Bothalia

A JOURNAL OF BOTANICAL RESEARCH

Vol. 30,1

May 2000

TECHNICAL PUBLICATIONS OF THE NATIONAL BOTANICAL INSTITUTE,

PRETORIA

Obtainable from the National Botanical Institute, Private Bag X101, Pretoria 0001, Republic of

South Africa. A catalogue of all available publications will be issued on request.

BOTHALIA

Bothalia is named in honour of General Louis Botha, first Premier and Minister of Agriculture of

the Union of South Africa. This house journal of the National Botanical Institute, Pretoria, is

devoted to the furtherance of botanical science. The main fields covered are taxonomy, ecology,

anatomy and cytology. Two parts of the journal and an index to contents, authors and subjects are

published annually.

Two booklets of the contents (a) to Vols 1-20 and (b) to Vols 21-25, are available.

STRELITZIA

A series of occasional publications on southern African flora and vegetation, replacing Memoirs of

the Botanical Survey of South Africa and Annals of Kirstenbosch Botanic Gardens.

MEMOIRS OF THE BOTANICAL SURVEY OF SOUTH AFRICA

The memoirs are individual treatises usually of an ecological nature, but sometimes dealing with

taxonomy or economic botany. Published: Nos 1-63 (many out of print). Discontinued after No. 63.

ANNALS OF KIRSTENBOSCH BOTANIC GARDENS

A series devoted to the publication of monographs and major works on southern African flora.

Published: Vols 14-19 (earlier volumes published as Supplementary volumes to the Journal of

South African Botany). Discontinued after Vol. 19.

FLOWERING PLANTS OF AFRICA (FPA)

This serial presents colour plates of African plants with accompanying text. The plates are prepared

mainly by the artists at the National Botanical Institute. Many botanical artists have contributed to

the series, such as Fay Anderson, Peter Bally, Auriol Batten, Gillian Condy, Betty Connell, Stella

Gower, Rosemary Holcroft, Kathleen Lansdell, Cythna Letty (over 700 plates), Claire Linder-

Smith and Ellaphie Ward-Hilhorst. The Editor is pleased to receive living plants of general interest

or of economic value for illustration.

From Vol. 55, twenty plates are published at irregular intervals.

An index to Vols 1-49 is available.

FLORA OF SOUTHERN AFRICA (FSA)

A taxonomic treatise on the flora of the Republic of South Africa, Lesotho, Swaziland, Namibia

and Botswana. The FSA contains descriptions of families, genera, species, infraspecific taxa, keys

to genera and species, synonymy, literature and limited specimen citations, as well as taxonomic

and ecological notes.

Contributions to the FSA also appear in Bothalia.

PALAEOFLORA OF SOUTHERN AFRICA

A palaeoflora on a pattern comparable to that of the Flora of southern Africa. Much of the informa-

tion is presented in the form of tables and photographic plates depicting fossil populations. Now

available:

Molteno Formation (Triassic) Vol. 1. Introduction. Dicroidium, 1983, by J.M. & H.M.

Anderson.

Molteno Formation (Triassic) Vol. 2. Gymnosperms (excluding Dicroidium), 1989, by J.M.

& H.M. Anderson.

Prodromus of South African Megafloras. Devonian to Lower Cretaceous, 1985, by J.M. &

H.M. Anderson. Obtainable from: A. A. Balkema Marketing, Box 3 17, Claremont 7735,

RSA.

Towards Gondwana Alive. Promoting biodiversity and stemming the Sixth Extinction, 1999,

by J.M. Anderson (ed.)

BOTHALIA

A JOURNAL OF BOTANICAL RESEARCH

Volume 30,1

Scientific Editor: G. Germishuizen

Technical Editor: B.A. Momberg

NASIONALE BOTANIESE

INSTITUUT

Private Bag X101 PRETORIA 0001

2000 -05- | 7 S

Pflvaatsak X 101 PRETORIA 0001

national botanical

INSTITUTE

NATIONAL

O T A N I C A L

INSTITUTE

2 Cussonia Avenue, Brummeria, Pretoria

Private Bag X101, Pretoria 0001

ISSN 0006 8241

May 2000

Editorial Board

D.F. Cutler

B.J. Huntley

P.H. Raven

J.P. Rourke

M.J. Werger

Royal Botanic Gardens, Kew, UK

National Botanical Institute, Cape Town, RSA

Missouri Botanical Garden, St Louis, USA

Compton Herbarium, NBI, Cape Town, RSA

University of Utrecht, Utrecht, Netherlands

Acknowledgements to referees

Archer, Mrs C., National Botanical Institute, Pretoria, RSA.

