, since the specific growth rate of E. crassipes at

TABU 10. Minimum limiting concentrations of N and 1’ in E. crassipes estimated

from yield coefficients (Yc). derived under N and 1’ limitation, compared w ith

minimum limiting concentrations of N and I’ in plants harvested from culture

Bothalia 15, 3 & 4 (1985)

701

each limiting nutrient concentration could be estab-

lished over a much longer growth period in culture.

This would eliminate any adverse effects on the

growth rate of E. crassipes arising through growth of

plants in N- or P-deficient cultures.

Although Umax values and Ks concentrations de-

rived for E. crassipes under N and P limitation va-

ried considerably, it should be possible to evaluate

their potential in modelling by using the most reli-

able values determined in culture in the Monod mo-

del to assess its predictive ability. This, in turn, may

serve as a basis for refinement of the model. In this

investigation, the Umax values measured for E.

crassipes were adversely influenced by the duration

of plant growth in deficient cultures required to in-

duce N or P limitation. It is suggested, therefore,

that, for purposes of testing the model and as a basis

for its refinement, the values determined in those ex-

periments where plants were grown for the shortest

spans in N- and P-deficient cultures are possibly

more reliable than those determined in other experi-

ments. The mean Ks concentrations derived for E.

crassipes under N and P limitation, on the other

hand, are possibly more reliable than the individual

concentrations determined, since they compare fa-

vourably with the critical limiting N and P concentra-

tions in the water for E. crassipes in the field.

The Yc values derived for E. crassipes under N

and P limitation showed little variation. They appear

reliable, since their reciprocals (dry mass basis) ex-

pressed as percentages adequately estimated the mi-

nimum limiting concentrations of N and P in E. cras-

sipes harvested from culture, i.e. when no further

significant increase in the total fresh mass of plants

at each level of N or P supplied was recorded. Since

the minimum limiting concentrations of N and P in

E. crassipes can be estimated from the respective Yc

values for these nutrients, it should be feasible to

predict the growth rate of E. crassipes in the field

from the limiting N or P concentrations in plants us-

ing the Droop model (Droop, 1968; Rhee, 1973).

ACKNOWLEDGEMENTS

We wish to thank Mr A. Zakwe and Mrs J. Schaap

for technical services rendered. Prof. D.F. Toerien

and Drs P.J. Ashton and M.C. Rutherford for their

valuable comments and criticisms and Mrs S.S.

Brink for typing the manuscript.

UITTREKSEL

Die N- en P-beperkte groeikinetika van Eichhornia

crassipes (Mart.) Solms wat in ’n kweekhuis gekweek

was, is ondersoek met die doel om 'n model te ont-

werp waarvolgens populasiegroottes, opbrengs,

groeitempo’s en frekwensies en die hoeveelheid van

die oes onder wisselende toestande van voedingsla-

ding en klimaat, vir hierdie plant voorspel kan word

om beide die voedings-elementinvoer en buitenspo-

rige groei in eutrofiese waterstelsels te beheer. Die ki-

netiese koeffisiente naamlik maksimum spesifieke

groeitempo (Umax), halfversadigingskonstante (Ks)

en opbrengskoeffisient (Yc) was van die N en P be-

perking in gerepliseerde lotkultuureksperimente ge-

meet. Umax waardes en Ks konsentrasies onder N be-

perking is van 5,37 tot 8,86% d ' en van 400 tot 1 506

pg N t‘ respektiewelik. Die afgelei onder P beper-

king het gewissel van 4,51 tot 10,89% d ' en van 41 tot

162 pg P U' respektiewelik. Yc waardes ( varsmassa-

basis) vir N bepaal, het gewissel van 1 660 tot 1 981

(87 tot 98 droemassabasis) en van 16 431 tot 18 671

(867 tot 980 droemassabasis) vir P. Die omgekeerdes

van Yc waardes (droemassabasis), as persentasies,

gee ’n voldoende aanduiding van die minimum beper-

kende N en P konsentrasies (% droemassa) in die

plantweefsel. Die afgeleide kinetiese koeffisiente word

met die van alge vergelyk. Die eksperimentele me-

todes gebruik en die resultate verkry, word krities

beoordeel.

REFERENCES

AMERICAN PUBLIC HEALTH ASSOCIATION, AMERI-

CAN WATER WORKS ASSOCIATION & WATER

POLLUTION CONTROL FEDERATION, 1975. Standard

methods for the examination of water and waste-water. 14th

edn. New York.

ASSOCIATION OF OFFICIAL AGRICULTURAL CHEM-

ISTS, 1975. Methods of analysis. 12th edn, 957 pp. Washing-

ton, D.C.: Assoc. Offic. Agric. Chem.

BHAGAT, S. K., FUNK, W. L. & JOHNSTONE, D. L., 1972.

Correlated studies of Vancouver Lake — Water quality pre-

diction study. Office of Research and Monitoring. U.S. En-

vironmental Protection Agency Report EPA-R2-72-111,

Washington.

BOCK, J. H., 1966. An ecological study of Eichhornia crassipes

with special emphasis on its reproductive biology. Ph.D. the-

sis, University of California, Berkeley.

BOCK, J. H., 1969. Productivity of the water hyacinth (Eichhor-

nia crassipes). Ecology 50: 460-464.

BOYD, C. E., 1970. Vascular aquatic plants for mineral nutrient

removal from polluted waters. Econ. Bot. 24: 95-103.

BOYD, C. E., 1976. Accummulation of dry matter, nitrogen and

phosphorus by cultivated water hyacinths. Econ. Bot. 30:

51-56.

BOYD, C. E. & SCARSBROOK, E., 1975. Influence of nutrient

additions and initial density of plants on production of water

hyacinth, Eichhornia crassipes. Aquat. Bot. 1: 253-261.

BOYD, C. E. & VICKERS, D. G., 1971. Variation in the ele-

mental content of Eichhornia crassipes. Hydrobiologia 38:

409^114.

CARTWRIGHT, B., 1972. The effect of phosphate deficiency on

the kinetics of phosphate absorption by sterile excised barley

roots, and some factors affecting the iron uptake efficiency

of roots. Communs Soil Sci. PI. Analys. 3: 313-322.

CENTER, T. D. & SPENCER, N. R., 1981. The phenology and

growth of water hyacinth ( Eichhornia crassipes (Mart.)

Solms) in a eutrophic north-central Florida lake. Aquat. Bot.

10,1: 1-32.

CHADWICK, M. J. & OBEID, M., 1966. A comparative study

of the growth of Eichhornia crassipes (Mart.) Solms and Pis-

tia stratiotes L. in water culture. J. Ecol. 54: 563-575.

CLOCK, R. M., 1968. Removal of nitrogen and phosphorus from

a secondary sewage treatment effluent. Ph.D. thesis, Univer-

sity of Florida.

COETZER, G. C., TOERIEN, D. F. & SCHOEMAN, F. R.,

1977. Silica, nitrogen and phosphorus requirements of some

South African diatoms. 7. Limnol. Soc. sth. Afr. 3,1:27-31.

CORNWELL, D. A.. ZOLTEK, J. Jr, PATRINELY, C. D.,

FURMAN, T. de S. & KIM, J. I., 1977. Nutrient removal

by water hyacinths. 7. Wat. Pollut. Control Fed. 49: 57-65.

CURRIE, D. J., 1982. Estimating Michaelis Menten parameters:

bias variance and experimental design. Biometrics 38:

907-919.

DODDEMA, H..TELEKAMP, G. P. & OTTEN, H., 1979. Up-

take of nitrate by mutants of Arabidopsis thaliana, disturbed

in uptake or reduction of nitrate. Physiologia PI. 45:

297-346.

DROOP, M. R., 1968. Vitamin B-12 and marine ecology. 4. The

kinetics of uptake, growth and inhibition in Monochrysis lu-

theri. J. mar. biol. Ass. U.K. 48: 689-733.

702

Bothalia 15, 3 & 4 (1985)

DUNIGAN. E. P., SHAMSUDDIN, Z. H. & PHELAN, R. A.,

1975. Water hyacinths tested for cleaning polluted water. La

Agric. 18: 12-13.

ENVIRONMENTAL PROTECTION AGENCY, 1974. Manual

of methods for chemical analysis of water and wastes. Report

no. EPA-625/6-74-003, Washington, D.C.: U.S. Environ-

mental Protection Agency Office ofTechnologicalTransfer.’

FITZGERALD, G. P., 1969. Field and laboratory evaluation of

bioassays for nitrogen and phosphorus with algae and aqua-

tic weeds. Limnol. Oceanogr. 14: 206-212.

FITZGERALD, G. P. & NELSON, T. C., 1966. Extractive and

enzymatic analysis for limiting or surplus phosphorus in al-

gae. J. Phycol. 2: 32-37.

GERLOFF. G. C. & SKOOG, F., 1954. Cell contents of nitrogen

and phosphorus as a measure of their availability for growth

of Microcystis aeruginosa. Ecology 35: 348-353.

GLASS, A.D.M., 1978. Regulation of potassium influx into in-

tact roots of barley by internal potassium levels. Can. J. Bot.

56: 1759-1764.

GOLDMAN, J. C., 1972. The kinetics of inorganic carbon limited

growth of green algae in continuous culture. Its relationship to

eutrophication. Ph.D. thesis, University of California, Ber-

keley.

GOLDMAN, J. C. & CARPENTER. E. J., 1974. A kinetic ap-

proach to the effect of temperature on algal growth. Limnol.

Oceanogr. 19,5: 756-766.

HALLER, W. T., KNIPLING, E. B. & WEST, S. H., 1970.

Phosphorus absorption by and distribution in water hya-

cinths. Proc. Soil Crop Sci. Soc. Fla 30: 64—68.

HALLER, W. T. & SUTTON, D. L., 1973. Effect of pH and

high phosphorus concentrations on growth of water hya-

cinth. Hyacinth Control J. 11: 59-61.

HALLER, W. T„ SUTTON, D. L. & BARLOWE, W. C., 1974.

Effects of salinity on growth of several aquatic macrophytes.

Ecology 54: 891-894.

HAMNER. K. C., LYON, C. B. & HAMNER, C. L., 1942. Ef-

fect of mineral nutrition on the ascorbic acid content of the

tomato. Bot. Gaz. 103: 586-616.

KNIPLING, E. B., WEST, S. H. & HALLER, W. T., 1970.

Growth characteristics, yield potential and nutrient content

of water hyacinths. Proc. Soil Crop Sci. Soc. Fla 30: 51-63.

LINEWEAVER. H. & BURK, D., 1934. The determination of

enzyme dissociation constants. J. Am. chem. Soc. 56: 658.

MACKENTHUN, K. M., 1964. Limnological aspects of recrea-

tional lakes. Washington D.C.: U.S. Government Printing

Office.

MACKENTHUN, K. M., 1965. Nitrogen and phosphorus in

water : an annotated bibliography of their biological effects.

Washington D.C.: U.S. Government Printing Office.

MALEK, I. & FENCL, Z., 1966. Theoretical and methodological

basis of continuous culture of micro-organisms. Prague: Cze-

choslovak Academy of Sciences.

MIDDELBROOKS, E. J., PORCELLA, D. B., PEARSON. E.

A., McGAUHEY, P. H. & ROHLICH. G. A., 1971. Bio-

stimulation and algal growth kinetics of waste water. J. Wat.

Pollut. Control Fed. 43: 454-473.

MINER, J. , WOOTEN, J. W. & DODD, J. D. , 1971. Water hya-

cinths to further treat anaerobic lagoon effluent. In Live-

stock Waste Management and Pollution Abatement. St Jo-

seph, Michigan: American Society of Agricultural Engin-

eers.

MUSIL, C. F., 1982. The use of growth kinetics in the develop-

ment of a predictive model of the growth of Eichhornia cras-

sipes (Mart.) Solms in the field. Ph.D. thesis, University of

Natal.

MUSIL, C. F. & BREEN, C. M., 1977. The application of

growth kinetics to the control of Eichhornia crassipes

(Mart.) Solms through nutrient removal by mechanical har-

vesting. Hydrobiologia 53,2: 165-171.

OAKS, A., WALLACE, W. & STEVENS, D., 1972. Synthesis

and turnover of nitrate reductase in corn roots. PI. Physiol.

50: 649-654.

PENFOUND, W. T., 1956. Primary production of vascular aqua-

tic plants. Limnol. Oceanogr. 1: 92-101.

PENFOUND, W. T. & EARLE, T. T., 1948. The biology of

water hyacinth. Ecol. Monogr. 18: 448-472.

RADFORD, P. J., 1967. Growth analysis formulae — their use

and abuse. Crop Sci. 7: 171-175.

RAYNER, A. A., 1967. A first course in biometry for agriculture

students. Pietermaritzburg: University of Natal Press.

RHEE, G-Y., 1973. A continuous culture study of phosphate up-

take, growth rate and polyphosphate in Scenedesmus sp. J.

Phycol. 9: 495-506.

RHEE, G-Y., 1974. Phosphate uptake under N limitation by Sce-

nedesmus sp. and its ecological implications. J. Phycol. 10:

470-475.

RHEE, G-Y., 1978. Effects of N : P atomic ratios and nitrate

limitation on algal growth, cell composition and nitrate up-

take. Limnol. Oceanogr. 23,1: 10-25.

RHEE, G-Y. & GOTHAM, I. J., 1981. The effect of environ-

mental factors on phytoplankton growth : light and the inter-

actions of light with nitrate limitation. Limnol. Oceanogr.

26,4: 649-659.

RICHARDSON, B., ORCUTT, D. M., SCHWERTNER, H.

A., MARTINEZ, C. L. & WICKLINE, H. E., 1969. Ef-

fects of nitrogen limitation on the growth and composition of

unicellular algae in continuous culture. Appl. Microbiol. 18:

245-250.

ROBBINS, W. R., 1946. Growing plants in sand cultures for ex-

perimental purposes. Soil Sci. 62: 3.

SAWADA, T., CHOHJI, T. & KUNO, S., 1978. Kinetic analysis

of unbalanced bacterial growth in temperature shift. In V.

W. Weekman, Jr. & D. Luss, Chemical reaction engineering,

163-172. Washington D.C.: American Chemical Society.

SCHWOERBEL, J. & TILLMANS, G. C., 1974. Stickstoffauf-

nahme aus dem Wasser und Nitratreductase-Aktivitat bei

submersen Wasserpflanzen Fontinalis antipyretica L. Arch.

Hydrobiol. Suppl. 47: 282-294.

SHELEF, G., 1968. Kinetics of algal systems in waste-treatment.

Light intensity and nitrogen concentration as growth-limiting

factors. Ph.D. thesis. University of California, Berkeley.

SHELEF, G., OSWALD, W. J. & GOLUEKE, C. G., 1968. Ki-

netics of algal systems in waste treatment. Light intensity and

nitrogen as growth limiting factors. SERL Report No 68-4,

University of California, Berkeley.

SHELEF, G., OSWALD, W. J. & GOLUEKE, C. G., 1970. As-

saying algal growth with respect to nitrate concentration by a

continuous flow turbidostat. Adv. Water Pollut. Res. 2: 1-9.

Oxford: Pergamon.

SMITH, F. A., 1973. The internal control of nitrate uptake into

excised barley roots with differing salt contents. New Phytol.

72: 769-782.

SOROKIN, C., 1960. Kinetic studies on temperature effects on

the cellular level. Biochim. biophys. Acta 38: 197-204.

SOROKIN, C., 1971. Calefaction and phytoplankton. BioScience

21: 1153-1159.

STEWART, K. M. & ROHLICH, G. A., 1967. Eutrophication

— a review. A report to the State Quality Control Board,

Sacremento, California.

STEYN, D. J., 1973. Die eutrofikasiepeile van vier Transvaalse

damme. M.Sc. verhandeling, Universiteit van Pretoria.

STUMM, W., 1974. Man's acceleration of hydrogeochemical cyc-

ling of phosphorus-eutrophication of inland and coastal

waters. Paper presented at the IWPC Conference, Salisbury,

April, 1974.

TOERIEN, D. F., 1972. Nutrient removal as a means of control of

Eichhornia crassipes. CSIR, National Institute for Water

Research, Pretoria. Report No. W3/24/1/16.

TOERIEN, D. F. & HUANG, C. H., 1973. Algal growth predic-

tion using kinetic constants. Wat. Res. 1: 1673-1681.

TOERIEN. D. F., HUANG, C. H., RADIMSKY, J., PEAR-

SON, E. A. & SCHERFIG, J., 1971. Provisional algal assay

procedures — final report. SERL Report No. 71-6, Univer-

sity of California. Berkeley.

TOERIEN, D. F., HYMAN, K. L. & BRUWER, M. J., 1975. A

preliminary trophic status classification of some South Afri-

can impoundments. Water S. A. 1,1: 15-23.

TOPIWALA, H. & SINCLAIR, C. G., 1971. Temperature re-

lationship in continuous culture. Biotechnol. Bioeng. Symp.

13: 795-813.

WAHLQUIST, H., 1972. Production of water hyacinth and re-

sulting water quality in earthen ponds. Hyacinth Control J.

10: 9-11.

WELCH, E. B., STURTEVANT, P. & PERKINS, M. A., 1978.

Dominance of phosphorus over nitrogen as limiter to phyto-

plankton growth rate. Hydrobiologia 57,3: 209-215.

WESTLAKE, D. F., 1963. Comparisons of plant productivity.

Biol. Rev. 38: 385-425.

WOLVERTON, B. C., 1975. Water hyacinths for removal of cad-

mium and nickel from polluted waters. NASA Technical

Memorandum, TM-X-72721.

Bothalia 15, 3 & 4 (1985)

703

WOLVERTON, B. C. & McDONALD, R. C., 1975a. Water

hyacinths and alligator weeds for removal of lead and mer-

cury from polluted waters. NASA Technical Memorandum,

TM-X-72723.

WOLVERTON, B. C. & McDONALD, R. C., 1975b. Water

hyacinths and alligator weeds for removal of silver, cobalt

and strontium from polluted waters. NASA Technical Mem-

orandum, TM-C-72727.

WOLVERTON, B. C. & McDONALD, R. C., 1975c. Water

hyacinths for upgrading sewage lagoons to meet advanced

waste water treatment standards. Part l. NASA Technical

Memorandum, TM-X-72729.

WOLVERTON, B. C. & McDONALD, R. C., 1975d. Water

hyacinths for upgrading sewage lagoons to meet advanced

waste water treatment standards. Part II. NASA Technical

Memorandum. TM-X-72730.

WOLVERTON. B. C. & McDONALD, R. C., 1976. Water hya-

cinths (Eichhornia crassipes) for removing chemical and pho-

tographic pollutants from laboratory waste waters. NASA

Technical Memorandum, TM-X-72731.

WOLVERTON, B. C. & McKOWN, M. M., 1976. Water hya-

cinths for removal of phenols from polluted waters. Aquat.

Bot. 2: 191-201.

YOUNT, J. L. & CROSSMAN, R. A. Jr. 1970. Eutrophication

control by plant harvesting. /. Wat. Pollut. Control Fed. 42:

173-183.

ZABAT, M. D., OSWALD, W. J., GOLUEKE, C. T. & GEE,

H., 1970. Kinetics of algal systems in waste treatment. Phos-

phorus as a growth limiting factor. Report of the Sanitary

Engineering Research Laboratory, University of California.

Bothalia 15, 3 & 4: 705-724 (1985)

The development from kinetic coefficients of a predictive model for the

growth of Eichhornia crassipes in the field. II. Testing and refining the

model under field conditions

C. F. MUSIL * and C. M. BREEN**

Keywords: Eichhornia crassipes, model validation, water nutrient concentrations

ABSTRACT

Kinetic coefficients derived for Eichhornia crassipes (Mart.) Solms under culture conditions of N and P limita-

tion were used in the Monod model to identify the limiting nutrient and to predict specific growth rates under

conditions of varying water nutrient concentration and air temperature. Predicted data were validated by compari-

son with specific growth rates measured for plants growing in loosely and densely crowded populations at two field

sites. The use of culture-derived maximum specific growth rates (Umax) in the model resulted in inaccurate predic-

tions of plant growth rates in loosely and densely crowded field populations. The use of field-derived Umax values

in the model, however, resulted in adequate predictions of plant growth rates in loosely crowded field populations.

The incorporation of radiant flux density (diffuse component of the radiant flux) and relative humidity into the

model considerably improved its accuracy of prediction. In all cases, specific growth rates were more accurately

predicted from the limiting total N or total P concentrations, than from other N or P fractions, in the water.

INTRODUCTION

Harvesting Eichhornia crassipes (Mart.) Solms

(water hyacinth) growing in eutrophied aquatic sys-

tems may constitute an effective means of removing

nutrients and controlling excessive growth of plants

(Boyd, 1970; Yount & Crossman, 1970). However,

to achieve maximum nutrient removal efficiency by

E. crassipes in a nutrient removal scheme, it is

necessary to establish the size of the population re-

quired to maintain desirable nutrient concentrations

in the water under varying conditions of nutrient

loading and climate, and the amounts and frequen-

cies of harvest required to control the population

size.

From the kinetic standpoint, it is theoretically

feasible to construct a mathematical model for E.

crassipes from which population sizes, yields, growth

rates and frequencies and amounts of harvest, under

varying conditions of nutrient loading and climate,

can be predicted to control nutrient inputs and ex-

cessive growth in eutrophied aquatic systems (Toe-

rien, 1972; Musil & Breen, 1977). Musil & Breen

(1985) measured the kinetic coefficients, maximum

specific growth rate (Umax), half saturation coeffi-

cient (Ks) and yield coefficient (Yc) for E. crassipes

growing in N and P tlimited batch cultures in a

greenhouse with the objective of developing a pre-

dictive model. This investigation was designed to

test the validity of and refine these culture deter-

mined kinetic coefficients for predicting growth rates

of E. crassipes under field conditions. In testing the

model, two assumptions were made:

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

** Botany Department, University of Natal, P.O. Box 375,

Pietermaritzburg 3200.

t Nutrients present at concentrations below those required for

maximum plant growth and hence restricting the growth rate.

(i) That the maximum specific growth rates

(Umax), derived under culture conditions of N and

P limitation (Musil & Breen, 1985), followed the

van’t Hoff rule, i.e. they approximately doubled for

each 10°C rise in the temperature, as has been de-

monstrated for various species of marine and fresh

water algae (Goldman, 1972; Goldman & Car-

penter, 1974).

(ii) That the specific growth rate (U) was not limi-

ted in a multiplicative or additive manner, but in a

threshold mode by the single nutrient in shorter sup-

ply. This principle, based on ‘Liebig’s law of the mi-

nimum’ that the maximum population size or maxi-

mum yield in plant material is controlled by the sin-

gle factor in shorter supply (Blackman, 1905), has

been observed to apply in the regulation of phyto-

plankton organic production by soluble nutrients

(Brandt, cited in Gran, 1912). More recently, this

has been extended to include the regulation or con-

trol of phytoplankton growth rate by the limiting nu-

trient (O’Brien, 1972). Droop (1974), for example,

has shown that the growth rate of Monochrysis lu-

theri growing under P and B12 limitation in culture is

not limited in a multiplicative pattern, but by the sin-

gle nutrient in shorter supply. Rhee (1978) also sur-

mized that the growth rate of Scenedesmus sp. is

limited in a threshold pattern.

Apart from air temperature, seasonal variations in

radiant flux density and relative humidity may influ-

ence E. crassipes growth rate under field conditions.

In view of this, two hypothetical, multiplicative ex-

pressions were investigated for correcting the pre-

dicted growth rates for the effects of radiant flux

density and relative humidity respectively. These are

defined as follows:

Umax = Ae-E/R(TxI) (i)

where Umax = maximum specific growth rate g

fresh mass g-1 d-1; T = absolute mean daily air tem-

perature °K; I = radiant flux density (diffuse compo-

706

Bothalia 15, 3 & 4 (1985)

nent of the radiant flux) MJ nr2 hr1; A = constant

day-1; E = activation energy cal. mole-1; R = uni-

versal gas constant cal. mole-1 °Kr1.

The first multiplicative expression assumed that

the effect of radiant flux density on E. crassipes

growth rate could not be considered independent of

the effect of temperature. This is in accordance with

the observed interaction between temperature and

light intensity in influencing both growth rates and

photosynthetic responses of algae (Sorokin, 1960;

Sorokin & Krauss, 1962; Maddux & Jones, 1964;

Smayda, 1969; Eppley, 1972; Harris & Lott, 1973)

and higher plant species (Pisek et al., 1973; Billings,

1974). Several environmental modellers (Di Toro et

al., 1971; Chen & Orlob, 1972; Park et al., 1975;

Kieffer & Enns, 1976) have used a multiplication of

independent light and temperature functions in phy-

toplankton population/productivity models, in many

instances without experimental evidence. Rodhe

(1948, 1978) suggested that the combined effects of

factors such as temperature, light and daylength on

the population dynamics of phytoplankton may be

more important than the effect of any single one fac-

tor.

Umax = T x i x rh) (ji)

where Umax = maximum specific growth rate g

fresh mass g-1 d-1; T = absolute mean daily air tem-

perature °K; I — radiant flux density (diffuse compo-

nent of the radiant flux) MJ nr2 fr1; RH = mean

daily relative humidity %; A = constant day-1; E —

activation energy cal. mole-1; R = universal gas con-

stant cal. mole-1 “K.-1.

The second multiplicative expression assumed

that the observed effect of relative humidity on E.

crassipes growth rate (Freidel et al., 1978) could not

be considered independent of the effect of tempera-

ture and radiant flux density. This is because of the

interaction between temperature, light intensity and

relative humidity in influencing transpiration rates

(Crafts et al., 1949) which in turn may indirectly in-

fluence growth rate, possibly by altering the water

potential (Slayter, 1967; Meidner & Sheriff, 1976) of

leaf cells. Cell and leaf growth are highly sensitive to

a reduced water potential, particularly as cell expan-

sion is caused by the action of turgor pressure upon

‘softened’ cell walls (Greacen & Oh, 1972). In fact,

Hsiao et al. (1976) show that even mild water stress

in mesophytic leaves, i.e. where the water potential

of leaf cells is reduced by only a few bars, can result

in a reduction in growth rate and the disruption of

several metabolic processes, including protein and

chlorophyll biosynthesis.

LOCALITIES AND METHODS

Sites

Two sites, characterized by different nutrient con-

centrations in the water, were selected for field esti-

mations of specific growth rates. These rates, meas-

ured periodically throughout the year, were com-

pared with those predicted by the use of culture-de-

rived kinetic coefficients (Musil & Breen, 1985) in

the Monod model. The sites are in the Durban Dis-

trict of Natal (Fig. 1) in the climatic region described

by Schulze (1965) as warm to hot and humid, subtro-

pical. The Maturation Pond 3 (MP3) site is enriched

by secondary treated waste-water effluent dis-

charged from the northern sewage treatment works

and the Botanic Gardens Lake (BGL) site by ferti-

lizer run-off.

Measurement of growth

Specific growth rates were measured by tagging

plants and introducing them for 12 to 14 day periods

into field populations (Bock. 1966; 1969). These

rates were measured in both loosely and densely

crowded field populations. Mitsch (1977) found that

the two different growth forms, viz marginal and

central forms (Musil, 1982), associated with these

two different population densities (Rao, 1920; Mc-

Clean, 1922; Lansdell, 1925; Bruhl & Gupta, 1927;

La Garde, 1930; Weber, 1950) exhibit different net

carbon uptakes and photosynthesis/respiration ratios

which suggests that they may have different growth

rates under similar field conditions.

Vegetatively propagated offsets (daughter plants)

possessing three pseudolaminae of the marginal

(fresh mass: ca 7 to 22 g) and central (fresh mass: ca

47 to 111 g) forms were collected at 12 to 14 day in-

tervals from loosely and densely crowded popula-

tions at each site. Plants were washed in site water,

to remove all extraneous particles, tagged and al-

lowed to drain for two minutes. They were shaken to

dislodge adhering water, their fresh masses recorded

on an electric, top-loading balance and introduced

into enclosures. Two sizes of enclosures were used.

Both were constructed of plastic-coated wire mesh

held in place by metal fencing posts driven into the

sediment. To minimize disturbance by wind and

wave action, they were located on the leeward side

of each water body in water ca 1,0 m deep.

Cylindrical enclosures with a diameter of ca 1 m

and a height of ca 1,5 m were used for containing

plants of the marginal form at each site. The water

area, ca 0,8 m2, contained within each enclosure

was adequate to accommodate an increase in the size

of the introduced population, over a 12 to 14 day

period, without causing the plants to become unduly

crowded. Forty plants of the marginal form collected

at each site were introduced into four enclosures (10

plants per enclosure) located at each site.

Two densely crowded populations (ca 6 m2) of

the central form were enclosed by wire mesh, ca 1,5

m high, at the MP3 site. This ensured that the plants

within were kept in densely crowded situations and

maintained in the central form. Thirty plants of the

central form collected at this site were inserted at

random (15 plants per enclosure) into the two en-

closed populations.

After 12 to 14 days, marginal and central forms

and their offsets were harvested from the enclosures

at each site. Care was taken not to separate the off-

sets from their respective parents. Plants were pick-

ed free of debris, washed and reweighed as de-

scribed above. Specific growth rates were calculated

for marginal and central forms, over each 12 to 14

Bothalia 15, 3 & 4 (1985)

707

FIG. 1. — Location of field sites. 1, maturation pond, northern sewage treatment works; 2, Botanic Gar-

dens Lake.

day interval, using the general growth equation (Ma-

lek & Fencl, 1966; Radford, 1967):

= In Xt - In Xo

t

where Xo = fresh mass at time = c(g); Xt = fresh

mass at time = t2(g); U = specific growth rate (g

fresh mass g-1 d-1); t = time period between time t2

and tj (days); €n = loge (natural logarithm).

Chemical analyses

Water samples were collected between llhOO and

14h00 from within the loosely and densely crowded

populations enclosed at each site, at the commence-

ment and termination of each growing interval.

Water samples were collected ca 20 cm below the

water surface, to avoid surface contamination, in 500

ml plastic bottles (Golterman, 1969) previously

cleaned with cone. HC1 and rinsed thoroughly in

deionised-distilled water. Bottles were sealed with

Parafilm (American Can Company, Greenwich,

Connecticut) and immediately transported in an in-

sulated container to the laboratory.

The following N and P fractions were analysed in

the water samples using published methods (En-

vironmental Protection Agency, 1974; American

Public Health Ass.: Standard methods, 1975). In fil-

tered samples (Rigler, 1964; Olsen, 1967; Golter-

man, 1969), nitrate-nitrogen (NO3-N) by colorime-

try after reduction to nitrite and soluble reactive

phosphorus (SRP) (Twinch & Breen, 1980) by colo-

rimetry using the molybdenum blue method. In un-

filtered samples, Kjeldahl nitrogen as ammonium

(NH4-N) after digestion of the samples by cone.

H2S04 in the presence of a mercury catalyst and

708

Bothalia 15, 3 & 4 (1985)

total phosphorus (total P) as SRP after digestion of

the samples with H2S04 and persulphate. Total ni-

trogen (total N) was calculated as the sum of Kjel-

dahl nitrogen and nitrate plus nitrite (NH4-N +

N03-N + N02-N).

The pH of the water was recorded daily between

llhOO and 14h00, beneath the loosely and densely

crowded populations enclosed at each site, with an

electrolytic probe connected to a portable pH meter.

Physical analyses

Over each growing interval, hourly measurements

of radiant flux density (diffuse component of the ra-

diant flux) and daily maximum, minimum and mean

air temperatures and relative humidities were ob-

tained from the nearby meteorological station at

Louis Botha Airport, Durban.

The diffuse component of the radiant flux was

chosen as a measure of the light as this includes a

greater proportion of the photosynthetically active

radiation. About one-third of the direct solar radia-

tion, often referred to as the global component, is

photosynthetically active compared with over two-

thirds for the diffuse component (Ross, 1975; Fitter

& Hay, 1981). Theoretical calculations have shown

that even under cloudless skies, the diffuse radiation

(D) may account for between one-third and three-

quarters of the total irradiance (T), and in a series of

FIG. 2. — Air temperatures, as daily averages over each growing interval, at two field sites.

MATURATION POND 3 BOTANIC GARDENS LAKE

FIG. 3. — Relative humidities, as daily averages over each growing interval, at two field sites.

Bothalia 15, 3 & 4 (1985)

709

FIG. 4. — Diffuse radiant fluxes, as hourly averages over each growing interval, at two field sites.

measurements Szeiez (1974) showed that the ratio

D/T was always greater than 0,5.

RESULTS AND DISCUSSION

Field data

Physical factors

Air temperatures, relative humidities and diffuse

radiant fluxes at the two sites are depicted in Figs 2,

3 & 4. They showed a seasonal pattern decreasing

progressively after summer (September to March)

through to winter (May to August). Mean daily air

temperatures and relative humidities at the two sites

ranged from 16,1 to 25,2°C and from 64 to 85% re-

spectively. Diffuse radiant fluxes ranged from 0,30

to 0,72 MJ m-2 h-i.

Chemical factors

Nitrogen and P concentrations in the water and

the variation in water pH beneath the enclosed,

loosely crowded populations at the two sites are

shown in Figs 5, 6 & 7.

Total N, NH4-N, SRP and total P concentrations

in the water at the BGL site were considerably lower

than those at the MP3 site, whereas N03-N concen-

FIG. 5. — Nitrogen (N03-N, NFL,-N and total N) concentrations in the water, averaged over each growing interval, beneath

loosely crowded populations enclosed at two field sites.

710

Bothalia 15, 3 & 4 (1985)

FIG. 6. — Phosphorus (SRP and total P) concentrations in the water, averaged over each growing interval, beneath loosely

crowded populations enclosed at two field sites.

FIG. 7. — The variation in water pH, as daily averages over each growing interval, beneath loosely crowded populations

enclosed at two field sites.

trations in the water were much higher. The latter

was attributed partly to enrichment by fertilizer run-

off and to a higher rate of nitrification at the BGL

site (Musil, 1982). There were no differences in the

N and P concentrations in the water beneath the

loosely and densely crowded populations enclosed at

the MP3 site (Musil, 1982).

At both sites, the water pH values were in close

proximity to pH 7,0. Only minor variations in the

water pH (pH 7,0 to 7,3 and pH 7,2 to 7,7 at the

BGL and MP3 sites respectively) occurred at these

two sites during the year. There were no differences

in the water pH beneath the loosely and densely

crowded populations enclosed at the MP3 site (Mu-

sil, 1982).

Specific growth rates

Specific growth rates of marginal and central

forms, growing in loosely and densely crowded pop-

ulations respectively, at the two sites are illustrated

in Fig. 8.

At both sites, specific growth rates of marginal

forms followed a distinct seasonal pattern with val-

ues decreasing progressively after summer (Septem-

ber to March) through to winter (May to August).

During 1978, the highest specific growth rates,

0,1698 and 0,1227 g fresh mass g-1 d-1 (16,98 and

12,27% d-1 at the MP3 and BGL sites respectively,

were measured during summer, in February, with

the lowest specific growth rates, 0,0526 and 0,0305 g

Bothalia 15, 3 & 4 (1985)

711

FIG. 8. — Specific growth rates of £. crassipes over each growing interval, at two field sites. Solid line = marginal forms

growing in loosely crowded populations (means of 40 replicates). Broken line = central forms, growing in densely

crowded populations (means of 30 replicates). No plants of the central form were produced during June, July and Au-

gust. Standard deviations of measured specific growth rates are shown by bars.

fresh mass gr1 cH (5,26 and 3,05% d-1) at these two

sites respectively, being measured for marginal

forms during midwinter, in June. Throughout 1978,

specific growth rates of marginal forms at the MP3

site were significantly higher (P=S 0,01), than those

at the BGL site (Musil, 1982) and reflected the

higher total N and total P concentrations in the

water at the MP3 site (Figs 5 & 6). In general, speci-

fic growth rates of marginal forms at both sites fell in

the range of specific growth rates (3,0 to 12,5% d-1)

reported by various authors (Seaman & Porterfield,

1964; Bock, 1969; Knipling et al., 1970; Morris,

1974; Boyd, 1976) for E. crassipes growing under

subtropical to tropical climates in other parts of the

world.

In contrast to marginal forms, specific growth

rates of central forms at the MP3 site did not show

any distinct seasonal pattern, since no plants of the

central growth form were produced during the mid-

winter months of June, July and August. The highest

specific growth rate, 0,0659 g fresh mass g-' d 1

(6,59% d'1), was measured for central forms during

summer, in December, 1977, with the lowest specific

growth rate, 0,0202 g fresh mass g-1 d-1 (2,02% d->),

being measured during winter, in May, 1978.

Throughout 1977 and 1978, specific growth rates of

central forms at the MP3 site were significantly

lower (17 to 35%, P=5 0,001) than those of marginal

forms (Musil, 1982).

The significantly lower specific growth rates meas-

ured in densely crowded field populations may be

partly related to the adverse effects of self shading

and intraspecific competition arising through over-

crowding in such populations, as well as to the intrin-

sic morphological limitations of plants of the central

form occurring in such populations. Center &

Spencer (1981) have shown that the production of E.

crassipes plants with elongate petioles (central

forms) results in a decline in the lamina area ratio

(LAR), i.e. in relatively less photosynthetic area per

unit of plant weight. The ratio of the lamina area to

plant weight is similar to the leaf area ratio of other

authors (e.g. Beevers & Cooper, 1964; Radford,

1967). Consequently, if it is assumed that photosyn-

thesis is proportional to the lamina area and respira-

tion to weight, the LAR should be an index of the

P/R ratio and an indicator of the growth potential,

i.e. net photosynthesis. Under these assumptions,

central forms with elongate petioles should have the

smallest potential for growth and consequently the

lowest specific growth rate.

Testing the model

Identifying the limiting nutrient

At each site, the nutrient limiting E. crassipes

growth rate was estimated from the average total N

and total P concentrations in the water using the

mean half saturation (Ks) concentrations of 976 pg

N G1 and 94,1 pg P G1, derived under culture

conditions of N and P limitation (Musil & Breen

1985), in the Monod model. For example, the aver-

age total N and total P concentrations determined in

the water at the MP3 site during 1978 were 20 746 pg

N 6-1 and 6 569 pg P G1 (Musil, 1982). The per-

centage of the maximum specific growth rate

(Umax) that E. crassipes would achieve at (i) the

average total N, (ii) the average total P concentrations

in the water at this site were estimated using the

Monod model as follows:

20 746

U = UmaX 976 + 20 746 X 100 (l)

= 95,5% Umax

U = UmaX 94,1 + 6 569 X 100 (ii)

= 98,6% Umax

712

Bothalia 15, 3 & 4 (1985)

The results show that at the MP3 site E. crassipes

would achieve a lower percentage of the Umax at

the average total N than at the average total P con-

centrations in the water which indicates that N was

the limiting nutrient.

At the BGL site, on the other hand, the average

total N and total P concentrations determined in the

water during 1978 were 10 206 pg N and 150 pg

P €_1 (Musil, 1982). Using the Monod model, it was

estimated that E. crassipes would achieve 91,3% and

61,4% of the Umax at the average total N and total

P concentrations in the water respectively at this

site, indicating that P was the limiting nutrient.

Predicting Umax for different temperatures

The Umax values derived for E. crassipes under

culture conditions of N and P limitation, at mean

daily air temperatures of 24°C and 28°C respectively,

were 0,0886 g fresh mass g~l d-1 for N and 0,1089 g

fresh mass g-1 d-1 for P (Musil & Breen, 1985). Using

these values, the Umax of E. crassipes, under condi-

tions of N or P limitation, may be predicted for other

temperatures according to the Van’t Hoff rule

(Tables 1 & 2) from the Arrhenius equations for the

exponential relationships between Umax(n) and

Umax(p) and temperature. Assuming a Q10 of 2,0,

these relationships are:

Umax(n) - 3,9151 x 10^59i67r 1

where Umax(n) = maximum specific growth rate for

N (g fresh mass g-] d-1); T = absolute mean daily air

temperature (°K).

Umax(p) - 6,5292 x KF^ossat 2

where Umax(p) = maximum specific growth rate for

P (g fresh mass g-1 d-1); T = absolute mean daily air

temperature (°K).

Comparison of predicted and measured specific

growth rates

Incorporating these equations and the Ks concen-

trations of 976 jug N U1 and 94,1 pg P (M into the

Monod model, specific growth rates were predicted

for E. crassipes, over each growing interval at the

MP3 and BGL sites, from the limiting N or P con-

centrations in the water and mean daily air tempera-

tures. These were then compared with measured

TABLE 1. — Maximum specific growth rates (Umax) predicted

for E. crassipes for various temperatures according to the

van’t Hoff rule. Predictions were based on the Umax value

of 0,0886 g fresh mass g"ld'l, derived under culture con-

ditions of N limitation at a mean daily air temperature of

24°C (Musil & Breen, 1985)

TABLE 2. — Maximum specific growth rates (Umax) predicted

for E. crassipes for various temperatures according to the

van’t Hoff rule. Predictions were based on the Umax value

of 0,1089 g fresh mass g" Id' 1, derived under culture con-

ditions of P limitation at a mean daily air temperature of

28°C (Musil & Breen, 1985)

specific growth rates. At the MP3 site, where N was

estimated to be the limiting nutrient, specific growth

rates were predicted from the various N (N03-N,

NH4-N and total N) fractions in the water using

Equation 1 in the Monod model. At the BGL site,

where P was estimated to be the limiting nutrient,

specific growth rates were predicted from the vari-

ous P (SRP and total P) fractions in the water using

Equation 2 in the Monod model. For example, the

mean daily air temperature and total N concentra-

tion in the water at the MP3 site over the growing

interval 1/2 to 16/2/78 were 24,9°C and 15 970 pg

N U1 respectively (Musil, 1982). The specific

growth rate (U) was predicted for E. crassipes for

this set of conditions as follows:

U = 3,9151 X 107e -5916/24,9 + 273,2 X U

976 + 15 970

= 0,0887 g fresh mass g-1 d-1 (8,87% d ').

Predicted specific growth rates and those meas-

ured for marginal and central forms, growing in

loosely and densely crowded populations respect-

ively, at the two sites are illustrated in Figs 9 and 10.

At the MP3 site, specific growth rates predicted

from total N, NH4-N and to a lesser extent from

N03-N concentrations in the water followed a fairly

similar seasonal pattern to those measured for mar-

ginal forms with values decreasing progressively

after summer (September to March) through to win-

ter (May to August). A similar relationship between

measured specific growth rates and those predicted

from total P concentrations in the water was ob-

tained at the BGL site. Specific growth rates pre-

dicted from SRP concentrations in the water at the

BGL site, however, were extremely variable and did

not follow any recognizable seasonal pattern.

In general, specific growth rates predicted from

the various N or P fractions in the water at the two

sites were significantly lower than those measured

for marginal forms. For example, of the 29 specific

growth rates predicted from total N concentrations

in the water at the MP3 site, only 3 (during April and

June) fell within the standard deviations of meas-

ured values. Of the 22 specific growth rates pre-

dicted from total P concentrations in the water at the

BGL site, only 6 (during March, April, June and

Bothalia 15, 3 & 4 (1985)

713

FIG. 9. — Specific growth rates predicted for E. crassipes from 2, total N; 3, NH4-N and 4, NO3-N concentrations in the water, over

each growing interval, at the maturation pond 3 site compared with those measured for 1, marginal forms, growing in loosely

crowded populations and 5, central forms, growing in densely crowded populations. Standard deviations of measured specific

growth rates are shown by bars. Umax values used in the Monod model derived under culture conditions of N limitation and

expressed as a function of air temperature.

July) fell within the standard deviations of measured

values. The differences between measured specific

growth rates and those predicted from other N

(NH4-N and NO3-N) or P (SRP) fractions in the

water at the two sites, however, were generally

larger than the differences between measured speci-

fic growth rates and those predicted from total N or

total P concentrations in the water. This suggests

that specific growth rates of E. crassipes in the field

may be more accurately predicted from total N or

total P concentrations, than from other N or P frac-

tions, in the water. At both sites, the smallest differ-

ences between measured specific growth rates and

those predicted from total N or total P concentra-

tions in the water occurred during midwinter in June

when relative humidities and diffuse radiant fluxes

were at their lowest levels (Figs 3 & 4). These results

suggest that the Umax values derived for E. cras-

sipes, under culture conditions of N and P limitation

(Musil & Breen, 1985), may have been depressed by

lowered relative humidities and radiant flux densi-

ties in the greenhouse, since specific growth rates

predicted from total N or total P concentrations in

the water at the two sites were significantly lower

than those measured for marginal forms, except

when diffuse radiant fluxes and relative humidities in

the field were low.

Relative humidities recorded in the greenhouse

during the experimental determination of kinetic co-

efficients for E. crassipes (Musil & Breen, 1985)

were lower than those recorded during summer for

the two field sites in the Durban area (Fig. 3). Mean

daily relative humidities in the greenhouse ranged

from 61 to 67%, whereas those recorded during

summer for the two field sites ranged from 74 to

85%. Since Freidel et al. (1978) have shown that E.

crassipes growth rate decreases with a decrease in

relative humidity, it would appear that the Umax

values derived for E. crassipes in culture were de-

pressed by lower relative humidities in the green-

house. In addition, it is possible that radiant flux

densities in the greenhouse limited E. crassipes

growth rate during the experimental determination

of kinetic coefficients for this plant. Measurements

of light intensity at midday in full sunlight inside and

outside the greenhouse showed that light intensity in

the greenhouse was attenuated by ca 37% (Musil.

1982).

No correction factor could be incorporated into

the model to amend the Umax values derived for E.

crassipes under lowered relative humidities and ra-

diant flux densities in the greenhouse, since the ra-

tios between measured and predicted specific growth

714

Bothalia 15, 3 & 4 (1985)

FIG. 10. — Specific growth rates predicted for E. crassipes from 2, total P and 3, SRP concentrations in the water, over each

growing interval, at the Botanic Gardens Lake site compared with those measured for 1, marginal forms, growing in loosely

crowded populations. Standard deviations of measured specific growth rates are shown by bars. Umax values used in the

Monod model derived under culture conditions of P limitation and expressed as a function of air temperature.

rates varied considerably during the year. For ex-

ample, at the MP3 site the ratios between specific

growth rates measured for marginal forms and those

predicted from total N concentrations in the water

ranged from ca 2,5 during summer to ca 1,1 during

winter. Similarly, at the BGL site, the ratios be-

tween measured specific growth rates and those pre-

dicted from total P concentrations in the water

ranged from ca 2,6 during summer to ca 1,1 during

winter.

In contrast to specific growth rates of marginal

forms, those of central forms growing in densely

crowded populations at the MP3 site were often sig-

nificantly lower than those predicted from the vari-

ous N fractions in the water (Fig. 9). For example,

only 2 of the 18 specific growth rates predicted from

total N concentrations in the water at this site fell

within the standard deviations of measured values.

No significance, however, was attached to the obser-

vation that the differences between specific growth

rates measured for central forms and those predicted

from total N concentrations in the water were

slightly larger than the differences between meas-

ured specific growth rates and those predicted from

other N (NH4-N and N03-N) fractions in the water.

The poor affinity obtained between predicted speci-

fic growth rates and those measured for central

forms may be attributed to the specific growth rate

of E. crassipes being depressed by intraspecific com-

petition and self shading in densely crowded field

populations. This suggests that, unless a correction

factor is introduced into the model to amend the

Umax for the density of the plant population, Umax

values derived for plants of the marginal growth

form in culture cannot be used in the Monod model

to predict specific growth rates of central forms

growing in densely crowded field populations.

Although Umax values derived for E. crassipes

under culture conditions of N and P limitation were

of little value in the Monod model for accurately

predicting specific growth rates of plants growing in

loosely or densely crowded field populations, it is

possible that more reliable estimates of Umax may

be derived for E. crassipes from the field data. At

the MP3 site, where N and P concentrations in the

water were very high, it was estimated from the

mean Ks concentrations, derived under culture con-

ditions of N and P limitation, that E. crassipes would

achieve ca 95,5 and 98,6% of its Umax at the aver-

age total N and total P concentrations in the water

respectively. It is evident, therefore, that, even

though N was considered to be the limiting nutrient

at this site, N and P concentrations in the water ap-

proached those saturating to E. crassipes growth

rate. Consequently, if it is assumed that specific

growth rates measured for plants of the marginal and

Bothalia 15, 3 & 4 (1985)

715

central forms at specific temperatures at the MP3

site closely approximate their respective Umax val-

ues, it should be possible to express the relation-

ships, if exponential, between these estimated Umax

values of marginal and central forms and air tem-

peratures in the form of Arrhenius equations. These

equations can then be incorporated into the Monod

model to predict specific growth rates of marginal

and central forms, from the limiting N or P concen-

trations in the water and mean daily air tempera-

tures, at other field sites. This should improve the

model’s accuracy of prediction for plants growing in

both loosely and densely crowded field populations.

The mean Ks concentrations derived for E. crassipes

under culture conditions of N and P limitation, on

the other hand, appear fairly reliable, since pre-

dicted specific growth rates generally followed a

similar seasonal pattern to those measured.

Refining the model

Deriving Umax in the field

Arrhenius plots of specific growth rates (estimated

Umax values) of marginal and central forms, ex-

pressed as natural logarithms (Loge), against the re-

ciprocals of absolute mean daily air temperatures at

the MP3 site (Figs 11 & 12), over the period Feb-

ruary to December, 1978, yielded linear relation-

ships with the correlation coefficients being high and

significant at P^ 0,01.

The regression equation exponentially relating

specific growth rates (estimated Umax values) of

marginal forms at the MP3 site to the reciprocals of

absolute mean daily air temperatures was:

U = 5,2631 x 108^54orr 3

where U = specific growth rate (estimated maxi-

mum specific growth rate) g fresh mass g-> d_1; T =

absolute mean daily temperature °K.

A Q10 value of 2,14, in the temperature range 15

to 25°C, was calculated from the above expression.