Barrington, Dr D.S., University of Vermont, Burlington, USA.

Bhat, Prof. R.B., University of Transkei, RSA.

Boucher, Prof. C., University of Stellenbosch, RSA.

Burgoyne, Ms P., National Botanical Institute, Pretoria, RSA.

Cook, Prof. C.D.K., University of Zurich, Switzerland.

Forster, P.I., Queensland Herbarium, Brisbane Botanic Gardens, Australia.

Goldblatt, Dr P, Missouri Botanical Garden, St Louis, USA.

Hammer, S., Sphaeroid Institute, Vista, USA.

Hanna, Dr W., Coastal Plain Experimental Station, Tifton, USA.

Ihlenfeldt, Prof. H-D., Waabs/Longholz, Germany.

Jordaan, Mrs M., National Botanical Institute, Pretoria, RSA.

Leistner, Dr O.A., National Botanical Institute, Pretoria, RSA.

Long, Dr D.G., Royal Botanic Garden, Edinburgh, UK.

Louw, Dr A., Nelspruit, RSA.

Manning, Dr J., National Botanical Institute, Cape Town, RSA.

Mucina, Prof. L., Kuwait University. Present address: University of Stellenbosch, RSA

Nelson, Dr E.C., Outwell, Wisbech, UK.

Oliver, Dr E.G.H., National Botanical Institute, Cape Town, RSA.

Phillipson, Dr P.B., Rhodes University, Grahamstown, RSA.

Rauh, Prof. W., Heidelberg, Germany.

Retief, Ms E., National Botanical Institute, Pretoria, RSA.

Rourke, Dr J.P, National Botanical Institute, Cape Town, RSA.

Smith, Prof. G.F., National Botanical Institute, Pretoria, RSA.

Steyn, Dr E.M.A., National Botanical Institute, Pretoria, RSA.

Thiede, Dr J., University of Cologne, Germany.

Thulin, Dr M.L., University of Uppsala, Sweden.

Van Jaarsveld, E.J., National Botanical Institute, Pretoria, RSA.

Van Wyk, Prof. A.E., University of Pretoria, RSA.

Van Wyk, Prof. B-E., Rand Afrikaans University, Johannesburg, RSA.

Verdcourt, Dr B., Royal Botanic Gardens, Kew, UK.

Yatskievych, Dr G., Missouri Botanical Garden, St Louis, USA.

CONTENTS

Volume 30,1

New genus, species and combinations in Bothalia 30,1 (2000) iv

1. Notes on the genus Frithia (Mesembryanthemaceae) and the description of a new species, F. humilis,

in South Africa. P.M. BURGOYNE, G.F. SMITH and F. DU PLESSIS 1-7

2. A review of generic concepts in the Stilbaceae. J.P ROURKE 9 ~~

3. Studies in the Sphaerocarpales (Hepaticae) from southern Africa. 2. The genus Sphaerocarpos and its

only local species, S. stipitatus. S.M. PEROLD 17 --2.4

4. FSA contributions 15: Piperaceae. K.L. IMMELMAN 25 — 3^

5. FSA contributions 16: Sphenocleaceae. W.G. WELMAN 31—33

6. Taxonomic studies in the Aizoaceae from South Africa: three new species and some new combina-

tions. C. KLAK 35 -U'i

7. Notes on African plants:

Agavaceae. Agave vivipara : a naturalised alien in southern Africa. E.M.A. STEYN and

G.F. SMITH 43

Asphodelaceae: Alooideae. The genus Poellnitzia included in Astroloba. J.C. MANNING and

G.F. SMITH 53

Asphodelaceae: Alooideae. Aloe delphinensis in Aloe sect. Lomatophyllum. PI. FORSTER ... 53 —

Ericaceae. Two new species of Erica from Western Cape, South Africa. E.G.H. OLIVER and

I.M. OLIVER 49-^3

Hyacinthaceae. Correction of a historical error in the taxonomic description of Urginea ciliata.

A.P DOLD and R. MOBERG 46 -U-8

8. Morphology and anatomy of the rhizome and frond in the African species of Polystichum (Pteropsida:

Dryopteridaceae). J.P. ROUX and A.E. VAN WYK 57 — 6 c?