This compares favourably with the Q10 value of 2,12

reported by Goldman (1972) for the exponential re-

lationship between the Umax values of various

species of fresh water algae and temperature, in the

range 19 to 39°C. It demonstrates that the effect of

air temperature on the Umax of E. crassipes (margi-

nal forms) follows the Van’t Hoff rule and confirms

the initial hypothesis made. In addition, the acti-

vation energy of 12,978 calories mole-1 calculated

for marginal forms from the above expression com-

pares favourably with the activation energy of 13,356

calories mole-1 reported by Goldman (1972) for al-

gae.

The regression equation exponentially relating

specific growth rates (estimated Umax values) of

central forms at the MP3 site to the reciprocals of

absolute mean daily air temperatures was:

U - 1,9932 x I05e-466i rv 4

where U = specific growth rate (estimated maxi-

mum specific growth rate) g fresh mass g-[ d_1; T =

absolute mean daily air temperature °K.

FIG. 11. — An Arrhenius plot of specific growth rates (estimated Umax values) of marginal forms (Loge) against the reciprocals of

absolute mean daily air temperatures at the maturation pond 3 site.

716 Bothalia 15, 3 & 4 (1985)

FIG. 12. — An Arrhenius plot of specific growth rates (estimated Umax values) of central forms (Log,,) against the reciprocals of

absolute mean daily air temperatures at the maturation pond 3 site.

A Q10 value of only 1,71 in the temperature range

15 to 25°C, was calculated from the above expres-

sion. This demonstrates that central forms growing

in densely crowded field populations show a propor-

tionally smaller increase in their specific growth rate

(estimated Umax) with a 10°C rise in the mean daily

air temperature than do marginal forms growing in

loosely crowded field populations. The lower Q10

value obtained for central forms was attributed to

their specific growth rate being depressed in densely

crowded field populations.

With respect to the above exponential expres-

sions, it should be pointed out, however, that recent

investigations have shown that plant growth in the

field can respond linearly rather than exponentially

to environmental temperature, which suggests that

the asymetric bell-shaped response to temperature

may not be as widely applicable as was formerly con-

sidered. Gallagher & Biscoe (1979), for example,

have demonstrated that, in the absence of water

stress, the expansion rate of barley leaves is directly

proportional to the temperature of the stem apex. It

remains to be seen, however, what the theoretical

interpretation of such linear responses can be.

Comparison of predicted and measured specific

growth rates

The usefulness of Equation 4 in the Monod model

could not be assessed, since no measurements of

specific growth rates of central forms were obtained

at the BGL site.

Substituting Equation 3 for Equation 1 in the

Monod model, specific growth rates were repre-

dicted for marginal forms at the BGL site from the

limiting P (SRP and total P) concentrations in the

water and mean daily air temperatures. Predicted

specific growth rates and those measured for margi-

nal forms are illustrated in Fig. 13.

Of the 22 specific growth rates predicted from

total P concentrations in the water at the BGL site,

14 (ca 64%) fell within the standard deviations of

measured specific growth rates. Predicted specific

growth rates generally followed the same seasonal

pattern as measured specific growth rates with va-

lues decreasing progressively after summer (Septem-

ber to March) through to winter (May to August).

Bell (1981) pointed out that, by analogy with the

correlation coefficient (r), a coefficient of variation

(R2) may be computed to test model outputs com-

pared with data defined as:

7 _ sum of squares of residuals

n SD2y

where n is the number of data points and SD2y is the

variance in predicted values. For values of R2 that

are high and approximate to 1, the fit is good; for

values of R2 that are low and approximate to 0, the

Bothalia 15, 3 & 4 (1985)

717

fit is poor. For values of R2 that are inbetween and

approximate to 0,5, the situation is uncertain. Dent

& Blackie (1979) recommended a simple regression

analysis between model outputs and data as paired

observations which produces the same value of R2 as

defined. The coefficient of variation (R2) calculated

between measured specific growth rates and those

predicted from total P concentrations in the water at

the BGL site had a value of R2 = 0,5321. Conse-

quently, the observed affinity between predicted and

measured values could not be regarded as significant

over the entire growing season. Specific growth rates

predicted from SRP concentrations in the water at

this site were extremely variable and did not follow

any recognizable seasonal pattern. Only one of the

predicted values fell within the standard deviations

of measured values.

The differences between measured specific growth

rates and those predicted from total P concentra-

tions in the water at the BGL site were considerably

smaller than the differences between measured

specific growth rates and those predicted from other

P (SRP) fractions in the water (Fig. 13). Similar re-

sults were obtained at sites where N was estimated to

be the limiting nutrient (Musil, 1982). These results

indicate that specific growth rates of E. crassipes in

the field are more accurately predicted from the

limiting total N or total P concentrations, than from

other N or P fractions, in the water which suggests

that the plant may utilize both inorganic and organic

forms of N and P for growth.

With respect to P, Jeschke & Simonis (1965) re-

ported that the main source of P for growth of aqua-

tic plants is in the form of inorganic phosphates.

However, specific growth rates of marginal forms at

the BGL site were more accurately predicted from

total P than from SRP concentrations in the water.

Consequently, it would appear that total P concen-

trations in the water at this site better reflected the

total amount of P available to plants during growth,

some of the P for growth of plants possibly being

provided by release of that bound to sediments as

well as to other soluble and insoluble fractions in the

water. For example, when P is added to lakes, it is

rapidly removed from solution by adsorption onto

sediments (Hepher, 1958; Hayes & Phillips, 1968).

This P is not rendered entirely unavailable since

sediment P and dissolved P exist in equilibrium (He-

pher, 1958; Pomeroy et al., 1965). The equilibrium

concentration increases with increased P content in

the sediment (Pomeroy et al., 1965). Consequently,

removal of P from the water by E. crassipes during

growth would displace the P equilibrium allowing

additional P to be released from the sediments into

the overlying water. Alternatively, it has been

shown that many zooplankton and phytoplankton

.1978

FIG. 13. — Specific growth rates predicted for marginal forms from 2, total P and 3, SRP concentiations in the water, over each

growing interval, at the Botanic Gardens Lake site compared with 1, measured specific growth rates. Standard deviations of

measured specific growth rates are shown by bars. Umax values used in the Monod model derived under field conditions and

expressed as a function of air temperature.

718

Bothalia 15. 3 & 4 (1985)

species excrete alkaline phosphatases which may ac-

celerate phytoplankton growth by supplying ortho-

phosphate from suitable organic esters (Berman,

1969; 1970; Jansson, 1976; Wynne & Gophen,

1981). The excretion of alkaline phosphatases by

higher aquatic plant species has not been reported in

the literature, though Wetzel (1969a, 1969b) has

shown that some species do excrete various dis-

solved organic compounds. One may, therefore,

speculate that species such as E. crassipes may ex-

crete alkaline phosphatases which would allow them

to utilize normally unavailable organic forms of P for

growth.

With respect to N, Sculthorpe (1967) suggested

that NH4-N does not serve as an N source for growth

of aquatic plants. However, several authors (Von

Schwoerbel & Tillmans, 1972; Toetz, 1971; 1973)

have subsequently shown a preference by aquatic

plants for NH4-N as an N source for growth. Best

(1980) demonstrated that, although NH4-N supplied

at low concentrations for a short period stimulated

the growth of Ceratophyllum demersum L. in cul-

ture, higher concentrations (in excess of 45 x 103 pg

NH4-N T1) applied for a prolonged period were

toxic. The ammonium-induced inhibition of growth

has been reported for several other aquatic macro-

phyte species (Mulligan et al., 1976). High NH4-N

concentrations in the water may repress the in-

duction of nitrate reductase as shown by Joy (1969)

and Orebamjo & Stewart (1975a, 1975b) in Lemna

minor L. Reduced nitrate reductase activity would

mean reduced N uptake and consequently a reduc-

tion in growth rate.

The largest differences between measured specific

growth rates and those predicted from total P con-

centrations in the water at the BGL site (Fig. 13)

occurred during the summer months (September to

March) when radiant flux densities (diffuse compo-

nent of the radiant flux) were high (Fig. 4). This sug-

gested that if the effects of radiant flux density plus

air temperature were incorporated into the model, it

might improve its accuracy of prediction.

Correcting Umax for radiant flux density

An Arrhenius plot of specific growth rates (esti-

mated Umax values) of marginal forms, expressed as

natural logarithms (Loge), against the products of

the reciprocals of absolute mean daily air temper-

atures and diffuse radiant fluxes at the MP3 site (Fig.

14), over the period February to December, 1978,

yielded a linear relationship. The correlation coeffi-

cient (r = 0,8329) obtained was highly significant

(PS? 0,001) and higher than that (r = 0,8089) ob-

tained for the regression of estimated Umax values

against temperature only (Fig. 13). The regression

equation obtained for this exponential relationship

was:

U = 0,2574^107/T x drf 5

where U = specific growth rate (estimated maxi-

mum specific growth rate) g fresh mass g1 d1; T =

absolute mean daily air temperature °K; DRF = dif-

fuse component of the radiant flux MJ m2 h1.

Substituting Equation 5 for Equation 3 in the

Monod model, specific growth rates were repre-

1/ (MEAN AIR TEMP. °K. x DIFFUSE RADIANT FLUX) x 104

FIG. 14. — An Arrhenius plot of specific growth rates (estimated Umax values) of marginal forms (Loge) against the products of the

reciprocals of absolute mean daily air temperatures and diffuse radiant fluxes at the maturation pond 3 site.

Bothalia 15, 3 & 4 (1985)

719

SUMMER WINTER SUMMER

FIG. 15. — Specific growth rates predicted for marginal forms from 2, total P and 3, SRP concentrations in the water, over each

growing interval, at the Botanic Gardens Lake site compared with 1, measured specific growth rates. Standard deviations of

measured specific growth rates are shown by bars. Umax values used in the Monod model derived under field conditions and

expressed as a function of air temperature and diffuse radiant flux.

dieted for marginal forms at the BGL site from the

limiting P (SRP and total P) concentrations in the

water, mean daily air temperatures and diffuse ra-

diant fluxes.

Predicted specific growth rates and those meas-

ured for marginal forms are illustrated in Fig. 15.

In general, specific growth rates were more accu-

rately predicted from the limiting total P concentra-

tions in the water at the BGL site using Equation 5

in the Monod model than Equation 3. In the former

example, 15 of the 22 specific growth rates predicted

from total P concentrations in the water (ca 68%)

fell within the standard deviations of measured

growth rates (Fig. 15), compared with 14 (ca 64%) in

the latter (Fig. 13). However, the coefficient of va-

riation (R2) calculated between measured and pre-

dicted values was considerably higher in the former

example (R2 = 0,7973) than in the latter (R2 =

0,5321). Furthermore, the differences between

measured and predicted values were also generally

much smaller in the former example than in the lat-

ter (Table 3). Similar results were obtained at sites

where N was estimated to be the limiting nutrient

(Musil, 1982). In both examples, specific growth

rates were more accurately predicted from total N or

total P concentrations than from other N or P frac-

tions in the water.

The largest differences between measured specific

growth rates and those predicted from total P con-

centrations in the water at the BGL site (Fig. 15)

occurred during midsummer (November to Feb-

ruary) when relative humidities were highest (Fig.

3). This suggested that if the effects of relative hu-

midity, in addition to air temperature plus radiant

flux density, were incorporated into the model, it

might further improve its accuracy of prediction.

Correcting Umax for relative humidity

An Arrhenius plot of specific growth rates (esti-

mated Umax values) of marginal forms, expressed as

natural logarithms (Loge), against the products of

the reciprocals of absolute mean daily air tempera-

tures, diffuse radiant fluxes and mean daily relative

humidities at the MP3 site (Fig. 16), over the period

February to December, 1978, yielded a linear re-

lationship. The correlation coeffient (r = 0,8567) ob-

tained was highly significant (P3= 0,001). It was

higher than that obtained for the regression of esti-

mated Umax values against the products of temper-

ature and diffuse radiant flux (r = 0,8329) or

temperature only (r = 0,8089). The regression equa-

tion obtained for this exponential relationship was:

U = 0 2247 e-7000^ x drf x rh

6

720

Bothalia 15, 3 & 4 (19851

TABLE 3. — A comparison of differences between measured

specific growth rates and those predicted from limiting total

P concentrations in the water at the Botanic Gardens Lake

site, where predicted values were calculated using Umax

values, in the Monod model, expressed as a function of:

A. air temperature,

B. air temperature and diffuse radiant flux

* smallest difference

where U = specific growth rate (estimated maxi-

mum specific growth rate) g fresh mass g-1 d-1; T =

absolute mean daily air temperature °K; DRF = dif-

fuse component of the radiant flux MJ m 2 h1; RH =

mean daily relative humidity.

Substituting Equation 6 for Equation 5 in the

Monod model, specific growth rates were repre-

dicted for marginal forms at the BGL site from the

limiting P (SRP and total P) concentrations in the

water, mean daily air temperatures, diffuse radiant

fluxes and mean daily relative humidities. Predicted

specific growth rates and those measured for margi-

nal forms are illustrated in Fig. 17.

In general, specific growth rates were predicted

with similar accuracy from the limiting total P con-

centrations in the water at the BGL site using Equa-

tion 6 in the Monod model compared with Equation

5. In the former example, 17 of the 22 specific

growth rates predicted from total P concentrations in

the water (ca 77%) fell within the standard devia-

tions of measured specific growth rates (Fig. 17),

compared with 15 (ca 68%) in the latter (Fig. 15).

However, the coefficients of variation (R2), and the

differences calculated between measured and pre-

dicted values (Table 4), were not very much differ-

ent in the former example (R2 = 0,7870) compared

with the latter (R2 = 0,7973). Similar results were

obtained at sites where N was estimated to be the

limiting nutrient (Musil, 1982). In both examples,

specific growth rates were more accurately predicted

from total N or total P concentrations, than from

other N or P fractions, in the water.

1 | ( MEAN AIR TEMP °K * DIFFUSE RADIANT FLUX x MEAN RELATIVE HUMIDITY) x 106

FIG. 16. — An Arrhenius plot of specific growth rates (estimated Umax values) of marginal forms (Log,,) against the products of the

reciprocals of absolute mean daily air temperatures, diffuse radiant fluxes and mean daily relative humidities at the maturation

pond 3 site.

Bothalia 15, 3 & 4 (1985)

721

FIG. 17. — Specific growth rates predicted for marginal forms from 2, total P and 3, SRP concentrations in the water, over each

growing interval, at the Botanic Gardens Lake site compared with 1, measured specific growth rates. Standard deviations of

measured specific growth rates are shown by bars. Umax values used in the Monod model derived under field conditions and

expressed as a function of air temperature, diffuse radiant flux and relative humidity.

The largest differences between measured specific

growth rates and those predicted from total P con-

centrations in the water at the BGL site (Fig. 17),

however, still occurred during midsummer (Novem-

ber to February). It appeared unlikely that the water

pH significantly influenced E. crassipes growth rate

at this site or at the MP3 site from where the Umax

values of marginal forms were derived. The re-

corded variation in the water pH at these two sites

(Fig. 7) being considerably smaller than the varia-

tion in pH of ca 1,2 pH units or greater, in the range

pH 3,0 to 8,2 shown by Chadwick & Obeid (1966) as

significantly influencing E. crassipes growth in cul-

ture. In addition, the water pH values at both sites

were in close proximity to pH 7,0 at which maximum

growth of E. crassipes occurs (Chadwick & Obeid,

1966). In view of this, two possible reasons are given

to explain why predicted specific growth rates were

significantly lower than those measured during mid-

summer:

(i) That the Ks concentrations derived for E. cras-

sipes under culture conditions of N and P limitation

(Musil & Breen, 1985) were temperature depen-

dent. In Chlorella pyrenoidosa and Oscillatoria

agardhii, Shelef et al. (1970) and Ahlgren (1978)

have shown that Ks varies with temperature. Simi-

larly, in Aerobacter aerogenes and Escherichia coli,

Topiwala & Sinclair (1971) and Sawada et al. (1978)

have also demonstrated that Ks changes with tem-

perature and an Arrhenius plot of the change is lin-

ear. Lower Ks concentrations in the water for the

specific limiting nutrient during midsummer, when

temperatures were highest, would have brought the

predicted specific growth rates closer to those meas-

ured. Consequently, it may be possible to describe

more accurately the effect of a limiting nutrient on

E. crassipes growth rate by expressing Ks as a func-

tion of temperature.

(ii) That the specific growth rates (estimated

Umax values) measured for marginal forms at the

MP3 site, particularly during midsummer, may have

been depressed by some toxic factor in the water.

High phytoplankton population densities in the

water at the MP3 site, particularly during midsum-

mer, may have resulted in the excretion by phyto-

plankton of some toxic factor in sufficiently high

concentrations to be inhibitory to E. crassipes

growth rate. Algae in some instances can severely

interfere with the growth of higher plants (Hewitt,

1966), possibly through the production of antibiotic

substances (Jorgensen, 1962) or toxic amino and car-

boxylic acids (Steinberg, 1947a; 1947b; Woltz &

Jackson, 1961; Woltz, 1963). Furthermore, the pres-

ence of phenolic acids, common by-products of de-

composition, in the secondary treated waste-water

effluent at the MP3 site may also have inhibited E.

crassipes growth rate. Glass (1973, 1974), for ex-

ample, has shown that phenolic acids severely inhibit

722

Bothalia 15, 3 & 4 (1985)

TABLE 4. — A comparison of differences between measured

specific growth rates and those predicted from limiting total

P concentrations in the water at the Botanic Gardens Lake

site, where predicted values were calculated using Umax

values, in the Monod model, expressed as a function of:

A. air temperature and diffuse radiant flux,

B. air temperature, diffuse radiant flux and relative humid

ity

* smallest difference

the uptake of P and K by barley. Consequently,

Umax values derived for E. crassipes at the MP3

site, under conditions of growth rate inhibition by

some toxic factor in the water, when incorporated

into the Monod model would have underestimated

the specific growth rates of plants at the BGL site.

This is what was observed during midsummer (Fig.

17).

Finally, it should be pointed out that at the BGL

site relative humidities were not substantially differ-

ent from the MP3 site, at which the Umax values of

E. crassipes were derived. However, at sites where

relative humidities were substantially different from

the MP3 site, Musil (1982) showed that specific

growth rates of marginal forms were generally more

accurately predicted from the limiting nutrient con-

centrations in the water using Equation 6 in the

Monod model than Equation 5.

CONCLUSIONS

Specific growth rates of E. crassipes growing in

loosely crowded field populations were adequately

predicted from the limiting total N or total P concen-

trations in the water using culture-derived half satu-

ration coefficients (Ks) and field-derived maximum

specific growth rates (Umax), expressed as a func-

tion of air temperature, diffuse radiant flux and rela-

tive humidity, in the Monod model. The possible

temperature dependency of Ks requires further in-

vestigation. This, however, will require precise

measurements of Ks, under conditions of N or P

limitation, at different temperatures in some type of

continuous flow culture system, as suggested by Mu-

sil & Breen (1985). The relationship between Ks and

temperature will then need to be mathematically for-

mulated and incorporated into the model. Whether

such refinements will improve the model’s accuracy

of prediction needs verification.

ACKNOWLEDGEMENTS

We wish to thank Mr A. Zakwe and Mrs J. Schaap

for technical services rendered; the City Engineer,

Durban for providing facilities at the northern sew-

age treatment works; Prof. D.F. Toerien, Drs P.J.

Ashton and M.C. Rutherford for their valuable

comments and criticisms and Mrs S.S. Brink for typ-

ing the manuscript.

UITTREKStL

Kinetiese koeffisiente afgelei vir Eichhornia cras-

sipes (Mart.) Solms onder kultuurtoestande met N en

P as beperkende faktore is in die Monodmodel ge-

bruik om die beperkende voedingselement te identifi-

seer en om die spesifieke groeitempo onder varie-

rende voedingselementkonsentrasies en lugtempera-

ture te voorspel. Voorspellings is korrek bewys deur

vergelyking met gemete spesifieke groeitempo’ s van

plantpopulasies wat of los of dig saamgepak by twee

proefpersele groei. Die gebruik van kultuurafgeleide

maksimum spesifieke groeitempo' s (‘Umax’) in die

model, het tot onakkurate voorspellings van plante-

groeitempo’s in los- en digsaamgepakte veldpopula-

sies gelei. Die gebruik van ‘Umax’e’ afgelei van veld-

toestande in die model, het egter voldoende voorspel-

lings van plantegroeitempo’s in losgepakte veldpopu-

lasies gegee. Die insluiting van stralingstroomdigtheid

(diffuse komponent van die stralingstroom) en rela-

tiewe vogtigheid in die model, het die akkuraatheid

van voorspelling aansienlik verbeter. In alle gevalle,

is spesifieke groeitempo’ s meer akkuraat van die be-

perkende totale N of totale P konsentrasies voorspel,

as van ander N of P fraksies, in die water.

REFERENCES

AHLGREN, G., 1978. Growth of Oscillatoria agardhii in che-

mostat culture. 2. Dependence of growth constant on tem-

perature. Mitt. int. Verein. theor. angew. Limnol. 21:

88-102.

AMERICAN PUBLIC HEALTH ASSOCIATION, AMERI-

CAN WATER WORKS ASSOCIATION AND WATER

POLLUTION CONTROL FEDERATION, 1975. Stan-

dard methods for the examination of water and waste-water ,

14th edn, pp. 1193. New York.

BEEVERS, L. & COOPER, J.P., 1964. Influence of tempera-

ture on growth and metabolism of ryegrass seedlings. I.

Seedling growth and yield components. Crop Sci. 4:

139-143.

BELL, C.J., 1981. The testing and validation of models. In D. A.

Rose & D.A. Charles-Edwards, Mathematics and plant

physiology , 229-309. London and New York: Academic

Press.

BERMAN, T., 1969. Phosphatase release of inorganic phospho-

rus in Lake Kinneret. Nature, Lond. 224: 1231-1232.

BERMAN, T., 1970. Alkaline phosphatase and phosphorus

availability in Lake Kinneret. Limnol. Oceanogr. 15:

663-674.

BEST, E.P.H., 1980. Effects of nitrogen on the growth and nitro-

genous compounds of Ceratophyllum demersum. Aquat.

Bot. 8,2 : 197-206.

Bothalia 15, 3 & 4 (1985)

723

BILLINGS, W.D., 1974. Arctic and alpine vegetation : plant ad-

aptations to cold summer climates. In J.D. Ives & R.G.

Barry, Arctic and alpine environments, 403-443. London:

Methuen.

BLACKMAN, F.F., 1905. Optima and limiting factors. Ann.

Bot. 19: 281-295.

BOCK, J.H., 1966. An ecological study of Eichhornia crassipes

with special emphasis on its reproductive biology. Ph.D.

thesis. University of California, Berkeley, 186 pp.

BOCK, J.H., 1969. Productivity of the water hyacinth (Eichhor-

nia crassipes). Ecology 50: 460-464.

BOYD, C.E., 1970. Vascular aquatic plants for mineral nutrient

removal from polluted waters. Econ. Bot. 24: 95-103.

BOYD, C.E., 1976. Accumulation of dry matter, nitrogen and

phosphorus by cultivated water hyacinths. Econ. Bot. 30:

51-56.

BRUHL, P. & GUPTA, J.S., 1927. Commentationes phytomor-

phologicae. IV. Eichhornia studies. III. On the production

of ripe seeds by artificial pollination of Eichhornia speci-

osa. J. Dep. Sci. Calcutta Univ. 8: 1-8.

CENTER, T.D. & SPENCER. N.R., 1981. The phenology and

growth of water hyacinth (Eichhornia crassipes (Mart.)

Solms) in a eutrophic north-central Florida lake. Aquat.

Bot. 10,1 : 1-32.

CHADWICK, M.J. & OBEID, M., 1966. A comparitive study of

the growth of Eichhornia crassipes (Mart.) Solms and Pis-

tia stratiotes L. in water culture. J. Ecol. 54: 563-575.

CHEN, C.W. & ORLOB, G.T., 1972. Ecological simulation of

aquatic environments. Final report of Office for Water Re-

sources Research (OWRR C-2044), U.S. Department of

Interior, Washington, 156 pp.

CRAFTS, A.S., CURRIER, H.B. & STOCKING, C.R., 1949.

Water in the physiology of plants. Waltham, Massachu-

setts: Chronica Botanica.

DENT, J.B. & BLACKIE, M.J., 1979. Systems simulation in

agriculture. Barking, USA: Applied Science Publishers.

DITORO, D.M., O’CONNER. D.J. & THOMANN, R.V.,

1971. A dynamic model of the phytoplankton populations

in the Sacremento-San Joaquim Delta. Adv. Chem. Ser.

106: 131-180.

DROOP, M.R., 1974. The nutrient status of algal cells in continu-

ous culture. J. mar. biol. Ass. U.K. 54: 825-855.

ENVIRONMENTAL PROTECTION AGENCY, 1974. Manual

of methods for chemical analysis of water and wastes. Re-

port no. EPA-625/6-74— 003. Washington, D.C.: U.S. En-

vironmental Protection Agency Office of Technological

Transfer.

FITTER, A.H. & HAY, R.K.M., 1981. Environmental physiol-

ogy of plants. London and New York: Academic Press.

EPPLEY, R.W., 1972. Temperature and phytoplankton growth

in the sea. Fish Bull. Calif. 70: 1063-1085.

FREIDEL, J., KOCH, W. & PHILIPP, O., 1978. Studies of the

biology and population dynamics of Eichhornia crassipes

(Mart.) Solms in the Sudan. Proceedings of the European

Weeds Research Society 5th Symp. on Aquatic Weeds.

GALLAGHER, J.N. & BISCOE, P.V., 1979. Field studies of

cereal leaf growth. III. J. exp. Bot. 117: 645-655.

GLASS, A. D M., 1973. Influence of phenolic acids on ion up-

take. I. Inhibition of phosphate uptake. PI. Physiol. 51:

1037-1041.

GLASS, A.D.M., 1974. Influence of phenolic acids on ion up-

take. III. Inhibition of potassium absorption. J. exp. Bot.

25: 1104-1113.

GOLDMAN, J.C., 1972. The kinetics of inorganic carbon limited

growth of green algae in continuous culture. Its relationship

to eutrophication. Ph.D. thesis. University of California,

Berkeley.

GOLDMAN, J.C. & CARPENTER, E.J., 1974. A kinetic ap-

proach to the effect of temperature on algal growth. Lim-

nol. Oceanogr. 19,5: 756-766.

GOLTERMAN, H.L., CLYMO, R.S. & OHNSTAD, M.A.M.,

1969. Methods for chemical analysis of fresh waters. 2nd

edn Oxford, Edinburgh, London, Melbourne: Blackwell

Scientific Publications.

GRAN, H.H., 1912. Pelagic plant life. In J. Murray & J. Hjort,

Depth of the ocean , 307-386. Weinhen: Cramer.

GREACEN, E.L. £ OH, J.S., 1972. Physics of root growth.

Nature, Lond. (New Biol.) 235: 24-25.

HARRIS, G.P. & LOTT, J.N. A., 1973. Light intensity and

photosynthetic rates in phytoplankton. J. Fish. Res. Bd

Can. 30: 1771-1778.

HAYES, F.R. & PHILLIPS, J.E., 1968. Lake water and sedi-

ment. IV. Radio-phosphorus equilibrium with mud, plants

and bacteria under oxidised and reduced conditions. Lim-

nol. Oceanogr. 3: 459 — 475.

HEPHER, B., 1958. On the dynamics of phosphorus added to

fish ponds in Israel. Limnol. Oceanogr. 3: 84-100.

HEWITT, E.J., 1966. Sand and water culture methods used in the

study of plant nutrition. Farnham Royal, Bucks, England:

Commonwealth Agricultural Bureau.

HSIAO, T.C., ACEVEDO, E., FERERES, E. & HENDER-

SON, D.W., 1976. Water stress, growth and osmotic ad-

justment. Phil. Trans. R. Soc. B273: 479-500.

JANSSON, M., 1976. Phosphatases in lake water: characteriza-

tion of enzymes from phytoplankton and zooplankton by

gel filtration. Science 194: 320-321.

JESCHKE, W.D. & SIMONIS, W., 1965. Uber die Aufnahme

von Phosphatund, Sulphationem durch Blatter von Elodea

densa und inhre Beeinflussung durch Licht, Temperatur

und Aussen-konzentration. Planta 67: 6-32.

JORGENSEN, E.G., 1962. Antibiotic substances from cells and

culture solutions in unicellular algae with special reference

to some chlorophyll derivatives. Physiologia PI. 15: 530.

JOY, K.W., 1969. Nitrogen metabolism of Lemna minor. I.

Growth, nitrogen sources and amino acid inhibition. PI.

Physiol. 44: 845.

KIEFER, D. A. & ENNS, T., 1976. A steady-state mode! of light-

temperature- and carbon-limited growth of phytoplankton.

In R.P. Canale, Modelling biochemical processes in aquatic

ecosystems, 319-336. Michigan: Ann Arbor Sci.

KNIPLING, E.B., WEST, S.H. & HALLER, W.T., 1970.

Growth characteristics, yield potential and nutrient con-

tent of water hyacinths. Soil Crop Sci Soc. Fla 30: 51-63.

LA GARDE, R.V., 1930. A plant that stopped navigation. Bull.

Mo. bot. Gdn 18: 48-51.

LANSDELL, K.A., 1925. The water hyacinth, Eichhornia speci-

osa Kunth. J. Dep. Agric. Un. S. Afr. 10: 24-28.

MADDUX, W.S. & JONES, R.F., 1964. Some interactions of

temperature, light intensity and nutrient concentrations

during the continuous culture of Nitzschia closterium and

Tetraselmis sp. Limnol. Oceanogr. 9: 79-86.

MALEK, I. & FENCL, Z., 1966. Theoretical and methodological

basis of continuous culture of micro-organisms. Prague:

Czechoslovak Academy of Sciences.

McLEAN, K., 1922. Water hyacinth. A serious pest in Bengal.

Agric. J. India 17: 23-40.

MEIDNER. H. & SHERIFF, D.W., 1976. Water and plants.

Glasgow: Blackie.

MITSCH, W.J., 1977. Water hyacinth (Eichhornia crassipes), nu-

trient uptake and metabolism in a north central Florida

marsh. Arch. Hydrobiol. 81: 188-210.

MORRIS, T.L., 1974. Water hyacinth Eichhornia crassipes

(Mart.) Solms: its ability to invade aquatic ecosystems of

Payne's Prairie Preserve. M.S. thesis, University of Flor-

ida, Gainesville, FL., 135 pp.

MULLIGAN, H.F., BARANOWSKY, A. & JOHNSON, R.,

1976. Nitrogen and phosphorus fertilization of aquatic vas-

cular plants and algae in replicated ponds. I. Initial re-

sponse to fertilization. Hydrobiologia 48: 109-116.

MUSIL, C.F. , 1982. The use of growth kinetics in the development

of a predictive model for the growth of Eichhornia crassipes

(Mart.) Solms in the field. Ph.D. thesis, University of Na-

tal.

MUSIL, C.F. & BREEN, C.M., 1977. The application of growth

kinetics to the control of Eichhornia crassipes (Mart.)

Solms through nutrient removal by mechanical harvesting.

Hydrobiologia 53,2: 165-171.

MUSIL, C.F. & BREEN, C.M., 1985. The development from

kinetic coefficients of a predictive model for the growth of

Eichhornia crassipes in the field. I. Generating kinetic co-

efficients for the model in greenhouse culture. Bothalia 15:

689-703.

O’BRIEN, W.J., 1972. Limiting factors in phytoplankton algae:

their meaning and measurement. Science 178: 616-617.

OLSEN, S., 1967. Recent trends in the determination of phos-

phate in the aquatic environment. In H.L. Golterman and

R.S. Clymo, Chemical environment in the aquatic habitat,

63-105. Proceedings of an IBP symposium, 10-16 October,

1966. Amsterdam: Royal Netherlands Academy of Sci-

ences.

724

Bothalia 15, 3 & 4 (1985)

OREBAMJO, T.O. & STEWART, G.R., 1975a. Ammonium re-

pression of nitrate reductase formation in Lemna minor L.

Plania 122: 27-36.

OREBAMJO, T.O. & STEWART, G.R., 1975b. Ammonium

inactivation of nitrate reductase in Lemna minor L. Planta

122: 37 — 44.

PARK, R.A., SCAVIA, D. & CLESCERI, N.L., 1975.

CLEANER: the Lake George Model. In C.S. Russel,

Ecological modelling in a resource management frame-

work, 49-82. Washington, D.C: Resources for the Future,

Inc.

PISEK, A., LARCHER, W., VEGIS, A. & NAPP-ZIN, K.,

1973. The normal temperature range. In H. Precht, J.

Christopherson, H. Hensel & W. Larcher, Temperature

and life , 102-194. Berlin: Springer-Verlag.

POMEROY, L.R., SMITH. E.E. & GRANT, C.M., 1965. The

exchange of phosphate between estuarine water and sedi-

ments. Limnol. Oceanogr. 10: 167-172.

RADFORD, P.J., 1967. Growth analysis formulae — their use

and abuse. Crop Sci. 7: 171-175.

RAO, P.S.J., 1920. The formation of leaf bladders in Eichhornia

speciosa Kunth (water hyacinth). J. Indian Bot. 1:

219-225.

RHEE, G-Y., 1978. Effects of N: P atomic ratios and nitrate limi-

tation on algal growth, cell composition and nitrate up-

take. Limnol. Oceanogr. 23,1: 10-25.

RIGLER, F.H., 1964. The phosphorus fractions and turnover

time of phosphorus in different types of lakes. Limnol.

Oceanogr. 9: 511-518.

RODHE, W., 1948. Environmental requirements of fresh-water

plankton algae. Symb. bot. upsal. 10: 1-149.

RODHE, W., 1978. Algae in culture and nature. Mitt. int. Ver-

ein. theor. angew. Limnol. 21: 7-20.

ROSS, J., 1975. Radiant transfer in plant communities. In J.L.

Monteith, Vegetation and the atmosphere. Vol. 1. Prin-

ciples, 13-55. London, New York, San Francisco: Aca-

demic Press.

SAW ADA, T., CHOHJI, T. & KUNO, S., 1978. Kinetic analysis

of unbalanced bacterial growth in temperature shift. In

V.W. Weekman, Jr & D. Luss, Chemical reaction engin-

eering, 163-172. Washington D.C.: American Chemical

Society.

SCHULZE, B.R., 1965. Climate of South Africa. Part 8. General

Survey. The Government Printer and Weather Bureau,

RSA, 330 pp.

SCHULTHORPE, C.O., 1967. The biology of aquatic vascular

plants. London: Arnold.

SEAMAN, D.E. & PORTERFIELD, W.A., 1964. Control of

aquatic weeds by the snail Marisa cornaurietis. Weeds 12:

87-92.

SHELEF, G., OSWALD. W.J. & GOLUEKE, C.G., 1970. As-

saying algal growth with respect to nitrate concentration by

a continuous flow turbidostat. Adv. Water Pollut. Res.

2:1-9. Oxford: Pergamon.

SLAYTER, R.O., 1967. Plant - water relationships. London and

New York: Academic Press.

SMAYDA, T.J., 1969. Experimental observations on the influ-

ence of temperature, light and salinity on cell division of

the marine diatom Detonula confervaceae (Cleve) Gran. J.

Phycol. 5: 150-157.

SOROKIN, C., 1960. Kinetic studies on temperature effects on

the cellular level. Biochim. biophys. Acta. 38: 197-204.

SOROKIN, C. & KRAUSS, R. W. , 1962. Effects on temperature

and illuminance on Chlorella growth uncoupled from cell

division. PI. Physiol. 37: Sl-M.

STEINBERG, R.A., 1947a. Growth responses to organic com-

pounds by tobacco seedlings in aseptic culture. J. agric.

Res. 75: 81.

STEINBERG, R.A., 1947b. Growth responses of tobacco seed-

lings in aseptic culture to diffusates of some common soil

bacteria. J. agric. Res. 75: 199.

SZEICZ, G., 1974. Solar radiation in crop canopies. J. appl.

Ecol. 11: 1117-1156.

TOERIEN, D.F., 1972. Nutrient removal as a means of control of

Eichhornia crassipes. CSIR, National Institute for Water

Research, Pretoria. Report No. W3/24/1/16.

TOETZ, D.W. , 1971. Diurnal uptake of NOj and NH4 by a Cera-

tophyllum community. Limnol. Oceanogr. 16: 819-822.

TOETZ, D.W., 1973. The kinetics of NH4 uptake by Ceratophyl-

lum. Hydrobiologia 41: 275-290.

TOPIWALA, H. & SINCLAIR, C.G., 1971. Temperature re-

lationship in continuous culture. Biotechnol. Bioeng.

Symp. 13: 795-813.

TWINCH, A. A. & BREEN, C.M., 1980. Advances in under-

standing phosphorus cycling in inland waters — their signifi-

cance for South African limnology. South African National

Scientific Programmes, Report No. 42.

VON SCHWOERBEL, J. & TILLMANS, G.C., 1972. Ammo-

nium-adaptation bei submersen phanerogamen in situ.

Arch. Hydrobiol. Suppl. 42: 139-141.

WEBER, H., 1950. Morphologische and anatomische Studien

iiber Eichhornia crassipes (Mart.) Solms. Abh. ath. Kl.

Akad. Mss. Mainz 6: 135-161.

WETZEL, R.G., 1969a. Excretion of dissolved organic com-

pounds by aquatic macrophytes. Bioscience 19: 539-540.

WETZEL, R.G., 1969b. Factors influencing photosynthesis and

excretion of dissolved organic matter by aquatic macro-

phytes in hard-water lakes. Verb. int. Verein. theor. angew.

Limnol. 17: 72-85.

WOLTZ, S.S., 1963. Growth modifying and anti-metabolite ef-

fects of amino acids on chrysanthemum. Plant Physiol.

38:93.

WOLTZ, S.S. & JACKSON, C.R., 1961. Production of yellow

strap leaf of chrysanthemum and similar disorders by

amino acid treatment. Plant Physiol. 36:197.

WYNNE, D. & GOPHEN, M., 1981. Phosphatase activity in

freshwater zooplankton. Oikos 37: 369-376.

YOUNT, J.L. & CROSSMAN, R.A. Jr, 1970. Eutrophication

control by plant harvesting. J. Wat. Pollut. Control Fed.

42: 173-183.

Bothalia 15, 3 & 4: 725-731 (1985)

The development from kinetic coefficients of a predictive model for the

growth of Eichhornia crassipes in the field. III. Testing a model for

predicting growth rates from plant nutrient concentrations

C. F. MUSIL * and C. M. BREEN**

Keywords: Eichhornia crassipes, model validation, plant nutrient concentrations

ABSTRACT

Kinetic coefficients derived for Eichhornia crassipes (Mart.) Solms under culture conditions of N and P limita-

tion and in the field were used in the Droop model to identify the limiting nutrient and to predict specific growth

rates under conditions of varying plant nutrient concentration, air temperature, diffuse radiant flux and relative

humidity. Predicted data were validated by comparison with specific growth rates measured for plants growing in

loosely crowded populations at one field site and compared with specific growth rates predicted from limiting

nutrient concentrations in the water at the same site using the Monod model. The results show that specific growth

rates were adequately predicted from the Droop model, but generally not as accurately predicted from this model

as from the Monod model.

INTRODUCTION

In developing a mathematical model for predict-

ing growth of Eichhornia crassipes (Mart.) Solms

(water hyacinth) under varying conditions of nutri-

ent loading and climate, in eutrophied aquatic sys-

tems (Toerien, 1972; Musil & Breen, 1977a; Musil &

Breen, 1985a), it is desirable to establish whether

growth rates are predicted with greater accuracy

from plant or water nutrient concentrations. Musil &

Breen (1985b) showed that specific growth rates of

E. crassipes in loosely crowded field populations

were reasonably accurately predicted from tlimiting

nutrient concentrations in the water using kinetic co-

efficients derived under culture and field conditions

in the Monod model.

Several researchers (Dymond, 1948; Boyd, 1969;

Boyd 1976) have shown that the chemical composi-

tion of E. crassipes is influenced by levels of nutri-

ents in the water, though attempts to correlate plant

nutrient concentrations with those in the water en-

vironment have yielded conflicting results (Boyd &

Vickers, 1971; Gosset & Norris, 1971). In algal cells,

however, consistent trends have been found be-

tween decreases in intracellular levels of limiting nu-

trients and increasing nutrient deficiency (Caperon,

1968; Droop, 1968; Fuhs, 1969; Davies, 1970; Cape-

ron & Meyer, 1972; Paasche, 1973). This has led

many researchers to conclude that cellular nutrient

concentrations, rather than external nutrient con-

centrations, control the growth rate. An internal nu-

trient reservoir which contains the raw material for

growth has been proposed, the size of which deter-

mines the rate of growth by analogy to enzyme-sub-

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag XI01, Pretoria 0001.

** Botany Department, University of Natal, P.O. Box 375,

Pietermaritzburg 3200.

t Nutrients present at concentrations below those required for

maximum plant growth and hence restricting the growth rate.

strate reactions (Caperon, 1968; Williams, 1971; Ca-

peron & Meyer, 1972). Laboratory investigations

have shown that growth rates of algal populations

depend on cellular levels of limiting nutrients. Mack-

ereth (1953), for example, observed that the growth

rate of the fresh water diatom Asterionella formosa

is related to its intracellular P level, whereas Cape-

ron (1968) showed that the growth rate of the marine

alga Isochrysis galbana could be hyperbolically re-

lated to the inferred cellular nitrate pool size.

The relationship between the specific growth rate

(U) and the cellular limiting nutrient concentration

can be described by the following hyperbolic equa-

tion (Muller, 1970):

U = Umax ^ — 32

Kq + (q - qo)

where qo = minimum intracellular concentration of

limiting nutrient or subsistence quota; q = measured

intracellular concentration of limiting nutrient;

Umax = maximum specific growth rate; Kq - half

saturation coefficient.

Droop (1968) and Rhee (1973) showed that the

half saturation (Kq) concentration for a specific limi-

ting nutrient in algal cells is equal to, or not signifi-

cantly different from, the minimum intracellular

concentration of the limiting nutrient. If Kq = qo or

q = 2qo when U = 0,5 Umax, the above equation

assumes the following simplified hyperbolic form

(Droop, 1968; Rhee, 1973):

U = Umax - — —

3

If it is postulated that the Kq concentration for a

specific limiting nutrient in E. crassipes is also equal

to the minimum concentration of the limiting nutri-

ent in plants, then this simplified equation (Droop

model) can be used to predict specific growth rates

of E. crassipes from the limiting nutrient (N or P)

concentrations in plants. The minimum limiting con-

726

Bothalia 15, 3 & 4 (1985)

centrations of N and P in E. crassipes can be derived

from the yield coefficient (Yc) values (dry mass ba-

sis) for these nutrients, when expressed as a recipro-

cal and a percentage (1/Yc x 100) (Musil & Breen,

1985a). This investigation was designed to test the

validity of the Droop model for predicting growth

rates of E. crassipes in the field from limiting nutri-

ent concentrations in plants. The accuracy with

which these growth rates were predicted were com-

pared with those predicted from limiting nutrient

concentrations in the water using the Monod model

(Musil & Breen, 1985b).

LOCALITIES AND METHODS

Sites

Two sites, namely the Maturation Pond 3 (MP3)

and Botanic Gardens Lake (BGL) sites character-

ized by different nutrient concentrations in the water

(Musil & Breen, 1985b), were selected for the field

estimations of specific growth rates. These rates

were measured periodically throughout the year and

compared with those predicted by the use of the

Droop model.

Measurement of growth

Specific growth rates were measured by tagging

plants and introducing them for 12 to 14-day growing

intervals into populations enclosed at each field site

(Musil & Breen, 1985b). These rates were measured

in both loosely and densely crowded field popula-

tions, since the two different growth forms, viz mar-

ginal and central forms (Musil, 1982), associated

with these two different population densities (Rao,

1920; McClean, 1922; Lansdell, 1925; Bruhl &

Gupta, 1927; La Garde, 1930; Weber, 1950) exhibit

different net carbon uptakes and photosynthesis/res-

piration ratios (Mitsch, 1977) which suggests that

they may have different growth rates under similar

field conditions.

Chemical analyses

Nitrogen and P concentrations were analysed in

plants harvested after each growing interval at each

site. Analyses were performed on samples compris-

ing both juvenile and mature plants (parents and ve-

getatively propagated offsets), since nutrient con-

centrations in E. crassipes often decline as plants age

(Boyd, 1969). Whole plants of the marginal and cen-

tral forms were analysed separately, since some nu-

trient concentrations in E. crassipes differ signifi-

cantly between the two growth forms and various

plant parts (Musil & Breen, 1977b). Ten plants, to-

gether with their respective offsets, were sampled at

random from each batch of marginal and central

forms harvested, after each growing interval, from

loosely and densely crowded populations enclosed at

each site. Marginal and central forms were rinsed

separately through three changes of deionised-dis-

tilled water and dried in a forced draft oven at 60°C.

The dry tissues were separated into three batches,

ground in a mill and redried at 60°C in the forced

draft oven to a constant weight.

Nitrogen and P concentrations were determined in

each batch of dry, ground plant tissues using pub-

lished methods (Association of Official Agricultural

Chemists, 1975). Nitrogen was determined by the

micro Kjeldahl method and P was determined colo-

rimetrically using ammonium molybdate after diges-

tion of the samples with nitric and perchloric acids

(Johnson & Ulrich, 1959).

Physical analyses

Hourly measurements of the diffuse component of

the radiant flux and maximum, minimum and mean

daily air temperatures and relative humidities were

obtained, over each growing interval, from the

nearby meteorological station (Musil & Breen,

1985b).

RESULTS AND DISCUSSION

Field data

The results of the physicochemical analyses of the

environment and measurements of specific growth

rates of plants of the marginal and central forms,

growing in loosely and densely crowded field popu-

lations respectively, at the two sites are presented in

Musil & Breen (1985b).

Nitrogen and P concentrations in marginal and

central forms at the two sites are shown in Figs 1 and

2. Nitrogen and P concentrations in marginal forms

at the MP3 site were significantly higher (P=S 0,001)

than in those at the BGL site (Table 1). They reflec-

ted the significantly higher specific growth rates

measured for marginal forms at this site and the

higher total N and total P concentrations in the

water (Musil & Breen, 1985b). Nitrogen and P con-

centrations in central forms at the MP3 site were sig-

nificantly lower (P^ 0,05) than in those of marginal

forms (Table 2). The significantly different N and P

concentrations determined in marginal and central

forms, however, could not be explained in terms of

the N and P concentrations in the water, since there

were no differences in these beneath loosely and

densely crowded populations enclosed at the MP3

site (Musil, 1982). These differences probably reflect

some aspect of the physiology of plants of different

growth form found under different degrees of

crowding. In fact, Mitsch (1977) found that large

water hyacinths (central forms) have a lower photo-

synthesis/respiration ratio than dwarf water hya-

cinths (marginal forms). She concluded that large

water hyacinths must put more of their captured

energy into the maintenance of their structures and

consequently would not be as efficient nutrient traps

per unit amount of plant mass as dwarf plants.

Testing the model

Identifying the limiting nutrient

At each site, the plant nutrient limiting E. cras-

sipes growth rate was estimated from the average N

and P concentrations in plants using the minimum

limiting concentrations (average values) of 1,10% N

and 0,11% P in E. crassipes (Musil & Breen, 1985a)

in the Droop model. It can be assumed (Musil &

Breen, 1985b) that the specific growth rate of E.

crassipes is limited not in a multiplicative or additive

manner, but in a threshold mode by the single nutri-

Bothalia 15, 3 & 4 (1985)

727

FIG. 1. — Nitrogen concentrations (means of 3 batches) in E. crassipes harvested after each growing interval at two field sites. Solid

line = marginal forms growing in loosely crowded populations. Broken line = central forms growing in densely crowded popu-

lations. No plants of the central form were produced during June, July and August.

FIG. 2. — Phosphorus concentrations (means of 3 batches) in E. crassipes harvested after each growing interval at two field sites.

Solid line = marginal forms growing in loosely crowded populations. Broken line = central forms growing in densely crowded

populations. No plants of the central form were produced during June, July and August.

ent (N or P) in shorter supply. For example, the av-

erage N and P concentrations determined in margi-

nal forms at the BGL site during 1978 were 3,34% N

and 0,51% P (Musil, 1982). The percentage of the

maximum specific growth rate (Umax) that marginal

forms would achieve at (i) the average N and (ii) the

average P concentrations in plants at this site were

estimated using the Droop model as follows:

it it 3,34-1,10 „

U = Umax x 100 (1)

3,34

= 67,1% Umax

U = Umax O’H x 100 (ii)

0,51

= 78,4% Umax

The results show that at the BGL site marginal

forms would achieve a lower percentage of the

Umax at the average N than at the average P con-

centrations in plants which indicates that N was the

limiting nutrient.

At the MP3 site, on the other hand, the average N

and P concentrations determined in marginal forms

during 1978 were 4,72% N and 1,19% P (Musil,

1982). Using the Droop model, it was estimated that

marginal forms would achieve 76,7% and 90,8% of

the Umax at the average N and P concentrations in

plants respectively at this site, indicating that N was

also the limiting nutrient. Similar results were ob-

tained from estimates based on the average N and P

concentrations in central forms at this site (Musil,

1982).

TABLE 1. — A statistical comparison of N and P concentrations

in marginal forms at the Botanic Gardens Lake (BGL) and

Maturation Pond 3 (MP3) sites (means over the period

1/2/78 to 6/12/78)

Bothalia 15, 3 & 4 (1985)

TABLE 2. — A statistical comparison of N and P concentrations

in marginal and central forms at the Maturation Pond 3

site (means over the period 3/1 1/77 to 6/12/78, but ex-

cluding the winter period 7/6/78 to 15/9/78 when no

plants of tiie central form were produced)

The estimated percentages of the Umax (76,7%

for N; 90,8% for P) that marginal forms would

achieve at the average N and P concentrations in

plants at the MP3 site, however, were lower than

those (95,5% for N; 98,6%for P) estimated from av-

erage total N and total P concentrations in the water

(Musil & Breen. 1985b). This may be explained by

the theoretical maximum specific growth rate

(U'max), at an infinite intracellular concentration of

the limiting nutrient, being always larger than the

predicted or observed maximum specific growth rate

(Umax). In real systems, the intracellular concen-

tration of the limiting nutrient (q) has a finite maxi-

mum value (qm) and cannot be infinite. The differ-

ence between Umax and U'rnax is determined by

the relationship between the storage capacity of cells

for the limiting nutrient (qm) and the minimum in-

tracellular concentration of the limiting nutrient (qo)

or the qo/qm ratio (Droop. 1973: 1974; Goldman &

McCarthy. 1978). For purposes of testing the model,

it was assumed, from estimates based on average

total N and total P concentrations in the water at the

MP3 site (Musil & Breen, 1985b), that specific

growth rates measured for marginal and central

forms, at specific temperatures, diffuse radiant

fluxes and relative humidities, at this site closely ap-

proximated their respective Umax values.

Comparison of predicted and measured specific

growth rates

Since no measurements of specific growth rates of

central forms were obtained at the BGL site (Musil

& Breen, 1985b), no comparisons could be made be-

tween these and those predicted by the use of the

Droop model.

The regression equation exponentially relating

specific growth rates (estimated Umax values) of

marginal forms at the MP3 site to the products of the

reciprocals of absolute mean daily air temperatures,

diffuse radiant fluxes and mean daily relative humid-

ities (Musil &. Breen, 1985b) was:

y = () ,2247e -7<'oot x drf x rh

where U = specific growth rate (estimated maxi-

mum specific growth rate) g fresh mass -> d 1 ; T =

absolute mean daily air temperature °K; DRF = dif-

fuse component of the radiant flux MJ irr- h1: RP1

= mean daily relative humidity.

Incorporating this equation and the minimum

limiting concentration of 1,10% N in E. crassipes

into the Droop model, specific growth rates were

predicted for marginal forms, over each growing in-

terval at the BGL site, from the limiting N concen-

trations in plants, mean daily air temperatures, dif-

fuse radiant fluxes and mean daily relative humid-

ities. These were then compared with measured

specific growth rates. For example, the N concentra-

tion determined in marginal forms and the mean

daily air temperature, diffuse radiant flux and mean

daily relative humidity recorded at the BGL site

over the growing interval 1/2 to 15/2/78 were 2,33%

N, 24,9°C. 0,6127 MJ m : h-' and 83,0% respectively

(Musil. 1982). The specific growth rate (U) was pre-

dicted for marginal forms for this set of conditions as

follows:

U

-700(1 (273.2 + 24.0) x 0.(3 127 x X3.0

0. 2247c

x

2,33 - 1.10

2.33

= 0,0747 g fresh mass g-> cF1

Predicted specific growth rates and those meas-

ured for marginal forms growing in loosely crowded

populations at the BGL site are illustrated

in Fig. 3.

Specific growth rates predicted from N concentra-

tions ih plants at the BGL site closely followed the

same seasonal pattern as measured specific growth

rates. Of the 22 predicted values, 13 (ca 59%) fell

within the standard deviations of measured values.

The coefficient of variation (Musil & Breen. 1985b)

calculated between measured and predicted specific

growth rates had a value of R: = 0,6866. Therefore,

the observed affinity between measured and pre-

dicted values could be regarded as significant over

the entire growing season.

The largest differences between measured and

predicted specific growth rates (Fig. 3) occurred dur-

ing midsummer (November to February) and mid-

winter (June and July) when the highest and lowest

air temperatures and diffuse radiant fluxes respec-

tively were recorded (Musil & Breen, 1985b). Two

possible reasons are given to explain the poor affin-

ity between predicted and measured values during

these periods:

(i) that the minimum limiting concentrations or

subsistence quotas (qo) of N and P in E. crassipes

were dependent on temperature and radiant flux

density. Rhee & Gotham ( 1981a. 1981b) have shown

that the subsistence quotas of N and P in algal cells

increase with decreasing temperature and irradi-

ance, indicating that at lower temperatures and irra-

diancc levels algal cells require more of the limiting

nutrient to grow at the same rate as they do at an

optimal temperature and irradiance. Goldman

(1979) also reported an increase in qo at low tem-

peratures in P limited Monochrysis lutherii, whereas

Bothalia 15, 3 & 4 (1985)

729

FIG. 3. — Specific growth rates predicted for marginal forms from 2, N concentrations in plants, over each growing interval at the

Botanic Gardens Lake site, compared with 1, measured specific growth rates. Standard deviations of measured specific growth

rates are shown by bars. Umax values used in the Droop model derived under field conditions and expressed as a function of air

temperature, diffuse radiant flux and relative humidity.

Davis (1976) noticed that the intracellular concen-

trations of limiting nutrients in the marine diatom

Skeletonema costatum were greater at suboptimal

light levels than those at light saturated growth. The

greater requirement for nutrients at lower tempera-

tures may reflect the cell’s need for more RNA to

synthesize the same amount of protein, as suggested

by Tempest & Hunter (1965). In fact, Chohji et al.

(1976) have shown that the efficiency of protein syn-

thesis, measured as the rate of protein synthesis per

unit weight of RNA, decreases at lower tempera-

tures. Higher minimum limiting concentrations of N

and P in marginal forms during midwinter, when

temperatures and diffuse radiant fluxes were at their

lowest levels (Musil & Breen, 1985b), would have

brought the predicted specific growth rates closer to

those measured. Similarly, lower minimum limiting

concentrations of N and P in marginal forms during

midsummer, when temperatures and diffuse radiant

fluxes were at their highest levels (Musil & Breen,

1985b), would also have brought the predicted speci-

fic growth rates closer to those measured. Conse-

quently, it may be possible to describe more accu-

rately the effects of limiting plant nutrients on E.

crassipes growth rate by expressing the minimum

limiting nutrient concentrations in E. crassipes as

functions of temperature and radiant flux density.

(ii) that the specific growth rates (estimated Umax

values) measured for marginal forms at the MP3

site, particularly during midsummer, may have been

depressed by some toxic factor in the water, as sug-

gested by Musil & Breen (1985b). This may also

partly explain why the predicted specific growth

rates were significantly lower than those measured

during midsummer.

Comparing specific growth rates predicted by the

Droop model with those predicted by the Monod

model (Musil & Breen, 1985b) showed that specific

growth rates of marginal forms at the BGL site were

generally not as accurately predicted from the limi-

ting nutrient concentrations in plants as from those

in the water. In the former example, only ca 59% of

the predicted values fell within the standard devia-

tions of measured values compared with ca 77% in

the latter. The coefficient of variation (Musil &

Breen, 1985b) calculated between predicted and

measured specific growth rates was also much lower

in the former example (R2=0,6866) than in the latter

(R2=0,7870). Furthermore, the differences between

predicted and measured values were generally much

larger in the former example than in the latter

(Table 3). Similar results were obtained at other

field sites (Musil, 1982).

CONCLUSIONS

Specific growth rates of E. crassipes growing in

loosely crowded field populations were adequately

730

Bothalia 15, 3 & 4 (1985)

TABLE 3. — A comparison of differences between measured

specific growth rates and those predicted from:

A. limiting nutrient (total P) concentrations in the water,

B. limiting nutrient (N) concentrations in plants,

at the Botanic Gardens Lake site. In both examples, pre-

dicted values were calculated using Umax values, in either

the Monod or Droop models, expressed as a function of

air temperature, diffuse radiant flux and relative humidity

Growing Differences

interval g fresh mass g'U'l

dates A B

* smallest difference

predicted from the limiting plant nutrient concentra-

tions using the minimum limiting nutrient concentra-

tions in E. crassipes, derived from yield coefficient

(Yc) values, and field-derived maximum specific

growth rates (Umax), expressed as a function of air

temperature, diffuse radiant flux and relative humid-

ity, in the Droop model. The possible dependency of

the minimum limiting nutrient concentrations in E.

crassipes on temperature and radiant flux density re-

quires further investigation. This, however, will re-

quire measurements of the variation in the Yc of E.

crassipes with changes in temperature and radiant

flux density, under culture conditions of N or P limi-

tation, with the combined effect of these factors on

the Yc being mathematically formulated. Whether

such refinements will improve the model’s accuracy

of prediction requires verification.

Specific growth rates were generally not as accu-

rately predicted from the limiting nutrient concen-

trations in plants using the Droop model as from

those in the water using the Monod model. How-

ever, on a seasonal (summer or winter) basis, it

should be possible to adequately predict E. crassipes

specific growth rate using the Droop model in situa-

tions where N and P concentrations in the water are

not known. Furthermore, if the relationships be-

tween the N and P concentrations in plants and those

in the water environment can be reliably established

and mathematically formulated, it will then be possi-

ble to predict the latter from the former, or vice

versa, and consequently the specific growth rates of

E. crassipes using both models. This should enhance

the usefulness of these models in designing an effec-

tive harvesting strategy for water hyacinths to con-

trol both nutrient inputs and excessive growth in eu-

trophied aquatic systems.

ACKNOWLEDGEMENTS

We wish to thank Mr A. Zakwe and Mrs J. Schaap

for technical services rendered; the City Engineer,

Durban for providing facilities at the northern sew-

age treatment works; Prof. D. F. Toerien and Drs P.

J. Ashton and M. C. Rutherford for their valuable

comments and criticisms and Mrs S. S. Brink for typ-

ing the manuscript.

UITTREKSEL

Kinetiese koeffisiente wat vir Eichhornia crassipes

(Mart.) Solms onder kultuurtoestan.de van N en P be-

perking en in die veld afgelei is, is in the Droopmodel

gebruik om die beperkende voedingstof te bepaal en

om die spesifieke groeitempo's onder toestande van

varierende plantvoedingstofkonsentrasie, lugtempe-

ratuur, diffuse stralingstroom en relatiewe vogtigheid

te voorspel. Voorspellings is gestaaf deur vergelyking

met spesifieke groeitempo’s gemeet vir plante wat in

losgepakte populasies by een veldpunt gegroei het en

deur vergelyking met spesifieke groeitempo's wat

voorspel is vanaf die beperkende voedingstof konsen-

trasies in die water op dieselfde plek, deur gebruik

van die Monodmodel. Die resultate toon dat spesi-

fieke groeitempo's voldoende met die Droopmodel

voorspel is, maar oor die algemeen minder akkuraat

met hierdie model as met die Monodmodel.

REFERENCES

ASSOCIATION OF OFFICIAL AGRICULTURAL CHEM-

ISTS. 1975. Methods of analysis , 12th edn, 957 pp. Washing-

ton. D C.: Assoc. Offic. Agric. Chem.

BOYD, C. E., 1969. The nutritive value of three species of water

weeds. Econ. Bot. 23: 123-127.

BOYD, C. E., 1976. Accumulation of dry matter, nitrogen and

phosphorus bv cultivated water hyacinths. Econ. Bot. 30:

51-56.

BOYD. C. E. & VICKERS, D. G., 1971. Variation in the ele-

mental content of Eichhornia crassipes. Hydrobiologia 38:

409-414.

BRUHL, P. & GUPTA, J. S., 1927. Commentationes phytomor-

phologicae. IV. Eichhornia studies. 111. On the production of

ripe seeds by artificial pollination of Eichhornia speciosa. J.

Dep. Sci. Calcutta Univ. 8: 1-8.

CAPERON, J., 1968. Population growth response of Isochrysis

galbana to nitrate variations at limiting concentrations. Ecol-

ogy 49: 866-876.

CAPERON, J. & MEYER J., 1972. Nitrogen limited growth of

marine phytoplankton. 1. Changes in populations character-

istics with steady-state growth rate. Deep Sea Res. 19:

601-618.

CHOH.IL T.. SAW ADA. T. & KUNO.S., 1976. Macromolecule

synthesis in Escherichia coli BB under various growth condi-

tions. Appl. Environ. Microbiol. 31: 864—869.

DAVIS, C. O., 1976. Continuous culture of marine diatoms

under silicate limitation. 2. Effects of light intensity on

growth and nutrient uptake of Skeletonema costatum. J.

Phycol. 12: 291-300.

DAVIES, A. G., 1970, Iron, chelation and the growth of marine

phytoplankton. 1. Growth kinetics and chlorophyll produc-

tion in cultures of the euryhaline flagellate Dunaliella tertio-

lecta under iron limiting conditions. ./. mar. biol. Ass. U.K.

50: 65-86.

DROOP, M. R., 1968. Vitamin B-12 and marine ecology. 4. The

kinetics of uptake, growth and inhibition in Monochrysis lu-

theri. J. mar. biol. Ass. U.K. 48: 689-733.

Bothalia 15, 3 & 4 (1985)

731

DROOP, M. R., 1973. Some thoughts on nutrient limitation in

algae. J. Phycol. 9: 264—272.

DROOP, M. R., 1974. The nutrient status of algal cells in con-

tinuous culture. J. mar. biol. Ass. U.K. 54: 825-855.

DYMOND, G. C., 1948. The water hyacinth — a Cinderella of

the plant world. In J. P. J. van Vuren, Soil fertility and sew-

age. pp.236. New York: Dover Publications.

FUHS, G. W., 1969. Phosphorus content and rate of growth in

the diatoms Cyclotella nana and Thalassiosira fluviatilis. J.

Phycol. 5: 312-321.

GOLDMAN, J. C., 1979. Temperature effects on steady-state

growth, phosphorus uptake, and the chemical composition

of a marine phytoplankter. Microb. Ecol. 5: 153-166.

GOLDMAN, J. C. & MCCARTHY, J. J., 1978. Steady state

growth and ammonium uptake of a fast-growing marine di-

atom. Limnol. Oceanogr. 23: 695-703.

GOSSET, D. R. & NORRIS, W. E., 1971. Relationship between

nutrient availability and content of nitrogen and phosphorus

in tissues of the aquatic macrophyte Eichhornia crassipes

(Mart.) Solms. Hvdrobiologia 38, 1: 15-28.

JOHNSON, C. M. & ULRICH, A., 1959. II. Analytical methods

for use in plant tissue analysis. Bull. Calif, agric. Exp. Stn.

766: 52-53.

LA GARDE, R. V., 1930. A plant that stopped navigation. Bull.

Mo. bot. Gdn 18: 48-51.

LANSDELL, K. A., 1925. The water hyacinth, Eichhornia speci-

osa Kunth. J. Dep. Agric. Un. S.Afr. 10: 24-28.

MACKERETH, F. J., 1953. Phosphorus utilization by Asterio-

nella formosa. J. exp. Bot. 4: 296-313.

McLEAN, K., 1922. Water hyacinth. A serious pest in Bengal.

Agric. J. India 17: 23-40.

MITSCH, W. J., 1977. Water hyacinth (Eichhornia crassipes),

nutrient uptake and metabolism in a north central Florida

marsh. Arch. Hydrobiol. 81: 188-210.

MUSIL, C. F. , 1982. The use of growth kinetics in the develop-

ment of a predictive model for the growth of Eichhornia cras-

sipes (Mart.) Solms in the field. Ph.D. Thesis, University of

Natal.

MUSIL, C. F. & BREEN, C. M., 1977a. The application of

growth kinetics to the control of Eichhornia crassipes

(Mart.) Solms through nutrient removal by mechanical har-

vesting. Hvdrobiologia 53, 2: 165-171.

MUSIL, C. F. & BREEN, C. M., 1977b. The influence of site

and position in the plant community on the nutrient distribu-

tion in, and content of Eichhornia crassipes (Mart.) Solms.

Hydrobiologia 53, 1: 67-72.

MUSIL, C. F. & BREEN, C. M., 1985a. The development from

kinetic coefficients of a predictive model for the growth of

Eichhornia crassipes in the field. I. Generating kinetic coef-

ficients for the model in greenhouse culture. Bothalia 15:

689-703.

MUSIL, C. F. & BREEN, C. M., 1985b. The development from

kinetic coefficients of a predictive model for the growth of

Eichhornia crassipes in the field. II. Testing and refining the

model under field conditions. Bothalia 15: 705-724.

MULLER, H., 1970. Das Wachstum von Nitzschia actinastroides

(Lemm) v. Gorr un Chemostatem bei limitierender Phos-

phatkonzentration. Ber. dt. bot. Ges. 83: 537-544.

PAASCHE, E., 1973. Silicon and the ecology of marine plankton

diatoms. 1. Thallassiosira pseudonana (Cyclotella nana)

grown in a chemostat with silicate as a limiting nutrient.

Mar. Biol. 19: 117-126.

RAO, P. S. J., 1920. The formation of leaf bladders in Eichhornia

speciosa Kunth (water hyacinth). J. Indian Bot. 1: 219-225.

RHEE, G-Y., 1973. A continuous culture study of phosphate up-

take, growth rate and polyphosphate in Scenedesmus sp. J.

Phycol. 9: 495-506.

RHEE, G-Y. & GOTHAM, I. J., 1981a. The effect of environ-

mental factors on phytoplankton growth: temperature and

the interactions of temperature with nutrient limitation.

Limnol. Oceanogr. 26: 635-648.

RHEE, G-Y. & GOTHAM, I. J., 1981b. The effect of environ-

mental factors on phytoplankton growth: light and the inter-

actions of light with nitrate limitation. Limnol. Oceanogr.

26: 649-659.

TEMPEST, D. W. & HUNTER, J. R., 1965. The influence of

temperature and pH value on the macromolecular composi-

tion of magnesium-limited and glycerol-limited Aerobacter

aerogenes growing in a chemostat. J. gen. Microbiol. 41:

267-273.

TOERIEN, D. F. , 1972. Nutrient removal as a means of control of

Eichhornia crassipes. CSIR, National Institute for Water

Research, Pretoria. Report no. W3/24/1/16.

WILLIAMS, F. M., 1971. Dynamics of microbial populations. In

B. C. Patten, Systems analysis and simulation in ecology 1:

197-267. London, New York: Academic Press.

WEBER, H., 1950. Morphologische and anatomische Studien

iiber Eichhornia crassipes (Mart.) Solms. Abh. math.-na-

turw. Kl. Akad. Wiss. Mainz 6: 135-161.

Bothalia 15, 3 & 4: 733-748 (1985)

The development from kinetic coefficients of a predictive model for the

growth of Eichhornia crassipes in the field. IV. Application of the

model to the Vernon Hooper Dam — a eutrophied South African

impoundment

C. F. MUSIL* and C. M. BREEN**

Keywords: Eichhornia crassipes, model application, harvesting, nutrient removal

ABSTRACT

A model developed for Eichhornia crassipes (Mart.) Solms was used to identify the limiting nutrient in the

Vernon Hooper Dam and to predict population sizes, yields, growth rates and frequencies and amounts of harvest

under varying conditions of nutrient loading and climate. Predicted data were used to evaluate the effectiveness of

harvesting measures currently being employed for controlling both nutrient inputs and the population size in this

impoundment. Predictions of the population size, before harvesting was initiated, generally compared favourably

with that estimated visually. Predictions of the quantities of P that could be removed daily by a 20 ha population

indicate that such a population in the impoundment could reduce P concentrations in the epilimnion during sum-

mer stratification to levels limiting for algae. This may explain the observed reduction in algal concentrations since

the introduction of harvesting. Estimates of the amounts and frequencies of harvest required to contain the pre-

dicted potential yields of a 20 ha population indicate that 100 metric tonnes of fresh water hyacinths harvested daily

from the impoundment, under present conditions of reduced nutrient loading, are adequate during winter, but not

during summer.

INTRODUCTION

Harvesting Eichhornia crassipes (Mart.) Solms

(water hyacinth) growing in eutrophied aquatic sys-

tems may constitute an effective means of removing

nutrients and controlling excessive growth of plants

(Boyd, 1970; Yount & Crossman, 1970). However,

to achieve maximum nutrient removal efficiency by

E. crassipes in a nutrient removal scheme, it is

necessary to establish the size of the population re-

quired to maintain desirable nutrient concentrations

in the water, under varying conditions of nutrient

loading and climate, and the amounts and frequen-

cies of harvest required to control the population

size.

Musil & Breen (1985a, 1985b, 1985c) developed,

tested and refined a model, incorporating culture

and field derived kinetic coefficients, for predicting

growth of water hyacinth in eutrophied aquatic sys-

tems. In this paper we demonstrate how this model

can be used to identify the limiting nutrient and pre-

dict population sizes, yields, growth rates and fre-

quencies and amounts of harvest, under varying con-

ditions of nutrient loading and climate, in the Ver-

non Hooper Dam. Predicted data are used to evalu-

ate the effectiveness of harvesting measures

currently being employed for controlling both nutri-

ent inputs and the population size in this impound-

ment.

STUDY SITE

The Vernon Hooper Dam at Ntshongweni in Na-

tal is located in a deep, rocky gorge below the origi-

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

** Botany Department, University of Natal, P.O. Box 375,

Pietermaritzburg 3200.

nal confluence of the Mlazi, Sterkspruit and Weke-

weke (Ugede) Rivers. A description of the catch-

ment and the physicochemical and hydrological

characteristics of the impoundment are presented in

Archibald & Warwick (1980). For many years this

impoundment has served as an important water sup-

ply for the city of Durban, but its continuation as a

water resource is jeopardized by poor water quality.

Effluent discharges from domestic sewage treatment

works and industrial complexes in the Mlazi and

Sterkspruit River catchments have resulted in a con-

siderably enriched impoundment with little opportu-

nity for effluent diversion. A history of the develop-

ment in the catchment area, which includes a record

of changes in water quality and subsequent action

taken by the Durban City Engineers Department, is

given by Howes (1976).

Before 1979, water hyacinth was observed occa-

sionally in the impoundment. The populations, how-

ever, were generally small and confined to the major

inlets with aggregations of plants occasionally being

wind-blown across the reservoir (C. G. M. Archi-

bald, pers. comm.t). Water quality at this time was

described by Howes (1983) as consisting of three

types, each requiring different chemical treatments.

These were; (i) ‘Normal’, (ii) ‘Algal laden' and (iii)

‘Manganese’ water. The latter results from anaer-

obic conditions during summer when severe stratifi-

cation occurs in the impoundment (Archibald &

Warwick, 1980).

During 1979, water hyacinth spread extensively

and covered 70 to 80% of the reservoir by March,

1980 (Everitt, 1980). Light wind action compressed

t C. G. M. Archibald, National Institute for Water Research,

Council for Scientific and Industrial Research, P.O. Box 17001,

Congella, Durban.

734

Bothalia 15, 3 & 4 (1985)

FIG. 1. — Map of Vernon Hooper Dam showing position of floating boom.

this population to ca 50% of the water surface area.

A marked improvement in the quality of the ‘Algal

laden’ water, noted by the Chemical Branch of the

Durban City Engineers Department, was attributed

to P removal by the water hyacinth population (A.

M. Howes, pers. comm.*). In view of this, the

**Durban City Engineers Department decided

A. M. Howes, Chemical Branch, City Engineers Department,

P.O. Box 680, Durban.

The Vernon Hooper Dam presently falls under the jurisdic-

tion of the Umgcni Water Board, P.O. Box 9, Pietermaritzburg,

3200.

against eradicating the water hyacinth population in

the impoundment by chemical control measures.

Instead, ca 20 ha of water hyacinths were retained

behind a floating boom in the Mlazi leg of the reser-

voir (Fig. 1) at the end of 1981 (Howes, 1983), and

the populations harvested regularly with the aid of a

mobile crane. Up to 100 metric tonnes of fresh water

hyacinths are harvested daily from the impoundment

(P. A. Larkan, pers. comm.1) which has resulted in

reduced water treatment costs (Howes, 1983).

P. A. Larkan, Umgeni Water Board, P.O. Box 9, Pietermaritz-

burg, 3200.

Bothalia 15, 3 & 4 (1985)

735

METHODS

Physicochemical measurements

Flow

Mean daily inflow rates (m3 s-1) for the impound-

ment were derived from flow-measuring weirs lo-

cated on the Mlazi, Sterkspruit and Wekeweke Riv-

ers, whereas outflow rates were derived from the

spillway, scour and dam outlet (draw). These rates,

measured fortnightly, were supplied by the National

Institute for Water Research (NIWR), Council for

Scientific and Industrial Research (CSIR), Durban,

and converted to monthly flows using the following

formula:

Monthly flow (M€) = mean daily flow rate per

month (m3 s-1) x number of days in month x

0,0864 x 103.

Total monthly inflow was calculated as the sum of

monthly flows of the Mlazi, Sterkspruit and Weke-

weke Rivers, whereas total monthly outflow was cal-

culated as the sum of monthly flows of the spillway,

scour and dam outlet.

Chemical analysis of water

Chemical analyses of water samples, collected

fortnightly from the Mlazi, Sterkspruit and Weke-

weke Rivers and from the outflowing water, were

carried out by the NIWR, CSIR, Durban. These

analyses were automated and derived from pub-

lished methods (Environmental Protection Agency,

1974; American Public Health Ass.: Standard Meth-

ods, 1975).

The following N and P fractions were analysed in

the water samples:

a) in filtered samples, nitrate-nitrogen (N03-N)

by colorimetry after reduction to nitrite and soluble

reactive phosphorus (SRP) (Twinch & Breen, 1980)

by colorimetry using the molybdenum blue method.

b) in unfiltered samples, Kjeldahl nitrogen as am-

monium (NH4-N) after digestion of the samples with

cone. H^S04 in the presence of a mercury catalyst

and total phosphorus (total P) as SRP after digestion

of the samples with H2S04 and persulphate. Total

nitrogen (total N) was calculated as the sum of Kjel-

dahl nitrogen and nitrate plus nitrite (NH4-N +

no3-n + N02-N).

Nutrient loading and release

Monthly flows of the Mlazi, Sterkspruit and We-

keweke Rivers were multiplied by the average

monthly total N and total P concentrations in the

water of each river and summed to give monthly

total N and total P inflow loads (inputs) for the im-

poundment. Total monthly outflows via the spill-

way, scour and dam outlet were multiplied by the

average monthly total N and total P concentrations

in the outflowing water to give monthly total N and

total P outflow loads (release).

Environment

Mean daily air temperature data for the Vernon

Hooper Dam, derived from monthly averages over

the period 1932 to 1946 (Climate of South Africa,

1954), were supplied by the Weather Bureau, De-

partment of Transport. No measurements of radiant

flux density or relative humidity were available for

the impoundment.

RESULTS AND DISCUSSION

Retention estimates for the impoundment

Chemical and biological transformations

Estimates of the proportions of N and P inflow

loads retained or lost by chemical and biological

transformations in the impoundment, i.e. through

sediment adsorption and denitrification, were based

on monthly inflow and outflow data and N and P bal-

ances for the impoundment for the period January to

December, 1976 (Table 1). Water hyacinth was

either absent, or present as small populations in the

reservoir during this period (C.G.M. Archibald,

pers. comm.). The differences between total N and

total P inflow and outflow loads, expressed as per-

centages of inflow loads, estimated the percentages

of N and P inflow loads retained or lost by chemical

TABLE 1. — Inflow and outflow data and total N and total P balances for the

Vernon Hooper Dam (monthly averages: January to December, 1976)

A negative value indicates export.

736

Bothalia 15, 3 & 4 (1985)

TABLE 2. — Inflow and outflow data and total N and total P balances for the

Vernon Hooper Dam (monthly averages: August, 1979 to July, 1 980)

* A negative value indicates export.

- data incomplete or unavailable.

and biological transformations in the impoundment

monthly and, although potentially available, were

assumed not to be readily available to water hya-

cinths for growth. A net export of N from the system

was evident during September, 1976. However, this

could not be attributed to a net loss of water from

the reservoir during this month since the recorded

inflow was larger than the recorded outflow (Table

1).

Chemical and biological transformations and water

hyacinth uptake

Monthly inflow and outflow data and N and P bal-

ances for the impoundment for the period August,

1979 to July, 1980, when an extensive cover of water

hyacinth was present (Everitt, 1980), are given in

Table 2. The differences between total N and total P

inflow and outflow loads, expressed as percentages

of inflow loads, estimated the percentages of N and

P inflow loads removed by water hyacinth plus those

retained or lost by chemical and biological trans-

formations in the impoundment monthly. Net ex-

ports of P from the system were evident during Janu-

ary and March, 1980. Again, these could not be

attributed to net losses of water from the reservoir

during these two months since recorded inflows were

larger than recorded outflows (Table 2).

Water hyacinth uptake

Subtracting the percentages of N and P inflow

loads estimated to have been retained, lost or re-

moved monthly in Table 1 from those in Table 2 and

multiplying these by the recorded monthly N and P

inflow loads for the period August, 1979 to July,

1980, gave estimates of the quantities of N and P re-

moved monthly by the water hyacinth population in

the reservoir. The results (Table 3) suggest that dur-

ing the periods March to April, 1980 and November,

1979 to June, 1980, N and P respectively were ex-

ported from the system. The estimated percentages

of N and P inflow loads removed by water hyacinth

plus those retained or lost by chemical and biological

transformations in the reservoir during these periods

being considerably lower than those estimated to

have been retained or lost by chemical and biological

transformations only. These apparent net exports of

N and P from the system, however, could not be

attributed to any large net losses of water from the

reservoir during these periods since recorded in-

flows, with the exception of May and June, 1980,

were generally considerably higher than recorded

outflows (Table 2).

The apparent net exports of P from the system

may be explained by a release of sediment-bound

P04-P during anoxic conditions produced in the im-

poundment by the extensive water hyacinth cover

(Everitt, 1980) and accentuated by summer stratifi-

cation in the reservoir (Archibald & Warwick,

1980). In fact, Everitt (1980) reported a dissolved

oxygen concentration of less than 0,5 mg C-^ppm)

in the water at the dam wall during March, 1980

which indicated anaerobiosis below the thermocline.

This suggestion is supported by the large (up to

1 100%) increase in the total P concentration in the

water during the period November, 1979 to June,

1980, even though P inflow loads during this period

were 25 to 50% lower than those during the prece-

ding period, i.e. August to October, 1979 (Table 3).

An increased P release from sediments under anoxic

conditions is well documented (Mortimer, 1941) and

Vollenweider (1972) has shown that oxygen deple-

tion is accompanied by a breakdown in the ability of

sediments to adsorb P04-P so that sediments act as a

source rather than a sink for P.

The apparent net exports of N from the system

could not be readily explained, since total N concen-

trations in the water relative to N inflow loads did

not show any significant differences during March

and April, 1980 compared with other months (Table

3). These apparent net exports of N from the system,

however, could have been due to increased rates of

N03-N loss via denitrification (Keeny, 1973; Chon

& Knowles, 1979), accentuated by anoxic conditions

produced in the reservoir by the extensive water

hyacinth cover (Everitt, 1980).

Bothalia 15, 3 & 4 (1985)

737

ON ON ,

<N VO 12 '

IS) Q0

c^- c-- rn

O ’t NO NO

“ n M m

^ V l Tf

lo (N ^

m in ^

n rH O 0\

- - VO ON

r: 2 On OO

VO O OO

' r- r-

S ^

^ Tj-

X ~ Tt VO

V£5 cn »— (

VO <N ro *— i

NO 0s- • on

On »0 .

^«OON

NO VO ro Ov

«" VU

O NO VO ^

CXC d

x .2 «

ill

Cd M «

>> O -C

£ Z

U 73

O &

5 C

1-°

73 _ 73

O 5 O VI

z s | ? I

g .5 >,.5 g

§ z z <2

' ' U—i c/3 <-»— < r-

^ o O O S

> l-H

o <L>

II

z .5

O §

•a ?< H -a

o. S

Ml °

X G

T3 .S

_§ ^

% °

C G ,

z 15

« E

p —

.5 o ;

»3j

o rt Xi

* « ©

>, o

'5i o

G O ^

« O o

o !S .£

E -o 2

« C <u

151

°if 1

<o > G

o

<-> C c

>,.£ E

xi £

So §

.5 | o

^ Cu

T) “*

o 3

> l_

O <D

£ aS

<D >

0- .£

03 Oh

E

Q Wh

Q U H

■i E ?

c •a S

. C aj

< D 73

Bothalia 15, 3 & 4 (1985)

738

The possibility of P inputs from internal sources

(sediments) and possible increases in rates of N loss

via denitrification, during those months when net ex-

ports of N and P from the system were not apparent,

introduce constraints to modelling the growth of

water hyacinth in the impoundment. This is because

it is not feasible to accurately estimate the propor-

tions of N and P inflow loads removed by the water

hyacinth population. However, during those months

where net exports of N and P from the system were

not apparent, the model was applied using estimates

made of the proportions of N and P inflow loads re-

moved by the water hyacinth population.

Predictions based on water nutrient concentrations

Limiting nutrient

The nutrient (N or P) limiting water hyacinth

growth rate in the impoundment was predicted from

total N and total P concentrations in the water

(monthly averages) using the half saturation (Ks)

concentrations of 976 fig N and 94,1 fig P fH,

derived for E. crassipes in culture (Musil & Breen,

1985a), in the Monod model. It can be assumed

(Musil & Breen, 1985b) that the specific growth rate

of E. crassipes is limited not in a multiplicative or

additive manner, but in a threshold mode by the

single nutrient (N or P) in shorter supply. Estimates

were based on total N and total P concentrations,

rather than on soluble N and P fractions, in the

water since specific growth rates of E. crassipes in

the field are more accurately predicted from these

concentrations (Musil & Breen, 1985b). For ex-

ample, the average total N and total P concentra-

tions in the water of the impoundment during Au-

gust, 1979 (Table 4) were 999 pg N €_1 and 45 pg P

€-• respectively. The percentages of the maximum

specific growth rate (Umax) that E. crassipes would

achieve at (i) this average total N and (ii) this aver-

age total P concentration in the water were esti-

mated using the Monod model as follows:

U = Umax

999

976 + 999

x 100

= 50,6% Umax

(i)

U = Umax x 100 (ii)

94,1 + 45

= 32,3% Umax

The results show that during August, 1979, E.

crassipes would achieve a lower percentage of the

Umax at the average total P than at the average total

N concentrations in the water, indicating that P was

the limiting nutrient.

Estimates of the limiting nutrient in the impound-

ment for the period August, 1979 to July, 1980, are

given in Table 4. The results show that during July,

1980 and for the period August to October, 1979, P

Nutrient present at concentrations below that required for max-

imum plant growth and hence restricting the growth rate.

was limiting for water hyacinth, whereas for the

period November, 1979 to June, 1980, N was limit-

ing. The change from P to N limitation after Oc-

tober, 1979 was reflected in the considerable in-

crease in the total P concentration in the water

(Table 4 ).

Specific growth rate

Specific growth rates of water hyacinth occurring

in loosely crowded populations in the impoundment

were predicted from mean daily air temperatures

(monthly averages) and limiting total N or total P

concentrations in the water (monthly averages) us-

ing the following mathematical expression (Musil &

Breen, 1985b):

U - 5,2631 x 108e“6540Ar x —

Ks + S

where U = specific growth rate g fresh mass g-1 d_1;

T = absolute mean daily air temperature °K; S =

limiting nutrient concentration pg €-1; Ks = half

saturation concentration pg

This expression, essentially a combination of the

Arrhenius and Monod equations, incorporates the

Ks concentrations of 976 pg N €_1 and 94,1 pg P

€_1. Specific growth rates of water hyacinth occur-

ring in densely crowded populations in the impound-

ment were estimated using a correction factor of

0,2236. This correction factor (Table 5) was derived

from a mean value of ratios calculated between

specific growth rates (estimated maximum specific

growth rates) measured for water hyacinths growing

in loosely and densely crowded field populations at a

nearby sewage maturation pond (Musil, 1982). For

example, during August, 1979 the mean daily air

temperature at the impoundment and the limiting

total P concentration in the water were 16,5°C and 45

pg P fH respectively (Table 4). Specific growth

rates (U) of water hyacinth occurring in (i) loosely

and (ii) densely crowded populations in the im-

poundment during this month were predicted as fol-

lows:

U = 5,2631 x ioV6540/273’2+ 16-5

y 45 ....(j)

94,1 + 45

= 0,0267 g fresh mass g-> d-1 (2,67% d-1)

tt c ^ ln8 -6540/273,2 + 16,5 ^ 45 ...(ii)

U = 5,2631 x 10 c x 94J + 45

= 0,0060 g fresh mass g-1 d-1 (0,60% d-1)

Predictions of specific growth rates of water hya-

cinth in the impoundment for the period August,

1979 to July, 1980 are given in Table 4. These ranged

from 0,0267 to 0,1024 g fresh mass g-1 d-1 (2,67 to

10,24% d_1) and from 0,0060 to 0,0229 g fresh mass

g-1 d-i (0,60 to 2,29% d-1) for loosely and densely

crowded populations respectively.

Predictions based on the proportions of N and P in-

flow loads removed by water hyacinth

Potential yield

The potential yields of water hyacinth in the im-

poundment were predicted from those proportions

of limiting N or P inflow loads estimated to have

TABLE 4. — Predicted growth rates, potential yields and population sizes of water hyacinth in the Vernon Hooper Dam (August, 1979 to July, 1980)

Bothalia 15, 3

& 4 (1985)

739

ON VO

CN O

o o

o' Z

00 ON

On OO On

mo VO ,

O O ^

' *- CN

on 4 o "

CO ON OO lo

On i-h vo

O ©

o' o'

r-H CO

VO CN

© o

o' o'

VO VO OO ^

OO ^ —

cm ^ ^

cn Z cn

ON CN

o ©

o' o'

^ vo ^ vq

m r-" 2 ts

m ^ ^ co

VO CO

oo'

ON ©

co r-

2^0^

£ oo ON ro'

cn r- Tf oo

On

CN CN

O CN

CN ^

oo °°

On CN

O O

o' o'

On <*!

<N 2

O O

CN On'

00 CO

o ©

o' o'

o o

o' o'

VO o

co' vo'

vo c--

r-' i-h

oo

cs00^^

^ Tj- ^

co P- oo'

Tf VO

r-~ vo

On rj-

ON

1 CS VO

'S ^

.S c

£ X

- £ -5 £

g pep

Vh o C CJ

^ CL) o <D

cl o cl

feo X

2 c c

i -2

^ JS

3 3

" CL CL

£ O O

Cjj CL CL

O T3 13

! O O

: ^ ^

£ £

g £

'o *a>

J Q

o cr

£ S

•3 .Kh

2 Zft. 2

cl a

C C 03 C C

O _0 3 O O

3 3 u 3 3

O- O-i ~45 Cl cl

o o

j CL CL

, TJ O

O O

Oh CL

^ 13 13

O U 13

~ =s 3=

42 o o

5 o S

d, E

>, >, o >, >, o

’oj 'o ^ 1u Id ^

O C o o C o

■ " O o

d •— 3 Q a)

O 13 13

£ £

42 o o

— I-H »— I

o o

>s >s

h* 44 E— ' 44 — J -t G-

E E 4

44 44 0- w

undeterminable

data incomplete or unavailable

Bothalia 15, 3 & 4 (1985)

TABLE 5. — Ratios calculated between specific growth rates (estimated maximum

specific growth rates) measured for E. crassipes growing in loosely and densely

crowded field populations in a maturation pond (Musil, 1982)

been removed daily by the water hyacinth popula-

tion using the yield coefficient (Yc) values (fresh

mass basis) of 1 768,5 for N and 17 248 for P derived

for E. crassipes in culture (Musil & Breen, 1985a).

The Yc expresses the relationship between mass of

plant material produced and mass of limiting nutri-

ent absorbed (Musil & Breen, 1985a). For example,

during August, 1979 the proportion of the limiting P

inflow load estimated to have been removed by the

water hyacinth population was 265 kg P (Table 3) or

265/31 kg P d_1. The potential yield (Xpy) of water

hyacinth during this month was predicted as follows:

Xpy = x 17 248

^ 31

= 147,4 metric tonnes of fresh water

hyacinths d-1

Predictions of the potential yields of water hya-

cinth in the impoundment for the period August,

1979 to July, 1980, are given in Table 4. These

ranged from 86,7 to 886,7 metric tonnes of fresh

water hyacinths d-E

Population size

Population sizes (loosely and densely crowded) re-

quired to produce the predicted potential yields of

water hyacinth in the impoundment were estimated

using the following form of the general growth equa-

tion (Malek & Fencl, 1966, Radford, 1967):

Xo + Xpy = Xoeut

where Xo = population size metric tonnes; Xpy =

potential yield metric tonnes d-1; u = specific growth

rate g fresh mass g-> cH; t = time interval between

initial biomass (Xo) and final biomass (Xo + Xpy)

days.

For example, population sizes (loosely and

densely crowded) required to produce the predicted

potential yield of water hyacinth during August,

1979 of 147,4 metric tonnes of fresh water hyacinths

d-1 at a predicted specific growth rate of (i) 0,0267

and (ii) 0,0060 g fresh mass g-1 d-1 for loosely and

densely crowded populations respectively, were esti-

mated as follows:

Xo + 147,4 = Xoe °.0267 x t (i)

Xo (loosely crowded) = 5,4 x 103 metric tonnes

of fresh water hyacinths

Xo + 147,4 = Xoe o.ooeo x t (ii)

Xo (densely crowded) = 24,5 x 103 metric of

fresh water hyacinths

Estimates of population sizes (loosely and densely

crowded) in the impoundment for the period Au-

gust, 1979 to July, 1980 are given in Table 4. These

ranged from 1,0 x 103 to 8,2 x 103 and from 4,6 x

103 to 38,3 x 103 metric tonnes of fresh water hya-

cinths for loosely and densely crowded populations

respectively.

Population area

Assuming stand densities (dry mass basis) of 2,21

and 21,3 metric tonnes ha-1 for loosely and densely

crowded populations respectively (Boyd & Scars-

brook, 1975) and a mean water content of water hya-

cinth of 94,75% (Penfound & Earle, 1948; West-

lake, 1963; Bock, 1969), it is possible to calculate the

areas occupied by the loosely and densely crowded

populations and express these as percentages of the

surface area of the impoundment, which at full sup-

ply level is 84 ha (Everitt, 1980). For example, the

estimated population sizes in the impoundment dur-

ing August, 1979 were 5,4 x 103 and 24,5 x 103 met-

ric tonnes of fresh water hyacinths for loosely and

densely crowded populations respectively. The areas

occupied by the populations in (i) loosely and (ii)

densely crowded situations were, therefore, calcu-

lated as follows:

Bothalia 15, 3 & 4 (1985)

741

— X (

2,21 100

= 128,3 ha (152,7% of the full surface area of the

impoundment)

24,5 x 1Q3^ 5,25

21,3 ' 100

= 60,4 ha (71,9% of the full surface area of the

impoundment)

(ii)

Estimates of the areas occupied by loosely and

densely crowded populations in the impoundment

for the period August, 1979 to July, 1980 are given in

Table 4. These ranged from 23,7 to 194,8 ha (28,2 to

231,9% of the full surface area of the impoundment)

and from 11,3 to 94,4 ha (13,4 to 112,4% of the full

surface area of the impoundment) for loosely and

densely crowded populations respectively.

Everitt (1980) visually estimated a 50% coverage

of the impoundment by water hyacinth during

March, 1980, when the population was compressed

by light wind action. However, the reservoir was

only ca 65% full at the time with the water surface

area covering ca 75% of that at full supply level (A.

M. Howes, pers. comm.). Consequently, the water

hyacinth population in the reservoir under crowded

conditions in actual fact covered only ca 31,5 ha or

37,5% of the full surface area of the impoundment.

This visual estimate of cover compares favourably

with the areas estimated to have been occupied by

densely crowded populations in the reservoir, when

expressed as percentages of the full surface area of

the impoundment (Table 4), during May, 1980

(36,9%), December, 1979 (39,0%), October, 1979

(41,7%) and even during September, 1979 (54,3%)

and June, 1980 (53,3%). It, however, does not com-

pare favourably with those during November, 1979

(13,4%), January, 1980 (112,4%) and February,

1980 (17,0%). Inaccurate estimates of the propor-

tions of N inflow loads removed by the water hya-

cinth population during these months may partly ex-

plain this.

Predictions based on a 20 ha population

Presently, ca 20 ha of water hyacinths are retained

in the impoundment behind a floating boom

(Howes, 1983). The population is maintained in

moderately crowded situations with up to 100 metric

tonnes of fresh water hyacinths being harvested daily

(P. A. Larkan, pers. comm.). The potential yields,

amounts and frequencies of harvest required and

quantities of N and P, i.e. assuming no luxury up-

take of these nutrients by the water hyacinths, that

could be removed by a 20 ha moderately crowded

population in the impoundment were predicted as

follows. Estimates were based on the period August,

1979 to July, 1980, since chemical and hydrological

data for the impoundment after this period was in-

complete. It was assumed that (i) the 20 ha moder-

ately crowded population had an average stand den-

sity of 11,7 metric tonnes ha-1 (dry mass basis), i.e. a

mean value of 2,21 and 21,3 metric tonnes ha-1 (dry

mass basis) for loosely and densely crowded popula-

tions respectively, (ii) the water hyacinths had a

mean water content of 94,75% and (iii) specific

growth rates were mean values of those predicted tor

loosely and densely crowded populations in the im-

poundment in Table 4.

Potential yield

The potential yields of a 20 ha moderately

crowded population, i.e. 20 x 11,7/5,25 x 100 =

4 457,1 metric tonnes of fresh water hyacinths, in the

impoundment were predicted using the general

growth equation (Malek & Fend, 1966; Radford,

1967). For example, the potential yield of 4 457,1

metric tonnes of fresh water hyacinths at an esti-

mated specific growth rate of 0,0163 g fresh mass g-1

d-1, i.e. a mean value of 0,0267 and 0,0060 g fresh

mass gr1 d-1 for loosely and densely crowded popula-

tions respectively (Table 4), during August, 1979

was:

Xpy = 4 457, le o.om x l _ 4 457,1

= 73,2 metric tonnes of fresh water

hyacinths d_1

Predictions of the potential yields of a 20 ha mod-

erately crowded population in the impoundment for

the period August, 1979 to July, 1980 are given in

Table 6. These ranged from 73,2 to 287,9 metric

tonnes of fresh water hyacinths d-T

Harvesting interval

Assuming that 100 metric tonnes of fresh water

hyacinths are harvested daily from the impound-

ment, the time required for a 20 ha moderately

crowded population to produce an additional 100

metric tonnes of fresh water hyacinths was estimated

using the general growth equation (Malek & Fencl,

1966; Radford, 1967) and is referred to as the har-

vesting interval, viz:

t = In Xt-ln Xo

U

where: Xt = final biomass (Xo + 100) metric

tonnes; Xo = initial biomass of 20 ha population

(moderately crowded) metric tonnes; U = specific

growth rate (moderately crowded population) g

fresh mass g-1 d-1; t = harvesting interval, i.e. time

interval between Xo and Xt days; €n = log^ (natu-

ral logarithm).

For example, it was predicted that a 20 ha moder-

ately crowded population, or 4 457,1 metric tonnes

of fresh water hyacinths, would produce 73,2 metric

tonnes of fresh plant material daily during August,

1979 at an estimated specific growth rate of 0,0163 g

fresh mass g-1 d-1 (Table 6). The time (t) required for

this population to produce an additional 100 metric

tonnes of fresh water hyacinths was:

t = In (4 457,1 + 100) - In (4 457,1)

0,0163

= 1,4 days

The harvesting interval in this example was 1,4 days.

After this period, one day’s removal would need to

be initiated and this would have to be repeated after

a further 1,4 days growth.

of a 20 ha population of water hyacinths confined in moderately crowded sitations

Bothalia 15, 3 & 4 (1985)

742

CO CO

VO Tf

ioOn

2

°„ <N 00

7^ vo vo

10 oo oo

CO vo

04 ON

D VO VO

VO OV

I

~ Q

« >

VO Oi

CO VO

— ' <D

x 2

E Ji

c _o

E 3

ft ? 13 TD

13 9>

0.2'“

E—. «

Si 2

z z s

mi2

O G

t/, c

<D 2

2 oZ

> -X

£ x

£ o

§ 5

2 ■= S .

2 z -5

— C

£ o tS

ft. H UJ

c J .2

o. n -n .

a. ^

m

OT3-H

5 s °

3 0-°

CTC

O C ^ Si 73

V ^ .5 (D (D

’ &+ 03

t 3 T3

- u dd a* c3

<3 T3 ^

£ £

- c ^ c 2 .a O

C3 .5 o .5 O D C

f -h I-, qi e

a negative value indicates export; ? undeterminable; - data incomplete or unavailable

Bothalia 15, 3 & 4 (1985)

743

Harvesting intervals estimated for a 20 ha moder-

ately crowded population in the impoundment for

the period August, 1979 to July, 1980 are given in

Table 6. The results indicate that during winter (May

to August) 100 metric tonnes of fresh water hya-

cinths harvested daily would more or less be ade-

quate to account for the predicted potential yields of

the population. The estimated harvesting intervals

ranging from ca 0,8 to 1,4 days. During summer

(September to April), however, this quantity of

fresh water hyacinths harvested daily would gener-

ally be insufficient. Approximately two to three

times as much fresh plant material (ca 161 ,5 to 287,9

metric tonnes) would generally need to be harvested

daily from the impoundment during summer to con-

tain the predicted potential yields of the population .

With respect to the amounts of harvest, it should

be pointed out that the 100 metric tonnes of fresh

water hyacinths apparently harvested daily from the

impoundment (P. A. Larkan. pers. comm.) are high

when compared with those reported in the literature

for various mechanical water hyacinth removal oper-

ations. Van Dyke (1971), for example, reported that

a stationary, land-based mechanical harvester proto-

type (Sarasota Weed and Feed Incorporation) was

only capable of removing an average of ca 5,9 metric

tonnes of fresh water hyacinths per hour of operat-

ing time, i.e. ca 47 metric tonnes of fresh water hya-

cinths per day assuming an 8 hour working day,

when time required for general maintenance and re-

pairs and that lost due to unfavourable weather was

taken into consideration. Similar results were ob-

tained by Phillippy & Perryman (1972b) using an

Aquamarine S-650 Shore Conveyor (Linder Indus-

trial Machine Company, Florida) where an average

of ca 52 metric tonnes of fresh water hyacinths were

removed per 8 hour working day of operating time.