9. The epidermis in Passerina (Thymelaeaceae): structure, function and taxonomic significance.

C.L. BREDENKAMP and A.E. VAN WYK 69 -

10. Vegetation of the coastal fynbos and rocky headlands south of George, South Africa. D.B. HOARE,

J.E. VICTOR, R.A. LUBKE and L. MUCINA 87-96

11. Checklist of plant species of the coastal fynbos and rocky headlands, south of George, South Africa.

J.E. VICTOR, D.B. HOARE and R.A. LUBKE 97 - to I

12. Miscellaneous notes:

Hyacinthaceae. Chromosome studies on African plants. 13. Lachenalia mutabilis, L. pustulata and

L. unicolor. J.J. SPIES, J.L. DU PREEZ, A. MINNAAR and R. KLEYNHANS 106 - 1 lo

Poaceae. Apomictic embryo sac development in Cenchrus ciliaris (Panicoideae). N.C. VISSER,

J.J. SPIES and H.J.T. VENTER 103 ^ / o*G

13. Obituaries:

Leslie Edward Wostall Codd (1908-1999). B. DE WINTER and G. GERMISHUIZEN Ill - 0 S'

Hugh Colin Taylor (1925-1999). D.J. MCDONALD, C. BOUCHER and E.G.H. OLIVER ... 115-110

Barbara Joan Jeppe (1921-1999): botanical artist extraordinaire. G.F. SMITH and E.M.A.

STEYN 119 -111

14. Book reviews 123

On 37 E2S©r4 , If ■ I'd . ({suruu.0 1-2. w-

G-, f ; (2 3,124-

6 LE N , R. , H • f, ' 1 7. t|

New genus, species and combinations in Bothalia 30,1 (2000)

Antimima excedens ( L.Bolus ) Klak, comb, nov., 35

Astroloba rubriflora ( L.Bolus ) G.F. Sm. & J.C. Manning, comb, nov., 53

Brownanthus fraternus Klak, sp. nov., 37

Brownanthus glareicola Klak, sp. nov., 37

Erepsia dunensis ( Sond .) Klak, comb, nov., 38

Erica columnaris E.G.H.Oliv., sp. nov., 50

Erica orthiocola E.G.H.Oliv., sp. nov., 49

Frithia humilis P.M.Burgoyne, sp. nov., 1

Hammeria meleagris ( L.Bolus ) Klak, comb, nov., 39

Kogelbergia Rourke, genus novum, 12

Kogelbergia phylicoides (A. DC.) Rourke, comb, nov., 13

Kogelbergia verticillata (Eckl. & Zeyli.) Rourke, comb, nov., 12

Scopelogena bruynsii Klak, sp. nov., 39

Stilbe gymnopharyngia ( Rourke ) Rourke, comb, nov., 11

Stilbe overbergensis Rourke, nom. nov., 1 1

Stilbe serrulata ( Hochst .) Rourke, comb, nov., 11

iv

Bothalia 30,1 : 1-7 (2000)

Notes on the genus Frithia (Mesembryanthemaceae) and the descrip-

tion of a new species, F. humilis , in South Africa

P.M. BURGOYNE*, G.F. SMITH* and F. DU PLESSIS**

Keywords: Frithia N.E.Br., Mesembryanthemaceae, new species, South Africa, summer rainfall mesembs, taxonomy, window plants

ABSTRACT

Frithia N.E.Br. (Mesembryanthemaceae), formerly thought to be a monotypic genus, has been found to comprise two

species. Populations from the eastern parts of the distribution range of Frithia pulchra N.E.Br. are recognised as a distinct

species, Frithia humilis P.M.Burgoyne. The genus has a limited distribution, although present in three provinces of South

Africa, namely Gauteng, North-West and Mpumalanga. The two window-leaved species are allopatric and morphological

differences between the roots, leaves, flowers, pollen, capsules and seeds are discussed. A formal description of the new

species, an identification key and a distribution map of the two species are provided.

INTRODUCTION AND HISTORICAL OUTLINE

The genus Frithia (N.E.Br.) was first mentioned in a

key by Brown (1925). At that time no species were

assigned to the genus and only later was a full descrip-

tion of Frithia pulchra given (Brown 1926). The genus

was named after Frank Frith (1872-1954), a railway ser-

vices gardener stationed at Park Station, Johannesburg.