Somewhat higher values, however, have been ob-

tained by Touzeau (1972), using an Aquamarine

H-650 Harvester combined with an S-650 Shore

Conveyor system (Linder Machine Company, Flor-

ida), where an average of ca 74 metric tonnes of

fresh water hyacinths were removed per 8 hour

working day of operating time and by Phillippy &

Perryman (1972a), using a modified, stationary,

land-based mechanical harvester prototype (Sara-

sota Weed and Feed Incorporation), where an aver-

age of ca 96 metric tonnes of fresh water hyacinths

were removed per 8 hour working day of operating

time. The latter was the highest value that could be

traced in the literature.

Nitrogen and phosphorus removal

The quantities of N and P that could be removed

daily by a 20 ha moderately crowded population in

the impoundment were estimated from the predicted

potential yields of the population (Table 6) using the

yield coefficient (Yc) values (fresh mass basis) of

1 768,5 for N and 17 248 for P. These were ex-

pressed as percentages of those proportions of N and

P inflow loads estimated not to have been retained

or lost by chemical and biological transformations in

the system, i.e. those proportions of N and P inflow

loads assumed to be readily available to the water

hyacinths for growth. For example, the predicted

potential yield of a 20 ha moderately crowded popu-

lation in the impoundment during August, 1979 was

73,2 metric tonnes of fresh water hyacinths d-1

(Table 6). The quantities of (i) N and (ii) P that

could be removed daily by the population during this

month were estimated as follows:

73,2 x 1 000

1 768,6

41,4 kg N d-i

(i)

73,2 x 1 000

17 248

4,2 kg P d-1 (ii)

During August, 1979, 20 218 kg N and 1 598 kg P

entered the impoundment (Table 2), or daily inflow

loads of 652,2 kg N d-1 and 51,5 kg P d_1, of which

26,3% with respect to N and 78,9% with respect to P

were estimated to be retained or lost by chemical and

biological transformations in the system (Table 1).

Consequently, the proportions of daily (i) N and (ii)

P inflow loads estimated to have been readily avail-

able to the water hyacinths for growth during Au-

gust, 1979 were:

652,2 x 100~26’3 = 480,7 kg N d-i .... (i)

100

51,5 x

100-78,9

100

= 10,9 kg P d-i (ii)

The predicted quantities of 41,4 kg N d_1 and 4,2 kg P

d_1 that could be removed by a 20 ha moderately

crowded population during August, 1979, expressed

as percentages of the estimated available N and P in-

flow loads of 480,7 kg N d-1 and 10,9 kg P d-1 were:

41,4 x 100 = 8,6% for N (i)

480,7

4,2 x 100 = 38,5% for P (ii)

10,9

Predictions of the quantities of N and P that could

be removed by a 20 ha moderately crowded popula-

tion in the impoundment for the period August,

1979 to July, 1980 are given in Table 6. The results

indicate that such a population could, at least during

most of the above-mentioned period, remove larger

quantities of P daily than those entering the system

that were readily available for water hyacinth

growth. The predicted quantities of P that could be

removed daily by the population, expressed as per-

centages of the estimated available P inflow loads,

ranging from 158,1 to 3 875,0% except during Au-

gust, September and October, 1979 when these

ranged from only 38,5 to 59,1% (Table 6). In con-

trast to P, the results indicate that the population

would generally remove smaller quantities of N daily

than those entering the impoundment that were

readily available for water hyacinth growth. The pre-

dicted quantities of N that could be removed daily by

the population, expressed as percentages of the esti-

mated available N inflow loads, ranging from 8,6 to

82,4% except during March and April, 1980 when

these ranged from 513,5 to 1 348,2% (Table 6).

The above estimates were based on the minimum

quantities of N and P that could be removed daily by

744

Bothalia 15, 3 & 4 (1985)

TABLL 7. A comparison of chemical treatment costs for the different water quality

types in the Vernon Hooper Dam, prior to and following the retention of

ca. 20 ha of water hyacinths behind a floating boom and the introduction of

harvesting, according to Howes (1983)

the population. Luxury uptake of N and P by the

water hyacinths during growth, however, would re-

sult in greater quantities of N and P removed by the

population than actually estimated.

Water quality

A comparison of chemical treatment costs re-

ported by Howes (1983) for the different ‘water

quality types' in the Vernon Hooper Dam, prior to

and after retention of ca 20 ha of water hyacinths

behind a floating boom and the introduction of har-

vesting are presented in Table 7. A reduction of 61%

in chemical treatment costs was achieved initially

through the introduction of harvesting. The cost re-

duction dropped to 45% during the second six

month period of harvesting due primarily to in-

creased nutrient loading and poor rainfall. Cost of

harvesting and disposal of water hyacinths varied be-

tween R600 and R1 000 per day which was initially

justified by savings in chemical treatment costs when

treating in excess of 37 Ml d-b Justification no longer

exists financially. However, the resultant reduction

in algal concentrations, i.e. improvement in quality

of the ‘Algal laden' water, is highly beneficial

(Howes, 1983).

Whether the ca 20 ha of water hyacinths presently

confined in the impoundment could reduce nutrient

concentrations in the water to levels limiting for al-

gae and account for the observed reduction in algal

concentrations is difficult to ascertain. The average

N : P ratio in the water is ca 25,5 (Archibald & War-

wick. 1980) suggesting that P may be the nutrient

most frequently limiting for algae in the impound-

ment. Furthermore, it was predicted that a 20 ha

moderately crowded population in the impound-

ment during the period August, 1979 to July, 1980

could generally remove larger quantities of P daily

than those entering the system that were readily

available for plant growth (Table 6). However, this

does not necessarily mean that the present popula-

tion could reduce P concentrations in the water of

the reservoir to levels limiting for algae. This would

be dependent on a number of factors, viz: (i) rate

and efficiency of P uptake by water hyacinths, (ii)

magnitude of P inflow loads, (iii) residence time of

inflowing water beneath the water hyacinth mat, (iv)

extent of mixing between inflowing water and reser-

voir water and (v) influence of the water hyacinth

population on chemical and biological transforma-

tions in the impoundment. It is clear that a large pro-

portion of the P entering the impoundment is re-

tained by adsorption onto sediments (Hepher, 1958;

Hayes & Phillips, 1968). This source of P is poten-

tially available to plants for growth, since sediment P

and dissolved P exist in equilibrium (Hepher, 1958;

Pomeroy et al., 1965). The equilibrium concentra-

tion increases with increased P content in the sedi-

ment (Pomeroy et al. , 1965). Removal of P from the

water by hyacinths during growth could, therefore,

displace the P equilibrium allowing additional P to

be released from sediments into the overlying water.

In addition, anoxic conditions that might be pro-

duced beneath the water hyacinth mat could also

provide conditions conducive for the release of sedi-

ment P (Morbmer. 1941; Vollenweider, 1972).

Available chemical and hydrological data for the

impoundment after July, 1980 are incomplete. How-

ever, they do indicate that since the end of 1981,

when ca 20 ha of water hyacinths were retained be-

hind a floating boom in the impoundment and har-

vesting was initiated (Howes, 1983), the magnitude

of the monthly P inflow loads have generally not

been very much different from those during 1979

and 1980 (Table 8). Consequently, if one extrapo-

lates from the predicted quantities of P that could be

removed by a 20 ha population, relative to the esti-

mated available inflow loads, for the period August.

1979 to July, 1980, it would appear that the 20 ha of

water hyacinths confined in the impoundment since

the end of 1981 could have removed those propor-

tions of P inflow loads not removed by processes

other than water hyacinth uptake in the system. Fur-

thermore, during summer the reservoir is stratified

and a well defined thermocline develops at a depth

of 6 to 8 m (Archibald & Warwick. 1980). There-

fore, one may speculate that the development of this

thermocline and consequent density gradient in the

impoundment could allow the water hyacinth popu-

lation to reduce P concentrations in the epilimnion

to levels that could be limiting for algae, at least dur-

ing summer when maximum algal growth rate and

production would be expected. Any P released from

sediments into the hypolimnion would theoretically

be restricted from diffusing into the epilimnion by

the thermocline. This may partly explain the ob-

served reduction in algal concentrations in the reser-

voir since the introduction of harvesting.

Bothalia 15, 3 & 4 (1985)

745

TABLE 8. — Inflow and P loading data for the Vernon Hooper Dam

(monthly averages)

- data incomplete or unavailable

Under conditions of increased inflow and P load-

ing, evident from monthly inflow and P loading data

for the impoundment for the period January to De-

cember, 1976 (Table 8), a larger population would,

however, generally be needed in the reservoir to re-

move those proportions of P inflow loads not re-

moved by processes other than water hyacinth up-

take in the system. This is evident from the predicted

specific growth rates, potential yields and N and P

removal potentials of a 20 ha moderately crowded

population in the impoundment for this period

(Table 9). The results indicate that such a population

would, at least during 8 months of the above-men-

tioned period, remove smaller quantities of P daily

than those entering the impoundment that were

readily available for plant growth. The predicted

quantities of P that could be removed daily by the

population, expressed as percentages of the esti-

mated available P inflow loads, ranging from 21,4 to

97,8%, except during March, June, September and

December, 1976 when they ranged from 104,3 to

115,2%. Using the model, the population sizes that

would be required in the impoundment to remove

the estimated available P inflow loads were pre-

dicted. These ranged from 20,6 to 93,4 ha, except

during March, June, September and December,

1976 when they ranged from 17,3 to 19,2 ha only

(Table 9). An example of the derivation is as fol-

lows: during January, 1976 the daily P inflow load

was 77,7 kg P d-1 of which 39,7 kg P d-1 was esti-

mated to have been readily available to plants for

growth (Table 9). The potential yield (Xpy) of water

hyacinth during this month would be:

Xpy = 39,7 x 17 248

= 684,7 metric tonnes of fresh water hya-

cinths d_1

The population size required to produce this poten-

tial yield at a predicted specific growth rate of 0,0449

g fresh mass g-1 d”1 for a moderately crowded popula-

tion (Table 9) would be:

Xo + 684,7 = Xoe 0-0449 x l

= 14 910,7 metric tonnes of fresh

water hyacinths

Assuming an average stand density (dry mass basis)

of 1 1,7 metric tonnes ha-1 for a moderately crowded

population and a mean water content of water hya-

cinth of 94,75%, the area occupied by the popula-

tion would be:

14 910,7

x 5^25

11,7 x 100

= 66,9 ha

CONCLUSIONS

Harvesting water hyacinth growing in eutrophied

aquatic systems directly addresses the problem of

nutrient enrichment of water and not only the ex-

cessive aquatic plant growth which is a manifestation

of the problem. In designing an effective harvesting

strategy for water hyacinth, the model serves as a

useful aid for identifying the limiting nutrient and

predicting population sizes, yields, growth rates and

frequencies and amounts of harvest, under varying

conditions of nutrient loading and climate, to control

both nutrient inputs and excessive growth in eutro-

phied aquatic systems. However, accurate predictive

estimates using the model will require the incorpor-

ation of mathematical expressions from which those

proportions of N and P inflow loads retained or lost

by chemical and biological transformations in such

systems can be predicted. Such mathematical ex-

pressions will also need to integrate the influence of

the water hyacinth population on these transforma-

tions. Furthermore, the relationship between maxi-

mum specific growth rate of water hyacinth and den-

sity of the population will need to be mathematically

formulated, since this presents a potential constraint

to the model’s application. It is clear that the nutri-

ent removal capacity of water hyacinth is a function

of the population size, its density and growth rate.

An inverse relationship exists between the two latter

TABLE 9. — Predicted yields, growth rates and nutrient removal potentials of a 20 ha population of water hyacinths confined in moderately crowded situations

in the Vernon Hooper Dam (January to December, 1976)

746

Bothalia 15,

<N

O _ —

00 * ON “

-r- ro

^ O' £

04 r-

co * »o

r- £2 ^

m

O

P-

OO vo

m *

0-

n ^ 1 in "

rsj vo m

jC <N

'n oo PI

tV oo vo

>''> O g

Tt (N H

' 1 <N

Tf VO

On vo

C>

^ VO

^iv 04

OA ro

oo ro

E -

” o

P

9 n

_ £

T £

60 C

^ O

03 c3

£ 3

£ ~ > "

C 5 £

•S j.5 j

COCO

O E-i fy E-"

CxO .x 6jQ

2 .9 c3

Coqo «

S “ » » 'm

y a, o p. +-

5 05 c <« 3

9 O t3 C

<D <-> <u ox)

03 0- 03 .S

73 S73.§ £

o o vi .5

huhu -j

o a,

4- <D

U a) £

gS'S

^ T3

2 ^

g.2 >-

c 60 « -b

c O'- tfl

2 S c £

M O C

*5 a> -£ .S

^0-90

' d 'd ^ ^

I* a) <D <D >

1 S £ £ "

to ;r

£ 5

v «

£•7 S ,S

^ ^ — c z

Z Z S .2 <~

•« Ml ° M °

o Pd C c c

•SB « “ 5 «=

•S ® f* ■» or

S^° O

T3 —

2 »

■ .y o

5 o .5 o xi C

73 £ 'P

UJ cl H W

i- X3

^ fl. S

°“c c

S ■« -s S

« S ° I

= O ~

•=

■o.E'Su

-2 Oh 91 'SI

'S *o

£ o P S

0- H W

E &-

M- bij a* 13

15 ^ T3 2

¥ x: a> •>

03 It' t-! ^

a. •—

X3 £

3 & 4 (1985)

Bothalia 15, 3 & 4 (1985)

747

factors, i.e. the higher the population density the

lower its specific growth rate (De Busk et al. , 1981).

However, the productivity of waterhyacinth, the

product of specific growth rate and density, defines

a bell-shaped curve with maximum productivities be-

ing achieved at intermediate densities (De Busk et

al., 1981). A regular harvesting programme could

maintain water hyacinth populations confined be-

hind floating booms in large water bodies in moder-

ately crowded situations and intermediate densities,

although the strict control of population density

would probably only be feasible on a small scale.

It would appear from the predictive estimates

made using the model that, under present conditions

of reduced inflow and nutrient loading in the Vernon

Hooper Dam which have persisted since the intro-

duction of harvesting, i.e. from ca December, 1981

to ca August, 1983, the ca 20 ha of water hyacinths

confined in the impoundment has been adequate to

remove those proportions of P inflow loads that are

readily available for plant growth and account for

the observed reduction in algal concentrations.

However, the 100 metric tonnes of fresh water hya-

cinths harvested daily from the impoundment, al-

though adequate during winter, would appear to be

insufficient during summer. It is estimated that

about two to three times as much fresh plant ma-

terial (ca 161 to 288 metric tonnes) would need to be

harvested daily from the impoundment during sum-

mer, under reduced nutrient loadings, to contain the

predicted potential yields of the population. Under

conditions of increased inflow and nutrient loading,

such as those prior to 1979, the population size and

the daily amounts of harvest would have to be in-

creased accordingly. These can be predicted from

the nutrient loading data using the model.

ACKNOWLEDGEMENTS

We wish to thank the National Institute for Water

Research of the Council for Scientific and Industrial

Research, in particular Mr C. G. M. Archibald, for

providing chemical and hydrological data for the

Vernon Hooper Dam; Prof. D. F. Toerien and Drs

P. J. Ashton and M. C. Rutherford for their valu-

able comments and criticisms; Mr A. M. Howes of

the Chemical Branch of the City Engineers Depart-

ment, Durban for permission to publish his results;

Messrs P. A. Larkan and B. N. T. Smith of the Um-

geni Water Board for information supplied and Mrs

S. S. Brink for typing the manuscript.

UITTREKSEL

’n Model wat vir Eichhornia crassipes (Mart.)

Solms ontwikkel is, is gebruik om die beperkende

voedingstof in die Vernon Hooperdam te identifiseer

en om bevolkingsgroottes, opbrengste, groeitempo's

en die frekwensies en hoeveelheid van oeste onder va-

rierende toestande van voedingstoflading en klimaat

te voorspel. Voorspelde data is gebruik om die doel-

treffendheid van oesmaatreels wat huidig vir die be-

heer van beide voedingstofinsette en die bevolkings-

grootte in hierdie dam gebruik word, te evalueer.

Voorspellings van die bevolkingsgrootte voor die oes

begin is, vergelyk oor die algemeen gunstig met die

wat op 'n visuele skatting gebaseer is. Voorspellings

van die hoeveelhede P wat daagliks deur 'n 20 ha be-

volking verwyder kan word, dui aan dat sodanige be-

volking in hierdie dam P konsentrasie in die epilim-

nion gedurende somerstratifikasie tot vlakke kan ver-

minder wat alge beperk. Dit kan die waargenome ver-

mindering in algkonsentrasies sedert die instelling van

oes verduidelik. Skattings van die hoeveelheid en fre-

kwensies van oeste benodig om te verklaar vir die

voorspelde potensiele opbrengste van 'n 20 ha bevol-

king, dui daarop dat die 100 metrieke ton vars water-

hiasinte wat tans daagliks uit die dam geoes word, on-

der huidige toestande van verminderde voedingstof-

lading, gedurende die winter voldoende is, maar nie

gedurende die somer nie.

REFERENCES

AMERICAN PUBLIC HEALTH ASSOCIATION, AMERI-

CAN WATER WORKS ASSOCIATION AND WATER

POLLUTION CONTROL FEDERATION, 1975. Standard

methods for the examination of water and waste-water, 14th

edn, pp. 1193. New York.

ARCHIBALD, C. G. M. & WARWICK, R. J., 1980. Vernon

Hooper Dam. In R. D. Walmsley & M. T. Batty. Limnology

of some selected South African impoundments, 187-194. Pre-

toria: CSIR Press.

BOCK, J. H., 1969. Productivity of the water hyacinth (Eichhor-

nia crassipes). Ecology 50: 460-464.

BOYD, C. E., 1970. Vascular aquatic plants for mineral nutrient

removal from polluted waters. Econ. Bot. 24: 95-103.

BOYD, C. E. & SCARSBROOK. E. , 1975. Influence of nutrient

additions and initial density of plants on production of water

hyacinth, Eichhornia crassipes. Aquat. Bot. 1: 253-261.

CHON, Y-K. & KNOWLES, R., 1979. Measurement of denitrifi-

cation in two freshwater sediments by an in situ acetylene

inhibition method. Appl. Environ. Microbiol. 37,6:

1067-1072.

DE BUSK, T. A., RYTHER. J. H., HANISAK, M. D. & WIL-

LIAMS, L. D., 1981. Effects of seasonality and plant density

on the productivity of some fresh-water macrophytes.

Aquat. Bot. 10,2: 133-142.

ENVIRONMENTAL PROTECTION AGENCY, 1974. Manual

of methods for chemical analysis of water and wastes. Report

No. EPA-625/6-74— 003. Washington, D.C.: U.S. Environ-

mental Protection Agency Office ofTechnologicalTransfer.

EVERITT, I., 1980. Water hyacinth — Shongweni Dam. Dur-

ban: City Engineer's Department.

HAYES. F. R. & PHILLIPS, J. E., 1968. Lake water and sedi-

ment. IV. Radio-phosphorus equilibrium with mud, plants

and bacteria under oxidized and reduced conditions. Lim-

nol. Oceanogr. 3: 459-475.

HEPHER. B.. 1958. On the dynamics of phosphorus added to

fish ponds in Israel. Limnol. Oceanogr. 3: 84-100.

HOWES, A. M., 1976. Report on the Shongweni catchment:

water quality and treatment. Durban: City Engineer’s De-

partment.

HOWES, A. M., 1983. Shongweni water treatment and hyacinth

control. Durban: City Engineer’s Department.

KEENY, D. R., 1973. The nitrogen cycle in sediment water sys-

tems. J. Environ. Qual. 2,1: 15—29.

MALEK, I. & FENCL. Z.. 1966. Theoretical and methodological

basis of continuous culture of micro-organisms. Prague: Cze-

choslovak Academy of Sciences.

MORTIMER, C. H., 1941. The exchange of dissolved substances

between mud and water in lakes. J. Ecol. 29: 280-329.

MUSIL, C. F., 1982. The use of growth kinetics in the develop-

ment of a predictive model for the growth o/Eichhornia cras-

sipes (Mart.) Solms in the field. Ph.D thesis. University of

Natal.

MUSIL, C. F. & BREEN. C. M.. 1985a. The development from

kinetic coefficients of a predictive model for the growth of

Eichhornia crassipes in the field. I. Generating kinetic coef-

ficients for the model in greenhouse culture. Bothalia 15:

689-703.

Bothalia 15, 3 & 4 (1985)

ML'SlL. C. F. & BREEN. C. M.. 1985b. The development from

kinetic coefficients of a predictive model for the growth of

Eichhornia crassipes in the field. II. Testing and refining the

model under field conditions. Bothalia 15: 705-724.

MUSIL. C. F. & BREEN, C. M.. 1985c. The development from

kinetic coefficients of a predictive model for the growth of

Eichhornia crassipes in the field. III. Testing a model for

predicting growth rates from plant nutrient concentrations.

Bothalia 15: 725-731.

PENFOUND. W. T. & EARLE. T. T., 1948. The biology of

water hyacinth. Ecol. Monogr. 18: 448-472.

PHILLIPPY, C. L. & PERRYMAN. J. M., 1972a. Mechanical

harvesting of water hyacinth (Eichhornia crassipes) in Gant

Lake Canal. Sumter County. Florida, pp.21. Florida Game

and Fresh Water Fish Commission. Unpubl.

PHILLIPPY, C. L. & PERRYMAN. J. M., 1972b. Mechanical

harvesting of water hyacinth (Eichhornia crassipes) in Trout

Lake. Lake County, Florida, pp.24. Florida Game and

Fresh Water Fish Commission. Unpubl.

POMEROY. L. R.. SMITH, E. E. & GRANT, C. M.. 1965. The

exchange of phosphate between estuarine water and sedi-

ments. Limnol. Oceanogr. 10: 167-172.

RADFORD. P. J., 1967. Growth analysis formulae — their uSe

and abuse. Crop Sci. 7: 171-175.

'TOUZEAU, L. F., 1972. Mechanical water hyacinth removal op-

erations. Aquamarine Corporation, Blufton, Florida, pp.16.

Florida Game and Fresh Water Fish Commission. Unpubl.

TWINCH. A. .1. & BREEN. C. M., 1980. Advances in under-

standing phosphorus cycling in inland waters — their signifi-

cance for South African limnology. South African National

Scientific Programmes, Report No. 42.

VAN DYKE, J. M., 1971. Mechanical harvesting of water hya-

cinth (Eichhornia crassipes) in Shell Creek Reservoir, Char-

lotte County. Florida, pp.25. Florida Game and Fresh Water

Fish Commission. Unpubl.

VOLLENWEIDER. R. A., 1972. Input-output models with

special reference to the phosphorus loading concept in limno-

logy. Paper given at Workshop Conference on Chemical -

ecological considerations for defining the goal of water pol-

lution control. Kaustanienbaum, Switzerland, April 19-21,

1972.

WEATHER BUREAU, RSA. 1954. Climate of South Africa.

Climate Statistics, Part 1. WB19. Pretoria: The Government

Printer.

WESTLAKE, E. F., 1963. Comparisons of plant productivity.

Biol. Rev. 38: 385-425.

YOUNT. J. L. & CROSSMAN. R. A., Jr. 1970. Eutrophication

control by plant harvesting. J. Wat. Pollut. Control Fed. 42:

173-183.

Bothalia 15, 3 & 4: 749-750 (1985)

Miscellaneous ecological notes

VARIOUS AUTHORS

SURVEY OF EXOTIC WOODY PLANT INVADERS OF THE TRANSVAAL — ADDENDUM

This survey [see Bothalia 15: 297-313 (1984)]

spanned four years from 1979 to 1982. The pilot

study area in the central Transvaal was surveyed

from mid April to mid May 1979. The rest of the

Transvaal was surveyed from March 1980 until Oc-

tober 1982. Table 1 indicates the degree squares that

were sampled and the month and year in which re-

cordings were done.

TABLE 1. — Degree squares sampled in the Transvaal and dates of recordings

Degree squares sampled

oooNor-'-ooosouosor^-oooNO’— < i/y r- oo

UUfhNNlNONNOUlNOnUNlNU

NMNOINOOINOOOOOOONOO

sO sO sO

fN N U

O t/Y sO O'

Ci n N N

(NOUN

Dates of recordings

L. HENDERSON*

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

PHYTOCAP. A FIELD-DATA CAPTURE PROGRAM FOR THE PHYTOTAB PROGRAM PACKAGE

Manual field recording of floristic data, for phyto-

sociological studies, entails re-encoding data on

computer encoding forms, transfer to magnetic tape

and finally loading onto the mainframe computer for

multi-variate analyses. The time taken from re-en-

coding to access on the mainframe computer can be

from one to six weeks. Classification of data while

fieldwork is in progress, therefore, becomes imprac-

ticable in the summer-rainfall areas because field-

work generally takes place in the relatively short

growth period and the delays in computer access

would limit fieldwork considerably if data were clas-

sified during this period. A serious disadvantage of

classifying vegetation only after completion of field-

work is that vegetation units are often either under-

sampled, resulting in invalid syntaxa, or over-

sampled resulting in wasted labour and expense.

Furthermore, the potential for errors is increased by

re-encoding the data.

These problems were overcome by using compu-

terized field-data capture where data is recorded di-

rectly onto a hand-held computer and loaded onto

the mainframe computer on return from the field.

Multivariate analysis techniques used are the PHY-

TOTAB program (Westfall et al ., 1982), which are

compatible with both the DECORANA (Hill,

1979a) and TWINSPAN (Hill, 1979b) programs.

Preliminary classification using either PHYTO20 or

TWINSPAN can then be available within a day of

returning from fieldwork.

The system used for field-data capture is the Sharp

PC 1 500 computer with an additional 8K expansion

memory module, printer/cassette interface, cassette

recorder and programmable RS232C interface. The

program PHYTOCAP, written in BASIC, is used

for recording floristic data in the PHYTOTAB

(Westfall et al., 1982) format. The program features

Bothalia 15, 3 & 4 (1985)

750

include automatic line number allocation, sample

and/or subsample numbers, alphanumeric species

codes, cover-abundance values, data pertinent to in-

dividual species, data display, data printing in two

formats, halting and continuing program execution,

line editing, saving data to tape and loading data

from tape. Furthermore, data input is verified for

errors such as sample number length, species omis-

sion and cover-abundance omission. The user is also

informed when five lines of memory are left. The

capacity of the computer is 70 lines which is approxi-

mately 14 releves or samples with 40 species per

sample, which is generally more than adequate for

the floristic data recorded in one day.

Field procedure includes quadrat location, qua-

drat demarcation, floristic sampling, voucher speci-

men collection and environmental parameter samp-

ling. The Sharp PC 1 500 computer is used for flor-

istic sampling. Species for which voucher specimens

exist are input as a four-letter genus code and a

three-letter species code. Species for which voucher

specimens are required are tagged with pre-num-

bered, specimen number, adhesive address labels

and input as a left-justified specimen number. Speci-

men collection is effected after completion of flor-

istic data recording. This process ensures a smooth

flow of data input and reduces the possibility of

species being overlooked. Environmental par-

ameters are recorded directly on a field data sheet. It

is envisaged that a second Sharp PC 1 500 be used

for recording environmental parameters and as a

standby machine. Memory capacity currently pre-

cludes the use of a single machine for both floristic

and environmental data.

After a day’s recording the computer is attached

to the printer/cassette interface, which remains in

the vehicle. A printout of the floristic data is ob-

tained of each sample for stapling to the field data

sheets which form the hardcopy for eventual perma-

nent safe-keeping at the Botanical Research Insti-

tute, Pretoria. The data are then transferred to C 15

cassettes, when the computer can be cleared for the

following data set.

Loading data to the Burroughs B 7 900 mainframe

is effected by means of a Burroughs ET 1 100 ter-

minal, RS232C interface and a transfer program

called DATATRAN written by S. J. Crafford. Data

are read from the cassette and simultaneously trans-

mitted at 300 Baud.

The advantages of this system of data recording

include the cost-saving production of preliminary

classifications for optimum sampling as well as the

labour- and cost-saving of not having to re-encode

data. The potential for errors is also reduced by the

reduction in data handling. Documentation and

taped copies of the program are available from the

author. Please forward a blank C 15 cassette for

copying.

ACKNOWLEDGEMENTS

The author thanks Dr J. C. Scheepers for com-

ments and suggestions and Mr S. J. Crafford for

writing the program DATATRAN.

REFERENCES

HILL. M. O., 1979a. DECORANA. A FORTRAN program for

detrended correspondence analysis and reciprocal averaging.

Unpublished report. Ecology and Systematics, Cornell Uni-

versity, Ithaca, New York. pp. 1—30.

HILL, M. O., 1979b. TWINSPAN. A FORTRAN program for

arranging multi-variate data in an ordered two-way table by

classification of the individuals and attributes. Unpublished

report. Ecology and Systematics, Cornell University, Ithaca,

New York. pp. 1-48.

WESTFALL, R. H., DEDNAM, G., VAN ROOYEN, N. &

THERON, G. K.. 1982. PHYTOTAB. A program package

for Braun-Blanquet tables. Vegetatio 49: 35-37.

R. H. WESTFALL*

* Botanical Research Institute, Department of Agriculture &

Water Supply, Private Bag X101, Pretoria 0001.

Bothalia 15, 3 & 4: 751-759 (1985)

New taxa, new records and name changes for southern African plants

STAFF OF THE NATIONAL HERBARIUM

PRECIS, the computerized herbarium specimen

database maintained at the Botanical Research Insti-

tute, records currently accepted names, recent syno-

nyms and literature references for all taxa of south-

ern African plants. This information is the basis for

the first and second editions of the List of Species of

Southern African Plants (Gibbs Russell et al . , 1984,

1985 and in preparation). However, these publica-

tions will soon be out of date as revisions and new

taxa are published and as taxa newly recorded for

southern Africa are collected in the field.

This section of Bothalia is initiated in order to re-

port annually the recent new taxa, new records and

name changes for southern African plants so that in

the future it will be possible for users to keep up to

date with the most recent application of plant

names.

The format is that of the second edition of the List

of Species for new taxa and for synonyms. Families

and genera are in the order of Dyer (1975, 1976) and

species are in alphabetical order. A name in current

use appears in capital letters with its PRECIS num-

bers that show its place in the List of Species. Syno-

nyms appear in lower case letters and each synonym

is entered twice, onbe indented below the name for

which it is a synonym and once in its alphabetical

place in the genus. New records are indicated by

quoting a specimen and its locality. Naturalized taxa

are shown with an asterisk following the name.

The contributor responsible for each family is in-

dicated by a superscript number after the family

name:

1 G. Germishuizen; 2 C. Reid; 3 E. Retief; 4 L.

Smook; 5 M. Welman.

G. E. GIBBS RUSSELL

SCHIZAEACEAE2

0000120 — MOHRIA

1. ROUX. 1984. JL S. AFR. BOT. 50: 435.

200 M. HIRSUTA J. P. ROUX

POACEAE 4

2500 E. RUPESTRIS NEES EX TRIN. SUBSP.

RUPESTRIS

2540 E. RUPESTRIS NEES EX TRIN. SUBSP.

TRICOSTATA (STAPF) GIBBS RUS-

SELL (= E. tricostata Stapf) 1

2560 E. SETACEA NEES SUBSP. DISTICHA

GIBBS RUSSELL

2580 E. SETACEA NEES SUBSP. SCABRA

(STAPF) GIBBS RUSSELL

( = E. setacea Nees var. scabra Stapf) 1

2600 E. SETACEA NEES SUBSP. SETACEA

2620 E. SETACEA NEES SUBSP. UNIFLORA

(BURCH. EX STAPF) GIBBS RUSSELL

( = E. uniflora Burch, ex Stapf) 1

E. setacea Nees var. scabra Stapf = E. SETA-

CEA SUBSP. SCABRA

E. tricostata Stapf = E. RUPESTRIS SUBSP.

TRICOSTATA

E. uniflora Burch, ex Stapf = E. SETACEA

SUBSP. UNIFLORA

9902110 — CORTADERIA

1. ROBINSON. 1984. S. AFR. J. BOT.

3:343.

50 C. JUBATA (LEM.) STAPF

RESTIONACEAE 2

0804000 — RESTIO

R. compressus Rottb. = PLATYCAULOS

COMPRESSUS

R. obtusissimus Steud. = NEVILLEA OBTU-

SISSIMUS

0804010 — PLATYCAULOS

1. LINDER. 1984. BOTHALIA 15: 11

500 P. COMPRESSUS (ROTTB.) LINDER

(= Restio compressus Rottb.)

0808000 — LEPTOCARPUS

L. rattrayi Pillans = HYDROPHILUS RAT-

TRAYI

0817010 — NEVILLEA

1. LINDER. 1984. BOTHALIA 15: 11

100 N. OBTUSISSIMUS (STEUD). LINDER

(= Restio obtusissimus Steud.)

0817020 — HYDROPHILUS

1. LINDER. 1984. BOTHALIA 15: 11.

100 H. RATTRAYI (PILLANS) LINDER

(= Leptocarpus rattrayi Pillans)

LILIACEAE 2

096900 — ANDROCYMBIUM

1. MULLER-DOBLIES. 1984. WILLDE-

NOWIA 14: 179

A. CILI.OLATUM SCHLTR. & KRAUSE

(= A. fenestratum Schltr. & Krause) 1

A. CRUCIATUM U. & D. MULLER-DOB-

LIES

A. DREGEI PRESL

(= A. exiguum Roessl.) 1

A. exiguum Roessl. + A. DREGEI

A. fenestratum Schltr. & Krause = A. CILIO-

LATUM

A. HENSSENIANUM U. & D. MULLER-

DOBLIES

A. POELTIANUM U. & D. MULLER-DOB-

LIES

A. VILLOSUM U. & D. MULLER-DOBLIES

A. VOGELII U. & D. MULLER-DOBLIES

9900890 — DIGITARIA

3370 D. RUKWAE CLAYTON East African species

collected in Natal. 2930 (Pietermaritzburg): 700

ChaseValley (-CB), Ellis 4517.

5000 D. VIOLASCENS LING * European species 850

collected in Natal. 2732 (Mtubatuba):

Lake St Lucia, northern end of False Bay 1000

(-CD), Ellis 4416.

9901600 — EHRHARTA

1. GIBBS RUSSELL. 1984. BOTHALIA 15:

145 & 149. 1450

E. dodii Stapf = E. RUPESTRIS SUBSP. DODII

1050 E. EBURNEA GIBBS RUSSELL 1850

2490 E. RUPESTRIS NEES EX TRIN. SUBSP.

DODII (STAPF) GIBBS RUSSELL 2470

(= E. dodii Stapf) 1 2490

752

Bothalia 15, 3 & 4 (1985)

(012000 — ERIOSPERMUM

1. PERRY. 1984. JL S. AFR. BOT. 50: 503.

750 E. BAYERI P. L. PERRY

1250 E. BREVISCAPUM MARLOTH EX P. L.

PERRY

2240 E. DESCENDENS MARLOTH EX P. L.

PERRY

2260 E. DESERTICOLUM MARLOTH EX P. L.

PERRY

2620 E. EXILIS P. L. PERRY

2640 E. FILICAULE MARLOTH EX P. L. PERRY

2660 E. FLABELLATUM MARLOTH EX P. L.

PERRY

3750 E. MINUTIFLORUM MARLOTH EX P. L.

PERRY

3850 E. NAMAQUANUM MARLOTH EX P. L.

PERRY

5350 E. PUSILLUM P. L. PERRY

1098000 — LACHENALIA

1. BARKER. JL S. AFR. BOT. 50: 535.

220 L. ALOIDES (L.F.) HORT. EX ASCHERS. &

GRAEBN. VAR. VANZYLIAE W. F. BAR-

KER

3720 L. PHYSOCAULOS W. F. BARKER

4725 L. SCHELPEI W. F. BARKER

1113000 — Asparagus = MYRSIPHYLLUM

A. consanguineus (Kunth) Bak. = MYRSI-

PHYLLUM FASCICULATUM

A. crispus Lam. = MYRSIPHYLLUM DE-

CLINATUM

A. declinatus L. = MYRSIPHYLLUM DE-

CLINATUM

A. fasciculatus Thunb. = MYRSIPHYLLUM

FASCICULATUM

A. juniperoides Engl. = MYRSIPHYLLUM JU-

NIPEROIDES

A. multituberosus R. A. Dyer = MYRSIPHYL-

LUM MULTITUBEROSUM

A. ovatus Salter = MYRSIPHYLLUM OVA-

TUM

A. ramosissimus Bak. = MYRSIPHYLLUM

RAMOSISSIMUM

A. scandens Thunb. = MYRSIPHYLLUM

SCANDENS

A. undulatus (L.f.) Thunb. = MYRSIPHYL-

LUM UNDULATUM

A. volubilis Thunb. = MYRSIPHYLLUM VOL-

UBILE

1113020 — MYRSIPHYLLUM

1. OBERMEYER. 1984. BOTHALIA

15:77.

100 M. ALOPECURUM OBERM.

300 M. DECLINATUM (L.) OBERM.

(= Asparagus crispus Lam.)

(= Asparagus declinatus L.)

400 M. FASCICULATUM (THUNB.) OBERM.

(= Asparagus consanguineus (Kunth) Bak.) 1

(= Asparagus fasciculatus Thunb.) 1

500 M JUNIPEROIDES (ENGL.) OBERM.

(= Asparagus juniperoides Engl.)

700 M. MULTITUBEROSUM (R. A. DYER)

OBERM.

(= Asparagus multituberosus R. A. Dyer)

800 M OVATUM (SALTER) OBERM.

(= Asparagus ovatus Salter)

900 M RAMOSISSIMUM (BAK.) OBERM.

(= Asparagus ramosissimus Bak.)

1000 M. SCANDENS (THUNB.) OBERM.

( Asparagus scandens Thunb.)

1 100 M. UNDULATUM (L.F.) KUNTH

(= Asparagus undulatus (L.f.) Thunb.)

1200 M. VOLUBILE (THUNB.) OBERM.

(= Asparagus volubilis Thunb.)

CYPERACEAE :

0456000 — CARPHA

1. REID & ARNOLD. 1984. BOTHALIA

15: 139.

250 C. FILIFOLIA REID & ARNOLD

COMMELINACEAE 2

0899000 — ANEILEMA

1. FADEN. 1984. BOTHALIA 15: 89.

150 A. ARENICOLA FADEN

170 A. BRUNNEOSPERMUM FADEN

320 A. INDEHISCENS FADEN SUBSP. LILACI-

NUM FADEN

IRIDACEAE 2

1306010 — CROCOSMIA

1. DE VOS. 1984. JL S. AFR. BOT.

50:463.

150 C. AUREA PLANCH. VAR. MACULATA

BAK.

250 C. FUCATA (HERB.) DE VOS

500 C. MATHEWSIANA (L. BOL.) GOLDBL. EX

DE VOS

800 C. X CROCOSMIIFLORA (LEMOINE EX

MORREN) N.E. BR.

1311000 — GLADIOLUS

1. GOLDBLAIT. 1984. JL S. AFR. BOT.

50: 449.

3920 G. DESERTICOLUS GOLDBL.

8550 G. NIGROMONTANUS GOLDBL.

10950 G. ROBUSTUS GOLDBL.

11150 G. ROGERSII BAK. VAR. VLOKII

GOLDBL

ORCHIDACEAE 2

1434000 — DISA

1. MACMURTRY. 1984. S.A. ORCH. JL

15: 91.

3150 D. MACULOMARRONINA MACMURTRY

SALICACEAE 1

189200 — POPULUS

200 P. CANESCENS (AIT.) SM. * European species

collected in eastern Cape. 3226 (Fort Beaufort):

Adelaide, between the farms Linton & Austrey

(-AD). L. Henderson 595.

ULMACEAE 1

1896000 — ULMUS

100 U. PARVIFOLIA JACQ. * Asian species col-

lected in central Transvaal. 2528 (Pretoria):

Fountains Valley Nature Reserve (-CC),

K. Duggan & L. Henderson 58.

200 U. PROCERA SALISB. European species

collected in central Transvaal 2528 (Preto-

ria): Fountains Valley Nature Reserve (-CC).

K. Duggan & L. Henderson 59.

MORACEAE 1

1913000 — MORUS

50 M. ALBA L. * Asian species collected in eastern

Transvaal. 2530 (Lvdenburg): Sabie, Loerie

Walk adjoining Castle Rock Caravan

Park (-BA). L. Henderson 645.

70 M. JAPONICA AUDIB. * Asian species col-

lected in eastern Transvaal. 2530 (Lvdenburg):

Barberton. 7 km from escarpment on road to

Snymansbult (-DB). L. Henderson 640.

1961999 — FICUS

1. VON BREITENBACH. 1982. JL DEN-

DROL. SOC. 2: 49.

1000 F. GLUMOSA DEL.

(= F. sonderi Miq.)

F. sonderi Miq. = F. GLUMOSA

Bothalia 15, 3 & 4 (1985)

753

PROTEACEAE 1

2036010 — VEXATORELLA

1. ROURKE. 1984. JL S. AFR. BOT. 50:

373.

100 V. ALPINA (SALISB. EX KNIGHT) ROURKE

200 V. AMOENA (ROURKE) ROURKE

300 V. LATEBROSA ROURKE

400 V. OBTUSATA (THUNB.) ROURKE SUBSP.

ALBOMONTANA (ROURKE) ROURKE

500 V. OBTUSATA (THUNB.) ROURKE SUBSP.

OBTUSATA

SANTALACEAE 1

2118000 — THESIUM

1. HILLIARD & BURTT. 1983. NOTES. R.

BOT. GDN EDINB. 41:309.

450 T. ALATUM HILLIARD & BURTT

3550 T. DECIPIENS HILLIARD & BURTT

4250 T. DURUM HILLIARD & BURTT

OPILIACEAE 1

2122000 — OPILIA

1. HIEPKO. 1982. WILLDENOWIA

12:161.

100 O. CAMPESTRIS ENGL. VAR. CAMPES-

TRIS

POLYGONACEAl

2195000 — RUMEX

1. RECHINGER. 1964. FL. EUROPAEA

1: 83.

50 R. ACETOSELLA L. *

AMARANTHACEAE 1

2335000 — ALTERNANTHERA

50 A. CARACASANA H.B.K. * South American

species collected in central Transvaal. 2528

(Pretoria): Faerie Glen Township (-CD),

SAG PI SAAB 1/50.

MESEMBRYANTHEMACEAE 1

2405030 — CONOPHYTUM

1. RAWE. 1982. CACTUS & SUCC. J.,

U.S. 54: 165, 218.

C. anomalum L. Bol. = C. ECARINATUM

VAR. CANDIDUM

C. approximatum Lavis = C. AUCTUM

FORMA APPROXIMATUM

112550 C. AUCTUM N.E. BR. FORMA APPROXI-

MATUM (LAVIS) RAWE

(= C. approximatum Lavis) 1

(= C. australe L. Bol.) 1

2600 C. AUCTUM N.E. BR. FORMA AUCTUM

(= C. phillippii L. Bol.) 1

C. australe L. Bol. = C. AUCTUM FORMA

APPROXIMATUM

C. candidum L. Bol. = C. ECARINATUM

VAR. CANDIDUM

C. complanatum L. Bol. = C. VIRIDICATUM

VAR VIRIDICATUM

C. craterulum Tisch. = C. VELUTINUM VAR.

CRATERULUM

C. ecarinatum L. Bol. var. angustum L. Bol = C.

ECARINATUM VAR. ECARINATUM

8750 C. ECARINATUM L. BOL. VAR. CANDI-

DUM (L. BOL.) RAWE

(= C. anomalum L. Bol.) 1

(= C. candidum L. Bol.) 1

(= C. niveum L. Bol.) 1

8800 C. ECARINATUM L. BOL VAR. ECARINA-

TUM

(= C. ecarinatum L. Bol. var. angustum

L. Bol.) 1

(= C. ecarinatum L. Bol. var. mutabile

L. Bol.) 1

C. ecarinatum L. Bol. var. mutabile L. Bol. = C.

ECARINATUM VAR. ECARINATUM

C. globuliforme Schick & Tisch. = C. MEYERI

VAR. GLOBULIFORME

C. koupense Tisch. = C. VIRIDICATUM VAR.

PISINNUM

18950 C. MEYERI N.E. BR. VAR. GLOBULI-

FORME (SCHICK & TISCH.) RAWE

(= C. globuliforme Schick & Tisch.) 1

19000 C. MEYERI N.E. BR. VAR. MEYERI

FORMA MEYERI

(= C. papillatum L. Bol.) 1

19020 C. MEYERI N.E. BR VAR. MEYERI

FORMA SEMILUNULUM (TISCH.) RAWE

(= C. semilunulum Tisch.) 1

19050 C. MEYERI N.E. BR. VAR. RAMOSUM

(LAVIS) RAWE

(= C. ramosum Lavis) 1

C. morganii Lavis = C. PEERSII VAR. PEER-

SII

C. multipunctatum Tisch. = C. PEERSII VAR.

MULTIPUNCTATUM

C. niveum L. Bol. = C. ECARINATL1M VAR.

CANDIDUM

C. papillatum L. Bol. = C. MEYERI VAR.

MEYERI FORMA MEYERI

24550 C. PEERSII LAVIS VAR. MULTIPUNCTA-

TUM (TISCH.) RAWE

(= C. steytlervillense Tisch.) 1

(= C. multipunctatum Tisch.) 1

24600 C. PEERSII LAVIS VAR. PEERSII

(= C. morganii Lavis ) 1

(= C. subglobosum Tisch.) 1

C. phillippii L. Bol = C. AUCTUM FORMA

AUCTUM

C. pisinnum N.E. Br. = C. VIRIDICATUM

VAR. PISINNUM

C. ramosum Lavis = C. MEYERI VAR.

RAMOSUM

C. semilunulum Tisch. = C. MEYERI VAR.

MEYERI FORMA SEMILUNULUM

C. stegmannianum L. Bol. = C. TRUNCATUM

VAR. TRUNCATUM

C. steytlervillense Tisch. = C. PEERSII VAR.

MULTIPUNCTATUM

C. subglobosum Tisch. = C. PEERSII VAR.

PEERSII

C. tischeri Schick. = C. VELUTINUM VAR.

VELUTINUM

C. truncatum (Thunb.) N.E. Br. forma renniei

(Lavis) Tisch. = C. TRUNCATUM VAR.

TRUNCATUM

33920 C. TRUNCATUM (THUNB.) N.E. BR. VAR.

TRUNCATUM

(= C. truncatum (Thunb.) N.E. Br. forma ren-

niei (Lavis) Tisch.) 1

(= C. stegmannianum L. Bol.) 1

(= C. wagnerianum Schwantes) 1

33970 C. TRUNCATUM (THUNB.) N.E. BR. VAR.

WIGGETTAE (N.E. BR.) RAWE

(= C. wiggettae N.E. Br.) 1

35350 C. VELUTINUM SCHWANT. VAR. CRATE-

RULUM (TISCH.) RAWE

(= C. craterulum Tisch.) 1

35400 C. VELUTINUM SCHWANT. VAR. VELUTI-

NUM

(= C. tischeri Schick) 1

35850 C. VIRIDICATUM N.E. BR. VAR. PISINNUM

(N.E. BR.) RAWE

(= C. koupense Tisch.) 1

(= C. pisinnum N.E. Br.) 1

C. viridicatum N.E. Br. var. punctatum N.E. Br.

= C. VIRIDICATUM VAR. VIRIDICATUM

35900 C. VIRIDICATUM N.E. BR. VAR. VIRIDI-

CATUM

(= C. complanatum L. Bol.) 1

(= C. viridicatum N.E. Br. var. punctatum

N.E. Br.) 1

C. wagnerianum Schwantes = C. TRUNCATUM

VAR. TRUNCATUM

754

Bothalia 15, 3 & 4 (1985)

C. wiggettae N.E. Br. = C. TRUNCATUM

VAR. WIGGETTAE

-.40s073 — MESEMBRYANTHEMUM

1. KOUTNIK & LAVIS. 1984. JL S. AFR.

BOT. 50: 369.

2120 M. HORRIDUM KOUTNIK & LAVIS

PORTULACACEAE 1

2419000 — PORTULACARIA

1. VAN JAARSVELD. 1984. JL S. AFR.

BOT 50’ 393

150 P. ARMIANA E.J. VAN JAARSVELD

NYMPHAEACEAE 1

2513000 — NYMPHAEA

I. LANDON. 1984. PHYTOLOGIA 55:

109.

150 N. CAPENSIS THUNB. VAR. ALBA K.

LANDON

CRASSULACEAE 1

3164000 — TYLECODON

1. VAN JAARSVELD. 1983. JL S. AFR.

BOT. 49: 305.

850 T. KRITZINGERI E.J. VAN JAARSVELD

ROSACEAE 1

3658000 — LISTIA

L. heterophylla E. Mey . = LOTONONIS LISTII

3662000 — ASPALATHUS

1. DAHLGREN. 1984. S. AFR. J. BOT. 3:

259.

16750 A. OLIVERI DAHLG.

3673000 — ARGYROLOBIUM

1. HILLIARD & BURTT. 1983. NOTES R.

BOT. GDN EDINB. 41: 308.

3550 A. SYMMOMONTANUM HILLIARD &

BURTT

3703000 — PSORALEA

1. STIRTON. 1983. JL S. AFR. BOT. 49:

329.

2250 P. IMPLEXA C.H. STIRTON

4450 P. TRULLATA C.H. STIRTON

3703010 — CULLEN

1. STIRTON. 1981. BOTHALIA 13: 317.

250 C. DRUPACEA (BUNGE) C.H. STIRTON

350 C. JAUBERTIANA (FENZL) C.H. STIRTON

500 C. PLICATA (DEL.) C.H. STIRTON

3703020 — BITUMINARIA

1. STIRTON. 1981. BOTHALIA 13: 318.

100 B. ACAULIS (STEV.) C.H. STIRTON

200 B. BITUMINOSA (L.) C.H. STIRTON

3333000 — COTONEASTER

100 C. PANNOSUS FRANCH. * Asian species

collected in central Transvaal. 2528 (Pretoria):

Rietvleidam, old Pretoria/Bronkhorstspruit

road between Mooiplaats and Cullinan turnoff

(-CD), K. Duggan & L. Henderson 31.