He was responsible for decorating railway platforms

from 1900 until his retirement in 1932 (Kroon 1997). In

1906, Olive Nation sent a specimen of a plant she dis-

covered near Rustenburg to Brown at Kew for identifi-

cation. The live specimen did not survive the journey, but

the remains were seen by Brown who regarded it as a

distinct entity. After Miss Nation died, a search for more

material proved unsuccessful. Some time later, a Mrs

Dobie of Rustenburg sent plants to Frank Frith, who took

the specimens to Brown at Kew while on a visit to

London, to create the African garden at the Wembley

Exhibition. Dobie’s specimen allowed Brown to finally

describe the genus.

De Boer (1968) published the name Frithia pulchra

var. minor in the Dutch journal Succulenta, but as no

type material was cited the name was invalid. Plants of

this variety are generally smaller than those included in

var. pulchra and are restricted to the eastern parts of the

distribution range of the genus. These differences were

again alluded to by Hardy & Fabian (1992). Zimmer-

mann (1996) confirmed the different characters of var.

minor, but gave no formal description or type validating

the varietal epithet. Although a number of short articles

have been written on Frithia (Brink 1985; Germishuizen

1975; Steffens 1988; Venter 1979, 1983), no in-depth

study has been done on the genus.

This paper reports on the taxonomic status of the

genus, and specimens from the eastern parts of its distri-

bution range are formally described as a new species,

Frithia humilis.

* National Botanical Institute, Private Bag X101, 0001 Pretoria.

** Department of Botany, University of Pretoria, 0002 Pretoria.

MS. received: 1999-03-15.

TAXONOMY

To prevent possible confusion with the illegitimate

name Frithia pulchra N.E.Br. var. minor de Boer, the

varietal epithet minor is not used at the specific rank. The

name of the new species is derived from the Latin

humilis, which means ‘smaller than others of its kind’.

Key to species

Window of leaf tips convex with no markings along margins;

leaves 15-25 mm long, blue-green or grey-green; flow-

ers bright magenta with gold, yellow or white centre,

25-35 mm diam.; growing west of Pretoria in the Rus-

tenburg area 1. F. pulchra

Window of leaf tips concave with crenulate markings along

margins; leaves shorter than 15 mm, brown-green or

dull green; flowers white with yellow centre, pale pink,

rarely entirely pink, petals sometimes tipped with pale

pink, 15-20 mm diam., when pollinated turn pale yel-

low or salmon-orange; growing east of Pretoria in the

vicinity of Bronkhorstspruit and Witbank 2. F. humilis

Frithia humilis P.M.Burgoyne, sp. nov.

Frithia pulchra N.E.Br. var. minor de Boer: 147, 148 (1968), nom.

illeg.

Plantae perennes nanae succulentae, radicibus succu-

lentis lateraliter ramosis, tempore mensium hibernium in

humum arenosum retractae per foliis contractilibus lon-

gitudinaliter vietis fiunt praesentia eorundum cavis relic-

tis indicata. Caulis simplex, brevis, ad 10 mm longus.

Folia spiral iter disposita, obscure virides ad brunneo-

virides, mensis aridis hibernis purpureo-suffusa, maxime

succulenta, ad 15 mm longa, cylindrica, idioblastis cera-

ceis distincte serialibus, apicibus foliorum fenestratis

cum centra concavo, maculisque conspicuis crenulatis

perimetro. Flores solitares, 15-20 mm diametro, albi vel

perdilute rosei, centro flavo, ante ad post meridiem ape-

rientes. Sepala 5, inaequalia, folia simulantia, tubum

brevem connata. Petala 20-30, in verticillis plures dis-

posita, apices plerumque acuminati, interdum rotundati.

Staminodia petaloidea ad filiformia, verticillos aliquot

staminum cingentia. Hypanthium basibus connatis

petalorum staminodium staminumque formatum. Necta-

2

Bothalia 30,1 (2000)

ria 5, libra, atroviridia, crenulata. Ovarium supra leviter

conicum, stigmata 5 vel 6, perbrevia. Fructus capsula 5-

vel 6-locularis, doliiformis, perfragilis, maturitate

rumpens, semina spargens; valvae ad positionem erec-

tam aperientes, alae valvarum absentes, margines val-

varum recurvati ubi omnis aperti, cristae turgescentes

pallide luteo-brunneae, apicibus divergentibus margin-

ibusque scissis, membranae tegentes ad regulam reduc-

tae. Semina rubro-brunnea, parva, tuberculis parvis tecta.

Florescentia a Decembri ad Februarium.

TYPE. — Gauteng, 2528 (Pretoria): Bronkhorstspruit

Dist., 29 km northeast of Bronkhorstspruit on tarred road

to Verena, then 3 km along road to Susterstroom, on Farm

Susterstroom, in sandy flat areas associated with rough

rocky outcrops, (-BD), Burgoyne 6693 (PRE, holo.).

Perennial, dwarf succulent with fleshy roots branch-

ing laterally; plants retracting into sandy soil by means of

contractile leaves shrinking lengthways during dry win-

ter months, leaving holes marking their presence. Stems

single, short, up to 10 mm long. Leaves arranged spiral-

ly, dull green to brown-green with a purple tinge in dry

winter months, highly succulent, up to 15 mm long,

cylindrical, covered by waxy idioblasts arranged in dis-

tinct rows, tips windowed, with concave centre and con-

spicuous crenulate markings along perimeter. Flowers

single, 15-20 mm diam., white or very pale pink, with

yellow centre, opening during mid-morning to mid-after-

noon. Pollinated flowers turn yellow or salmon-orange.

Sepals 5, unequal, resembling leaves, united to form a

short tube. Petals 20-30 per flower, arranged in several

whorls, tips mostly acuminate, sometimes rounded.

Staminodes petaloid to filiform, surrounding several

whorls of stamens. Hypanthium formed by fused bases

of petals, staminodes and stamens. Nectaries 5, free, dark

green, crenulate. Ovary slightly conical above; stigmas 5

or 6, very short. Fruit a capsule, 5- or 6-locular, barrel-

shaped, very fragile, breaking up when ripe and then dis-

persing seeds; valves opening to an upright position,

valve wings absent, valve margins recurved when fully

open; expanding keels light yellow-brown, parallel, with

diverging tips and torn margins; covering membranes

reduced to a ledge; closing bodies absent. Seeds red-

brown, small, covered by small tubercles. Flowering

time: December-February (summer in the southern

hemisphere).

As the formerly monotypic genus Frithia now has

two species, the type species of the genus is Frithia pul-

chra N.E.Br.

SPECIMENS EXAMINED

All specimens held at PRE.

Burgoyne 6692 , 6693, 6694, 6694b, 6696, 6698, 6699 (2), 6699b,

6699c (1).

Crundall PRE5498J (2).

Dyer 4774 ( 1 ). Dyer & Verdoorn 3922 ( 1 ).

Gilfdlan 7272 (2).

Jacobsen 758 (1).

Rose Innes 167 (1).

Van PRE54978 ( 1 ). Venter 2997 (2).

Young 38395 ( 1).

DISTRIBUTION AND HABITAT

Frithia is one of the few genera in the Mesem-

bryanthemaceae exclusive to the summer rainfall region

of South Africa. Other mesemb genera with a distinctly

summer rainfall distribution include Neohenricia , Mos-

sia and Khadia, while Delosperma, Hereroa, Lithops,

Chasmatophyllum, Nananthus and Stomatium may occur

in summer rainfall areas but also have wider distribu-

tions.

Previously thought to be a monotypic genus and a

Magaliesberg endemic, an enlarged Frithia still has a

restricted distribution. Populations of these miniature

window plants have been found in two disjunct regions,

in the North-West between Rustenburg and the Harte-

beespoort Dam in the west, and in an area between

Bronkhorstspruit (Gauteng) and Witbank (Mpumalanga)

in the east (Figure 1). The two areas are roughly 150 km

apart, and so far, no specimens of either species have

been collected in the intervening area.

Both species of Frithia grow in very shallow soils

derived from coarse sediments: quartzites of the

Magaliesberg Formation of the Pretoria Group of the

Transvaal Supergroup in the case of Frithia pulchra and

sandstones of the Irrigasie Formation of the Ecca Group

of the Karoo Supergroup for F. humilis. Rocks in both

areas are very rough, porous and weather to form a very

coarse gravel.

Frithia pulchra mostly grows exposed on rock plates,

the roots anchored in cracks between the coarse

quartzites. This substrate reaches very high temperatures

in summer. Plants are also found in coarse gravel and are

not restricted to rocky outcrops. F. humilis grows pre-

dominantly in shallow sand along the rims of large, flat,

rock plates. Temperatures of the substrate are probably

lower as more organic matter is present, insulating the

plant bodies against heat and dessication.

Both species grow at altitudes ranging from 1 368 m

to 1 616 m, and rainfall varies between 700 and 800 mm

per annum. Winters are cold and dry and severe frost

occurs in the areas where the plants grow.