3333010 — PYRACANTHA

100 P. ANGUSTIFOLIA (FRANCH.) SCHNEID. *

Asian species collected in central Transvaal

2627 (Potchefstroom): Roodepoort, Rooikrans.

Maraisburg/Magaliesberg road crossing the

Muldersdrifseloop (-BB), K. Duggan & L.

Henderson 55.

200 P. COCCINEA M.J. ROEM. * European and

Asian species collected in central Transvaal.

2528 (Pretoria): Wingate Park, Kempton Park/

Wingate Park road crossing the Sesmylspruit

(-CD), K. Duggan & L. Henderson 7.

3396000 — PRUNUS

150 P. PERSICA (L.) BATSCH. * Asian species col-

lected in Transvaal. 2530 (Lydenburg): Sabie,

Loerie Walk adjoining Castle Rock Caravan

Park (-BA), L. Henderson 644.

200 P. SEROTINA EHRH. * North American

species collected in central Transvaal. 2528

(Pretoria): tributary of Morelettaspruit near

Valley Farm Agricultural Holdings (-CD),

L. Henderson 532.

FABACEAE 1

3446000 — ACACIA

93550 A. PODALYRIIFOLIA A CUNN. * Australian

species collected in central Transvaal. 2528

(Pretoria): Verwoerdburg, Irene, along the

Irene/Bapsfontein road (-CC), K. Duggan & L.

Henderson 64.

3602000 — SOPHORA

200 S. JAPONICA L. * Asian species collected in

central Transvaal. 2528 (Pretoria): tributary

of Morelettaspruit near Valley Farm Agricultu-

ral Holdings (-CD), L. Henderson 533.

3657000 — LOTONONIS

1. POLHILL. 1976. IN HEYWOOD, BO-

TANICAL SYSTEMATICS, 324.

5150 L. LISTII POLHILL

(= Listia heterophylla E. Mey.) 1

3703030 — OTHOLOBIUM

1. STIRTON. 1983. JL S. AFR. BOT. 48:

337

100 O. PICTUM C.H. STIRTON

200 O. RUBICUNDUM C.H. STIRTON

3733000 — ROBINIA

100 R. PSEUDO-ACACIA L. * A central American

species collected in Transvaal. 2530 (Lyden-

burg): 13 km from Dullstroon on road to Ly-

denburg (-AB), G. Germishuizen 82.

3877000 — MUCUNA

1. WILMOT-DEAR. 1984. KEW BULL. 39:

63.

250 M. PRURIENS (L.) DC. VAR. UTILIS

(WALL. EX WIGHT) BAK. EX BURCK *

(= Stizolobium deeringianum Bort) 1

3877010 — Stizolobium = MUCUNA

S. deeringianum Bort = MUCUNA PRURIENS

VAR. UTILIS

3898000 — ERIOSEMA

1. STIRTON. 1983. JL S. AFR. BOT. 49:

451.

999 E. NAVICULARE C.H. STIRTON

GERANIACEAE 1

3928000 — PELARGONIUM

1. V.D. WALT. 1984. S. AFR. J. BOT. 3:

256.

7270 P. GREYTONENSE J.J.A. v.d. WALT

LINACEAE 3

3945000 — LINUM

1. ROGERS. 1981. NORD. J. BOT. 1: 711.

Bothalia 15, 3 & 4 (1985)

755

POLYGALACEAE 3

4273000 — POLYGALA

1. CHODAT. 1912. BOT. JB. 48: 331.

1850 P. ESTERAE CHOD.

EUPHORBIACEAE 3

4309000 — DRYPETES

1. RADCLIFFE-SMITH. 1978. KEW

BULL. 32: 477.

200 D. GERRARDII HUTCH. VAR. GERRAR-

DII

250 D. GERRARDII HUTCH. VAR. TOMEN-

TOSA RADCLIFFE-SMITH

500 D. RETICULATA PAX

4498000 — EUPHORBIA

1. KOUTNIK. 1984. S. AFR. J. BOT. 3: 262

E. chamaesyce L. * = CHAMAESYCE PROS-

TRATA *

E. chamaesycoides B. Nord. = CHAMAESYCE

CHAM AES YCOIDES

E. eylesii (Rendle) Koutnik = CHAMAESYCE

EYLESII

E. glanduligera Pax = CHAMAESYCE GLAN-

DULIGERA

E. hirta L. = CHAMAESYCE HIRTA

E. hypericifolia L. = CHAMAESYCE

HYPERICIFOLIA

E. inaequilatera Sond. var. inaequilatera =

CHAMAESYCE INAEQUILATERA

E. inaequilatera Sond. var. perennis N.E. Br. =

CHAMAESYCE INAEQUILATERA.

E. livida E. Mey. ex Boiss. = CHAMAESYCE

LIVIDA

E. maculata L. = CHAMAESYCE MACU-

LATA

17575 E. MISCELLA LEACH

E. mossambicensis (Klotzsch & Garcke) Boiss. =

CHAMAESYCE MOSSAMBICENSIS

E. neopolycnemoides Pax & K. Hoffm. =

CHAMAESYCE NEOPOLYCNEMOIDES

E. nutans Lag. = CHAMAESYCE NUTANS

E. pergracilis P.G. Mey. = CHAMAESYCE

PERGRACILIS

E. schlechteri Pax = CHAMAESYCE

SCHLECHTERI

E. serpens H.B.K. = CHAMAESYCE SER-

PENS

E. tettensis Klotzsch = CHAMAESYCE TET-

TENSIS

E. zambesiana Benth. = CHAMAESYCE

ZAMBESIANA

4498020 — CHAMAESYCE

1. KOUTNIK. 1984. S. AFR. J. BOT.

3:262.

100 C. CHAMAESYCOIDES (B. NORD.) KOUT-

NIK

(= Euphorbia chamaesycoides B. Nord.) 1

200 C. EYLESII (RENDLE) KOUTNIK

(= Euphorbia eylesii Rendle) 1

300 C. GLANDULIGERA (PAX) KOUTNIK

(= Euphorbia glanduligera Pax) 1

400 C. HIRTA (L.) MILLSP.

(= Euphorbia hirta L.) 1

500 C. HYPERICIFOLIA (L.) MILLSP.

(= Euphorbia hypericifolia L.) 1

600 C. INAEQUILATERA (SOND.) SOJAK

(= Euphorbia inaequilatera Sond. var.

inaequilatera) 1

(= Euphorbia inaequilatera Sond. var. perennis

N.E. Br.) 1

700 C. LIVIDA (E. MEY. EX BOISS.) KOUTNIK

(= Euphorbia livida E. Mey. ex Boiss.) 1

800 C. MACULATA (L.) SMALL

(= Euphorbia maculata L.) 1

900 C. MOSSAMBICENSIS (KLOTZSCH &

GARCKE) KOUTNIK

(= Euphorbia mossambicensis Klotzsch &

Garcke) 1

1000 C. NEOPOLYCNEMOIDES (PAX & K.

HOFFM.) KOUTNIK

(= Euphorbia neopolycnemoides Pax &

Hoffm.) 1

1100 C. NUTANS (LAG.) SMALL

(= Euphorbia nutans Lag.) 1

11200 C. PERGRACILIS (P.G. MEY.) KOUTNIK

(= Euphorbia pergracilis P.G. Mey.) 1

11300 C. PROSTRATA (AIT.) SMALL *

(= Euphorbia chamaesyce L. *) 1

11400 C. SCHLECHTERI (PAX) KOUTNIK

(= Euphorbia schlechteri Pax) 1

11500 C. SERPENS (H.B.K.) SMALL

(= Euphorbia serpens H.B.K.) 1

11600 C. TETTENSIS (KLOTZSCH) KOUTNIK

(= Euphorbia tettensis Klotzsch) 1

11700 C. ZAMBESIANA (BENTH.) KOUTNIK

(= Euphorbia zambesiana Benth.) 1

CELASTRACEAE 3

4626000 — MAYTENUS

1. VAN WYK. 1984. S. AFR. J. BOT. 3:

115.

50 M. ABBOTTII VAN WYK

COMBRETACEAE 3

5538000 — COMBRETUM

1. VAN WYK. 1984. S. AFR. J. BOT. 3:

125.

2975 C. VENDAE VAN WYK

MELASTOMATACEAE 3

5659000 — DISSOTIS

1. FERNANDES. 1955. BOL. SOC. BROT.

29: 54.

800 D. PULCHRA A. & R. FERNANDES

ERICACEAE 3

6237000 — ERICA

1. OLIVER. 1984. S. AFR. J. BOT. 3: 268.

50 E. ABELLI E.G.H. OLIVER

E. absinthoides Thunb. = PHILIPPIA ABSIN-

THOIDES

E. xeranthemifolius Salisb. = ACROSTEMON

XERANTHEMIFOLIUS

624000 — PHILIPPIA

1. OLIVER. 1984. S. AFR. J. BOT. 3: 270.

20 P. ABSINTHOIDES (THUNB.) E.G.H.

OLIVER

(= Erica absinthoides Thunb.) 1

40 P. ALTICOLA E.G.H. OLIVER

150 P. ELSIEANA E.G.H. OLIVER

170 P. ESTERHUYSENIAE E.G.H. OLIVER

SUBSP. ESTERHUYSENIAE

180 P. ESTERHUYSENIAE E.G.H. OLIVER

SUBSP. SWARTBERGENSIS E.G.H.

OLIVER

250 P. IRRORATA E.G.H. OLIVER

350 P. NOTHOLEEANA E.G.H. OLIVER

450 P. PETROPHILA E.G.H. OLIVER

470 P. PROCAVIANA E.G.H. OLIVER

6243030 — ACROSTEMON

1. OLIVER. 1984. S. AFR. J. BOT. 3: 267.

A. fourcadei L. Guthrie = SIMOCHEILUS

FOURCADEI

700 A. HIRSUTUS (THUNB.) KLOTZSCH

(= A. incurvus (Klotzsch) Benth.) 1

A. incurvus (Klotzsch) Benth. = ACROSTE-

MON HIRSUTUS

1200 A. XERANTHEMIFOLIUS (SALISB.) E.G.H.

OLIVER

(= Erica xeranthemifolius Salisb.) 1

A. viscidus N.E. Br. = ARACHNOCALYX

VISCIDUS

6243035 — ARACHNOCALYX

1. OLIVER. 1984. S. AFR. J. BOT. 3: 268.

Bothalia 15, 3 & 4 (1985)

100 A. CERERIS COMPTON

200 A. VISCIDUS (N.E. BR.) E.G.H. OLIVER

(= Acrostemon viscidus N.E. Br.) 1

6244000 — SIMOCHEILUS

1. OLIVER. 1984. S. AFR. J. BOT. 3: 284.

1000 S. FOURCADEI (L. GUTHRIE) E.G.H.

OLIVER

(= Acrostemon fourcadei L. Guthrie) 1

1300 S HIRTUS (KLOTZSCH) E.G.H. OLIVER

(= Octogonia hirta Klotzsch) 1

(= S. hirsutus Benth.) 1

S. hirsutus Benth. = S. HIRTUS

S. klotzschianus Benth. var. glabrifolius N.E. Br.

= SIMOCHEILUS PUBERULUS

S. klotzschianus Benth. var. klotzschianus =

SIMOCHEILUS PUBERULUS

2050 S. PUBERULUS (KLOTZSCH) E.G.H.

OLIVER

(= S. klotzschianus Benth. var. glabrifolius

N.E. Br.) 1

(= S. klotzschianus Benth. var. klotzschianus) 1

(= Plagiostemon puberulus Klotzsch) 1

(= Simocheilus klotzschianus Benth.) 1

6244010 — Octoaonia = SIMOCHEILUS

6. hirta Klotzsch = SIMOCHEILUS HIRTUS

6244020 — Plagiostemon = SIMOCHEILUS

P puberulus Klotzsch = SIMOCHEILUS PU-

BERULUS

6245020 — LEPTERICA

L. tenuis (Benth.) N.E. Br. = SCYPHOGYNE

TENUIS

6246020 — SCYPHOGYNE

1. OLIVER. 1984. S. AFR. J. BOT. 3: 282.

S. biconvexa N.E. Br. = SCYPHOGYNE

DIVARICATA

S. brownii Compton = SCYPHOGYNE DIVA-

RICATA

S. burchellii N.E. Br. = SCYPHOGYNE URC-

EOLATA

400 S. CAPITATA (KLOTZSCH) BENTH.

(= Scyphogyne capitata (Klotzsch) Benth. var.

brevifolia N.E. Br.) 1

(= Scyphogyne capitata (Klotzsch) Benth. var.

capitata) 1

(= Scyphogyne viscida N.E. Br.) 1

S. capitata (Klotzsch) Benth. var. brevifolia N.E.

Br. = SCYPHOGYNE CAPITATA

S. capitata (Klotzsch) Benth. var. capitata =

SCYPHOGYNE CAPITATA

600 S. DIVARICATA (KLOTZSCH) BENTH.

(= S. biconvexa N.E. Br.) 1

(= S. brownii Compton) 1

(= S. glandulifera N.E. Br.) 1

(= S. rieidula N.E. Br. var. brevicilata N.E.

Br.) 1

(= S. rigidula N. E. Br. var. rigidula)

S. glandulifera N.E. Br. = SCYPHOGYNE

DIVARICATA

1200 S. MUSCOSA (AIT.) STEUD.

(= S. schlechteri N.E. Br.) 1

S rigidula N.E. Br. var. breviciliata N.E. Br. =

SCYPHOGYNE DIVARICATA

S rigidula N.E. Br. var. rigidula = SCYPHO-

GYNE DIVARICATA

S. schlechteri N.E. Br. = SCYPHOGYNE MUS-

COSA

1750 S TENUIS (BENTH.) E.G.H. OLIVER

(= Lepterica tenuis (Benth.) N.E. Br.) 1

S. trimera N.E. Br. = SCYPHOGYNE

URCEOLATA

1900 S. URCEOLATA (KLOTZSCH) BENTH.

(= S. burchellii N.E. Br.) 1

(= S. trimera N.E. Br.) 1

S viscida N.E. Br. = SCYPHOGYNE CAPI-

TATA

6248020 — NAGELOCARPUS

1. OLIVER. 1984. S. AFR. J. BOT. 3: 270.

N. ciliatus (Benth.) Bullock = NAGELOCAR-

PUS SERRATUS

200 N. SERRATUS (THUNB.) BULLOCK

(= N. ciliatus (Benth.) Bullock) 1

GENTIANACEAE 1

6481000 — SEBAEA

1. HILLIARD & BURTT. 1984. NOTES R.

BOT. GDN EDINB. 41: 304.

3350 S. PLEUROSTIGMATOSA HILLIARD &

BURTT

3550 S. RADIATA HILLIARD & BURTT

6503000 — CHIRONIA

1. HILLIARD. 1984. NOTES R. BOT.

GDN EDINB. 41: 303.

50 C. ALBIFLORA HILLIARD

PERIPLOCACEAE '

6747000 — RAPHIONACME

1. RE1IEF & VENTER. 1983. S. AFR. J.

BOT. 2: 326.

150 R. DYERI RETIEF & VENTER

ASCLEPIADACEAE '

6778000 — SCHIZOGLOSSUM

1. KUPICHA. 1984. KEW BULL. 38: 599.

S. aciculare N.E. Br. = STENOSTELMA CA-

PENSE

S. addoense N.E. Br. = ASPIDOGLOSSUM

GRACILE

S. altissimum Schltr. = ASPIDOGLOSSUM IN-

TERRUPTED

S. anomalum N.E. Br. = MIRAGLOSSUM

ANOMALUM

S. araneiferum Schltr. = ASPIDOGLOSSUM

ARANEIFERUM

1000 S. ATROPURPUREUM E. MEY. SUBSP.

ATROPURPUREUM

1025 S. ATROPURPUREUM E. MEY. SUBSP.

TRIDENTATUM (SCHLTR.) KUPICHA

1050 S. ATROPURPUREUM E. MEY. SUBSP. VI-

RENS (E. MEY.) KUPICHA

(= S. euphorbioides E. Mey.) 1

(= S. tridentatum Schltr.) 1

(= S. virens E. Mey.) 1

S. atrorubens Schltr. = SCHIZOGLOSSUM BI-

DENS SUBSP. ATRORUBENS

S. auriculatum N.E. Br. = ASPIDOGLOSSUM

GLANDULIFERUM

S. barberiae Schltr. = ASPIDOGLOSSUM IN-

TER RUPTUM

S. baumii Schltr. ex N.E. Br. = SCHIZOGLOS-

SUM BIDENS SUBSP. ATRORUBENS

S. biauriculatum Schltr. = ASPIDOGLOSSUM

DELAGOENSE

1550 S. BIDENS E. MEY. SUBSP. ATRORUBENS

(SCHLTR.) KUPICHA

(= S. atrorubens Schltr.) 1

(= S. baumii Schltr. ex N.E. Br.) 1

1600 S. BIDENS E. MEY. SUBSP. BIDENS

(= S. truncatum Schltr.) 1

1610 S. BIDENS E. MEY. SUBSP. GALPINII

(SCHLTR.) KUPICHA

(= S. galpinii Schltr.) I

1620 S. BIDENS E. MEY. SUBSP. GRACILE

KUPICHA

(= S. diversum N.E. Br.) 1

(= S. umbellatum Schltr.) 1

1630 S. BIDENS E. MEY. SUBSP. HIRTUM KUPI-

CHA

1640 S. BIDENS E. MEY. SUBSP. PACHYGLOS-

SUM (SCHLTR.) KUPICHA

(= S. pachyalossum Schltr. var. abbreviatum

N.E. Br.) 1

(= S. pachvglossum Schltr. var. pachy-glos-

sum) 1

Bothalia 15, 3 & 4 (1985)

757

1650 S. BIDENS E. MEY. SUBSP. PRODUCTUM

(N.E. BR.) KUPICHA

(= S. pachyglossum Schltr. var. productum

N.E. Br.) 1

S. biflorum (E. Mey.) Schltr. var. biflorum = AS-

PIDOGLOSSUM BIFLORUM

S. biflorum (E. Mey.) Schltr. var. concinnum

. N.E. Br. = ASPIDOGLOSSUM BIFLORUM

S. biflorum (E. Mey.) Schltr. var. integrum N.E.

Br. = ASPIDOGLOSSUM BIFLORUM

S. bolusii Schltr. = ASPIDOGLOSSUM GRA-

CILE

S. bowkeriae N.E. Br. = ASPIDOGLOSSUM

GRACILE

S. buchananii N.E. Br. = ASPIDOGLOS-

SUM GRACILE

S. burchellii N.E. Br. = ASPIDOGLOSSUM

GRACILE

S. capense (Schltr.) Huber = STENOSTELMA

CAPENSE

S. carinatum Schltr. = ASPIDOGLOSSUM

CARINATUM

S. ciliatum Schltr. = ASPIDOGLOSSUM FAS-

CICULARE

S. commixtum N.E. Br. = ASPIDOGLOSSUM

GLANDULIFERUM

S. conrathii Schltr. = ASPIDOGLOSSUM BIF-

LORUM

S. consimile N.E. Br. = ASPIDOGLOSSUM

HETEROPHYLLUM

S. contracurvum N.E. Br. = ASPIDOGLOS-

SUM OVALIFOLIUM

3300 S. CORDIFOLIUM E. MEY.

(= S. cordifolium E. Mey. var. centralis N.E.

Br.) 1

(= S. divaricatum N.E. Br.) 1

(= S. hirsutum Turcz.) 1

S. cordifolium E. Mey. var. centralis N.E. Br. =

SCHIZOGLOSSUM CORDIFOLIUM

S. corniculatum (E. Mey.) R.A. Dyer = STE-

NOSTELMA CORNICULATUM'

S. davyi N.E. Br. = MIRAGLOSSUM DAVYI

S. decipiens N.E. Br. = SCHIZOGLOSSUM

STENOGLOSSUM SUBSP. FLAVUM

S. delagoense Schltr. = ASPIDOGLOSSUM

DELAGOENSE

S. dissimile N.E. Br. var. dissimile = ASPIDO-

GLOSSUM DISSIMILE

S. dissimile N.E. Br. var. pubiflorum N.E. Br. =

ASPIDOGLOSSUM DISSIMILE

S. divaricatum N.E. Br. = SCHIZOGLOSSUM

CORDIFOLIUM

S. diversum N.E. Br. = SCHIZOGLOSSUM BI-

DENS SUBSP. GRACILE

S. dregei N.E. Br. = ASPIDOGLOSSUM GRA-

CILE

4300 S. ELINGUE N.E. BR. SUBSP. ELINGUE

4350 S. ELINGUE N.E. BR. SUBSP. PURPUREUM

KUPICHA

S. euphorbioides E. Mey. = SCHIZOGLOSSUM

ATROPURPUREUM SUBSP. VIRENS

S. excisum Schltr. = ASPIDOGLOSSUM BI-

FLORUM

S. exile Schltr. = ASPIDOGLOSSUM GRA-

CILE

S. filifolium Schltr. = ASPIDOGLOSSUM

GRACILE

S. filipes Schltr. = ASPIDOGLOSSUM GRA-

CILE

S. flanaganii Schltr. = ASPIDOGLOSSUM

FLANAGANII

5000 S. FLAVUM SCHLTR.

(= S. flavum Schltr. var. lineare N.E. Br.) 1

S. flavum Schltr. var. lineare N.E. Br. = SCHI-

ZOGLOSSUM FLAVUM

S. aalpinii Schltr. = S. BIDENS SUBSP.

GALPINII

5400 S. GLABRESCENS SCHLTR.

(= S. glabrescens Schltr. var. longirostre

(Schltr.) N.E. Br.) 1

S. glabrescens Schltr. var. longirostre (Schltr.)

N.E. Br. = SCHIZOGLOSSUM GLABRES-

CENS

S. glanduliferum Schltr. = ASPIDOGLOSSUM

GLANDULIFERUM

S. grandiflorum Schltr. = ASPIDOGLOSSUM

GRANDIFLORUM

5900 S. HAMATUM E. MEY.

(= S. hamatum E. Mey. var. elegans N.E.

Br.) 1

(= S. hamatum E. Mey. var. pallidum N.E.

Br.) 1

S. hamatum E. Mey. var. elegans N.E. Br. =

SCHIZOGLOSSUM HAMATUM

S. hamatum E. Mey. var. pallidum N.E. Br. =

SCHIZOGLOSSUM HAMATUM

S. harveyi N.E. Br. = ASPIDOGLOSSUM

HETEROPHYLLUM

S. heterophvllum (E. Mey.) Schltr. var. hetero-

phyllum = ASPIDOGLOSSUM HETERO-

PHYLLUM

S. heterophyllum (E. Mey.) Schltr. var. majus

N.E. Br. = ASPIDOGLOSSUM HETERO-

PHYLLUM

S. heterophyllum (E. Mey.) Schltr. var. schinzia-

num N.E. Br. = ASPIDOGLOSSUM HET-

EROPHYLLUM

6450 S. HILLIARDIAE KUPICHA

S. hirsutum Turcz. = SCHIZOGLOSSUM COR-

DIFOLIUM

S. hirtiflorum N.E. Br. = ASPIDOGLOSSUM

GLABRESCENS

S. interruptum (E. Mey.) Schltr. = ASPIDO-

GLOSSUM INTERRUPTUM

S. lamellatum Schltr. = ASPIDOGLOSSUM

LAMELLATUM

S. loreum S. Moore = ASPIDOGLOSSUM

GLABRESCENS

S. macowanii N.E. Br. var. macowanii = ASPI-

DOGLOSSUM GRANDIFLORUM

S. macowanii N.E. Br. var. tugelense N.E. Br. =

ASPIDOGLOSSUM GRANDIFLORUM

S. masaicum N.E. Br. = ASPIDOGLOSSUM

MASAICUM

S. monticola Schltr. = ASPIDOGLOSSUM

GRACILE

7700 S. NITIDUM SCHLTR.

(= S. wallacei Schltr.) 1

S. ovalifolium Schltr. = ASPIDOGLOSSUM

OVALIFOLIUM

S. pachyglossum Schltr. var. abbreviatum N.E.

Br. = SCHIZOGLOSSUM BIDENS SUBSP.

PACHYGLOSSUM

S. pachyglossum Schltr. var. pachyglossum =

SCHIZOGLOSSUM BIDENS SUBSP. PA-

CHYGLOSSUM

S. pachyglossum Schltr. var. productum N.E. Br.

= SCHIZOGLOSSUM BIDENS SUBSP.

PRODUCTUM

S. parcum N.E. Br. = ASPIDOGLOSSUM

GRACILE

S. parile N.E. Br. = ASPIDOGLOSSUM

GLANDULIFERUM

S. parvulum Schltr. var. parvulum = ASPIDO-

GLOSSUM GRACILE

S. parvulum Schltr. var. sessile N.E. Br. = ASPI-

DOGLOSSUM GRACILE

S. pilosum Schltr. = MIRAGLOSSUM PILO-

SUM

S. pulchellum Schltr. = MIRAGLOSSUM PUL-

CHELLUM

S. pumilum Schltr. = ASPIDOGLOSSUM

OVALIFOLIUM

S. randii S. Moore = ASPIDOGLOSSUM RES-

TIOIDES

S. restioides Schltr. = ASPIDOGLOSSUM

RESTIOIDES

S. robustum Schltr. var. inandense N.E. Br. =

ASPIDOGLOSSUM OVALIFOLIUM

S. robustum Schltr. var. pubiflorum N.E. Br. =

ASPIDOGLOSSUM OVALIFOLIUM

Bothalia 15, 3 & 4 (1985)

S robustum Schltr. var. robustum = ASPIDO-

GLOSSUM OVALIFOL1UM

S. schlechteri N.E. Br. = ASPIDOGLOSSUM

GLABRESCENS

9800 S. STENOGLOSSUM SCHLTR. SUBSP. FLA-

VUM (N.E. BR.) KUP1CHA

(= S. decipiens N.E. Br.) 1

9180 S. STENOGLOSSUM SCHLTR. SUBSP.

LATIFOLIUM KUPICHA

9820 S STENOGLOSSUM SCHLTR. SUBSP. STE-

NOGLOSSUM

S. striatum Schltr. = ASPIDOGLOSSUM OVA-

LIFOLIUM

S. strictum Schltr. = ASPIDOGLOSSUM BIF-

LORUM

S. tomentosum Schltr. = ASPIDOGLOSSUM

GRACILE

S. tricuspidatum Schltr. = ASPIDOGLOSSUM

CARINATUM

S. tridentatum Schltr. = SCHIZOGLOSSUM

ATROPURPUREUM SUBSP. VIRENS

S. tridens N.E. Br. = ASPIDOGLOSSUM

GLABRESCENS

S. truncatum Schltr. = SCHIZOGLOSSUM BI-

DENS SUBSP. BIDENS

S. tubulosum Schltr. = ASPIDOGLOSSUM

BIFLORUM

S. umbellatum Schltr. = SCHIZOGLOSSUM

BIDENS SUBSP. GRACILE

S. uncinatum N.E. Br. = ASPIDOGLOSSUM

UNCINATUM

S. unicum N.E. Br. = ASPIDOGLOSSUM

GLABRESCENS

S. verticillare Schltr. = MIRAGLOSSUM VER-

TICILLARE

S. virens E. Mey. = SCHIZOGLOSSUM

ATROPURPUREUM SUBSP. VIRENS

S. virgatum (E. Mey.) Schltr. = ASPIDOGLOS-

SUM VIRGATUM

S. wallacei Schltr. = SCHIZOGLOSSUM NITI-

DUM

S. woodii Schltr. = ASPIDOGLOSSUM

WOODI!

6778010 — ASPIDOGLOSSUM

1. KUPICHA. 1984. KEW BULL. 38: 599.

100 A ARANEIFERUM (SCHLTR.) KUPICHA

(= Schizoglossum araneiferum Schltr.) 1

200 A. BIFLORUM E. MEY.

(= Schizoglossum biflorum (E. Mey.) Schltr.

var. biflorum) 1

(= Schizoglossum biflorum (E. Mey.) Schltr.

var. concinnum (Schltr.) N.E. Br.) 1

(= Schizoglossum biflorum (E. Mey.) Schltr.

var. integrum N.E. Br.) 1

(= Schizoglossum conrathii Schltr.) 1

(= Schizoglossum excisum Schltr.) 1

(= Schizoglossum strictum Schltr.) 1

(= Schizoglossum tubulosum Schltr.) 1

300 A. CARINATUM (SCHLTR.) KUPICHA

(= Schizoglossum carinatum Schltr.) 1

(= Schizoglossum tricuspidatum Schltr.) 1

400 A. DELAGOENSE (SCHLTR.) KUPICHA

(= Schizoglossum biauriculatum Schltr.) 1

(= Schizoglossum delagoense Schltr.) 1

500 A. DEMISSUM KUPICHA

600 A. DISSIMILE (N.E. BR.) KUPICHA

(= Schizoglossum dissimile N.E. Br. var. dissi-

mile) 1

(= Schizoglossum dissimile N.E. Br. var.

pubiflorum N.E. Br.) 1

700 A. FASCICULARE E. MEY.

(= Schizoglossum ciliatum Schltr.) 1

800 A. FLANAGANII (SCHLTR.) KUPICHA

(= Schizoglossum flanaganii Schltr.) 1

900 A. GLABRESCENS (SCHLTR.) KUPICHA

(= Schizoglossum hirtiflorum N.E. Br.) 1

(= Schizoglossum loreum S. Moore)

(= Schizoglossum schlechteri N.E. Br.) 1

(= Schizoglossum tridens N.E. Br.) 1

(- Schizoglossum unicum N.E. Br.) 1

1000 A. GLANDULIFERUM (SCHLTR.) KUPI-

CHA

(= Schizoglossum auriculatum N.E. Br.) 1

(= Schizoglossum commixtum N.E. Br.) 1

(= Schizoglossum glanduliferum Schltr.) 1

(= Schizoglossum parile N.E. Br.) 1

1100 A GRACILE (E. MEY.) KUPICHA

(= Schizoglossum addoense N.E. Br.) 1

(= Schizoglossum bolusii Schltr.) 1

(= Schizoglossum bowkeriae N.E. Br.) 1

(= Schizoglossum buchananii N.E. Br.) 1

(= Schizoglossum burchellii N.E. Br.) 1

(= Schizoglossum dregei N.E. Br.) 1

(= Schizoglossum exile (Decne.) Schltr.) 1

(= Schizoglossum filifolium Schltr.) 1

(= Schizoglossum filipes Schltr.) 1

(= Schizoglossum monticola Schltr.) 1

(= Schizoglossum parcum N.E. Br.) 1

(= Schizoglossum parvulum Schltr. var. parvu-

lum) 1

(= Schizoglossum parvulum Schltr. var. sessile

N.E. Br.) 1

(= Schizoglossum tomentosum Schltr.) 1

1200 A. GRANDIFLORUM (SCHLTR.) KUPICHA

(= Schizoglossum grandiflorum Schltr.) 1

(= Schizoglossum macowanii N.E. Br. var.

macowanii) 1

(= Schizoglossum macowanii N.E. Br. var. tu-

gelense N.E. Br.) 1

1300 A. HETEROPHYLLUM E. MEY.

(= Schizoglossum consimile N.E. Br.) 1

(= Schizoglossum harveyi N.E. Br.) 1

(= Schizoglossum heterophyllum (E. Mey.)

Schltr. var. heterophyllum) 1

(= Schizoglossum heterophyllum (E. Mey.)

Schltr. var. majus N.E. Br.) 1

(= Schizoglossum heterophyllum (E. Mey.)

Schltr. var. schinzianum N.E. Br.) 1

1400 A. INTERRUPTUM (E. MEY.) BULLOCK

(= Schizoglossum altissimum Schltr.) 1

(= Schizoglossum barberiae Schltr.) 1

(= Schizoglossum interruptum (E. Mey.)

Schltr.) 1

1500 A. LAMELLATUM (SCHLTR.) KUPICHA

(= Schizoglossum lamellatum (Schltr.) Kupi-

cha) 1

1550 A. MASAICUM (N.E. BR.) KUPICHA

(= Schizoglossum masaicum N.E. Br.) 1

1600 A. OVALIFOLIUM (SCHLTR.) KUPICHA

(= Schizoglossum contracurvum N.E. Br.) 1

(= Schizoglossum ovalifolium Schltr.) 1

(= Schizoglossum pumilum Schltr.) 1

(= Schizoglossum robustum Schltr. var. inan-

dense N.E. Br.) 1

(= Schizoglossum robustum Schltr. var. pubi-

florum N.E. Br.) 1

(= Schizoglossum robustum Schltr. var. robus-

tum) 1

(= Schizoglossum striatum Schltr.) 1

1700 A. RESTIOIDES (SCHLTR.) KUPICHA

(= Schizoglossum randii S. Moore) 1

(= Schizoglossum restioides Schltr.) 1

1800 A. UNCINATUM (N.E. BR.) KUPICHA

(= Schizoglossum uncinatum N.E. Br.) 1

1900 A. VALIDUM KUPICHA

2000 A. VIRGATUM (E. MEY.) KUPICHA

(= Schizoglossum virgatum (E. Mey.) Schltr.) 1

2100 A. WOODII (SCHLTR.) KUPICHA

(= Schizoglossum woodii Schltr.) 1

6778020 — MIRAGLOSSUM

1. KUPICHA. 1984. KEW BULL. 38: 599.

100 M. ANOMALUM (N.E. BR.) KUPICHA

(= Schizoglossum anomalum N.E. Br.) 1

200 M. DAVYI (N.E. BR.) KUPICHA

(= Schizoglossum davyi N.E. Br.) 1

300 M. LAEVE KUPICHA'

400 M. PILOSUM (SCHLTR.) KUPICHA

(= Schizoglossum pilosum Schltr.) 1

500 M. PULCHELLUM (SCHLTR.) KUPICHA

(= Schizoglossum pulchellum Schltr.) 1

Bothalia 15, 3 & 4 (1985)

759

600 M. SUPERBUM KUPICHA

700 M. VERTICILLARE (SCHLTR.) KUPICHA

(= Schizoglossum verticillare Schltr.) 1

6787000 — STENOSTELMA

1. KUPICHA. 1984. KEW BULL. 38: 599.

100 S. CAPENSE SCHLTR.

(= Schizoglossum aciculare N.E. Br.) 1

(= Schizoglossum capense (Schltr.) Huber) 1

200 S. CORNICULATUM (E. MEY.) BULLOCK

(= Schizoglossum corniculatum (E. Mey.)

R.A. Dyer) 1

6787010 — PACHYCARPUS

1. SMITH. 1984. NOTES R. BOT. GDN

EDINB. 41: 301.

100 P. LEBOMBOENSIS D.M.N. SMITH

6885000 — STAPELIA

1. LEACH. 1984. S. AFR. J. BOT. 3: 169.

1050 S. BAYLISSII LEACH

6650 S. KOUGABERGENSIS LEACH

8425 S. MONTANA LEACH VAR. GROSSA

LEACH

8450 S. MONTANA LEACH VAR. MONTANA

9450 S. OBDUCTA LEACH

10625 S. PRAETERMISSA LEACH VAR. LU-

TEOLA LEACH

10650 S. PRAETERMISSA LEACH VAR. PRAE-

TERMISSA

ASTERACEAE 5

8740000 — ERLANGEA

1. WILD. 1977. KIRKIA 10: 362.

300 E. MISERA (OLIV. & HIERN) S. MOORE

(= E. schinzii O. Hoffm.)

E. schinzii O. Hoffm. = E. MISERA

8900000 — ASTER

A. caffrorum Less. = M1CROGLOSSA CAF-

FRORUM

8919000 — FELICIA

F. caffrorum (Less.) Nees = MICROGLOSSA

CAFFRORUM

8921000 — MICROGLOSSA

1. GRAU. 1976. MITT. BOT. STSAMML.,

MUNCH. 12: 399.

50 M. CAFFRORUM (LESS.) GRAU

(= Aster caffrorum Less.) 1

(= Felicia caffrorum (Less.) Nees)

8926000 — CONYZA

'1. WILD. 1975. KIRKIA 10: 60.

700 C. GOUANII (L.) WILLD.

(= C. hochstetteri Sch. Bip. ex A. Rich.)

C. hochstetteri Sch. Bip. ex A. Rich. = C.

GOUANII

8987010 — GALEOMMA (= Eriosphaera)

1. HILLIARD. 1983. FSA 33, PART 7,

FASC. 2: 13.

100 G. OCULUS-CATI (L.F.) RAUSCHERT

(= Gnaphalium oculus-cati L.f. )

(= Eriosphaera oculus-cati (L.f.) Less.)

200 G. STENOLEPIS (S. MOORE) HILLIARD

(= Gnaphalium stenolepis S. Moore)

(= Eriosphaera stenolepis (S. Moore) Hilliard

& Burtt)

8992000 — GNAPHALIUM

G. oculus-cati L.f. = GALEOMMA OCULUS-

CATI

G. stenolepis S. Moore = GALEOMMA STE-

NOLEPIS

9002000 — Eriosphaera = GALEOMMA

E. oculus-cati (L.f.) Less. = GALEOMMA

OCULUS-CATI

E. stenolepis (S. Moore) Hilliard & Burtt =

GALEOMMA STENOLEPIS

9049000 — ANAGLYPHA

A. latifolia S. Moore = ANISOPAPPUS LATI-

FOLIUS

9096000 — ANISOPAPPUS

1. HILLIARD & BURTT. 1979. NOTES R.

BOT. GDN EDINB. 37: 294.

350 A. LATIFOLIUS (S. MOORE) B.L. BURTT

(= Anaglypha latifolia S. Moore)

9098000 — OSMITOPSIS

1. BREMER. 1976. BOT. NOTISER 129:

21.

900 O. PINNATIFIDA (DC.) BREMER SUBSP.

SERRATA BREMER

9351000 — COTULA

1. COMPTON. 1941. JL S. AFR. BOT. 7:

189.

2175 C. PEDICELLATA COMPTON

9366000 — PENTZIA

1. HUTCHINSON. 1916. KEW BULL.

1916: 241.

1950 P. NANA BURCH.

(= P. quinquefida (Thunb.) Less. var. nana

(Burch.) Harv.) 1

P. quinquefida (Thunb.) Less. var. nana (Burch.)

Harv. = P. NANA

9405010 — CRASSOCEPHALUM

1. HILLIARD & BURTT. 1979. NOTES R.

BOT. GDN EDINB. 37: 295.

1900 C. SARCOBASIS (DC.) S. MOORE

(= Gynura sarcobasis DC.)

9405020 — GYNURA

G. sarcobasis DC. = CRASSOCEPHALUM

SARCOBASIS

REFERENCES

DYER, R.A., 1975. The genera of southern African flowering

plants, Vol. 1, Dicotyledons. Pretoria: Botanical Research

Institute.

DYER, R.A.. 1976. The genera of southern African flowering

plants, Vol. 2, Monocotyledons. Pretoria: Botanical Re-

search Institute.

GIBBS RUSSELL, G.E. & THE STAFF OF THE NATIONAL

HERBARIUM., 1984. List of species of southern African

plants. Mem. bot. Surv. S. Afr. 48.

GIBBS RUSSELL, G.E., REID, C., VAN ROOY, J. &

SMOOK, L., 1985. List of species of southern African

plants, edn. 2, Part 1, Bryophyta, Pteridophyta, Gymnos-

permae, Monocotyledonae. Mem. bot. Surv. S. Afr. 51.

GIBBS RUSSELL, G.E., WELMAN. W.G., GERMISHUI-

ZEN, G., RETIEF, E. & PIENAAR, B.J.. in prep. List of

species of southern African plants, edn. 2, Part 2, Dicotyle-

donae. Mem. bot. Surv. S. Afr.

REVIEW OF THE WORK OF THE BOTANICAL RESEARCH INSTITUTE, 1983/1984

1st April 1983 — 31st March 1984

CONTENTS

Introduction 761

Reports of sections 761

Staff list 769

Publications by the staff 773

INTRODUCTION

The increased demand for information on plants

and vegetation by agriculturists, biologists, palaeon-

tologists and many other users of plant information

is a vindication of the foresight shown by the Botan-

ical Research Institute which has for many years

placed a strong emphasis on taxonomic and vegeta-

tion research. The demands made for this type of

information are often in strong contrast to the lack

of acknowledgement of the importance of taxonomic

and vegetation research on the part of the users of

such information. This dualistic attitude will have to

change drastically if taxonomic and vegetation ecol-

ogy are to receive sufficient support to meet the fu-

ture demand. The development of the Flora of

Southern Africa subproject as a co-operative effort,

which is national as well as international in scope,

thus drawing in support from outside sources, has

ensured progress which otherwise would not have

been possible. Insufficient support by the state has,

however, denied the FSA the national status which,

by common consent, it merits. The creation of a sep-

arately funded national co-operative programme to

ensure the speedy completion of the FSA is long

overdue.

The continuing deterioration of the plant cover of

southern Africa is a cause of great concern. The

creation of a Vegetation Ecology Section in 1981 fol-

lowed by the establishment of a small Experimental

Ecology Section in 1983 were significant advances.

The ecologists in these research teams are continu-

ing to make a contribution towards generating the

botanical knowledge required as a basis on which

sound veld management practices can be estab-

lished.

In addition to these contributions, botanical re-

search is being advanced on a broad front but many

of the activities lack depth due to the subcritical size

of some of the research teams.

The use of computers as an aid in botanical re-

search is steadily gaining momentum. An important

product of the PRECIS data bank is the List of

Species of Southern African Plants. It comprises

about 24 000 species and infraspecific taxa and is the

first list of its kind ever compiled in this country.

Research productivity has been maintained at a

very high level as reflected by the 108 publications

produced. A milestone was reached with the publi-

cation of Bothalia 14, 3 & 4 (1983) which presents

the proceedings of the AETFAT International Con-

gress held in Pretoria in 1982. The proceeedings

cover almost 700 pages and comprise 115 contribu-

tions.

Two events worth noting, which took place during

the report year, were the Biennial General Staff

Meeting (Fig. 1) held in Pretoria from 16-18 August

1983 and a party (Fig. 2) held on 20 March 1984 to

say farewell to Cythna Letty, doyenne of South Afri-

can botanical artists, who was leaving Pretoria after

many years of residence in that city.

REPORTS OF THE SECTIONS

HERBARIUM SERVICES SECTION

The four herbaria of the Institute continued to

identify plants and provide information for a wide

range of collectors including officers of the Institute,

various State and Provincial departments. Universi-

ties and the public, both in the Republic of South

Africa and in neighbouring states.

National Herbarium, Pretoria (PRE)

The herbarium section continued to be adminis-

tered in an acting capacity by Mrs E. van Hoepen as

curator and by Miss W. G. Welman (finances).

A total of 18 994 specimens was identified. During

the year 60 loans (4 772 specimens) were sent to

other institutes and 31 loans (892 specimens) were

received. Some 29 430 duplicates were sent out and

3 611 duplicate specimens were received in ex-

change.

Collecting expeditions were undertaken during

the year mainly to expand the collections and to pro-

vide better coverage of the entire country. Some

parts, as shown by a computer print-out, had so far

hardly been visited by plant collectors. Field trips

were made to the Karoo south of Beaufort West, the

north-western Transvaal, eastern Transvaal, the Na-

tal Drakensberg, South West Africa/Namibia and to

various localities closer to Pretoria, in the latter case

to look for specific taxa. Because of the continuing

drought, only selected areas were visited.

Installaton of the modular steel cabinets to replace

old herbarium cabinets continued. A number of

cabinets for the fossil collection are still awaiting

installation, pending building alterations to the fossil

herbarium. The lichen collection has been moved to

B16, a room formerly used for housing photocopy-

ing machines. The installation of air-conditioning in

the herbarium has been promised for the near fu-

ture.

Visitors to the National Herbarium during the

year numbered about 650. Among the visitors from

other countries were Messrs B. Halliwell (collecting)

and G. LI. Lucas, from Kew; Prof. O. H. Vlok,

Wurzburg (Lichens), Prof. O. Almborn and Dr I.

Bothalia 15, 3 & 4 (1985)

FIG. 1. — Staff of the Botanical Researcn Institute on the occasion of the Biennial General Staff Meeting held from 16-18 August

1983. 1, Mr D. F. M. Venter; 2, Mr D. H. Dry; 3, Mrs E. van Hoepen; 4, Dr D. J. B. Killick (Deputy Director); 5, Dr B. de

Winter (Director); 6, Dr O. A. Leistner; 7, Dr J. C. Scheepers; 8, Dr G. E. Gibbs Russell; 9, Mr M. J. Wells; 10, Mrs S. D.

Hewitt; 11, Mr J. H. A. Jordaan; 12, Mr H. C. Taylor; 13, Dr L. E. W. Codd; 14, Dr I. C. Verdoorn; 15, Dr R. P. Ellis; 16, Mr

T. H. Arnold; 17, Mr C. Boucher; 18, Mr E. G. H. Oliver; 19, Dr P. J. Weisser; 20, Mrs L. R. Filter; 21, Mrs K. P. Clarke; 22,

Mrs M. Dednam; 23, Mrs M. P. M. C. van der Merwe; 24, Mrs B. F. Lategan; 25. Dr H. M. Anderson; 26, Mrs J. C. Mogford;

27, Mrs J. H. Jooste; 28, Mrs C. M. van Wyk; 29, Miss C. Reid; 30, Mrs M. J. A. W. Crosby; 31, Mr P. Gonsalves; 32, Miss L.

Smook: 33, Mrs T. M. Creffield: 34, Mrs M. M. Loots; 35, Mrs H. du Plessis; 36, Mrs A. M. Verhoef; 37, Mrs S. S. Brink; 38,

Mrs R. C. Holcroft; 39, Mrs D. M. C. Fourie; 40, Mrs M. Heymann; 41, Mrs B. A. Momberg; 42, Mrs J. J. van Niekerk; 43,

Mrs B. J. Vermeulen; 44, Mrs S. M. Perold; 45, Mrs B. J. Pienaar; 46, Miss W. G. Welman; 47, Dr C. F. Musil; 48, Mrs C. van

Niekerk; 49, Mrs K. J. Musil; 50, Mrs S. M. Thiart; 51, Mr B. D. Schrire; 52, Mr D. J. McDonald; 53, Mrs J. I. M. Grobler;

54. Mrs W. .1. G. Roux; 55, Mrs J. B. Hoffmann; 56, Mrs S. J. C. Burger; 57, Miss E. Retief; 58, Miss W. J. Geldenhuys; 59,

Mrs P. W. van der Helde; 60, Mrs A. C. Potgieter; 61, Miss G. C. Condy; 62, Mrs H. Joffe (Plant Protection Research

Institute); 63, Mrs J. Rautenbach; 64, Mrs J. C. P. Spangenberg; 65, Mrs A. C. Fellingham; 66, Mrs G. S. Nel; 67, Dr J. M.

Anderson; 68, Mrs I. A. Ebersohn; 69, Mr M. G. O’Callaghan; 70, Miss L. Henderson; 71. Mr A. A. Balsinhas; 72, Mr P. P.

J. Herman; 73, Mrs J. M. Mulvenna; 74, Mrs M. Jordaan; 75. Mr M. D. Panagos; 76, Mrs E. B. Evenwel; 77, Mrs H. Ebert-

sohn; 78, Dr H. F. Glen; 79, Mrs R. Botha; 80, Mr J. J. Spies; 81, Mr G. Germishuizen; 82, Mrs J. Gerke; 83, Mr R. H.

Westfall; 84, Mrs E. Potgieter; 85, Miss A. P. Backer; 86, Miss H. M. Schreiider; 87, Mr F. A. Brusse; 88, Mrs U. C. W.

Schoeman; 89. Mrs D. .1. Gerber; 90, Miss K. L. Immelman; 91. Mr H. J. de Villiers; 92, Mr G. B. Deall.

Karnefeldt, Lund (Lichens); Prof. H.-D. Ihlenfeld,

Hamburg (Mesembryanthemaceae, Pedaliaceae);

Mr B. L. Burtt, Edinburgh (various families); Dr R.

E. Magill, Missouri Botanic Gardens, St Louis

(Mosses) and Mr A. P. M. de Kruif, Wageningen

(collecting). South African visitors included staff

and students from several universities, from state de-

partments, nature conservation and Philatelic Ser-

vices. Many of them spent some time in the herbar-

ium working on various projects.

All the Agricultural Researchers in the Herbar-

ium Section were involved in checking species names

for publication of the first edition of the 'List of

Species of Southern African Plants’, which was pub-

lished in Memoirs of the Botanical Survey of South

No is in February 1 984 , and have now started

work on the second edition, which will include syno-

nyms.

Wing A: Miss C. Reid is responsible for identifica-

tion of Pteridophytes and all monocotyledons with

the exception of Poaceae. She has continued her stu-

dies on Cyperaceae. She completed a review of Cyr-

tanthus for Herbertia. Miss L. Smook, who is respon-

sible for Poaceae, has continued work on her facet

on collecting grasses to fill in gaps in our collections

from under-collected areas. She also worked on her

card indices of common names and updated names

of grasses.

Wing B : Mr G. Germishuizen, who is responsible

for this wing, and for the spirit collection housed

therein, has a special interest in Fabaceae. He was

on his own in the wing for the greater part of the

year.

Wing C: Miss E. Retief is in control of this wing

and of the fruit and seed collections. She has begun

Bothalia 15, 3 & 4 (1985)

763

FIG. 2. — Farewell party to Cythna

Letty held in the grounds of

the Pretoria National Botan-

ical Garden on 20th March

1984. Left to right: Mrs I. A.

Ebcrsohn, Mrs C. A. Bester,

Dr D. J. B. Killick (Deputy

Director), Miss M. D. Gunn

(former Librarian), Cythna

Letty (Mrs C. Forssman), Dr

L. E. W. Codd (Director,

1963-1973), Dr I. C. Ver-

doorn (former Curator of the

National Herbarium) and Dr

R. A. Dyer (Director,

1944-1963).'

working on a study of fruit, seed, and seedlings of

the family Cucurbitaceae in southern Africa. Miss

Retief serves on the Seminar Committee, which has

been responsible for very interesting lunch-hour film

and slide shows enjoyed by all.

Mr P. P. J. Herman assisted in Wing C, mainly

with identifications. He was responsible for numer-

ous translations from English to Afrikaans. During

the year he organized and carried out the sorting and

sending out of a huge backlog of duplicates which

had accumulated over five years.

Wing D\ Miss W. G. Welman continues as re-

gional extractor for Excerpta Botanica. She is in con-

trol of Wing D, and is responsible for identifications

mainly in Asteraceae. Miss Welman is working on a

publication on Karoo shrubs in the Orange Free

State region.

Mrs M. J. A. W. Crosby, who assists with identifi-

cations in this Wing, also administers the personnel

fund.

Cryptogams : Mr J. van Rooy, curator of the moss

herbarium, was on study leave during 1983 and ob-

tained his B.Sc. Hons degree. He is now working on

Bryaceae for Flora of Southern Africa.

Mrs S. M. Perold continued her work on Riccia-

ceae which involves a great deal of SEM work. She

remained in charge of the SEM, attended two three-

day courses on SEM maintenance and optimizing its

performance. During the year, she produced 2 912

SEM micrographs for various staff members. In ad-

dition, she prepared 599 slides for Dr R. E. Magill’s

work on Meteoriaceae.

Mr F. A. Brusse, curator of the lichen herbarium,

continued to expand the collection, which is now

housed in B16 and expanded into three extra cabi-

nets. He has prepared three papers on a new species

in Parmelia and one in Toninia and on a new genus

Corynecystis, all of which are in the press.

Service Room : The herbarium service room,

through which all specimens handled in the herbar-

ium wings pass, continues to be administered by Mrs

M. Dednam. A microwave oven has been installed

for quick sterilization of specimens needing urgent

identification.

Natal Herbarium , Durban (NH)

A total of 3 858 specimens was identified, and 450

visitors and 8 student groups were received. New ac-

cessions to the herbarium numbered 2 876 and 93

specimens were sent out on loan.

Mr B. D. Schrire, the officer in charge of the Unit,

has continued his research on Desmodieae (Faba-

ceae). Mrs M. Jordaan, Senior Technician, deals

with the majority of identifications. She was assisted

by Dr V. G. Coetzee, who retired at the end of

March 1984.

Renovation of the buildings was completed and a

highly informative and attractive Museum was de-

veloped in the front building which used to be the

private house of Dr J. Medley Wood. Items on the

history of botanical exploration in Natal as well as on

past curators of the Herbarium and Durban Botanic

Gardens are displayed together with old furniture

and paintings.

A species list of all the plants in the Herbarium

garden has also been completed.

Albany Museum , Grahamstown (GRA)

1 872 specimens were identified and 908 visitors

and 6 groups (120 persons) dealt with. Accessions

totalled 1 162, and 265 specimens were sent out on

loan.

Mrs E. Brink, who is in charge of the herbarium

and information service, is assisted by Dr A. F. M.

G. Jacot Guillarmod, who works part-time.

Miss G. V. Britten retired at the end of February

after nearly 63 years of service, and Mrs M. Furlong

has been appointed in the technical assistant’s post.

The Botanical Research Unit in Grahamstown is

housed in the Albany Museum and deals with more

764

Bothalia 15, 3 & 4 (1985)

visitors than either of the other units. It is used as a

reference centre for visiting botanists and other sci-

entists not only from the Republic of South Africa,

but also from other African countries as well as from

overseas.

While building extensions are in progress, the en-

tire herbarium has been housed in temporary quar-

ters. The move to the new, permanent wing is antici-

pated to take place before the end of 1984. Half the

existing garden around the Museum had to be

cleared to make way for the new extensions, and liv-

ing specimens of only the rarest and most valuable

plants have been kept.

During the year six articles were published; a

paper was read at a symposium Towards an en-

vironmental plan for the southern Cape’; the staff

participated in checklisting for the fynbos Biome

project, in research on famine relief in the Transkei,

and in establishing the status of endangered species.

Government Herbarium , Stellenbosch (STE)

A total of 2 846 specimens was identified, 332 visi-

tors were dealt with, accessions to the herbarium

numbered 7 180 and 5 062 specimens were sent out

on loan.

Mrs L. van Zyl, who was in charge of the herbar-

ium, resigned at the end of December 1983. Mrs C.

M. van Wyk and Mrs A. C. Fellingham assisted with

identifications and general herbarium work.

Miss W. J. Geldenhuys, technical assistant, was

transferred to Pretoria in February 1984, where she

is now working in the personnel section of Adminis-

tration. Her place has been taken by Miss J. Fourie.

Members of the Unit staff undertook several col-

lecting trips and a number of valuable specimens

were collected.

FLORA RESEARCH SECTION

Flora of Southern Africa (FSA)

Thanks to support from the Department in the

form of a research contract and a fellowship the

Flora of Southern Africa subproject was further pro-

moted both nationally and internationally. The in-

itiation of a volume on lichens deserves special men-

tion in this context.

A special session on the FSA was held during the

congress of the South African Association of Botan-

ists in January 1984 and news on the FSA was in-

cluded in Forum Botanicum, the newsletter of the

Association.

Three fascicles were published: (1) Vol. 7,2,2 (M.

P. de Vos) dealing with Syringodea and Romulea of

the Iridaceae, (2) Vol. 21,1 (late Prof. H. Wild) on

Tiliaceae, and (3) Vol. 33,7,2 (Prof. O. M. Hilliard)

dealing with part of the Gnaphaliinae of the Astera-

ceae. Two fascicles are in press and will be published

during 1985: (1) Vol. 4,2 dealing with Xyridaceae to

Juncaceac and (2) Vol. 28,4 on Lamiaceae. The vol-

ume on Pteridophyta by Prof. E. A. C. L. E.

Schelpe and Vol. 14 on Crassulaceae by Dr H. R.

Tblken arc at an advanced stage of editing and will

go to press during 1984.

A catalogue of South African green, brown and

red marine algae, compiled by Prof. S. C. Seagrief of

Rhodes University in Grahamstown, went to press

and will be published soon.

Institute members reported as follows on progress

made with research fascicles on volumes of the FSA :

Lichens: A volume on lichens is being planned. The

project is co-ordinated by Dr O. Almborn of

Lund, who worked in South Africa as a research

fellow of the Department. Mr F. A. Brusse of

the Institute and more than 20 overseas re-

searchers are co-operating on the project.

Bryophyta: The major part of fascicle 2, covering

Funariales and Bryales was completed by Dr R.

E. Magill before he returned to the Missouri Bo-

tanical Garden. Work on the outstanding family

Bryaceae was continued by Mr J. van Rooy and

should be completed soon. Work on fascicle 3 is

also in progress.

Vol. 2.: Poaceae. It was agreed in principle that in a

few large families, such as the grasses, it would

be permissible to publish Prodromus or Con-

spectus volumes in advance of the full Flora ac-

count.

Register of names and types for Poaceae: Photoco-

pies of original descriptions of about 1 200 grass

names were pasted on A4 paper and filed. It was

found that in Poaceae (as in Mesembryanthema-

ceae) there are about four times as many names

in literature as there are accepted taxa.

Oryzoideae, Centostecoideae and Bambusoideae:

Dr G. E. Gibbs Russell divided the genus Ehr-

harta into seven groups and the vegetative and

spikelet characters of each group were tabu-

lated. The treatment of species with bulbous

bases is largely complete.

Vol. 4: Restionaceae. The generic revision dealing

with the 19 genera distinguished in the family

has been completed by Dr H. P. Linder and has

been submitted to Bothalia. The paper includes

chapters on palynology, culm anatomy, morpho-

logy, phytochemistry and testa morphology. A

conspectus listing about 320 species and dealing

with nomenclature, typification, keys, new de-

scriptions and new combinations is near comple-

tion. Descriptions of 55 new species were pre-

pared and 83 new combinations were made.

Xyridaceae, Eriocaulaceae, Commelinaceae, Ponte-

deriaceae and Juncaceae: Fascicle 4,2 dealing

with these families, and prepared largely by

Mrs A. A. Mauve, has gone to press.

Vol. 5: Asparagaceae. A revision of Myrsiphyllum ,

dealing with 12 species, was prepared by Mrs

Mauve and sent to press. Work on Protaspara-

gus was largely completed by the same re-

searcher. The genus is divided into two subge-

nera and in southern Africa 64 species were

found of which 17 are new.

Vol. 8: Orchidaceae. The genus Disa , revised by Dr

Linder, was adapted to Flora format.

Vol. 11: Mesembryanthemaceae. Draft revisions of

the minor genera Astridia, Acrodon, Bergeran-

Bothalia 15, 3 & 4 (1985)

765

thus, Disphyma, Carruanthus, Cerochlamys,

Eberlanzia and Rhomb ophy llum were drawn up

by Dr H. F. Glen.

Vol. 16: Fabaceae-Desmodieae. An account of this

group is being written up in thesis form by Mr B.

D. Schrire.

Vol. 21: Tiliaceae. The revision compiled by the late

Prof. H. Wild and updated by Dr L. E. Codd

was published as fascicle 21,1.

Vol. 25: Ericaceae. Mr E. G. H. Oliver completed

the revision of Scyphogyne in which he recog-

nizes 11 species. Ericinella with three species in

the Cape outside Capensis and one in Malawi

was finalized. Philippia was also finalized and a

start was made on the revision of Salixis and

Coccosperma. Some work was done on Erica.

Vol. 28: Lamiaceae. An account of the family, writ-

ten by Dr Codd, was sent to press as fascicle

28,4. Leonotis was contributed by Mr M. Iwars-

son of Uppsala.

Vol. 30: Acanthacea t-Justicia. The final write-up of

the genus was completed by Miss K. L. Immel-

man and submitted to the editor. Twenty-two

species and six subspecies are distinguished in

Southern Africa. Siphonoglossa and Aulojusti-

cia , two small closely related genera, have also

been investigated and it is planned that they, to-

gether with Justicia and Monechma (recently

completed by Mrs J. Munday of Witwatersrand

University), form a fascicle of the Flora.

Ceropegia and related genera

A semi-popular account of Ceropegia,

Brachystelma and Riocreuxia by Dr R. A. Dyer was

published by A. A. Balkema.

History of plant collecting

Dr Codd has collected biographical notes on a

total of 44 collectors whose names appeared in Bo-

tanical Exploration of Southern Africa without sup-

porting information. In addition, 27 biographies

have been gathered of collectors not originally in-

cluded in the book. It is proposed to publish this sup-

plementary information in Bothalia.

Pretoria Flora

A further 93 pages (57 English, 30 Afrikaans) of

camera-ready copy were produced at the Institute.

Some 50 typeset pages were sent to the Language

Bureau for translation into Afrikaans.

Palaeoflora of Southern Africa

The first volume on the Molteno Formation, writ-

ten by Drs J. M. and H. M. Anderson, which deals

with the ubiquitous gymnosperm genus Dicroidium,

was published by A. A. Balkema. Volume 2 on the

other gymnosperms of the formation is nearing com-

pletion. A companion volume, Prodromus of South-

ern African fossil floras (Devonian to Lower Creta-

ceous), will go to the printers shortly.

Liaison Officer, Kew

The present incumbent, Dr Linder, furnished in-

formation on taxonomic and related subjects to re-

searchers in southern Africa and overseas. His re-

search centred mainly on the Restionaceae.

PLANT STRUCTURE AND FUNCTION SECTION

This section is still accommodated in Velcich

House and the laboratories are functioning effi-

ciently despite early problems. Mrs R. Botha left us

during October 1983 after completing her studies on

applied grass leaf anatomy. She is to be replaced by

Mr P. P. J. Herman who is to concentrate on wood

anatomy. Miss A. Alberts successfully completed

her National Diploma in Agriculture.

Comparative grass leaf anatomy

During August 1983, Dr T. R. Soderstrom of the

Smithsonian Institution, Washington spent three

weeks with Dr R. P. Ellis discussing and evaluating

their joint work on bamboo leaf anatomy. Over 275

bamboo taxa have been studied anatomically and

the indications from this study are that the classifica-

tion of the Bambusoideae is in need of revision.

Anatomical criteria appear to be very useful at the

tribal and sub-tribal levels but do not yet contribute

significantly to generic delimitation as the sample is

still too small.

Grass identification by epidermal structure for her-

bivore food preference studies

Mrs Botha has completed her studies on the epi-

dermal variation of three widely distributed South

African grasses. Her results were-submitted as a the-

sis for which she was awarded an M.Sc. degree by

the University of Pretoria. This work clearly demon-

strated that the abaxial leaf blade epidermal features

are not representative of the epidermis of the whole

plant and that the adaxial leaf blade epidermis, the

leaf sheath and culm epidermis differ considerably

from the abaxial leaf epidermis which is usually used

as a reference. This observation, therefore, ques-

tions one of the basic assumptions for the technique

of microscopic faecal analysis.

Cytogenetic studies

The cytotaxonomic study of Lantana camara has

been completed by Mr J. J. Spies. The results of this

study show that this species is still actively evolving

and it was found to be impracticable to define infras-

pecific taxa. Any such classification will be artificial

with continual hybridization occurring between the

groups. A group of plants was identified which are

more advanced and may possibly represent the basis

for the formation of subspecific taxa.

Mrs H. du Plessis is now giving her attention to a

similar study of Rubus. Chromosome numbers of 11

of the 17 species, found in South Africa, have been

determined. Five of these species have been found

to be polyploids. This high incidence of polyploidy

probably results from hybridization.

Mr Spies and Miss Alberts have just commenced a

study of the chromosome numbers of the Poaceae

Bothalia 15, 3 & 4 (1985)

766

and have tested various techniques. Seed of most of

the species of Sorghum has been received and a cyto-

taxonomic study of this genus will be undertaken.

Co-operation with the Institute for Ecological Re-

search (Potchefstroom University for C.H.E. and

the National Transport Commission) continues with

the study of reproduction of Eragrostis curvula. Mrs

J. C. P. Spangenberg has been seconded to the Bo-

tanical Research Institute to assist with this work.

DATA SUB-SECTION

Data Sub-Section serves the needs of the Institute

for electronic data-processing, using the Burroughs

7800 computer of the Department of Agriculture

and a Hewlett-Packard 9845B microcomputer be-

longing to the Institute. Dr G. E. Gibbs Russell is

Data Officer and is responsible for liaison between

Data and the needs of individual Sections of the In-

stitute, as well as for broad applications of the Her-

barium database, PRECIS. The datametrician, Mr

N. P. Barker, is seconded to the Institute from Bi-

ometry and Datametrical Services. He supervises

the entry of new specimens into PRECIS, the cor-

rect running of all the PRECIS programs, and pro-

vides information from the system, both for routine

Herbarium applications and for special research re-

quests. Mrs J. C. Mogford is in charge of quality

control for PRECIS, in both newly-entered and in

backlog specimens. Encoding and punching new

specimen informatiorf and updates to existing speci-

mens is done by Mrs J. H. Jooste, Mrs E. B. Even-

wel and Mrs H. P. van der Westhuizen.

VEGETATION ECOLOGY SECTION

The most significant recent development concern-

ing ecological research has been the division of the

former Ecology Section into two Sections, viz the

Vegetation Ecology Section and the Experimental

Ecology Section. The activities of these two sections

are, accordingly, reported on separately.

Transvaal bushveld studies

The full-scale study of the plant ecology of the

Waterberg bushveld by Mr R. H. Westfall is in the

planning and reconnaissance phase. Messrs West-

fall, M. D. Panagos and J. F. van Blerk have devel-

oped a method for field identification of taxa. Com-

puterized field-data capture has also been developed

and is nearly operational. A technique being tested

for pattern recognition in vegetation-enhanced satel-

lite imagery shows promise for the Waterberg area,

which should greatly facilitate stratification of the

area for sampling.

Transvaal forest survey

Mr G. B. Deall has refined the results of his study

in the Sabie area. He has distinguished a total of 62

syntaxa (plant communities and variants) occurring

throughout eight broad physiographic zones. These

syntaxa are grouped into four ecological-formation

classes on the basis of physiognomic and ecological

criteria, thus:

(a) Forest and mesic Thicket of the Mistbelt and

Low Country (20 syntaxa);

(b) Woodland and xeric Thicket of the Low Coun-

try (22 syntaxa);

(c) Woodland of the Humid Mistbelt (7 syntaxa);

and

(d) Grassland of the Humid Mistbelt (13 syntaxa).

Coastal studies

The suitability of aerial photographs for monitor-

ing emergent and submerged macrophytes has been

studied by Dr P. J. Weisser, using material from the

Wilderness Lakes. He has also revised the manu-

script on conservation priorities in the dune area be-

tween Richards Bay and Mlalazi Estuary which is

nearing completion for publication. He completed a

reconnaissance of conservation priorities in Reserve

7, Zululand (between Tugela River and Mlalazi Est-

uary) and has submitted his report to the Committee

for Controlled Utilization of Natural Resources (Zu-

luland Coast).

Data collected on the Cape south and west coasts

have been prepared for a book on Dry Coastal

Ecosystems in the Ecosystems of the World series.

In one chapter, Messrs H. C. Taylor and C. Boucher

describe the main communities of the dunes and

cliffs between Cape Town and Port Elizabeth, dis-

cuss their relationships and summarize information

on exploitation and conservation of the coastal

ecosystems. In another chapter by Mr Boucher and

Miss A. le Roux, the strand zone plant communities

found along the Cape coast between Algoa Bay and

the Orange River mouth have been described and

their distributions have been mapped. The com-

munities form natural groups which can be corre-

lated with major inland vegetation boundaries, de-

spite the expected overriding influence of the close

proximity to the sea.

During this year, Mr M. G. O’Callaghan under-

took field trips to eight estuarine systems and pub-

lished articles on three systems under the auspices of

the Estuarine and Coastal Research Unit (CSIR).

He also conducted much field work at Kleinmond,

sampling the aquatic and semi-aquatic vegetation.

He is preparing two articles discussing the vegetation

of the rivers entering False Bay.

Cape fynbos studies

(a) Vegetation survey of the Cape of Good Hope

Nature Reserve

The data gathered in monitoring the encroach-

ment of invasive alien woody plants over the period

1966-1980, have been analysed and written up as

two papers by Mr Taylor and Mr and Mrs I. A. W.

MacDonald. In 1966, three Australian acacias (Aca-

cia cyclops, A. saligna and A. longi folia) and Pinus

pinaster were the most serious invaders. In

1976-1980 P. pinaster, which had been selectively

cleared in the interim, was less serious, but the other

invaders had increased both their densities and dis-

tribution ranges.

(b) The vegetation of the Cedarberg

Mr Taylor started work on an ecological study of

the mountain fynbos vegetation of the Cedarberg

State Forest and the adjoining Mountain Catchment

Bothalia 15, 3 & 4 (,1985)

767

Areas at a semi-detailed scale to establish the practi-

cal significance of these communities.

(c) Structural classification of mountain fynbos

Mr B. M. Campbell is finalizing the research facet

to develop a structural-functional classification of

mountain fynbos. A series of research papers ema-

nating from this work has been published during the

recent past. The definitive publication is being pre-

pared as a Botanical Survey Memoir.

(d) The vegetation of Swartboschkloof, Jonkershoek

Mr D. J. MacDonald has completed his study of

the fynbos and forest communities of Swartbosch-

kloof. He used the TABSORT computer pro-

gramme package to analyse the data. Sixteen Moun-

tain Fynbos communities, grouped into three

groups, and five Forest communities grouped into

two groups have been identified. The data of a pre-

vious study of the same area have been re-analysed

and interpreted in the context of this study. Two

publications on this work are being prepared.

(e) A semi-detailed regional ecological study of the

western lowland fynbos

Mr Boucher has analysed and mapped the plant

communities found along east-west aligned transects

through the western Cape coastal foreland on 46

maps with a scale of 1: 10 000. These maps have

been reduced to a scale of 1: 50 000 for comparison

with the land-type survey maps of the area, when

these become available during 1984.

Monitoring vegetation change

Long-term monitoring of bushveld vegetation,

under various treatments, has begun at the Matimba

Power Station, near Ellisras in the north-western

Transvaal, where a 4 ha plot has been set aside for

observation. Two 10 m x 10 m permanent plots have

been selected (one in the open, the other in more

dense vegetation) and surveyed by Messrs Panagos

and Westfall using 1 m: quadrats.

Misses A. Stadler and A. P. Backer are monitor-

ing vegetation of the Siyaya Catchment near Mtun-

zini. Natal, by marking 392 trees and laying out 21

permanent plots. After one year of monitoring

about 15% of the marked trees had died, mainly be-

cause of the effects of the cyclones Demoina and Im-

boia. Tree growth rates over the last year range from

0,28 to 2,06 m in height.

EXPERIMENTAL ECOLOGY SECTION

In 1983, the Experimental Ecology Group was

formally recognized as a fully-fledged section with a

new enlarged post structure. The section, under Dr

M. C. Rutherford, is consequently expanding its re-

search activities and several new research facets are

in the advanced planning stage. For one of these

facets, a Plant Exploration Section’s post is being

fully utilized within the Experimental Ecology Sec-

tion in an area of common research interest. Re-

search results for the reporting period are limited to

those from four currently registered facets, grouped

as follows:

Plant production studies in savanna

Research on primary production within the South

African Savanna Ecosystem Project has led to an in-

vestigation of how to extrapolate production data

from primary data sites. Dr M. C. Rutherford has

been testing, in the north-western Transvaal, a re-

gional savanna productivity model which predicts

mean annual production of savanna components

under both natural conditions and radical land use

changes. Main modifications to the model have been

application of improved soil factors. Current model

resolution is adequate to express spatial variation of

production on a medium mapping scale.

Ecophysiological research in fynbos

A series of experiments has been undertaken by

Miss F. M. Pressinger to optimize germination in

Protea repens for use in competition experiments.

Cold treatment prior to sowing resulted in increased

and earlier germination, whereas dilute solution of

hydrogen peroxide caused earlier but reduced num-

ber of germinations. Experimental measures of com-

petition between neighbouring Acacia saligna seed-

lings showed that water stress and high mortalities

occur in seedlings less than four days old. Biomass

and leaf area of older seedlings were greatest and

water potential least negative at low population den-

sities combined with relatively high water additions.

Aquatic ecology

Dr C. F. Musil’s work on the classification of the

water plant communities of Natal has been com-

pleted and the account of the results is in press. His

major work on growth of Eichhornia crassipes has

also been completed. Four papers dealing with a mo-

del for predicting growth of E. crassipes in eutro-

phied aquatic systems have been submitted for pub-

lication. These concern:

(a) generating kinetic coefficients for the model

in green house culture;

(b) testing and refining the model under field con-

ditions;

(c) testing a model for predicting growth rates

from plant nutrient concentrations; and

(d) applying the model to a eutrophied South

African impoundment.

PLANT EXPLORATION SECTION

The section, under Mr M. J. Wells has continued

to concentrate its attention on food plant research

and on related aspects such as the conservation of

germ plasm. Good progress was made on the devel-

oping food plants projects, as well as on projects that

are due for completion or are being phased out i.e.

work on barrier plants, timber sources and weeds.

Taxonomy of Ficinia

Several new species of Ficinia have been distin-

guished and are being written up by Mr T. H. Ar-

nold.

National Weed list

Extra categories and information, including extra

common names, have been added to the list. Check-

Bothalia 15, 3 & 4 (1985)

768

inn and preparation for publication by Mrs H. Joffe

and Mr Wells is almost complete. Three papers

analvsinn the weed list have been published [Botha-

lia 14,3 & 4: 945-948; 961-966; 967-970 (1983)].

Timber use

The results of a survey by Miss C. A. Liengme of

timber use by the Tsonga of Gazankulu have been

published [Bothalia 14,2: 245-258 (1983)].

Ethnobotany

A survey by Miss Liengme of ethnobotanical re-

search in southern Africa, including an extensive

bibliography, has been published [Bothalia 14,3 & 4:

621-630 (1983)]. It shows that the peoples whose

plant uses are least well known are the: Swazi,

Transvaal Ndebele, Mpukushu and Khoi. The plant

aspect most in need of study is wood use.

Barrier plants

Information-gathering on 220 priority indigenous

barrier species was completed. A paper by Miss L.

Henderson on barrier plants in South Africa has

been published [Bothalia 14,3 & 4: 635-640 (1983)]

and a more comprehensive work detailing the

characteristics of 504 garden hedge, security barrier

and windbreak plants is in preparation.

Conservation of germ plasm

Although the drought curtailed collecting, plant-

ings were made of Citrullus lanatus and Sorghum bi-

color, and these yielded 624 seed collections that

were either stored in the seed bank and/or distrib-

uted to other researchers. This work was carried out

by Mrs K. J. Musil and Mrs J. Hoffman under the

guidance of Mr Arnold.

Crop plants of African origin

The severe drought precluded field work, but de-

tailed analysis of the material already collected was

continued, and is already enabling Mr Arnold to cat-

egorize and distinguish races of crop plants with

greater ease and certainty.

Indigenous food plants

Data on a further 516 species from 58 families, has

been added to the national food-plant data bank by

Mr A. A. Balsinhas. Particular attention was paid

to Khoisan (Hottentot and Bushman) food-plants,

which are the subject of a paper being prepared by

Mr Arnold and Mr Wells for the Kew International

Conference of Economic Plants for Arid Lands

(KICEPAL), London, 23-27 July 1984.

Water conservation gardening

A report was produced by Mrs D. M. C. Fourie on

water conservation gardening, listing both exotic

and indigenous species of trees, shrubs, climbers,

bedding plants and ground covers that are hardy and

not water-thirsty. The public demand for the report

was so great that it is now being produced as a

departmental publication.

PRETORIA NATIONAL BOTANICAL GARDEN

After six years without a Curator we are glad to

report that Mr D. H. Dry has assumed responsibility

for the Garden.

Despite the hampering effects of the drought,

good progress has been made. 737 new plant collec-

tions were received and accessioned by the records

team, including a group of South West African

plants brought in by Mr D. S. Hardy. Some of these

went directly to the ‘rare and endangered’ house. A

nature trail was constructed along the ridge with

vantage-points overlooking various biome areas.

The ground in front of the main Reynolds Gate was

landscaped and a fine large specimen of Aloe baine-

sii established there. 220 Proteaceae were planted to

improve the fynbos, and new beds created round the

dam in the south-west corner. A start was made with

developing a watercourse down the koppie to the

lowveld section.

It is with deep regret that we record the death of

Mr J. Admiraal. He was in charge of the Garden

from 1952 to 1977, being appointed Curator in 1965,

and was responsible for the development of the Pre-

toria National Botanical Garden during most of its

formative stages.

PUBLICATIONS

During the year the following were published: Bo-

thalia 14,2 and 14,3 & 4 (Proceedings of the 1982

AETFAT Congress), Flora of Southern Africa 21,1

and 33,7,2; Flowering Plants of Africa/ Blomplante

van Afrika 47,3 & 4; Memoirs of the Botanical Sur-

vey of South Africa 48; Palaeoflora of Southern

Africa, Molteno Formation 1; and Dr Dyer’s Cero-

pegia, Brachystelma and Riocreuxia. The latter two

works were published by the private sector. Alto-

gether, 108 papers were published by staff in Insti-

tute, local and overseas journals.

Bothalia 15, 3 & 4 (1985)

769

BOTANICAL RESEARCH INSTITUTE

Scientific, Technical and Administrative Staff

(31st March 1984)

Director

B. de Winter, M.Sc., D.Sc. (Taxonomy of Poaceae,

especially Eragrostis and of Hermannia; plant ge-

ography)

Deputy Director

D. J. B. Killick, M.Sc., Ph.D., F.L.S. (General tax-

onomy, nomenclature, mountain ecology and

editing)

Assistant Director

D. Edwards, M.Sc., Ph.D. (Ecological methodo-

logy, aquatic plants, remote sensing, vegetation

structure and physiognomy)

ADMINISTRATION

Senior Provisioning Ad-

ministration Officer ....

State Accountant

Provisioning Administra-

tion Officer

Senior Provisioning Ad-

ministration Clerks

Provisioning Administra-

tion Clerk

Personal Secretary to

Director

Senior Registration

Clerks

Accounting Clerk

Receptionist

Typists

Senior Technician: Edi-

torial Assistant to

Deputy Director

D. F. M. Venter

Mrs J. Rautenbach

Mrs D. J. Gerber

Miss W. J. Geldenhuys

Mrs S. Swanepoel

Mrs I. A. Ebersohn

Mrs C. van Niekerk

Mrs M. M. Loots

Mrs T. M. Creffield

Mrs I. J. Joubert*

Mrs C. A Bester*

Mrs H. S. Hacker

Mrs S. S. Brink

Mrs J. Gerke*

Mrs S. M. Thiart*

Mrs M. P. M. C. van

der Merwe*

Mrs J. M. Mulvenna

Mrs B. A. Momberg,

B.Sc.*

HERBARIUM SERVICES SECTION

Officer-in-Charge

(Acting) Mrs E. van Hoepen.

M.Sc.

National Herbarium, Pretoria (PRE)

Senior Agricultural

Researcher Mrs E. van Hoepen,

M.Sc. (Acting Cura-

tor; Supervision of

identifications and

enquiries)

*Half-day

Wing A (Pteridophytes-Monocotyledons)

Agricultural Researcher Miss C. Reid, B.Sc.

Hons (Cyperaceae)

Principal Research

Technician Miss L. Smook, B.Sc.

(Poaceae)

Technical Assistant Mrs S. J. C. Burger

Wing B (Piperaceae-Oxalidaceae)

Senior Agricultural

Researcher G. Germishuizen,

M.Sc. (Polygona-

ceae)

Agricultural Researcher Mrs P. M. Olivier,

B.Sc. Hons (Until

June 1983)

Technical Assistant Mrs H. H. Kotze*

(Until July 1983)

Mrs A. C. Potgieter*

(From Aug. 1983)

Wing C (Linaceae-Asclepiadaceae)

Senior Agricultural

Researcher

Agricultural Researcher

Technical Assistants

Miss E. Retief M.Sc.

P. P. J. Herman, M.Sc.

Mrs J. I. M. Grobler*

Mrs M. Heymann*

Wing D (Convolvulaceae-Asteraceae)

Senior Agricultural

Researcher

Senior Research

Technician

Technical Assistants

Cryptogamic Herbarium

Assistant Agricultural

Researchers

Technical Assistants

Services

Principal Research

Technician

Technical Assistant

Miss W. G. Welman,

M.Sc.

Mrs M. J. A. W.

Crosby B.Sc.*

Mrs R. Loggenberg*

(Until July 1983)

Mrs D. C. Bothma

(From Nov. 1983

until April 1984)

F. A. Brusse, M.Sc.

(Lichens)

J. van Rooy, B.Sc.

Hons (Musci)

Mrs L. R. Filter*

Mrs P. W. van der

Helde

Mrs S. M. Perold,

B.Sc.* (S.E.M. and

laboratory work,

Ricciaceae)

Mrs M. Dednam*

(Herbarium records,

loans and ex-

changes, etc.)

Bothalia 15, 3 & 4 (1985)

Typist Mrs A. M. Verhoef

Laboratory Assistants J. Phahla (Mounting)

C. Letsoala (From

Sept. 1983)

G. Lephaka (Prepara-

tion and packing)

Natal Herbarium. Durban (NH)

Assistant Agricultural

Researcher

Principal Research

Technician

Senior Research

Technician

Clerical Assistants

Laboratory Assistants

B. D. Schrire, B.Sc.

Hons (Curator, Fa-

baceae and general

identifications)

Mrs V. G. Coetzee,

Ph.D.*

Mrs M. Jordaan,

B.Sc.*

Mrs M. K. Lynch*

(Until June 1983)

Mrs H. E. Noble*

(From July 1983)

C. Buthelezi

A. Ngwenya

Albany Museum Herbarium, Grahamstown

(GRA)

Agricultural Researcher

Assistant Agricultural

Researcher

Technical Assistant

Laboratory Assistant

Mrs E. Brink, B.Sc.

(Curator, general

identifications)

Mrs A. F. M. G. Jacot

Guillarmod D.Sc.*

Miss G. V. Britten*

(Until Feb. 1984)

Mrs M. Furlong*

(From March 1984)

A. Booi

Government Herbarium, Stellenbosch (STE)

Agricultural Researchers

Principal Research

Technician

Technical Assistant

Clerical Assistant

Laboratory Assistant

Mrs L. van Zyl, M.Sc.

(Curator, general

identifications; until

Dec. 1983)

Mrs C. M. van Wyk,

M.Sc.

Mrs A. C. Fellingham,

B.Sc.

Miss W. J. Geldenhuys

(Until Jan. 1984)

Miss J. Fourie

C. Paulse

J. Leith (From July

1983)

FLORA RESEARCH SECTION

Officer-in-Charge O. A. Leistner, M.Sc.

D.Sc., F.L.S.

Flora of Southern Africa Team

Assistant Director O. A. Leistner, M.Sc.

D.Sc., F.L.S. (Gen-

eral taxonomy)

Senior Agricultural

Researchers

Agricultural Researchers

Research Technician

Graphic Artists

Technical Assistant

E. G. H. Oliver, M.Sc.

(Taxonomy of Eri-

caceae)

J . M . Anderson ,

M.Sc., Ph.D. (Pala-

eobotany, plant

geography)

H. F. Glen, M.Sc.,

Ph.D., F.L.S. (Nu-

merical taxonomy,

Mesembryanthema-

ceae)

G. E. Gibbs Russell,

B.S., Ph.D. (Taxo-

nomy, especially

grasses; compute-

rized data proces-

sing)

Miss K. L. Immel-

man, M.Sc. (Taxo-

nomy, especially

Justicia)

H. P. Linder, B.Sc.,

Ph.D. (Liaison Offi-

cer, Kew; taxonomy

especially Orchida-

ceae and Restiona-

ceae)

L. E. W. Codd,

M.Sc., D.Sc. (Taxo-

nomy, especially La-

miaceae; history of

plant collecting)

Mrs A. A. Mauve,

M.Sc. (Taxonomy,

especially Monoco-

tyledons)

Mrs H. M. Ander-

son, M.Sc. Ph.D.*

(Palaeobotany)

Mrs R. C. Holcroft*

Miss G. C. Condy,

M. A.

Mrs W. J. G. Roux*

PLANT STRUCTURE AND FUNCTION

SECTION

Officer-in-Charge R. P. Ellis, M.Sc.,

D.Sc.

COMPARATIVE PLANT ANATOMY

Assistant Director

Agricultural Researcher

Technical Assistant

R. P. Ellis, M.Sc.,

D.Sc. (Anatomy of

South African

grasses)

Mrs R. Botha, M.Sc.

(Applied grass anat-

omy; until Novem-

ber 1983)

Mrs H. Ebertsohn

(Microtechnique)

Bothalia 15, 3 & 4 (1985)

771

Cytogenetics VEGETATION ECOLOGY SECTION

Agricultural Researcher

Research Technicians

Senior Librarian.

Librarian

Library Assistant

Data Officer

Datametrician....

Research Technician

Technical Assistants.

J. J. Spies, M.Sc. (Cy-

togenetics of Lan-

tana, Rubus and

grasses)

Mrs H. du Plessis,

B.Sc. (Microtechni-

que of Lantana and

Rubus)

Miss A. Alberts, Nat.

Dip. Agric. (Cyto-

genetics of grasses)

Mrs J. C. P. Spangen-

berg, B.Sc.** (Em-

bryo sac studies of

Eragrostis curvula )

Mrs E. Potgieter

B. Libr.t

Mrs U. M. Carte B.A.,

H.C.D.L., Lib. Dip.

C. T.t

Mrs B. F. Lategant

G. E. Gibbs Russell,

B.S., Ph.D. (Regis-

ters of scientific

names, regional

phytogeography)

N. P. Barker, B.Sc.

Hons+ (Database

manager for PRE-

CIS, system man-

ager for Hewlett

Packard, computer

taxonomy, pollina-

tion ecology)

Mrs J. C. Mogford,

B.Sc. Hons (Quality

control of

PRECIS)*

Mrs J. H. Jooste (Chief

encoder for

PRECIS)

Mrs E. B. Evenwel

(Quality control and

update encoding for

PRECIS)*

Mrs H. P. van der

Westhuizen (Data-

Capture and encod-

ing for PRECIS)*

Officer-in-Charge

Assistant Director

Senior Agricultural

Researchers

Agricultural Researchers

Assistant Agricultural

Researchers

J. C. Scheepers, M.Sc.,

D.Sc.

J. C. Scheepers, M.Sc.,

D.Sc. (Vegetation

ecology, especially

of forest/woodland/

grassland relation-

ships; conservation

and land-use plan-

ning; phytogeogra-

phy)

C. Boucher, M.Sc.

(O/C Botanical Re-

search Unit, Stellen-

bosch; lowland fyn-

bos ecology and

phytosociology; con-

servation and land-

use planning;

Braun-Blanquet ap-

proach and tech-

niques)

H. C. Taylor, M.Sc.

(Mountain fynbos

and forest ecology;

Braun-Blanquet ap-

proach and tech-

niques; conserva-

tion)

P. J. Weisser, Ph.D.

(Air-photo interpre-

tation and mapping;

reedswamp ecology;

Zululand coast dune

vegetation; conser-

vation)

D. J. McDonald,

M.Sc. (Mountain

fynbos ecology and

phytosociology ;

Braun-Blanquet ap-

proach and tech-

niques)

R. H. Westfall, M.Sc.

(Ecology and phyto-

sociology of Trans-

vaal bushveld; eco-

logical data and lit-

erature storage, re-

trieval and proces-

sing; syntaxonomic

nomenclature)

G. B. Deall, B.Sc.

Hons (Vegetation

ecology of forest/-

woodland/grassland

interrelationships)

M. G. O’Callaghan,

B.Sc. Hons (Estua-

rine ecology and

phytosociology)

** National Transport Commission, Institute for Ecological Re-

search, Potchefstroom University for C.H.E.

t Library Services, Department of National Education.

+ Datametrics

Photographic Services

Photographer Mrs A. J. Romanowski

Mary Gunn Library

Data Subsection

Bothalia 15, 3 & 4 (1985)

Research Technicians

Miss A. P. Backer,

B.Sc. (Ecological

data processing and

presentation; eco-

logical literature;

nature conservation;

air-photo interpreta-

tion and cartogra-

phy)

Miss M. Morley, B.Sc.

Agric. (Ecological

data processing and

presentation; eco-

logical literature; es-

tuarine and fynbos

vegetation; air-

photo interpretation

and cartography)

M. D. Panagos, N.

Dipl. Agric. (Bot.

Res.). (Computer

science; data proces-

sing; sampling and

monitoring vegeta-

tion and environ-

ment)

C. W. Ries, B.Sc.

(Bushveld and grass-

land ecology; eco-

logical literature;

pasture science; ve-

getation/substratum

relationships)

Mrs J. Schaap,

H.P.E.D.

(Draughtsmanship

and cartography;

artwork, layout and

design)

Miss A. Stadler, B.Sc.

(Ecological data

processing and pres-

entation; ecological

literature; nature

conservation; air

photo interpretation

and cartography)

J. F. van Blerk, B.Sc.

(Grassland ecology;

ecological literature;

pasture science)

Mrs B. J. Vermeulen,

B.Sc. For. (Nat.

Cons.). (Ecological

data banking; infor-

mation systems; syn-

taxonomic nomen-

clature)

EXPERIMENTAL ECOLOGY SECTION

Officer-in-Charge M. C. Rutherford,

M.Sc., Ph.D., Dipl.

Datamet.

Assistant Director M. C. Rutherford,

M.Sc., Ph.D., Dipl.

Senior Agricultural

Researcher

Assistant Agricultural

Researchers

Research Technician

Technical Assistant..

Datamet. (Primary

production ecolo-

gy of terrestrial

ecosystems, particu-

larly in savanna; ex-

perimental ecologi-

cal studies in fynbos

and Karoo)

C. F. Musil, M.Sc.

Ph.D. (Aquatic eco-

physiology; repro-

ductive ecophysio-

logy in fynbos)

Miss F. M. Pressinger,

B.Sc. Flons (Eco-

physiological studies

in fynbos; compe-

tition and stress in

fynbos ecosystems)

G. W. Davis, M.Sc.

(Transformations of

fynbos systems by

the wild flower pick-

ing industry)

A. P. Flynn, B.Sc.

(Fynbos ecology)

Mrs J. M. Watermeyer

(Technical and ad-

ministrative assist-

ance in savanna,

fynbos and Karoo

research)

PLANT EXPLORATION SECTION

Officer-in-Charge .

Assistant Director

Agricultural Researchers

Research Technicians

M. J. Wells, M.Sc.

(Weeds research,

botanical horticul-

ture, fynbos utiliza-

tion and conserva-

tion)

T. H. Arnold, M.Sc.

(Plant utilization,

taxonomy of crop

plants — Sorghum,

conservation of

germ plasm, and

threatened indigen-

ous plants)

Miss L. Henderson,

B.Sc. Hons (Cover

and barrier plants,

woody exotic inva-

ders)

M rs K. J. Musil,

B.Sc. Hons (Conser-

vation of germ

plasm, woody exotic

invaders)

A. A. Balsinhas

(Indigenous food

plant data bank)

Bothalia 15, 3 & 4 (1985)

773

Liaison Officer

Technical Assistants

PRETORIA NATIONAL BOTANICAL

GARDEN

Mrs D. M. C. Fourie,

B.Sc.* (Scientific in-

formation service,

identification of

exotics)

Mrs J. B. Hoffman,

B.Sc. (Crop plants

of African origin)

Dip. Data* (Garden

records)

Chief Research Techni-

cian (Curator)

Agricultural Research

Technicians

Pupil Research Techni-

cian

Technical Assistant

Farm Foreman

D. H. Dry, NTC

(Hort.) Dipl.

H. J. de Villiers, NTC

(Hort.) Dipl. Rec.

P.A. (Development

of savanna biome)

D. S. Hardy (Nursery

supervision, succu-

lents and orchids)

Miss S. C. Kruger,

Nat. Dipl. (Hort.)

K. D. Panagos

Miss J. A. Taussig,

Nat. Dipl. (Hort.)

H. N. J. de Beer

L. C. Steenkamp

PUBLICATIONS BY THE STAFF

ANDERSON, J. M. & ANDERSON, H. M., 1983. Vascular

plants from the Devonian to Lower Cretaceous in southern

Africa. Bothalia 14: 337-344.

ANDERSON, J. M. & ANDERSON, H. M., 1983. Palaeoflora

of southern Africa, Molteno Formation (Triassic). Vol. 1, 1:

Introduction. Vol. 1, 2: Dicroidium, pp.227. Rotterdam:

Balkema.

ARNOLD, T. H., 1984. Comments on primitive South African

crop sorghums and the evolution of sorghum races in Africa.

Bothalia 14: 587-594.

ARNOLD, T. H. & MUSIL, K. J., 1984. A preliminary survey of

primitive crops cultivated in the northern Transvaal of South

Africa. Bothalia 14: 595-601.

BALSINHAS, A. A., 1983. The weeds of abandoned cotton

fields in Mozambique. Bothalia 14: 971-975.

BOTHA, R., 1983. 'n Anatomiese ondersoek van die epidermis

van Brachiaria serrata (Thunb.) Stapf Eragrostis capensis

(Thunb.) Trin., en Themeda triandra Forssk. M.Sc. proef-

skrif, Universiteit van Pretoria.

BOUCHER, C., 1983. Floristic and structural features of the

coastal foreland vegetation south of the Berg River, western

Cape Province, South Africa. Bothalia 14: 669-674.

CAMPBELL, B. M., 1983. Montane plant environments in the

Fynbos Biome. Bothalia 14: 283-298.

CODD, L. E., 1983. Southern African species of Mentha (Lamia-

ceae). Bothalia 14: 169-175.

CODD, L. E., 1983. The genus Tetradenia Benth. (Lamiaceae).

Bothalia 14: 177-183.

CODD, L. E., 1983. The correct author citation for Pachypodium

succulentum. Bothalia 14: 219.

CODD, L. E., 1983. A new name for Hartogia Thunb. ex L. E.

Bothalia 14: 219.

CODD, L. E., 1983. A new subspecies in Ocimum. Bothalia 14:

219-220.

CODD, L. E., 1983. A new species of Corchorus. Bothalia 14:

221-222.

CODD, L. E., 1983. Obituary: Jan Erens (1911-1982). Bothalia

14: 303-304.

CODD, L. E., 1983. Polygala fruticosa Berg. Flower. PI. Afr. 47,

1. 1861 .

CODD, L. E., 1983. Macrorungia iongistrobus C.B.C1. Flower.

PL Afr. 47, 1. 1862.

CODD, L. E., 1983. Monodora junodii Engl. & Diels. Flower.

PL Afr. 47, 1. 1870.

CODD, L. E., 1983. Aptosimum procumbens (Lehm.) Steud.

Flower. PL Afr. 47, 1. 1880.

CODD, L. E., 1984. Review: Flora Zambeziaca Vol. VII. 1. J1 S.

Afr. Bot. 50: 143-144.

CODD, L. E. & GUNN, M. D., 1983. More notes on plant col-

lectors. Veld Flora 69: 145-147.

DYER, R. A., 1983. Ceropegia, Brachystelma and Riocreuxia, in

southern Africa, pp.242. Rotterdam: Balkema.

DYER, R. A., 1983. Ceropegia connivens R. A. Dyer forma

angustata R. A. Dyer. Flower. Pl. Afr. 47, 1. 1872.

EDWARDS, D., 1983. Review: Distribution and ecology of vas-

cular plants in a tropical rain forest — forest vegetation in

Ghana by J. B. Hall and M. D. Swaine. Bothalia 14: 323.

EDWARDS, D., 1983. A broad-scale structural classification of

vegetation for practical purposes. Bothalia 14: 705-712.

EDWARDS, D., DE VOS, W. H., HARTKOPF, D., HAT-

TINGH, D. J., SCHEEPERS, J. J. & WILBY, A. F., 1983.

Monitoring of veld burns using satellite imagery. Proc.

Grassld Soc. sth. Afr. 18: 131-134.

ELLIS, R. P., 1983. Leaf anatomy and taxonomy of Lintonia nu-

tans (Chloridiodeae: Poaceae). S. Afr. J. Bot. 2: 162-167.

ELLIS, R. P., 1983. Leaf anatomy of the South African Dantho-

nieae (Poaceae). VIII. Merxmuellera decora, M. Lupulina

and M. rufa. Bothalia 14: 197-203.

ENGELBRECHT, A. J., EDWARDS, D. & KILLICK,

D. J. B., 1983. An ecological bibliography for southern

Africa. Bothalia 14: 765.

FABIAN, A. & GERMISHUIZEN, G., 1983. Transvaal Wild

Flowers, pp.292. Johannesburg: Macmillan.

GERMISHUIZEN, G., 1983. Cassia burttii Bak f. Flower. PL

Afr. 47, 1. 1868.

GERMISHUIZEN, G., 1983. Canavalia virosa (Roxb.) Wight A.

Flower. Pl. Afr. 47, t.1869.

GERMISHUIZEN, G., 1983. Review: Trees and shrubs of the

Etosha National Park by C. Berry. Bothalia 14: 321.

GIBBS RUSSELL, G. E., 1983. The taxonomic position of C3

and C4 Alloteropsis semialata (Poaceae) in southern Africa.

Bothalia 14: 205-213.

GIBBS RUSSELL, G. E., 1983. Review: Index Herbariorum

Edn 7. Bothalia 14: 322-323.

GIBBS RUSSELL, 1983. Correlaton between evolutionary his-

tory, flowering phenology, growth form and serai status for

important veld grasses. 5. Afr. J. Bot. 2: 175-180.

GIBBS RUSSELL, G. E., 1983. Register of names and types in

Poaceae: a computerized index for southern Africa. Bothalia

14: 943-944.

GIBBS RUSSELL, G. E., GERMISHUIZEN, G., HERMAN,

P., OLIVIER, P., PEROLD, S. M„ REID, C., RETIEF,

E. , SMOOK, L., VAN ROOY, J. & WELMAN, W. G.,

1984. List of species of southern African plants. Mem. bot.

Surv. S. Afr. 48: 1-144.

GIBBS RUSSELL, G. E. & ROBINSON, E. R., 1983. Specia-

tion environments and centres of species diversity in south-

ern Africa: 2. Case studies. Bothalia 14: 1007-1012.

GLEN, H. F. , 1983. Gardenia thunbergia Thunb. Flower. Pl. Afr.

47, 1. 1863.

Bothalia 15, 3 & 4 (1985)

GLEN, H F., 1983. Pleiospilos dekenahii (N.E. Br.) Schwant.

Flower. PI. Afr, 47, 1. 1864.

GLEN, H. F., 1983. Gibbaeum pubescens (Lettsom ex Haw.)

subsp. shandii (N.E. Br.) Glen. Flower. PI. Afr. 47, 1. 1865 A.

GLEN, H. F., 1983, Pleiospilos nelii Schwant. Flower. PI. Afr.

47, t.l865B.

HALL, A. V., DE WINTER, B., FOURIE, S. P. & ARNOLD,

T. H., 1984. Threatened plants in southern Africa. Biol.

Conserv. 28: 5-20.

HARDY, D. S., 1983. A preliminary list of the succulent plants

of Venda, their Venda names and domestic use. Aloe 20:

66-67.

HARDY, D. S., 1983. For the Love of an Island. I. Aloe 20:

16-18.

HARDY, D. S., 1983. For the Love of an Island. II. Aloe 20:

42-45.

HARDY, D. S.. 1983. For the Love of an Island. III. Aloe 20:

56-58.

HENDERSON, L., 1983. Barrier plants in South Africa. Botha-

lia 14: 635-639.

IMMELMAN, K. L., 1983. Review: South African parasitic flow-

ering plants by J. Visser. Bothalia 14: 323-324.

JACOT GUILLARMOD, A., 1983. Recovery of eastern Cape

heathland after fire. Bothalia 14: 701-704.

JACOT GUILLARMOD, A., 1983. A weed, yet beautiful and

useful! The Rose Bulletin (Royal National Rose Society),

Autumn, 1983. pp. 22-24.

JACOT GUILLARMOD, A., 1983. Cosmos, so very beautiful!

The Naturalist 27, 3: 3-4.

JACOT GUILLARMOD, A., 1984. Special report — Rare and

endangered. Albany Museum Report, Jan. 1982. March

1983: 25-27.

JACOT GUILLARMOD, A. & EVA, P., 1983. Limnological

bibliography for Africa south of the Sahara 31: 1-30. Gra-

hamstown: Limnological Society of southern Africa.

JACOT GUILLARMOD, A. & EVA, P., 1983. Limnological

bibliography for Africa south of the Sahara 32: 1-27. Gra-

hamstown: Limnological Society of southern Africa.

JACOT GUILLARMOD, A. & EVA, P., 1983. Limnological

bibliography for Africa south of the Sahara 33: 1-24. Gra-

hamstown: Limnological Society of southern Africa.

JACOT GUILLARMOD, A. & EVA, P., 1983. Limnological

bibliography for Africa south of the Sahara 34: 1-42. Gra-

hamstown: Limnological Society of southern Africa.

JARMAN, M. L., JARMAN, N. G. & EDWARDS, D., 1983.

Remote sensing and vegetation mapping in South Africa.

Bothalia 14: 271-282.

KELLERMAN, T. S., COETZER, J. A. W., SCHNEIDER, D.

J. & WELMAN, W. G., 1983. Photosensitivity in South

Africa. 111. Ovine hepatogenous photosensitivity caused by

the plant Athanasia trifurcata L. (Asteraceae). Onderste-

poortJ. vet. Res. 50: 45-53.

KNOX, M. D. E. & BRUSSE, F. A., 1983. New Xanthoparme-

liae (Lichenes) from southern and central Africa. Jl S. Afr.

Bot. 49: 143-159.

LEISTNER, O. A., 1983. Review: Trees and shrubs of the Cape

Peninsula by E. Moll and L. Scott. Bothalia 14: 321.

LEISTNER, O. A., 1983. Review: Flore des Mascareignes. Bo-

thalia 14: 321-322.

LEISTNER, O. A., 1983. Progress report on the Flora of South-

ern Africa (FSA). Bothalia 14: 1018.

LEISTNER, O.A., 1983. The Pretoria Flora. Bothalia 14: 1019.

LIENGME, C. A., 1983. A study of wood use for fuel and build-

ing in an area of Gazankulu. Bothalia 14: 245-258.

LIENGME, C. A., 1983. A survey of ethnobotanical research in

southern Africa. Bothalia 14: 621-630.

LINDER, H. P.. 1983. The historical phytogeography of the Disi-

nae (Orchidaceae). Bothalia 14: 565-570.

MAGILL, R. E., GIBBS RUSSELL, G. E., MORRIS, J. W. &

GONSALVES, P., 1983. PRECIS, the Botanical Research

Institute herbarium data bank. Bothalia 14: 481^495.

MOONEY, H. A., KUMMEROW, J., MOLL, E. J., ORSHAN,

G., RUTHERFORD, M. C. & SOMMERVILLE, J. M.,

1983. Plant form and function in relation to nutrient gra-

dients. In J. A. Day, Mineral nutrients in mediterranean

ecosystems. South African National Scientific Programmes

Report No. 71: 55-76. Pretoria: CSIR.

OBLRMEYER, A. A., 1983. Protasparagus Oberm, nom. nov.:

new combinations. 5. Afr. J. Bot. 2: 243-244.

OBERMEYER, A. A., 1983. Commelina bella Oberm. Flower.

PI. Afr. 47, 1. 1876.

O’CALLAGHAN, M. G., 1983. Flora. In A.E.F. Heydorn & J.

R. Grindley, Estuaries of the Cape. Report No. 19, Groot

(Wes) (CMS23) and Sout (CMS22): 24-26. Stellenbosch:

CSIR.

O'CALLAGHAN, M. G., 1983. Flora. In A. E. F. Heydorn & J.

R. Grindley, Estuaries of the Cape. Report No. 20, Groot

Brak (CMS3): 26-30. Stellenbosch: CSIR.

O’CALLAGHAN, M. G. , 1983. Flora. In A. E. F. Heydorn & J.

R. Grindley, Estuaries of the Cape. Report No 21, Bree

(CSW22): 22-24. Stellenbosch: CSIR.

O’CALLAGHAN, M. G., 1983. Biology. In A. E. F. Heydown

& J. R. Grindley, Estuaries of the Cape. Report No. 22,

Swartvlei (CMS11): 24-31. Stellenbosch: CSIR.

OLIVER, E. G. H., LINDER, H. P. & ROURKE, J. P., 1983.

Geographical distribution of present-day Cape taxa and

their phytogeographical significance. Bothalia 14: 427^140.

RETIEF, E., 1983. A new species of Raphionacme (Periploca-

ceae) from the Orange Free State. 5. Afr. J. Bot. 2: 326-328.

RUTHERFORD, M. C., 1983. Growth rates, biomass and distri-

bution of selected woody plant roots in Burkea africana

-Ochna pulchra savanna. Vegetatio 52: 45-63.

RUTHERFORD, M. C., 1983. Herbaceous standing crop in re-

lation to surface and subsurface rockiness. Bothalia 14:

259-264.

SCHEEPERS, J. C., 1983. Progress with vegetation studies in

South Africa. Bothalia 14: 683-690.

SCHEEPERS, J. C., 1983. The present status of vegetation con-

servation in South Africa. Bothalia 14: 991-995.

SCHRIRE, B. D., 1983. Centenary of Natal Herbarium, Durban

882-1982. Bothalia 14: 223-236.

SPECHT, R. L., MOLL, E. J., PRESSINGER, F. & SOMMER-

VILLE, J., 1983. Moisture regime and nutrient control of

seasonal growth in Mediterranean ecosystems. In F. J.

Kruger, D. T. Mitchell & J. U. M. Jarvis, Mediterranean-

type ecosystems, 120-132. Berlin, Heidelberg, New York,

Tokyo: Springer-Verlag.

SITES, J. J., 1984. Determination of genome homology in poly-

ploids. S. Afr. J. Sci. 80: 44-46.

SPIES, J. J. & HARDY, D. S., 1983. A karyotypic and anatom-

ical study of an unidentified liliaceous plant. Bothalia 14:

215-217.

STIRTON, C. H., 1983. Nocturnal petal movements in the Aster-

aceae. Bothalia 14: 1002-1006.

TAYLOR, H. C., 1983. The vegetation of the Cape of Good

Hope Nature Reserve. Bothalia 14: 779-784.

TILMAN, D., BOND, W. J., CAMPBELL, B. M., KRUGER,

F. J., LINDER, H. P.,SCHOLTZ, A., TAYLOR, H. C. &

WITTER, M., 1983. Origin and maintenance of plant

species diversity. In J. A. Day, Mineral nutrients in mediter-

ranean ecosystems. South African National Scientific Pro-

grammes Report No. 71: 125-135. Pretoria: CSIR.

TOLKEN, H. R., 1983. Kalanchoe sexangularis N.E.Br. Flower.

PI. Afr. 47, t.1878.

TOLKEN, H. R., 1983. Cotyledon tomentosa Harv. subsp. to-

mentosa. Flower. PI. Afr. 47, 1. 1877.

VAN WYK, A. E. & BOTHA, R., 1984. The genus Eugenia

(Myrtaceae) in southern Africa: ontogeny and taxonomic

value of the seed. 5. Afr. J. Bot. 3: 63-80.

VERDOORN, I. C., 1983. Diary of an expedition — 1937. Part

4. Aloe 20: 28-29.

VERDOORN, I. C., 1983. Diary of an expedition. Part 6.

(Should read Part 5). Aloe 20: 40-41.

VERDOORN, I. C., 1983. Crinum firmifolium Bak.f. Flower.

PI. Afr. 47, 47, t.1874.

VERDOORN, I. C., 1983. Crinum asiaticum L. Flower. PI. Afr.

47, 1. 1875 .

WEISSER, P. J. & BACKER, A. P., 1983. Monitoring beach

and dune advancement and vegetation changes 1937-1977 at

the Farm Twinstreams, Mtunzini, Natal, South Africa. In A.

McLachlan & T. Erasmus, Sandy beaches as ecosystems. The

Hague: Junk.

WEISSER, P. J. & MULLER, R., 1983. Dune vegetation dy-

namics from 1937 to 1976 in the Mlalazi-Richards Bay area

of Natal, South Africa. Bothalia 14: 661-667.

WEISSER, P. J. & STADLER, A., 1983. Suitability of aerial

photographs for monitoring emergent and submerged

macrophyte vegetation in the Wilderness Lakes, South

Africa. Proc. Int. Symp. Aquat. Macrophytes, 298-305. Nij-

megen: Faculteit Wiskunde en Natuurwetenschappen, Ka-

tholieke Universiteit Nijmegen.

Bothalia 15, 3 & 4 (1985)

775

WELLS, M. J., 1983. Review: Medicinal plants r>f the West In-

dies by E. S. Ayensu. Bothalia 14: '

WELLS, M. J., ENGELBRECHT, V. M., BALSINHAS, A. A.

& STIRTON, C. H., 1983. Weed flora of South Africa. 1:

major grouDings. Bothalia 14: 9as *548.

WELLS, M. J., ENGELBRECHT, V. M., BALSINHAS, A. A.

& STIRTON, C. H., 1983. Weed Bora of South Africa. 2:

power shifts in the veld. Bothalia 14: 961-966.

WELLS, M. J., ENGELBRECHT, V. M., BALSINHAS, A. A.

& STIRTON, C. H., 1983. Weed flora of South Africa. 3:

more power shifts in the veld. Bothalia 14: 967-970.

WESTFALL, R. H., EVERSON, C. S. & EVERSON, T. M.,

1983. The vegetation of the protected plots at Thabamhlope

Research Station. S. Afr. J. Bot. 2: 15-25.

WESTFALL, R. H., VAN ROOYEN, N. & THERON, G. K.,

1983. The plant ecology of the farm Groothoek, Thabazimbi

District. 1. Ordination. Bothalia 14: 785-790.

WESTFALL, R. H., VAN ROOYEN, N. & THERON, G. K.,

1983. Veld condition assessments in Sour Bushveld. Proc.

Grassld Soc. sth. Afr. 18: 73-76.

WESTFALL, R. H., VAN ROOYEN, N. & THERON, G. K.,

1983. A homogeniety index based on species diversity in

Sour Bushveld. Bothalia 14: 299-301.

WESTMAN, W. E., BOUCHER, C„ CAMPBELL, B. M.,

COWLING, R. M., LAMONT, B., LINDER, H. P.,

NOBLE, R. G. & VAN WILGEN, B. W., 1983. The struc-

ture and dynamics of plant communities. In J. A. Day,

Mineral nutrients in mediterranean ecosystems. South Afri-

can National Scientific Programmes Report No. 71: 77-89.

Pretoria: CSIR.

Bothalia 15, 3 & 4 (1985)

111

Book Reviews

A REVISION OF THE GENUS HAEMANTHUS L. (AMA-

RYLLIDACEAE) by DEIRDRE SNIJMAN. Journal of South

African Botany, Supplementary Volume No. 12, 24 April 1984.

Pp. 139 with 23 colour plates by Ellaphie Ward-Hilhorst, 1 colour

plate by Fay Anderson, 9 black and white illustrations and 23 dis-

tribution maps. Price R18 excl. GST.

The genus Haemanthus is a moderately sized genus of bulbous

plants found exclusively in southern Africa. The flowers are

placed in a dense umbel, and surrounded by erect or spreading

bracts; they are often dark red or scarlet, from which the genus

has derived its scientific name (literally meaning ‘blood-flower’).

Species of Haemanthus were among the first plants from the Cape

to be brought to Europe, and the earliest account of a species of

Haemanthus was published by the Flemish botanist 1 'Obel in

1605. Linnaeus included in his concept of this genus the original

Cape Haemanthus, together with another species from eastern

Cape called Haemanthus puniceus, and thus initiated a long tra-

dition of uniting the strictly South African species of Haemanthus

with other, predominantly tropical species. Through various divi-

sions and subdivisions of the genus, this union continued until

1976, when it was broken by Friis and Nordal, who distinguished

between the principally South African Haemanthus, and the prin-

cipally tropical genus Scadoxus. In fact, the two genera are well

separated by characters from the bulb and the leaves, as well as in

chromosome number.

The present work gives a comprehensive revision of the genus

Haemanthus in this new restricted sense. It is based on at least five

years of field study, during which the author, Ms Deirdre Snij-

man, has studied almost all the South African species in the field.

The book includes chapters on morphology (with an interesting

discussion on bulb-morphology), and a chapter on phytogeogra-

phy, in which an interesting pattern of local endemism is revealed.

There is also a chapter on ecology, including a summary of the

little which is known about pollination biology and seed dispersal,

an interesting section on evolution, and a full taxonomic revision,

for which South African herbaria, British herbaria (Kew, British

Museum and Edinburgh) and the herbarium of the Missouri Bo-

tanical Garden, have been examined. Included are also two keys

to species and subspecies, one based mainly on vegetative charac-

ters, and another based mainly on floral characters. In the taxo-

nomic part a total of 21 species and eight subspecies have been

recognized, of which six species and all the subspecies have been

established by the author, either in this revision or in previous

publications.

Of great interest in the introduction is a new informal classifica-

tion of the genus, based on characters of the bulb, leaves, and

inflorescences, as well as phytogeography. This is a much more

biologically meaningful classification than the one which has hith-

erto been used, and which has been based on the position of the

involucral bracts only. One comment here might be of interest for

future work: it seems that some species of the Haemanthus cocci-

neus group, i.e. H. pumilio, H. lanceifolius, and H. amarylloides,

form a rather distinct subgroup on account of colouring and very

short corolla tubes, and differ markedly from typical members of

the Haemanthus coccineus group.

Another very interesting new discovery is the concentration of

local endemics in Namaqualand. This underlines that Haeman-

thus is basically a dry country genus, as opposed to the mesophytic

tropical African Scadoxus, from which Snijman suggests that Hae-

manthus has evolved. Strictly speaking, the two genera are more

likely to have evolved from a common ancestor occurring in the

summer rain areas of tropical or southern Africa, as was recently

stated in a paper by I. Nordal and T. Duncan in Nordic J. Bot. 4:

145-153 (1984). However, all seem to agree that the two genera

are indeed related, monophyletic taxa.

Major problems in the study of Haemanthus are to correlate

old, fragmentary type specimens with modern material, or to cor-

relate previously published plates, often of essential nomenclatu-

ral importance, with modern herbarium specimens and natural

populations. Together, these two problems constitute the tradit-

ional obstacles to the herbarium worker on this genus, but the

troubles here have been overcome by the author's extensive field

study and cultivation of plants from the wild. The identification of

the old plates, often made from cultivated specimens in European

gardens, is particularly difficult, and proper identifications re-

quire a considerable degree of familiarity with the plants.

The identification of a large plate published by Jacquin in 1804,

and named H. amarylloides, is an interesting case in point. Since

1888 it has been assumed that this plate illustrated a plant distrib-

uted from the eastern Cape to Natal, but the author has convin-

cingly shown that the plate should instead be identified with a

previously unnamed species from Namaqualand. And that is not

the only instance where the author has provided new insight in

taxonomic or nomenclatural questions of long standing in Hae-

manthus taxonomy. The identification of several other of the Jac-

quin plates has given interesting results, for instance the one of H.

lanceifolius, a name which could not be correlated with existing

material, but which the author has shown is the correct name for

another previously unnamed species from Namaqualand.

An equally intricate problem in the study of Haemanthus is the

existence of many isolated populations. This may sometimes

make recognition of species difficult, and, as correctly stated by

the author, introduce an unavoidably subjective element in the

taxonomy. It is pointed out in the revision that the two species H.

amarylloides and H. pubescens are taxa where subdivision may be

necessary after further study, and that the species H. coccineus

and H. sanguineus are taxa where many local races are connected

by intermediates. These complex patterns lead inevitably to com-

plications when they have to be expressed in a taxonomical hier-

achy with a limited number of levels, such as in the division of H.

amarylloides, where subsp. amarylloides and subsp. polyanthus

would appear to be closer related to each other than to the third

subspecies, subsp. toximontanus. In another part of the genus, the

variation accepted in H. barkerae is considerable, so that distinc-

tion of the local endemic H. tristis on the specific level does not

appear to be fully warranted.

The illustrations of the book are worth separate mention. Fif-

teen of the species and seven of the subspecies are illustrated. The

reasons behind the choice of species illustrated are not clearly

stated, but availability of material must inevitably have been a

major consideration. As it is, there are two plates of H. humilis

subsp. humilis, two plates of H. crispus, H. barkerae, and //. coc-

cineus, but unfortunately no illustration of H. nortieri, which has

never been well illustrated. A few other species, which have pre-

viously been well illustrated by Jacquin or in Curtis’s Botanical

Magazine, are not illustrated in this work either.

The colour plates are very informative and pleasingly com-

posed, particularly considering the cumbersome plants they illus-

trate. In Plate 9, showing the unusual H. deformis, and Plate 18,

of Linnaeus’s original species, H. coccineus, Ellaphie Ward-Hil-

horst has had to deal with difficult tasks which have been admir-

ably solved. Reproduction of the plates would appear to be good,

and the colours true.

In conclusion, it is a valuable and handsome book, which will

be of equal interest to botanists and horticulturalists. Botanists

will be interested in the evolutionary or phytogeographical intri-

cacy of the genus, and find the book useful as the first comprehen-

sive guide to the identification of Haemanthus in about 100 years.

The horticulturalists will find much information on the identity

and natural habitats of Haemanthus, and both will enjoy the

beautiful illustrations. Because of the many local endemics, the

book will also be of interest from a conservationist’s point of view.

In short, I can fully recommend this book to anybody interested in

African Amaryllidaceae. or in bulbs in general.

IB FRIIS*

institute of Systematic Botany, University of Copenhagen.

Gothersgade 140, DK-1123 Copenhagen K, Denmark.

GUIDE TO THE PRICES OF ANTIQUARIAN AND SEC-

ONDHAND BOTANICAL BOOKS (1979-1982): FLOW-

ERING PLANTS compiled by L. VOGELENSANG. Leiden:

Boerhaave Press (P.O. Box 1051 — 2302 BB Leiden, Nether-

lands). 1983. ISBN 90 70153 17 3. Pp. xiv + 760; 240 x 170 X 40

mm. Price Dfl. 85,00.

The publishers of this gargantuan book catalogue provide the

following information: ‘This guide covers more than 8 000 titles of

books, monographs and reprints selected from the catalogues of

antiquarian and secondhand booksellers (listed with full addresses

on pages vii-xiv) from all over the world, covering the period

1979-1982. Russian literature is excluded. Bibliographic details of

each item are recorded, including edition, pages, figures and

778

Bothalia 15, 3 & 4 (1985)

plates which have been carefully checked. For each title the price

is' given at which it was offered for sale during the year indicated'

(but it is not mentioned which bookseller offered the title in ques-

tion). ‘If two prices are given for a book, these are the highest and

the lowest price for which it was offered in one year. All prices are

given in West German (DM) and American ($) currency.’

Entries are arranged alphabetically according to author, except

works not associated with a particular author, such as floras,

which are arranged according to title. No subject index or other

guide to subjects is provided.

The work was obviously compiled mainly for botanical librari-

ans, booksellers and collectors but its wealth of reliable informa-

tion on botanical publications old and new will prove useful to a

much larger circle of users. Readers seeking a wider coverage of

the vast field of botanical literature should acquire a copy of Prit-

zel’s Thesaurus literaturae botanicae [offered as a reprint for DM

198,00 ($86,00) during 1982] and relevant recent reference works

such as A. A. Bullock’s Bibliography of South African Botany (not

offered in the present guide).

O. A. LEISTNER

FLORA OF AUSTRALIA (Vol. 22) edited by ALEX S.

GEORGE. Canberra: Australian Government Publishing Service.

1984. Pp 239, line drawings and 16 colour photographs; 257 maps.

Price US $29,00 (soft cover edn), US $33,00 (case bound edn),

post free.

The Flora of Australia Project after ‘a long gestation and diffi-

cult birth’ has come into existence under the auspices of the Bu-

reau of Fauna and Flora in Canberra. So far four volumes of this,

the successor to Bentham’s 1863 Flora Australiensis, have been

published, the first one just in time for the XIII International Bo-

tanical Congress held in Sydney in August 1981.

The arrangement of volumes of this Flora follows the Cronquist

system, one of the recently proposed phylogenetic systems. I

really wonder whether the use of this system is important for the

production of the Flora. Presumably the Editorial Committee is

recommending that all Australian botanists become familiar with

this system. The endpapers to the volumes can at least help here

with their listing of the families. Will Australian herbaria be re-

arranged to match the Floral I note that the family Cornaceae,

which should be included in Vol. 22, will probably be included

under Grossulariaceae in Vol. 10.

Vol. 1, the Introduction, is a very useful one, giving the back-

ground to the project. It includes an annotated bibliography of

Australian botanical literature, an excellent chapter on the origin

and evolution of the flora, a discussion of systems of classifica-

tions, a key to all families and finally a partially illustrated gloss-

ary. The other volumes to have appeared so far are Vol. 8 cover-

ing Lecythidales to Batales (19 families including Capparaceae,

Cucurbitaceae and Cruciferae (not Brassicaceae?)) and Vol. 29

with Solanaceae.

I have Vol. 22, the fourth volume in the series, in front of me. It

covers Rhizophorales to Celastrales in which 17 families (includ-

ing Santalaceae, Loranthaceae, Viscaceae and Celastraceae) with

67 genera and 225 species are treated in rather typical flora style.

The following are given: type, synonomy, diagnostic description,

geographical distribution, up to five cited specimens and the occa-

sional sentence of comments, all very brief. I heartily approve of

the single spread of text across each page. For each specific and

subspecific taxon a distribution map is given and these are

grouped together on full pages of 15 maps to a page. On these

maps Australia is rather small and the very localised distributions

are sometimes difficult to find (in some cases these are circled).

The arrangement of the genera and species follows the system

chosen by each author to show natural affinities.

Most genera are represented by a line drawing which have been

executed by different artists. Unfortunately their styles are very

different and the way plates have been composed varies consider-

ably. Some plates depict up to eight species per plate whereas

others show just one large fruiting branch. It would appear that

then: is no rigid editorial policy for plates. An interesting innovat-

ion for a Flora of this nature is the inclusion of colour photo-

graphs. The present volume contains photographs of 16 species.

Some of these are very good but others show little and are of

doubtful value. I would have liked to have seen a good selection

"l (axonomically and photographically outstanding photographs.

There is a pleasant coloured frontispiece, a painting of an in-

cluded species which is repeated on the dust jacket.

The volume contains an appendix in which new taxa, new com-

binations and lectotypifications are given extremely briefly ‘for

economy reasons’. The reader is referred to the main text for the

fuller, albeit still very brief, treatment. At the beginning of each

genus and each family a useful list of references to recent works

covering the group is given. There is also a glossary supplemen-

tary to the one given in Vol. 1.

The work is well printed on good quality paper and is well

bound. The review copy is a hard covered edition and this looks

and feels like a good book. I understand that all volumes are

available in hard or soft covered editions. Previous volumes ran at

A$12.50, A$34 and A$25 for the hard covered copies and A$9.50,

A$29 and A$22 for the soft covered copies. No printing errors

were noticed.

There are only a few points that I do not like in this volume.

The type used in the headings of each species is not pleasing. Ro-

man is used for the author and italics for the reference. I would

have preferred these the other way round. I would have wel-

comed the inclusion of dotted leader lines to help the reader’s

eyes in the keys. I would also have preferred the characters in the

descriptions to have been italicised. I find the mixture of centi-

metres and millimetres in the descriptions rather confusing but

was pleasantly surprised to see the retention of the decimal point

rather than the comma. I find the citation of the types unsatisfac-

tory; neither the abbreviations, typography, punctuation nor the

lack of brackets around herbarium abbreviations are to my liking.

My congratulations to the contributors under the Executive

Editorship of my friend Alex George and to the Editorial Com-

mittee for the work done so far on this very worthwhile project.

This volume has 10 contributors including one from Europe.

When completed, 49 volumes will have covered the vascular

flora of Australia, which is estimated in Vol. 1 to be probably

25 000 species. In the brochure it is stated that ‘the series is plan-

ned to be published over a twenty-year period’! With 682 species

published after four years we can see their Flora taking 146 years

to produce. But they do have some 300 members of the Austra-

lian Systematic Botany Society. In contrast we in South Africa

with a flora of comparable size have only 76 systematically orien-

tated botanists. So the Flora of Southern Africa is perhaps not

doing too badly.

E. G. H. OLIVER

MEDICINAL PLANTS OF NORTH AFRICA by LOUTFY

BOULOS. Algonac: Reference Publications. 1983. Pp. 186, 103

line drawings. Price US $39,95.

This is the first of the recent influx of books on medicinal plants

that I am tempted to keep for myself, rather than donating it to

our Institute’s library.

It is the third in a series of works on ‘medicinal plants of the

world’ in which each volume has shown improvements. The im-

provements to the current volume include: common names in the

text; more information on plant uses; text references to the origin

of data e.g. ‘Rabat drug market’, and more illustrations. Of par-

ticular importance is the author’s first hand knowledge of the

plants, their uses and users. It makes this book something more

than the usual impersonal compilation from the literature.

Medicinal plants included in the book are those indigenous, cul-

tivated or naturalized in the North African region encompassing

Morocco, Algeria, Tunisia, Libya and Egypt. Altogether 369

species, belonging to 97 plant families are dealt with. For each

species the following information is provided: scientific name,

well-known synonyms, distribution in N Africa, origin, uses and

vernacular names. About 30% of the species, representing about

45% of the families, are illustrated.

There is no comment on the frequency of occurrence or use of

the species, their chemical content or efficacy, and the distribu-

tion data refers only to countries i.e. is too broad to reflect habi-

tat. The illustrations are good and it is a pity that more species

could not be pictured.

It is interesting to note that although there is only ± 1% overlap

in species coverage between this book and the earlier volume by

Ayensu on the Medicinal Plants of West Africa (which deals with

Bothalia 15, 3 & 4 (1985)

779

only 187 specimens), there is a ± 4% overlap with Kokwaro’s

book on the Medicinal Plants of East Africa (which deals with

1 300 species); and a ± 8% overlap with Ayensu’s Medicinal

Plants of the West Indies. The overlapping species are mostly

weeds, led by such hardy perennials as Ricinus communis and

Foeniculum vulgare.

About 20% of the North African medicinal plant species occur

in South Africa, but mostly as crop plants or escapes, and rela-

tively few of them are recorded by Watt and Breyer-Brandwijk as

being used here for medicinal purposes.

Dr. Boulos’s book is completed by: Notes and explanations (on

taxonomic limits etc.); Glossary (of medical terms); Bibliography

(very selective, including 66 references); Medicinal index (very

useful, but not flawless e.g. only 4 out of the 7 plants listed under

‘weight adding’ are credited with this use in the text); Common

names index (including ± 3 800 names); and Index to species.

This is a well balanced and informative publication. With it the

author should go some way toward achieving his stated objective:

‘to bridge the gap between folk and modern medicine’. What is

more certain is that by assembling the plethora of basic informa-

tion in one place, and presenting it in a well organised format, Dr.

Boulos will encourage research by botanists, pharmacologists and

chemists as well as by medical men.

M. J. WELLS

PLANTS OP THE CAPE PLORA, A DESCRIPTIVE CATA-

LOGUE by PAULINE BOND & PETER GOLDBLATT. Jour-

nal of South African Botany Supplementary Volume No. 13,

1984. Pp. xi + 455, map, 8 colour photographs. Price R18,00.

Bond & Goldblatt’s Catalogue of the Cape flora is the first com-

prehensive treatment of the species in this floristically rich region.

The boundaries of the Cape flora are based on the occurrence of

the fynbos vegetation types as shown by Acocks (1953) map of the

Veld Types of South Africa, but the species of other vegetation

types that also lie within the area, such as karoo and forest, are

included as well. Earlier works that include the Cape flora have

either covered a much larger area, such as Flora Capensis, or only

a small portion of the Cape flora, such as Adamson & Salter’s

Flora of the Cape Peninsula.

All the more than 8 500 species of seed plants known from this

area are dealt with. Families, genera and species are arranged al-

phabetically within the three major groups Gymnospermae,

Monocotyledonae and Dicotyledonae. The authors have suc-

ceeded in compressing a surprisingly large amount of information

into the two or three printed lines allowed for each species: name,

author, common name, short description, month of flowering,

habitat notes and distribution both within and outside the Cape

region.

The necessarily brief format for the treatment of each species

has resulted in the exclusion of features that would have made the

book more useful. Species descriptions are often too short to be

diagnostic, so that identification of species in large genera is sel-

dom possible. There are no sequential numbers for families, gen-

era and species, so ecologists and other field workers lack a

readily available code for recording taxa. Infraspecific taxa are

omitted entirely, so that the true complexity of many difficult

plant groups is not apparent.

On the other hand, in their natural desire to make their work as

up-to-date as possible, the authors have included many ‘ined.’

names that were in press at the time the book was published. The

wisdom of this is debatable. The obvious advantage of including

these names is that all known species are accounted for in their

most recently determined genus. The disadvantage (to taxonom-

ists at least) of the use of these names before they are validly pub-

lished is that one more step must be added when checking for

nomenclatural accuracy in preparing future treatments for groups

within the Cape flora. If any of these ‘ined.’ names should not be

published, after having been admitted to general use through the

Catalogue, long-term confusion will result. Indeed, possibilities

for mix-up are foreshadowed even within the Catalogue itself.

Species formerly in Asparagus are in the text correctly placed in

their ‘new’ genus Protasparagus. However, in the introduction the

name Asparagopsis is used, which was at one time considered for

these species but not taken up because it was found to be a later

homonym. Despite the potential for confusion, the authors de-

serve credit for actively adopting a method to overcome, even if

only on the short term, the horrible problem of name changes that

so quickly outdates taxonomic work.

The introductory sections of the book outline some of the

physical and biotic factors in the Cape flora region, and relate

these to the great species richness observed. The outstanding

species diversity and high level of endemism found in the Cape

flora is compared to that of other parts of the world, especially

areas of Mediterranean climate. The largest families and genera

in the Cape flora are listed, and an appendix gives the number of

genera and species as well as the number of endemic genera and

species in the Cape flora. The colour photographs illustrate some

of the most striking plant communities.

The introduction suffers from minor mistakes. For example,

there are a number of misprints, at least two references are not

included in the literature cited, and the names of Acocks’s veld

types are capitalized in some places and not in others. More se-

riously, the section on the origin and evolution of the Cape flora

appears to be based on work published during the 1970’s, and is

not up to date with more recent thinking of contributors to the

Fynbos Biome Project of the National Programme for Environ-

mental Sciences. It is hoped that this omission is not the symptom

of a sad lack of communication between those cataloguing the

flora and those studying the vegetation.

However, in relation to the enormous amount of information

about the species that has been synthesized from herbaria, litera-

ture and experts on various plant groups, these criticisms in detail

are mere motes in the eye. The authors have struck a successful

balance in their species treatments, so that although the informa-

tion is complete enough to be useful and simple enough to be eas-

ily understandable to the general user, the book is still a con-

venient size to be used in the field.

G. E. GIBBS RUSSELL

GUIDE FOR AUTHORS

INTRODUCTION

Bothalia is a medium for the publication of botanical papers

dealing with the flora and vegetation of southern Africa. Papers

submitted for publication in Bothalia should conform to the gen-

eral style and layout of recent issues of the journal (from Vol. 14

onwards) and may be written in either English or Afrikaans.

TEXT

Manuscripts should be typed, double-spaced on one side of uni-

formly-sized A4 paper" having at least a margin of 30 mm all

round. Latin names of plants should be underlined to indicate

italics. Metric units are to be used throughout. Manuscripts

should be submitted in duplicate to the Editor, Bothalia, Private

Bag X101, Pretoria 0001.

KEYWORDS

Up to six keywords (more correctly, index terms) should be

provided. The following points should be borne in mind when se-

lecting keywords:

1, Keywords should be unambiguous, internationally acceptable

words and not recently-coined little-known words; 2, they should

be in a noun form and verbs should be avoided; 3, they should not

consist of an adjective alone; adjectives should be combined with

nouns; 4, they should not contain prepositions; 5, use the singular

form for processes and properties e.g. evaporation; 6, use the plu-

ral form for physical objects e.g. augers; 7, location (province

and/or country); taxa (species, genus, family) and vegetation type

(community, veld type, biome) should be used as keywords; 8,

keywords should be selected hierarchically where possible, e.g.

both family and species should be included; 9, they should include

terms used in the title; 10, they should answer the following ques-

tions: 10.1, what is the active concept in the document (activity,

operation or process); 10.2, what is the passive concept or object

of the active process (item on which the activity, operation or pro-

cess takes place); 10.3, what is the means of accomplishment or

how is the active concept achieved (technique, method, appara-

tus, operation or process); 10.4, what is the environment in which

the active concept takes place (medium, location) and 10.5, what

are the independent (controlled) and dependent variables?; 11,

questions 10.1 to 10.3 should preferably be answered in the title.

ABSTRACT

A short abstract of 100-200 words in both English and Afri-

kaans should be provided. In the abstract the names of new

species and new combinations should not be underlined.

FIGURES

Black and white drawings, including graphs, should be in jet-

black Indian ink preferably on bristol board or plastic film. Lines

should be bold enough to stand reduction. Lettering should be

done using some printing device such as stencilling, letraset, etc.

It is recommended that drawings should be twice the size of the

final reduction.

Photographs submitted should be of good quality, glossy, sharp

and of moderate, but not excessive contrast. Photograph mosaics

should be composed by the authors themselves: the component

photographs should be mounted neatly on a white card base leav-

ing a narrow gap between each print; number the prints using

some printing device.

Figures should be planned to fit, after reduction, into a width of

80 mm, 118 mm or 165 mm with a maximum vertical length of 240

mm. The number of each figure and the author’s name should be

written on the back of the figure using a soft pencil. Captions for

figures should be collected together and typed on a separate page

headed Captions for Figures. A copy of each caption should be

attached to the base of each figure.

It is recommended that in taxonomic revisions dot maps should

be included as figures to show the distribution of the taxa.

GIDS VIR SKRYWERS

INLEIDING

Bothalia is ’n medium vir die publikasie van plantkundige arti-

kels wat handel oor die flora van suidelike Afrika. Artikels wat

voorgele word vir publikasie in Bothalia behoort ooreen te stem

met die algemene styl en rangskikking van onlangse uitgawes van

die tydskrif (vanaf Vol. 14). Dit mag in Engels of in Afrikaans

geskryf word.

TEKS

Manuskripte moet getik wees in dubbelspasiering slegs op een

kant van ewegroot A4-papier, met reg rondom ’n rand van min-

stens 30 mm breed. Latynse name van plante moet onderstreep

word om aan te dui dat dit kursief gedruk moet word. Metrieke

eenhede moet deurgaans gebruik word. Manuskripte moet in

tweevoud ingedien word by die Redakteur, Bothalia, Privaatsak

X101, Pretoria 0001.

SLEUTELWOORDE

Tot ses sleutelwoorde (meer presies, indeksterme) moet ver-

skaf word. Wanneer sleutelwoorde geselekteer word moet die

volgende punte in gedagte gehou word:

1, Sleutelwoorde moet ondubbelsinnige, internasionaal aanvaar-

bare woorde wees en nie onlangs-geskepte minder bekende

woorde nie; 2, hulle moet naamwoorde wees en werkwoorde

moet vermy word; 3, hulle moet nie net uit ’n byvoeglike naam-

woord bestaan nie; byvoeglike naamwoorde moet met ’n naam-

woord saamgevat word; 4, hulle moet nie voorsetsels bevat nie; 5,

gebruik die enkelvoud vir werkwyses en eienskappe bv. verdam-

ping; gebruik die meervoud vir fisiese voorwerpe bv. handbore; 7,

plek (provinsie en/of land); taksa (spesies, genus, familie) en

plantegroeitipe (gemeenskap, veldtipe, bioom) moet as sleutel-

woorde gebruik word; 8, sleutelwoorde moet waar moonlik vol-

gens rangorde gekies word, bv. beide familie en spesies moet in-

gesluit word; hulle moet terme wat in die titel gebruik word

insluit; 10, hulle moet die volgende vrae beantwoord: 10.1, wat is

die aktiewe konsep in die dokument (aktiwiteit, handeling of

proses); 10.2, wat is die passiewe konsep of doel van die aktiewe

proses (item waarop die aktiwiteit, handeling of proses plaas-

vind); 10.3, deur middel waarvan word dit uitgevoer of hoe word

die aktiewe konsep bereik (tegniek, metode, apparaat, handeling

of proses); 10.4, wat is die omgewing waarin die aktiewe konsep

plaasvind (medium, plek) en 10.5, wat is die onafhanklike (ge-

kontroleerde) en afhanklike variante; 11, vrae 10.1 tot 10.3 moet

verkieslik in die titel beantwoord word.

UITTREKSEL

'n Kort uittreksel van 100 — 200 woorde moet voorsien word,

beide in Engels en Afrikaans. In die uittreksel moet die name van

nuwe soorte en nuwe kombinasies nie onderstreep word nie.

AFBEELDINGS

Wit en swart tekeninge, insluitende grafieke, moet met pik-

swart Indiese ink geteken word, verkieslik op ‘bristol board’ of

plastiekfilm. Lyne moet dik genoeg wees om verklein te kan

word. Letterwerk moet gedoen word met gebruik van een of an-

der hulpmiddel soos letraset of ’n sjabloon. Dit is wenslik dat te-

keninge tweemaal so groot as die uiteindelike verkleining sal

wees.

Foto’s wat ingedien word, moet van hoe kwaliteit wees — glan-

send, skerp en van matige maar nie oordrewe kontras. Fotomo-

saieke moet deur die skrywer self saamgestel word: die afsonder-

like foto’s moet netjies monteer word op ’n stuk wit karton met ’n

smal strokie tussen die foto’s; nommer die foto’s met behulp van

een of ander druk-hulpmiddel.

Afbeeldings moet so beplan word dat hulle na verkleining sal

pas in ’n breedte van 80 mm, 118 mm of 165 mm met 'n maksi-

mum vertikale lengte van 240 mm. Die nommer van elke afbeel-

ding sowel as die skrywer se naam moet op die rugkant van die

afbeelding geskryf word met ’n sagte potlood. Onderskrifte vir

afbeeldings moet bymekaar getik word op ’n afsonderlike bladsy

met die opskrif Onderskrifte vir Afbeeldings. ’n Afskrif van elke

onderskrif moet aan die onderkant van elke afbeelding vasgeheg

word.

Dit word aanbeveel dat in taksonomiese hersienings, kolkaarte

gebruik word (as figure gemerk), om die verspreiding van die

taksa aan te dui.

781

INDEX— INDEKS

Acanthaceae, 107

Acorus calamus, 547

Acriulus Ridl., 513

subgenus Scleria (Berg.) C.B. Cl., 513

section Acriulus (Ridl.) C.B. Cl., 513

greigifolius Ridl., 527

madagascariensis Ridl., 528

titan C.B. CL, 528

Africa, 7, 11, 77

African, 387

Afromontane Forest, 217

Alien plants, 297

Amaryllidaceae, 545

Amaryllis

flexuosus Jacq., 546

humilis Jacq., 546

Amethystanthus Nakai, 8

Anatomy, 11, 30, 153, 561, 567, 573, 579

Aneilema

arenicola Faden, 94

brunneospermum Faden, 91

calceolus Brenan, 98

clarkei Rendle, 98

dregeanum sensu Compton non Kunth, 91, 96

dregeanum Kunth, 90

subsp. dregeanum, 90

subsp. mossambicense Faden, 90

var. galpinii C.B. Clarke, 91

indehiscensFacfert, 96

subsp. indehiscens, 97

subsp. lilacinum Faden, 91

petersii (Hassk.) C.B. Clarke, 96

schlechteri K. Schum., 91

schlechteri sensu Brenan, 91

tanaense Faden, 98

Anthochortus Nees, 66, 484

capensis Esterhuysen, 484

crinalis (Mast.) Linder, 486

ecklonii Nees, 486

graminifolius (Kunth) Linder, 486

insignis (Mast.) Linder, 487

laxiflorus (Nees) Linder, 487

Anthony, Nicola C. New combinations in Herschelianthe (Oxchid-

aceae), 554

Anthony, Nicola C. & Schelpe, E.A.C.L.E. A checklist of the

Pteridophytes of the ‘ Flora of Southern Africa' region, 541

Anthony, Nicola C. & Schelpe, E.A. Two new taxa and a new

combination in southern African Pteridophyta (Pteridophyta),

554

Anthony, Nicola C. & Schelpe, E.A. X Pleopodium — a putative

intergeneric fern hybrid from Africa(Pteridophyta), 555

Apomixis, 167

Araceae, 547

Argyrolobium transvaalense - Combretum molle Open Woodland,

678

Aristida diffusa — Combretum molle Open Woodland, 673

Askidiosperma Steud, 64, 431

albo-aristatum (Pillans) Linder, 431

andreaeanum (Pillans) Linder, 431

capitatum Steud., 431

chartaceum (Pillans) Linder, 431

subsp. alticolum Esterhuysen, 432

esterhuyseniae (Pillans) Linder, 432

insigne (Pillans) Linder, 432

longiflorum (Pillans) Linder, 432

nitidum (Mast.) Linder, 432

paniculatum (Mast.) Linder, 432

rugosum Esterhuysen, 432

Asparagaceae, 548

Asparagopsis Kunth, 78

consanguinea Kunth, 87

decumbens (Jacq.) Kunth, 86

scandens (Thunb.) Kunth, 86

schlechtendalii Kunth, 87

Asparagus

sect. Myrsiphyllum (Willd.) Bak., 77

asparagoides sensu Jessop, 77, 79, 80, 82

asparagoides (L.) Wight, 78

asparagoides (L.) Druce, 78

consanguineus (Kunth) Bak., 87

crispus Lam., 86

declinatus L., 86

decumbens Jacq., 86

fasciculatus Thunb., 87

flexuosus Thunb., 86

juniperoides Engl., 84

klinghardtianus Dinter, 83

krausii Bak., 80

krausianus (Kunth) MacBride, 80

kuisibensis Dinter, 78

medeoloides (L. f.) Thunb., 78

var. falciformis (Kunth) Bak., 78

medeoloides (Thunb.) Bak., 78

var. angustifolius (Mill.) Bak., 78

multitub erosus R.A. Dyer, 77

ovatus Salter, 79

pectinatus Delile, 86

ramosissimum Bak., 87

scandens Thunb., 86

var. deflexus Bak., 87

undulatus (L. f.) Thunb., 83

volubilis Thunb., 82

Aspidiaceae, 554

Asplenium splendens - Celtis africana Kloof Forest, 666

Association analysis, 245

Asthenatherum glaucum (Nees) Nevski, 153

Barker, N.P. Evidence of a volatile attractant in Ficus ingens

(Moraceae), 607

Basal cover, 241

Biographies, 631

Blechnaceae, 555

Blechnum australe L.

var. aberrans N.C. Anthony & Schelpe, 555

Boeckhia Kunth, 67, 488

laevigata Kunth, 489

striata Kunth, 492

Book reviews, 339, 111

Braun-Blanquet, 245, 655

Broad-leaved thicket, 259, 283-285

Brusse, F. New species and combinations in Parmelia (Lichenes)

from southern Africa, 315

Brusse, F. Corynecystis, a new lichen genus from the Karoo,

South Africa (Heppiaceae, Lichenes), 552

Calopsis Beauv. ex Desv., 65, 464

adpressa Esterhuysen, 465

andreaeana (Pillans) Linder, 465

aspera (Mast.) Linder , 465

burchellii (Mast.) Linder, 465

clandestina Esterhuysen, 465

dura Esterhuysen, 466

esterhuyseniae (Pillans) Linder, 466

festucacea Kunth, 470

Filiform is (Mast.) Linder, 467

fruticosa (Mast.) Linder, 46 7

gracilis (Mast.) Linder, 467

hyalina (Mast.) Linder, 467

782

Calopsis continued

hirtella Kunth, 470

impolita (Kunth) Linder , 467

incurvata Pillans, 470

incurvata (Thunb.) Kunth, 494

levynsiae (Pillans) Linder , 467

marlothii (Pillans) Linder, 467

membranacea (Pillans) Linder, 469

monostylis (Pillans) Linder, 469

muirii (Pillans) Linder, 469

neglecta Hochst., 453

nudiflora (Pillans) Linder, 469

oxylepis Kunth, 470

peronata Kunth, 470

paniculata (Rottb.) Desv., 469

pulchra Esterhuysen, 469

ramiflora (Nees) Kunth, 469

rigida (Mast.) Linder, 470

rigorata (Mast.) Linder, 470

triticea (Rottb.) Kunth, 462

viminea (Rottb.) Linder, 470

Calorophus Labill.

anceps (Mast.) Kuntze, 436

asper Kuntze, 465

burchellii (Mast.) Kuntze, 469

digitatus (Thunb.) Kuntze, 487

filiformis (Mast.) Kuntze, 467

gracilis (Mast.) Kuntze, 467

laxiflorus (Nees) Kuntze, 487

tenuis (Mast.) Kuntze, 486

virgatus (Mast.) Kuntze, 444

CannomoisBetfwv. ex Desv., 65, 480

acuminata (Kunth) Pillans, 481

aristata Mast. , 480

cephalotes Desv., 482

complanatus Mast., 481

congesta Mast. , 480

dregei Pillans, 482

nitida (Mast.) Pillans, 481

parviflora (Thunb.) Pillans, 481

schlechteri Mast., 481

scirpoides (Kunth) Mast. , 482

var. minor Pillans, 481

var. primosii Pillans, 48 1

simplex Kunth, 481

spicatus Mast., 481

virgata (Rottb.) Steud., 482

Canopy cover, 241

Cape Fynbos Shrublands, 191

Cape of Good Hope Nature Reserve, 245, 259

Cape Transitional Small-leaved Shrublands, 203

Carbon isotope ratios, 587

Carpha filifolia Reid & Arnold, 139

Ceratocaryum Nees, 65, 479

argenteum Nees ex Kunth, 479

decipiens (N.E. Br.) Linder, 479

fimbriatum (Kunth) Linder, 479

fistulosum Mast., 480

xerophilum (Pillans) Linder, 480

Checklist, 131, 297, 541

Chloridoideae, 587

Chondropetalum Rottb., 64, 427

acockii Pillans, 428

aggregatum (Mast.) Pillans, 428

albo-aristatum Pillans, 431

andreaeanum Pillans, 431

capitatum (Steud.) Pillans, 431

chartaceum (Pillans) Pillans, 43 1

decipiens Esterhuysen, 428

deustum Rottb., 429

ebracteatum (Kunth) Pillans, 429

esterhuyseniae Pillans, 432

hookerianum (Mast.) Pillans, 429

insigne Pillans, 432

longiflorum Pillans, 432

macrocarpum (Kunth) Pillans, 431

marlothii (Pillans) Pillans, 429

microcarpum (Kunth) Pillans, 429

mucronatum (Nees) Pillans, 430

nitidum (Mast.) Pillans, 432

nudum Rottb., 430

paniculatum (Mast.) Pillans, 432

rectum (Mast.) Pillans, 430

tectorum (L. f.) Rafin., 430

Chromosomes, 597

Chromosome numbers, 591

Classification, 11, 505

Codd, L.E. The genus Tetradenia Benth. (Lamiaceae). II. Mala-

gasy Republic, 1

Codd, L.E. The genus Isodon (Schrad. ex Benth.) Spach in Africa

and a new genus Rabdosiella Codd (Lamiaceae), 7

Codd, L.E. A new species of Plectranthus (Lamiaceae), 142

Codd, L.E. & Gunn, Mary. Additional biographical notes on plant

collectors in southern Africa, 631

Coleochloa setifera - Combretum molle Open Woodland, 675

Combretum molle — Landolphia capensis Closed Woodland, 674

Commelinaceae, 89

Computerization of taxonomic literature, 125

Conspectus, 387

Corynecystis5rwsse, 552

capensis Brusse, 552

Cover meter, 241

Cowling, R.M. A syntaxonomic and synecological study in the

Humansdorp region of the Fynbos Biome, 175

Craspedolepis Steud., 65, 437

verreauxii Steud., 446

Cucullifera Nees, 65, 480

dura Nees, 430

Cyperaceae, 139,505

Cytology, 161, 167, 591, 597

C4, 153

Danthonieae, 153, 561, 567, 573, 579

Diplocalymma Spreng., 107

Diplospory, 167

Disa

forficaria H. Bol., 554

lutea Linder, 553

newdigateae L. BoL, 554

patens (L. f.) Thunb., 553

schlechteriana H. Bol., 554

tenuifolia Sw., 553

tripartita Lindl., 554

Distribution, 101, 297

Distribution maps, 131

Distribution ranges, 541

Dorsal epithelium, 117, 531

Douglas, K.H. The identity of Nerine flexuosa (Amaryllidaceae),

545

Douglas, K.H. The relationship between Nerine flexuosa and N.

humilis (Amaryllidaceae), 545

Dovea Kunth, 64, 431

aggregata Mast., 428

binata Steud., 489

bolusii Mast., 429

chartacea Pillans, 431

cylindrostachya Mast., 430

ebracteata Kunth, 429

hookeriana Mast., 429

macrocarpa Kunth, 431

marlothii Pillans, 429

microcarpa Kunth, 429

mucronata (Nees) Mast., 430

783

Doxez Kunth continued

nitida Mast., 432

nuda (Rottb.) Pillans, 430

paniculata Mast., 432

racemosa (Poii.) Mast., 426

recta Mast., 430

rigens Mast., 429

tectorum (L. f.) Mast., 430

thyrsoidea Mast., 427

Dracaena

medeoloides L. f., 78

undulata L. f., 83

volubilis L. f., 82

Drakensberg, Natal, 139, 550

Duthie, 531

Ecology, 175, 229, 241, 245, 259, 293, 294, 655, 689, 705, 725,

733, 749

Ehrharta

dodii Stapf, 150

eburnea Gibbs Russell, 145

rupestris Nees ex Trin., 150

subsp. dodii (Stapf) Gibbs Russell, 150

subsp. rupestris, 150

subsp. tricostata (Stapf) Gibbs Russell, 150

setacea Nees, 150

subsp. disticha Gibbs Russell, 151

subsp. scabra (Stapf) Gibbs Russell, 151

subsp. setacea Stapf, 150

subsp. uniflora (Burch, ex Stapf) Gibbs Russell, 151

var. scabra Stapf, 151

tricostata Stapf, 150

uniflora Burch, ex Stapf, 151

Eichhomia crassipes (Mart.) Solms, 689, 705, 725, 733

Elegia L„ 63, 418

acuminata Mast., 427

altigena Pillans, 420

amoena Pillans, 426

asperiflora (Nees) Kunth, 420

var. lacerata (Pillans) Pillans, 420

atratiflora Esterhuysen, 420

bella Pillans, 426

caespitosa Esterhuysen, 421

capensis (Burm. f.) Schelpe, 421

ciliata Mast., 420

coleura Nees ex Mast., 421

cuspidata Mast., 422

deusta (Rottb.) Kunth, 429

dregeana Kunth, 420

elongata Mast., 430

equisetacea (Mast.) Mast., 422

esterhuyseniae Pillans, 422

var. dispar Pillans, 422

exilis Mast., 421

extensa Pillans, 422

fastigiata Mast. , 422

fenestrata Pillans, 422

filacea Mast., 422

flstulosaAwnr/z, 423

var. parviflora Pillans, 423

fucata Esterhuysen, 423

fusca N.E. Br., 426

galpinii N.E. Br., 423

glauca Mast., 420

gracilis N.E. Br., 423

grandis (Nees) Kunth, 424

grandispicata Linder, 424

hutchinsonii Pillans, 424

intermedia (Steud.) Pillans, 424

juncea L., 424

var. geniculata Pillans, 425

lacerata Pillans, 420

membranacea (Nees) Kunth, 424

mucronata (Nees) Kunth, 430

muirii Pillans, 425

neesii Mast., 425

nuda (Rottb.) Kunth, 430

obtusiflora Mast., 426

panicoides Kunth, 430

paniculata Pers., 482

parviflora (Thunb.) Kunth, 481

parviflora Pillans, 423

var. filacea (Mast.) Pillans, 423

var. rigida (Mast.) Pillans, 426

pectinata Pillans, 426

persistens Mast. , 425

prominens TY/Jans, 425

propinqua (Nees) Kunth, 425

var. equisetacea Mast., 422

var. minor Mast., 425

racemosa (Poir.) Pers., 425

rehmannii Mast., 423

rigida Mast., 426

spathacea Mast. , 426

var. attemiata Pillans, 426

squamosa Mas?., 426

stipularis Mast. , 426

stokoei Pillans, 427

thyrsifera (Rottb.) Pers., 427

thyrsoidea (Mast.) Pillans, 427

vaginulata Mast. , 427

verreauxii Mast. , 427

verticillaris (L. f.) Kunth, 421

Elide Medicus, 78

Ellis, R.P. Leaf anatomy of the South African Danthonieae

(Poaceae). IX. Asthenatherum glaucum, 153

Ellis, R.P. Leaf anatomy of the South African Danthonieae

(Poaceae). X. Pseudopentameris, 561

Ellis, R.P. Leaf anatomy of the South African Danthonieae

(Poaceae). XI. Pentameris longiglumis and Pentameris sp.

nov., 567

Ellis, R.P. Leaf anatomy of the South African Danthonieae

(Poaceae). XII. Pentameris thuarii, 573

Ellis, R.P. Leaf anatomy of the South African Danthonieae

(Poaceae). XIII. Pentameris macrocalycirta and P. obtusifolia,

579

Elsholtzia sect. Rabdosia Blume, 8

Embryo sac, 161, 167

Endemic species, -6 13

Endomelas Rafin., 107

Equisetum

capense Burm. f., 421

Eragrostideae, 587

Eragrostis curvula Complex, 167

Eragrostis pallens - Andropogon apendiculatus Grassland, 680

Eragrostis racemosa - Trachypogon spicatus Grassland, 682

Erica - Trachypogon Community, 199

Ericaceae, 550

Erythrina lysistemon - Celtis africana Kloof Forest, 663

Euclea crispa - Combretum molle Closed Woodland, 671

Evergreenness, 229

Fabaceae, 551

Faden, R.B. New taxa of Aneilema R. Br. (Commelinaceae) from

southern and tropical East Africa, 89

Ficus ingens (Miq.) Miq., 607

Fig, 607

Flora, 613

Floristic classification, 655

Forest, 229

Franklin Hennessy, E.F. The genus Scleria in southern Africa,

505

Fynbos, 229, 245, 259, 264-282

Fynbos Biome, 175

784

Gamma diversity, 131

Gas chromatography, 607

Geographical distribution, 591

Geraniaceae, 345

Geraniospermum

angulosum (Mill.) Kuntze, 349

capitatum (L.) Kuntze, 353

cucullatum (L.) Kuntze, 349

denticulatum (Jacq.) Kuntze, 361

radula (Cav.) Kuntze, 365

ribifolium (Jacq.) Kuntze, 369

scabrum (L.) Kuntze, 367

terebinthinaceum (Cav.) Kuntze, 364

tomentosum (Jacq.) Kuntze, 383

Geranium

acerifolium Cav., 349, 35 1

angulosum MilL, 349

asperum (Ehrh. ex Willd.) Poir., 364

balsameum (Jacq.) Poir., 367

betulinum L., 352

capitatum L., 353

cordifolium Cav., 379

crataegifolium Roth, 359

crispum Berg., 376

cucullatum L., 349, 350

var. fimbriatum Burm. f., 349

denticulatum (Jacq.) Poir., 361

glutinosum Jacq., 359

graveolens (L’Herit.) Thunb., 364

hermanniifolium Berg., 378

hispidum L. f., 381

lanatum Thunb., 379

papilionaceum L., 382

quercifolium sensu Cav., 356

quercifolium L. f., 358

radula Roth, 364

radula Cav., 365

revolutum Jacq., 365

ribifolium (Jacq.) Poir., 369

ribisioides Burm. f., 355

scabrum L., 367

terebinthinaceum Murray, 356

terebinthinaceum Cav., 364

tomentosum (Jacq.) Poir., 383

viscosum Scop., 359

vitifolium L., 355

Gibbs Russell, G.E. A new species of Ehrharta, 145

Gibbs Russell, G.E. Notes on species of Ehrharta with a short first

sterile lemma, 149

Gibbs Russell, G.E. Analysis of the size and composition of the

southern African flora, 613

Gibbs Russell, G.E. & Glen, H.F. Register of names and types: a

comparison between Mesembryanthemaceae and Poaceae, 125

Gibbs Russell, G.E., Retief. E. & Smook, L. Intensity of plant

collecting in southern Africa, 131

Grass root pattern, 293

Grassland, 241

Grassy Fynbos, 194

Growth kinetics, 689

Harvesting, 733

Hecatris Salisb., 77

asparagoides (L.) Salisb., 78

Heksacentris Nees, 107

Helichrysum nudifolium - Protea roupelliae Sparse Woodland,

681

Henderson, L. Survey of exotic woody plant invaders of the

Transvaal — Addendum, 749

Henderson, L. & Musil, K.J. Exotic woody plant invaders of the

Transvaal, 297

Heppiaceae (Lichenes), 552

Herbarium, 131

Herschelia lugens var. nigrescens Linder, 554

Herschelianthe

barbata (L. f.) N.C. Anthony, 554

forficaria (H. Bol.) N.C. Anthony, 554

lugens (H. Bol.) Rauschert

var. nigrescens ( Linder ) N.C. Anthony, 554

newdigateae (L. Bol.) N.C. Anthony, 554

schlechteriana (H. Bol.) N.C. Anthony, 554

spathulata (L. f.) Rauschert

subsp. tripartita (Lindl.) N.C. Anthony, 554

Heteropogon contortus - Combretum molle Closed and Open

Woodlands, 676

Homalocheilos J.K. Morton, 8

ramosissimum (Hook, f.) J.K. Morton, 8

Hybridization, 591, 597

Hydrophilus Linder, 66, 484

rattrayi (Pillans) Linder, 66, 484

Hypodiscus Nees, 67, 488

albo-aristatus (Nees) Mast., 488

var. oliverianus (Mast.) Pillans, 488

sltermns Pillans, 489

argenteus (Thunb.) Mast. , 489

aristatus (Thunb.) Krauss, 489

var. bicolor Mast., 489

var. protractus (Mast.) Pillans, 489

binatus (Steud.) Mast., 489

capitatus Mast., 488

dodii Mast., 494

duplicatus Hochst., 488

eximius Mast., 490

gracilis Mast., 493

laevigatus (Kunth) Linder, 489

montanus Esterhuysen, 489

neesii Mast. , 490

nitidus Nees ex Mast., 481

oliverianus Mast., 488

paludosus Pillans, 492

parkeri Pillans, 492

procurrens Esterhuysen, 490

protractus Mast., 489

rigidusMrsf., 492

rugosus Mast., 492

squamosus Esterhuysen, 492

striatus (Kunth) Mast., 492

sulcatus Pillans, 493

synchroolepis (Steud.) Mast., 493

tristachyus Mast., 492

willdenowia (Nees) Mast., 493

zeyheri Mast., 492

Hypolaena R. Br.

anceps Mast., 436

aspera Mast., 465

bachmannii Mast. , 462

browniana Mast., 444

burchellii Mast., 469

conspicua Mast., 445

crinalis (Mast.) Pillans, 486

decipiens N.E. Br., 444

diffusa Mast., 463

digitata (Thunb.) Pillans, 487

eckloniana Nees ex Mast., 487

filiformis Mast., 467

foliosa Mast., 486

gracilis Nees ex Mast., 467

graminifolia (Kunth) Pillans, 486

hyalina Mast., 467

impolita (Kunth) Mast., 467

incerta Mast., 487

laxiflora Nees, 487

mahonii N.E. Br., 452

membranaceae Mast., 445

785

purpurea Pillans, 487

schlechteri Mast., 454

spathulata Pillans, 487

stokoei Pillans, 487

subtilis Mast., 444

tabularis Pillans, 486

tenuis Mast., 486

tenuissima Pillans, 415

virgata Mast., 444

Hypoporum Nees, 513

subgenus Hypoporum (Nees) C.B. Cl. , 513

section Hypoporum (Nees) Endl., 513

humile Nees, 522

nutans (Willd. ex Kunth) Nees, 522

pergracile Nees, 521

Immelman, K.L. Flowering in Kirkia wilmsii Engl., 15 1

IschyrolepiSiSfewd., 63, 397

affinis Esterhuysen, 401

arida (Pillans) Linder, 402

caespitosa Esterhuysen , 402

capensis (L.) Linder, 402

cincinnata (Mast.) Linder, 403

coactilis (Mast.) Linder, 403

curvibracteata Esterhuysen, 403

curviramis (Kunth) Linder, 404

distracta (Mast.) Linder, 404

duthieae (Pillans) Linder, 404

eleocharis (Mast.) Linder, 404

esterhuyseniae (Pillans) Linder, 404

feminea Esterhuysen, 404

fraterna (Kunth) Linder, 405

fuscidula (Pillans) Linder, 405

gaudichaudiana (Kunth) Linder, 405

gossypina (Mast.) Linder, 405

helenae (Mast.) Linder, 406

hystrix (Mast.) Linder, 406

kaiooica Esterhuysen, 406

laniger (Kunth) Linder, 406

leptoclados (Mast.) Linder, 407

longiaristata Pillans ex Linder, 407

macer (Kunth) Linder, 407

marlothii (Pillans) Linder, 407

monanthos (Mast.) Linder, 407

nana Esterhuysen, 409

nubigena Esterhuysen, 409

ocreata (Kunth) Linder, 410

paludosa (Pillans) Linder, 410

papillosa Esterhuysen, 410

pratensis Esterhuysen, 411

pygmaea (Pillans) Linder, 413

rivula Esterhuysen, 413

rottboellioides (Kunth) Linder, 413

sabulosa (Pillans) Linder, 413

schoenoides (Kunth) Linder, 413

setiger (Kunth) Linder, 413

sieberi (Kunth) Linder, 414

sporadica Esterhuysen, 414

subverticellata St eud. ,415

tenuissima (Kunth) Linder, 415

triflora (Rottb.) Linder, 415

unispicata Linder, 415

vilis (Kunth) Linder, 417

virgea (Mast.) Linder, 417

wallichii (Mast.) Linder, 417

wittebergensis Esterhuysen, 417

Isodon (Schrad. ex Benth.) Spach, 7, 8

ramosissimus (Hook, f.) Codd, 8

temifolius (D. Don) Kudo, 10

Isoetaceae, 555

Isoetes

capensis Duthie

var. stephansenii (Duthie) Schelpe & N.C. Anthony, 555

stephansenii Duthie, 555

Kaffrarian Succulent Thicket, 216

Kaffrarian Thicket, 213

Keys : Aneilema sp. (Commelinaceae), 90

Anthochortus sp. (Restionaceae), 484

Askidiosperma sp. (Restionaceae), 431

Calopsis sp. (Restionaceae), 464

Cannomois sp. (Restionaceae), 480

Carpha sp. (Cyperaceae), 142

Ceratocaryum sp. (Restionaceae), 479

Chondropetalum sp. (Restionaceae), 428

Ehrharta sp. (Poaceae), 149

Elegia sp. (Restionaceae), 419

Hypodiscus sp. (Restionaceae), 488

Ischyrolepis sp. (Restionaceae), 399

Mastersiella sp. (Restionaceae), 487

Myrsiphyllum sp. (Liliaceae), 78

Nevillea sp. (Restionaceae), 482

Platycaulos sp. (Restionaceae), 434

Protasparagus & Myrsiphyllum (Liliaceae), 78

Restio sp. (Restionaceae), 437

Restionaceae, 45, 394

Rhodocoma sp. (Restionaceae), 478

Scleria (Cyperaceae), 513, 514, 515

Staberoha sp. (Restionaceae), 395

Tetradenia sp. (Lamiaceae), 1

Thamnochortus sp. (Restionaceae), 471

Thunbergia sp. (Acanthaceae), 109

Willdenowia sp. (Restionaceae), 493

Kirkia wilmsii Engl. ,151

Knysna Afromontane Forest, 218

Kranz, 153

Kranz syndrome, 587

Lamiaceae, 1, 7, 142

Lamprocaulos Mast., 63

grandis (Nees) Mast., 424

neesii (Mast.) Mast., 425

schlechteri Gilg-Ben., 425

Lantana camara L., 161

Leaf anatomy, 153, 561, 567, 573, 579

Leaf variation, 101

Lepidanthus Nees, 67, 488

willdenowia Nees, 493

Leptocarpus R. Br.

andreaeanus Pillans, 465

asper (Mast.) Pillans, 465

brachiatus Mast., 442

burchellii Mast., 465

cymosus Mast., 444

distichus (Rottb.) Pillans, 445

divaricatus Mast., 462

ejuncidus (Mast.) Pillans, 446

esterhuyseniae Pillans, 466

festucaceus (Kunth) Mast., 470

fruticosus Mast., 467

gracilis (Mast.) Pillans, 467

hyalinus (Mast.) Pillans, 467

impolitus (Kunth) Pillans, 467

incurvatus Pillans, 470

incurvatus (Thunb.) Mast., 494

intermedius Pillans, 444

levynsiae Pillans, 467

marlothii Pillans, 467

membranaceus Pillans, 469

modestus (Kunth) Mast., 479

monostylis Pillans, 469

muirii Pillans, 469

neglectus (Hochst.) Mast., 453

nudiflorus Pillans, 469

oxylepis (Kunth) Mast., 470

786

paniculatus (Rottb.) Mast., 469

parkeri Pillans, 446

peronatus (Kunth) Mast., 470

var. hirtellus (Kunth) Mast., 470

ramosissimus Pillans, 467

rattrayi Pillans, 484

rigidus Mast., 470

rigoratus Mast., 470

var. simulans Pillans, 470

secundus Pillans, 460

stokoei Pillans, 457

vimineus (Rottb.) Pillans, 470

vai. hirtellus (Kunth) Pillans, 470

Leucoploeus Nees, 67, 488

argenteus Nees ex Mast., 489

Lichenes, 315,552

Liliaceae, 77

Linder, H.P. A phylogenetic classification of the genera of the

African Restionaceae, 1 1

Linder, H.P. Conspectus of the African species of Restionaceae,

387

Linder, H.P. A new name for Disa patens (Orchidaceae), 553

Malagasy Republic, 1

Marchantiales, 117, 531

Mastersiella Gilg-Ben., 66, 487

browniana (Mast.) Gilg-Ben., 444

diffusa (Mast.) Gilg-Ben., 463

digitata (Thunb.) Gilg-Ben., 487

foliosa (Mast.) Gilg-Ben., 486

hyalina (Mast.) Gilg-Ben., 467

laxiflora (Nees) Gilg-Ben., 487

purpurea (Pillans) Linder, 487

spathulata (Pillans) Linder, 487

Mast fruiting, 229

Medeola

angustifolia Mill., 78

asparagoides L., 78

Mediterranean-type shrublands, 175

Melianthaceae, 143, 145

Melianthemum, 145

Melianthus

dregeanus Sond. , 144

var. insignis (Kuntze) Phill. & Hofmeyer, 145

subsp. dregeanus, 145

subsp. insignis (Kuntze) Tansley , 145

gariepinus Merxm. & Roessler, 144

insignis Kuntze, 145

pectinatus Harv. , 143

subsp. gariepinus (Merxm. & Roessler) Tansley, 144

subsp. pectinatus, 144

sibiricus/W/. ex Georgi, 145

trimenianus Hook, f., 143

Meiosis, 597

Mesanthus Nees, 65, 480

macrocarpus Nees, 481

Mesembryanthemaceae, 125

Method, 241

Model application, 733

Model development, 689

Model validation, 705

Moraceae, 607

Morphology, 13, 506

Mundulea pondoensis Codd, 552

Musil, C.F. & Breen, C.M. The development from kinetic coef-

ficients of a predictive model for the growth of Eichhornia

crassipes in the field. 1. Generating kinetic coefficients for the

model in greenhouse culture, 689

Musil, C.F. & Breen, C.M. The development from kinetic coef-

ficients of a predictive model for the growth of Eichhornia

crassipes in the field. II. Testing and refining the model under

field conditions, 705

Musil, C.F. & Breen, C.M. The development from kinetic coef-

ficients of a predictive model for the growth of Eichhornia

crassipes in the field. III. Testing a model for predicting growth

rates from plant nutrient concentrations, 725

Musil, C.F. & Breen, C.M. The development from kinetic coef-

ficients of a predictive model for the growth of Eichhornia

crassipes in the field. IV. Application of the model to the

Vernon Hooper Dam — a eutrophied South African impound-

ment, 733

Myrsiphyllum Willd., 77

alopecurum Oberm. , 85

angustifolium (Mill.) Willd., 78

asparagoides (L.) Willd., 78

declinatum (L.) Oberm., 86

falciforme Kunth, 78

fasciculatum (Thunb.) Oberm., 87

Igramineum Kunth, 80

juniperoides (Engl.) Oberm., 84

kraussianum Kunth, 80

multituberosum (R.A. Dyer) Oberm., 77

ovatum (Salter) Oberm. , 7 9

ramosissimum (Bak.) Oberm., 87

scandens (Thunb.) Oberm., 86

undulatum (L. f.) Kunth, 83

volubile (Thunb.) Oberm., 82

Name changes, 147

Nematanthus Nees, 67, 493

ecklonii Nees, 494

Nerine

flexuosa (Jacq.) Herb., 545

humilis (Jacq.) Herb., 546

pulchella Herb., 546

tulbaghensis W.F. Barker, 546

Nevillea Esterhuysen & Linder, 66, 482

obtusissimus (Steud.) Linder, 66, 482

singularis Esterhuysen, 484

New combinations, 10, 64, 77, 79, 82, 84, 86, 87, 106, 143 — 145,

150, 151, 315, 320, 321, 387, 402-407, 410, 413-415,417,

431, 432, 436, 437, 444, 460, 465-467, 469, 470, 475, 479,

480, 486, 487, 489, 494, 552, 554, 555, 751

New names, 457, 463, 553, 751

New tax a, 5, 9, 64, 66, 85, 89-91, 94, 96-98, 101, 104, 117,

139, 142, 145, 151, 315, 317-320, 387, 396, 401-404,407,

409-411, 413-415, 417, 420, 421, 423, 424, 428, 432, 434,

443, 447, 449, 450, 453, 455, 457-460, 462, 463, 465, 466,

469, 472, 473, 477, 484, 489, 490, 492, 495, 531, 534, 550,

552,554,555,557, 751

Nomenclature, 387

Nutrient removal, 733

Nutrition, 229

Obermeyer, A. A. Revision of the genus Myrsiphyllum Willd., 77

Obermeyer, A. A. The genus Protasparagus in southern Africa

(Asparagaceae), 548

Oliver, E.G.H. A new species of Philippia from the Drakensberg

(Ericaceae), 550

Ophryoscleria Nees, 514

subgenus Scleria (Berg.) C.B. Cl., 513

section Ophryoscleria (Nees) C.B. Cl., 514

Ophrys patens L. f., 553

Orchidaceae, 553, 554

Orchis barbata L. f., 554

Osyris lanceolata-Celtis africana Kloof Forest, 666

Pachycarpus schinzianus-Combretum molle Open Woodland, 678

Palynology, 1 1, 46

Panarello, H.O. & Sanchez, E. The Kranz syndrome in the Era-

grostoideae (Choridoideae, Poaceae) as indicated by carbon

isotopic ratios, 587

Panicum maximum-Combretum molle Closed Woodland, 670

Parmelia

aggregata (D. Knox) Brusse, 320

astricta Brusse, 315

787

cedrus-montana (Brusse) Brusse, 320

clivorum Brusse, 315

conturbata Mull. Arg.

vax. exornata Zahlbr., 321

dysprosa (Brusse & D. Knox) Brusse, 321

exornata (Zahlbr.) Brusse, 321

karoo (D. Knox & Brusse) Brusse, 321

leucostigma (Brusse) Brusse, 321

marroninipuncta Brusse, 315

patula Brusse, 317

scitula Bmsse, 317

spargens Brusse, 318

unctula Brusse 319

vernicosa Brusse, 320

Pelargonium, 345, 346

section Pelargonium, 346

acerifolium L’Herit., 35 1

angulosum (Mill.) L’Herit., 349, 351

var. acerifolium (L’Herit.) Harv., 351

asperum Ehrh. ex Willd., 364

balsameum Jacq., 367

betulaefolium Salisb., 352

betulinum (L.) L’Herit., 351

capitatum (L.) L’Herit., 353

citronellum J.J.A. v.d. Walt, 370

cordatum L’Herit., 379

var. lanatum (Thunb.) Harv., 379

var. rubrocinctum (Link) Harv., 379

cordifolium (Cav.) Curtis, 379

corymbosum Turcz., 383

crispum (Berg.) L’Herit., 376

var. hermanniifolium (Berg.) Harv., 378

var. latifolium L’Herit., 376

var. majus DC., 376

cucullatum (L.) L’Herit., 348, 349, 350

subsp. cucullatum, 349

subsp. strigifolium Volschenk, 351

subsp. tabulare Volschenk, 350

denticulatum Jacq., 361

dregeanum Turcz., 379

dummondii Hook, f., 353

englerianum Knuth, 374

erectum Knuth, 359

georgense Knuth, 352

glutinosum (Jacq.) L’Herit., 359

graveolens L’Herit., 364

greytonense J.J.A. v.d. Walt, 375

hermanniifolium (Berg.) Jacq., 378

hispidum (L. f.) Willd., 381

intermedium Knuth, 364

karrooense Knuth, 358

lanatum (Thunb.) DC., 379

micranthum Eckl. & Zeyh., 383

multifidum Salisb., 365

panduriforme (sphalm.panduraeforme) Eckl.&Zeyh., 356

papilionaceum (L.) L’Herit., 382

penicillatum Willd., 352

populifolium Eckl. & Zeyh., 369

pseudoglutinosum Knuth, 363

var. scabridum (Knuth) Knuth, 363

quercifolium sensu L’Herit., 356

quercifolium (L. f.) L’Herit ., 358

var. pinnatifidum L’Herit., 358

xzAem H.E. Moore, 365

radula (Cav.) L’Herit., 365

ribifolium Jacq., 369

rubrocinctum Link, 379

scabroide Knuth, 373

scabrum (L.) L’Herit., 366

var. balsameum (Jacq.) Harv., 367

schonlandii Knuth, 369

sublignosum Knuth, 372

terebinthinaceum (Cav.) Desf., 364

tomentosum Jacq., 383

uniondalense Knuth, 363

var. scabridum Knuth, 363

vitifolium (L.) L’Herit., 355

Pentameris, 567, 573

longiglumis (Nees) Stapf, 567

macrocalycina (Steud.) Schweick., 579

obtusifolia (Hochst.) Schweick., 579

thuarii Beauv. , 573

Pentaschistis, 573

Penthea patens (L. f.) Lindl., 553

Phenolic compounds, 229

Philippia drakensbergensisE’.G.T/. Oliver, 550

Phyllocomos Mast., 66, 484

insignis Mast., 487

Phylogeny, 60

Phytocap, 749

Phytogeography, 131, 613

Phytosociology, 245, 655

Phytotab, 655, 749

Plant collectors, 631

Plant collecting, 131

Plant community, 245, 259

Plant families, 613

Plant invaders, 297, 749

Plant nutrient concentrations, 725

Platycaulos Linder, 64, 434

acutus Esterhuysen, 434

anceps (Mast.) Linder, 436

callistachyus (Kunth) Linder, 436

cascadensis (Pillans) Linder, 436

compressus (Rottb.) Linder, 64, 436

depauperatus (Kunth) Linder, 436

major (Mast.) Linder, 436

subcompressus (Pillans) Linder, 437

Plectranthus

sect. Isodon Schrad. ex Benth., 8

bullatus Robyns & Lebrun, 8

calycinus Benth., 10

var. pachystachyus (Briq.) T. Cooke, 10

dolomiticus Codd, 142

hoslundioides Bak., 8

pachystachyus Briq., 10

paniculatus Bak., 8

pyramidatus Giirke, 10

ramosissimus Hook, f., 8

schimperi Vatke, 8

ternifolius D. Don, 10

whytei Bak., 8

Pleopeltis/7.5./T. ex Willd. X Polypodium L., 557

macrocarpa forma sinuata (Sim) Schelpe, 557

XPleopodium Schelpe & N.C. Anthony, 557

simianum Schelpe & N.C. Anthony, 557

Pleuremidis Rafin., 107

Poaceae, 125, 145, 147, 149, 153, 167, 561, 567, 573, 579, 587

Pollinator attraction, 607

Polyploidy, 591, 597

Polypodium lanceolatum var. sinuatum Sim, 557

Polystichum

monticola N.C. Anthony & Schelpe, 554

pungens sensu Sim, 554

Protasparagus, 77, 78, 548

subgenus Africani, 548

subgenus Protasparagus, 548

Protea caffra— Combretum molle Open Woodland, 679

Protea-Clutia Community, 199

Pseudopentameris, 561, 567

Pteridophyta, 541, 554, 555

Rabdosia (Blume) Hasskarl, 8

788

Rabdosia sensu Codd, 9

calycina (Benth.) Codd, 10

ternifolia (D. Don) Haia, 10

Rabdosiella Codd, 7, 9

calycina (Benth.) Codd, 10

ternifolia (D. Don) Codd, 10

Reid, C. & Arnold, T.H. A new species of Carpha from the Natal

Drakensberg, South Africa, 139

Restio Rottb., 64, 437

acuminatus Thunb., 430

acuminatus Kunth, 481

acockii Pi/fans, 441

albo-aristatus Nees, 488

alticola Pillans, 441

ambiguus Mast, 441

ameles Steud., 415

anceps (Mast.) Pillans, 436

araneosus Mast., 415

arcuatus Mast., 441

argenteus Thunb., 489

aridus Pillans, 402

aristatus Thunb., 489

aspericaulis Pillans, 436

asperiflorus Nees, 420

aureolusP//to«s. 441

bifariu s Mast. ,441

bifidus Thunb., 441

bifurcus Nees ex Mast. , 442

bigeminus Nees ex Mast., 452

bolusii Pillans, 442

brachiatus (Mast.) Pillans, 442

brownianus (Mast.) Pillans, 444

brunneus Pillans, 442

burchellii Pz'/fons, 442

callistachyus Kunth, 436

c&piXiaiis Kunth, 443

cascadensis Pillans, 436

cemuus L. f., 396

chondropetalum Nees, 429

cincinnatus Mast., 403

cirratus Mast., 405

coactilis Mast., 403

colliculospermus Linder, 443

communis Pillans, 443

comosus N.E. Br., 479

compressus Rottb., 436

var. major Mast., 436

concolor Steud., 436

con fusus Pillans, 443

consimilis Mast., 414

conspicuus (Mast.) Pillans, 445

corneolus Esterhuysen, 443

crinalis Mast., 486

curviramis Kunth, 404

cuspidatus Thunb., 402

cymosus (Mast.) Pillans, 444

debilis Nees, 444

var. subulatus (Mast.) Pillans, 444

decipiens (N.E. Br.) Linder, 444

degenerans Pillans, 444

depauperatus Kunth, 436

dichotomus L., 402

digitatus Thunb., 487

dimorphostachyus Mast., 454

dispar Mast., 445

distachyos Rottb., 396

distans Pillans, 445

distichus Rottb. , 445

distractus Mast., 404

divaricatus Mast., 414

dodii Pillans, 445

var. purpureus Pillans, 445

duthieae Pillans, 404

echinatus Kunth, 445

ecklonii Mast., 470

egregius Hochst., 445

var. nutans Mast., 445

ejuncidus Mast., 446

elatus Mast., 405

elegans Poir, 482

elegia Murray, 425

eleocharis Nees ex Mast., 404

elongatus Thunb., 474

erectus Thunb. 474

esterhuyseniae Pillans, 404

exilis Mast. , 446

fastigiatus Nees ex Mast., 436

ferruginosus Link ex Kunth, 405

festuciformis Nees ex Mast, (‘festucaeformis’) , 446

filicaulis Pillans, 446

filiformis Poir. , 446

var. monostachyus (Mast.) Mast., 446

var. oligostachyus (Mast.) Mast., 446

foliosus N.E. Br., 479

fourcadei Pillans, 447

fragilis Esterhuysen, 447

fratemus Kunth, 405

fruticosus Thunb. 479

fuirenoides Kunth, 414

furcatus Nees ex Mast., 442, 454

fuscidulus Pillans, 405

fusiformis Pillans, 447

galpinii Pillans, 447

gamotianus Kunth, 446

var. monostachyus Steud. ex Mast., 446

var. oligostachyus Mast., 446

gaudichaudianus Kunth, 405

var. luxurious Pillans, 405

var. microstachyus Nees ex Mast., 405

giganteus (Kunth) N.E. Br., 479

glomeratus Thunb., 494

gossypinus Mast., 405

graminifolius Kunth, 486

grandis Spreng. ex Nees, 424

haxveyi Mast. , 447

helenae Mast., 406

humilis Pillans, 417

hystrix Mast., 406

imbricatus Thunb., 396

implexus Mast., 443

implicatus Esterhuysen, 447

impolitus Kunth, 467

inconspicuus Esterhuysen, 449

incurvatus Pillans, 470

incurvatus Thunb., 494

ingens Esterhuysen, 449

insignis Pillans, 45 0

intermedins Steud., 424

intermedins Kunth, 414

intricatus Mast., 455

inveteratus Esterhuysen, 450

involutus Pillans, 452

junceus (L.) Nees, 425

kunthii Steud., 415

laniger Kunth, 406

var. distractus (Mast.) Pillans, 404

leptoclados Mast., 407

leptostachyusATunt/r, 452

luceanus Kunth, 405

lucens Poir., 475

var. minor Mast., 475

ludwigii Steud., 415

789

macer Kunth, 407

macowanii Pillans, 479

madagascariensis Cherm. , 452

var. humbertii Cherm., 452

mahonii (N.E. Br.) Pillans, 452

major (Mast.) Pillans, 436

marlothii Pillans, 407

var. parviflorus Pillans, 407

mastersii F. Muell., 436

membranaceus Nees, 424

micansTVees, 452

miser Kunth, 452

monanthos Mast., 407

monostachyus Steud., 446

montanus Esterhuysen, 453

mucronatus Nees, 430

multicurvus N.E. Br., 446

multiflorus Spreng., 453

var. tuberculatus Pillans, 453

neesii Mast., 414

nodosus Pillans, 453

nudus (Rottb.) Nees, 430

nutans Thunb., 476

nutans Steud., 415

nuwebergensis Esterhuysen, 45 3

oblongus Mast., 486

ob scurus Pillans, 454

obtusissimus Steud., 484

occultus (Mast.) Pillans, 454

ocreatus Kunth, 410

oligostachyus Kunth, 446

pachystachyus Ai/nf/z, 454

paludosus Pillans, 410

paniculatus Rottb., 469

pannosus Mast., 462

papyraceus Pillans, 455

parviflorus Thunb., 481

patens Most., 455

pauciflorus Poir., 470

peculiaris Esterhuysen, 455

pedicellatusMzst. , 455

penicillatus Mast., 407

perplexusKiinf/i, 455

var. gracilis Mast., 443

perseverans Esterhuysen, 457

pillansii Linder, 457

polystachyus Kunth, 436

pondoensis Mast., 469

praeacutusMwf., 457

praefixus Mast., 436

procurrens Mast., 406

productus Mast., 406

propinquus Nees, 425

protractus Mast., 453

pseudoleptocarpus Kunth, 441

punctulatus Nees ex Mast., 445

pulvinatus Esterhuysen, 457

pumilus Esterhuysen, 458

purpurascens Nees ex Mast. , 458

pusillus Pillans, 452

pycnostachyus Mast., 414

pygmaeus Pillans, 413

quadratusMzsf., 458

quinquefarius Nees, 459

racemosus Poir., 426

ramiflorus Nees, 469

rams Esterhuysen, 459

rhodocoma Mast., 478

rottboellioides Kunth, 413

rupicola Esterhuysen, 459

sabulosus Pillans, 413

sarocladus Mast. , 460

scaber Mast., 460

scaberulus N.E. Br., 460

scariosus Thunb., 474

schlechteri (Mast.) Pillans, 454

schoenoides Kunth, 413

scopa Thunb., 482

scoparius Kunth, 414

scopula Mast., 455

secundus (Pillans) Linder, 460

sejunctus Mast. , 460

setiger Kunth, 413

sieberi Kunth, 414

var. schoenoides (Kunth) Pillans, 413

var. venustulus (Kunth) Pillans, 414

similis Pillans, 460

singularis Esterhuysen, 460

sonderianus Mast., 455

sparsus Mast., 461

spicifer Poir., 397

spicigerus Thunb. 478

spiculatus Mast., 414

spinulosus Kunth, 436

sprengelii Mast., 402

squarrosus Poir., 402

stereocaulisMzsf., 461

stokoei Pillans, 46 1

strictus N.E. Br., 461

strobilifer Kunth, 461

subcompressus Pillans, 437

subfalcatus Nees ex Mast., 414

subtilis Nees ex Mast. , 46 1

subulatus Mast., 444

subverticillatus (Steud.) Mast., 415

sulcatus Kunth, 493

synchroolepis Steud., 493

tabularis Pillans, 460

tectorum L. f., 430

tenuissimus Kunth, 415

tetragonus Thunb., 461

tetrasepalus Steud., 496

thamnochortus Thunb., 475

thyrsifer Rottb., 427

trichocaulis Mast., 455

triflorus Rottb., 415

triticeus7?orrZ>., 462

tuberculatus Pillans, 462

umbellatus Thunb., 396

vaginatus Thunb., 397

vallis-simius Linder , 462

venustulus Kunth, 414

verrucosus Esterhuysen, 462

versatilis Linder, 463

verticillaris L. f., 421

vi His Kunth, 417

vimineus Rottb., 470

virgatus Rottb., 482

virgeus Mast., 417

wallichii Mast., 417

xyridioides Kunth, 45 9

zuluensis Linder, 463

zwartbergensis Pillans, 464

Restionaceae, 11, 387

Retief, E. & Reyneke, W.F. The genus Thunbergia in southern

Africa, 107

Review of the work of the Botanical Research Institute, 1982/83,

323; 1983/84,761

Revision, 77, 107

Rhodocoma Nees, 65, 478

capensis Nees ex Steud. , 478

fruticosa (Thunb.) Linder, 479

790

Rhodocoma Nees continued

gigantea (Kunth) Linder , 479

Riccia, 117, 531

alatospora Volk & Perold, 5 34

duthieae Volk <& Perold, 531

parvo-areolata Volk & Perold, 117

villosa Steph. ex Brunnthaler, 120

section Pilifer, 117, 531

Roots, 229

Rosaceae, 101, 591, 597

Rubus, 101, 591, 597

subgenus Eubatus, 591

subgenus Idaeobatus, 591

longepedicellatus (C.E. Gust.) C.H. Stirton, 106

ludwigii Eckl. & Zeyh., 101

subsp. ludwigii, 102

subsp. spatiosus C.H. Stirton, 104

rhodacantha E. Mey., 101

rigidus Sm.

var. longepedicellatus C.E. Gust., 106

var. rigidus f. subinermis C.E. Gust., 106

IRuscus volubilis Thunb., 82

Rutherford, M.C. & Westfall, R.H. Sectors of the Transvaal

Province of South Africa, 294

Savanna, 655

Schizolepis Nees, 514

subgenus Scleria (Berg.) C.B. Cl., 513

section Schizolepis (Nees) C.B. Cl. , 514

Schmidia Wight., 107

Schoenus capensis L., 402

Schrire, B.D. A new combination in Tephrosia (Fabaceae), 55 1

Scientific names, 125

Scleria Berg. , 5 05 , 5 1 2

subgenus Scleria (Berg.) C.B. Cl. , 513

section Scleria (Berg.) Endl. ,513

achtenii De Wild., 526

acriulus C.B. Cl., 528

f. leopoldiana C.B. Cl. ex De Wild., 528

angusta Nees ex Kunth, 528

aquatica Cherm., 523

aterrima (Ridl.) Napper, 522

atrosanguinea Hochst. ex Steud., 518

bojeri C.B. Cl., 522

buchananii Boeck., 518

bullbifera Hochst. ex A. Rich., 518

caespitosa Welw. ex Ridl., 519

canaliculato-triquetra Boeck., 526

var. clarkeana Pierart, 526

catophylla C.B. Cl., 522

cenchroides Kunth, 522

centralis Cherm., 526

cervina Ridl., 526

dieterlenii Turrill, 521

diurensis Boeck., 526

dregeana Kunth, 519

dumicola Ridl., 525

flexuosa sensu Robinson, 521

foliosa Hochst. ex A. Rich., 525

friesii Kiik., 528

greigiifolia (Ridl.) C.B. CL, 527

hirtella Sw. sensu Boeck., 522

hirtella Sw.

var. aterrima Ridl., 522

var. chondrocarpa Nelmes, 522

var. tuberculata C.B. Cl., 522

holcoides Kunth, 519

junciformis Welw., 517

lacustris Wright, 523

lagoensis Boeck., 526

longigluma Kiik., 526

longispiculata Nelmes, 518

macrantha Boeck., 526

mayottensis C.B. Cl., 526

melanomphala Kunth, 526

meyeriana Kunth, 519

mollis Kunth, 522

moritziana Boeck., 526

natalensis C.B. Cl., 523

nutans Wild, ex Kunth, 522

nyasensis sensu Nelmes, 5 26

oryzoides Presl, 527

pergracilis (Nees) Kunth, 521

var. brachystachys Nelmes, 521

perrieri Cherm., 525

poiformis7?etz., 527

rehmannii C.B. CL, 517

schliebenii Gross, 518

schweinfurthiana sensu Hutch. & Dalz., 521

sch weinfurth iana Boeck., 518

setulosa Boeck., 519

sobolifer E.F. Franklin, 521

subintegriloba De Wild., 526

substriatoalveolata De Wild., 526

thomasii Pierart, 518

transvaalensis E.F. Franklin, 523

unguiculata E.A. Robinson, 525

vanderystii De Wild., 526

verdickii De Wild., 5 18

veseyfitzgeraldii E.A. Robinson, 5 18

welwitschii C.B. Cl., 517

var. tuberculata Cherm., 517

woodii C.B. CL, 5 16

Sclerophylly, 229

Sectors of the Transvaal Province, 294

Seed coat, 11, 30

Serapias patens (L. f.) Thunb., 553

Setaria megaphylla-Combretum molle Closed Woodland, 672

Simaroubaceae, 151

Smook, L. & Gibbs Russell, G.E. Name changes and additional

species of southern African Poaceae, 147

Soils, 229

South Coast Dune Fynbos, 201

South Coast Renosterveld, 205

South-eastern Cape, 175

South-eastern Mountain Fynbos, 193

Southern Africa, 89, 101, 107, 131,315,505,531,541,548,631

Southern Cape, 229

South-west Cape, 117

Species diversity, 131,613

Spies, J.J. Embryo sac development in some South African

Lantana species (Verbenaceae), 161

Spies, J.J. & Du Plessis, H. The genus Rubus in South Africa.

I. Chromosome numbers and geographical distribution of

species, 591

Spies, J.J., Du Plessis, H. & Liebenberg, H. The genus Rubus in

South Africa. II. Meiotic chromosome behaviour, 597

Sporangia, 531

Spores, 117, 531

Sporophytes, 117

Staberoha Kunth, 63, 396

aemula (Kunth) Pillans, 396

banksiiF/7/aws, 396

caricina (Mast.) Dur. & Schinz, 474

cernua (L. f.) Dur. & Schinz, 396

distachyos (Rottb.) Kunth, 396

disticha (Rottb.) Dur. & Schinz, 445

gracilis (Mast.) Dur. & Schinz, 474

imbricata (Thunb.) Kunth, 396

var. stenoptera (Kunth) Dur. & Schinz, 396

multispicula Pillans, 396

ornata Esterhuysen, 396

remota Pillans, 397

791

stenoptera Kunth, 396

stokoei Pillans, 397

vaginata (Thunb.) Pillans , 397

Staff of the National Herbarium. New taxa, new records and

name changes for southern African plants, 751

Stirton, C.H. Notes on the germs Rubus in southern Africa, 101

Structural classification, 655

Subtropical Transitional Thicket, 209

Survey, 297

Syntaxonomy, 175

Tansley, S.A. & Schelpe, E.A.C.L.E. Two new combinations in

Melianthus, 143

Tansley, S.A. & Schelpe, E.A.C.L.E. An excluded species in

Melianthus , 145

Taxonomic revision, 345

Taylor, H.C. A vegetation survey of the Cape of Good Hope

Nature Reserve. I. The use of association-analysis and Braun-

Blanquet methods, 245

Taylor, H.C. A vegetation survey of the Cape of Good Hope

Nature Reserve. II. Descriptive account, 259

Taxonomy, 1, 7, 11, 77, 89, 101, 107, 117, 139, 142, 143, 145,

147, 149, 345, 387, 505, 531, 541, 545, 547, 548, 550-555,

751

Tephrosia pondoensis (Codd) Schrire, 552

Terminalia sericea-Combretum molle Closed Woodland, 672

Tetradenia Benth., 1

fruticosa Benth., 1

goudotii Briq. , 2

hildebrandtii Briq., 2

nervosa Codd, 5

Thallus, 531

ThamnochortusBerg., 65, 471

acuminatus Pillans, 472

aemulus Kunth, 396

arenarius Esterhuysen, 472

argenteus (Thunb.) Kunth, 489

argenteus Pillans, 473

bachmanniiAfosL , 473

bromoides Kunth, 475

burchellii Mast., 474

canescens Mast., 473

caricinus Mast., 474

cernuus (L. f.) Kunth, 396

cinereus Linder, 473

comptonii Pillans, 476

consanguineus Kunth, 476

dichotomus (L.) Spreng., 402

dichotomus Mast., 475

dichotomus Pillans

var. hyalinus Pillans, 475

distichus (Rottb.) Mast., 445

dumosus Mast., 473

ecklonianus Kunth, 475

ellipticus Pillans, 474

elongatus (Thunb.) Mast., 474

erectus (Thunb.) Mast., 474

floribundus Kunth, 474

fratemus Pillans, 474

fruticosus Berg. , 474

var. glaber Mast., 474

giganteus Kunth, 479

glaber (Mast.) Pillans , 474

gracilis Mast. , 474

guthrieae Pillans, 475

imbricatus (Thunb.) Mast., 396

var. stenopterus (Kunth) Mast., 396

insignis Mast. , 475

levynsiae Pillans, 475

lewisiae Pillans, 475

lucens (Poir.) Linder, 475

mastersii Gand., 497

maximus Kuntze, 478

membranaceus Mast., 445

micans (Nees) Kunth, 452

modestus Kunth, 479

muirii Pillans, 476

muticus Pillans, 478

nervosus Pillans, 475

nutans (Thunb.) Pillans, 476

obtusus Pillans, 476

occultus Mast., 454

paniculatusA/izsf, 476

papillosus Pillans, 475

papyraceus Pillans, 476

pellucidus Pillans, 416

piketbergensis Pillans, 478

platypteris Kunth, 476

plumosus Pillans, 475

pluristachyus Mast. , 476

pulcher Pillans, 477

punctatusPB/aws, 477

rigidu s Esterhuysen ,477

robustus Kunth, 482

scabridus Pillans, 477

scariosus (Thunb.) Spreng., 474

schlechteri Pillans, 478

scirpiformis Mast., 474

scirpoides Kunth, 482

similis Pillans, 478

spicigerus (Thunb.) Spreng., 478

sporadicus Pillans, 478

stokoei Pillans, 478

striatus Hochst., 478

strictus Kunth, 481

sulcatus Mast., 473

umbellatus (Thunb.) Kunth, 396

virgatus (Rottb.) Kunth, 482

Thamnochortus-Erica Community, 199

Thamnochortus-Tristachya Community, 201

Themeda-Passerina Community, 199

Themeda triandra-Combretum molle Open Woodland, 677

Thunbergia Retz., 107

alata Sims, 114

amoena C.B. Cl., 114

aspera Nees, 111

var. parvifolia Sond., Ill

atriplicifolia E. Mey. ex Nees, 110

var. kraussii C.B. Cl., Ill

aureaALB. Br., 115

bachmannii Lindau, 111

var. minor C.B. Cl., Ill

baurii Lindau, 111

capensis Retz., Ill

cordibracteolata C.B. CL, 111

dregeana Nees, 114

dregeana sensu Van der Schijff, 113

flavohirta Lindau, 111

fragrans sensu E. Mey., 1 14

galpinii Lindau, 1 1 1

hirta Sond., 113

hirtistyla C.B. CL, 111

natalensis Hook., 108

neglecta Sond. , 113

pondoensis Lindau, 113

purpurata Harv. ex C.B. Cl., 113

venosa C.B. Cl., 109

xanthotricha Lindau, 111

Transvaal, 294, 297, 749

Tropical East Africa, 89

Type section, 345

Type specimens, 125

Typification, 387

792

Ubbink, B. & Bxedenkamp, GJ. A new record of Acorns calamus

in South Africa (Araceae), 547

Van Daalen, J.C. Distinguishing features of forest species on

nutrient-poor soils in the Southern Cape, 229

Van der Walt, J.J.A. A taxonomic revision of the type section of

Pelargonium L’Herit. (Geraniaceae), 345

Vegetation, 655

Vegetation characterization, 175

Vegetation dynamics, 175

Vegetation map of the Cape of Good Hope Nature Reserve, 287

Vegetation survey, 245, 259

Verbenaceae, 161

Volatile compound, 607

Volk, O.H. & Perold, S.M. Studies in the liverwort genus Riccia

(Marchantiales) from the south-west Cape, 117

Volk, O.H. & Perold, S.M. Studies in the genus/?/ccza (Marchant-

iales) from southern Africa. 1. Two new species of the section

Pilifer : R. duthieae and R. alatospora, 531

Vorster, T.B. & Liebenberg, H. Classification of embryo sacs in

the Eragrostis curvula Complex, 167

Wasp pollinator, 607

Water nutrient concentrations, 705

Westfall, R.H. Phytocap. A field-data capture programme for

the phytotab program package, 749

Westfall, R.H. & Drewes, R. Grass root pattern in an Orange Free

State floodplain, 293

Westfall, R.H. & Panagos, M.D. A cover meter for canopy and

basal cover estimations, 241

Westfall, R.H., Van Rooyen, N. & Theron, G.K. The plant ecol-

ogy of the farm Groothoek, Thabazimbi District. II. Classif-

ication, 655

Willdenowia Thunb., 67, 493

affinis Pillans, 493

aiescens Kunth, 493

argentea (Kunth) Hieron., 479

bolusii Pillans, 494

brevis Nees ex Mast., 493

compressa Thunb., 481

cuspidata Mast., 494

decipiens N.E. Br., 479

ecklonii (Nees) Kunth, 494

ecklonii (Nees) Dur. & Schinz, 486

esterhuyseniae Pillans, 479

fimbriata Kunth, 479

fistulosa (Mast.) Pillans, 480

fratema N.E. Br., 495

galpinii N.E. Br., 495

glomerata (Thunb.) Linder , 494

humilis Nees ex Mast. , 494

incurvata (Thunb.) Linder, 494

lucaeana Kunth, 494

neglecta Steud., 494

peninsularis N.E. Br., 495

purpurea Pillans, 494

rugosa Esterhuysen , 495

simplex N.E. Br., 489

stokoei Pillans, 495

striata Thunb., 494

sulcata Mast. , 495

teres Thunb., 495

xerophila Pillans, 480

Woodland, 241

Woody invader, 297

Xanthoparmelia

aggregata D. Knox, 320

cedrus-montana Brusse, 320

dysprosa Brusse, 321

exornata (Zahlbr.) Brusse & D. Knox, 321

karoo D. Knox & Brusse, 321

leucostigma Brusse, 321

i

-• \