FIGURE 1. — Known distribution of Frithia pulchra ■, and F. humi-

lis, • .

Bothalia 30,1 (2000)

3

Other species often associated with both species of

Frithia are the fern ally Selaginella dregei and the

legume Indigofera melanadenia. Species sometimes

found associated with either species of Frithia are the

succulents Anacampseros subnuda subsp. subnuda ,

Crassula lanceolata subsp. transvaalensis,' C. setulosa

var. setulosa and Mossia intervallaris. Monocots like

Microchloa kunthii, Anthericum calyptocarpum together

with an extremely minute and monophyllous species of

Drimia , are also found in these habitats. The habitat of

Frithia pulchra tends to be drier than that of F. humilis,

the soils where the latter grow, having a higher organic

content, sometimes resembling peat, and thus retaining

moisture better.

FRITHIA IN HORTICULTURE

Frithia pulchra and F. humilis differ widely in their

horticultural history. F. humilis was introduced to the

Dutch seed trade by de Boer thirty years ago and a few

of the plants dating from that introduction are still alive.

This species is obviously quite tolerant and it responds to

water more eagerly than does F. pulchra. The latter has

been in continuous cultivation since the late 1920’s but

the plants are usually not long-lived, easily succumbing

to rot. Both species can mature in a few months from

seed under favourable conditions, and in this respect they

are typical of a Delosperma alliance. The two species

readily hybridize (S. Hammer pers. comm.) producing

fertile offspring, with a variety of flower colours includ-

ing orange and bright pink. Many attempts made by

Hammer to hybridize Frithia with other genera ( Delo-

sperma, Drosanthemum , Dorotheanthus, Lithops and

Fenestraria ) have always failed, not even producing the

‘dummy’ (empty) fruits which often result from such dis-

junct liaisons.

MORPHOLOGICAL CHARACTERS

Habit

Both species are dwarf perennials with thickened

roots. The stems are much reduced and during periods of

drought the plants retract into the sandy soil. This has

been ascribed to contractile roots, but no such roots are

present in either species. However, in Frithia the cells of

the leaves are arranged in columnar, axial rows and when

moisture is lost and the cells shrink, the tangential walls

contract. Artificially induced dessication using silica gel,

indicated that a leaf may contract to up to one third of its

length (Figure 2B). This causes the plants to retract into

the soil, a mechanism which renders protection to the

plants during times of drought (Figure 3C). Retraction

into the ground is thus achieved by means of ‘contractile

leaves’, not contractile roots.

Roots

Roots of Frithia pulchra differ from those of F.

humilis in being more fibrous, possibly because of the

drier conditions prevailing in its habitat and the strategy

to insulate the plants against the heat of surrounding

rocks in summer.

FIGURE 2. — Frithia pulchra: A, turgid leaf when conditions are

favourable; B, shrunken leaf under drought conditions C,

Neohenricia sibbettii, leaf. A, Burgoyne 6699c ; B, Burgoyne

6694b: C, Burgoyne 6786b. Scale bars: 1 mm.

4

Bothalia 30,1 (2000)

FIGURE 3. — Habit of Frithia: A, F. pulchra: B, F. humilis', C, plants

of Frithia humilis retract underground during periods of

drought. Scale bars: A, 10 mm; B, 30 mm; C, 10 mm.

Leaves

Borne spirally, the leaves of both species are cylindri-

cal with windowed tips and are covered by an epidermal

layer of waxy idioblasts arranged in distinct rows (Figure

2A). Leaves in adult plants of F. pulchra are longer

(15-25 mm) than in F. humilis (shorter than 15 mm ).

Leaf colour also varies slightly between the two species,

those of F. pulchra having a bluish tinge, whereas those

FIGURE 4. — Surface of leaf tips of Frithia: A, F. pulchra showing

convex tips and no markings; B, F. humilis with concave cen-

tre and crenulate markings on margins. A, Burgoyne 6699c ; B,

Burgoyne 6694b. Scale bars: 1 mm.

of F. humilis are tinged brown or purple. The windowed

tips of the leaves, however, differ conspicuously between

the two species. Windows of F. pulchra are convex when

turgid, slightly concave when flaccid (Figure 4A) and

those of F. humilis are concave even when turgid, with

crenulate markings (Figure 4B) along the margins. These

leaf differences were also noted by Zimmermann (1996).

Flowers

Flowers in both species are borne singly and on very

short stalks or are stalkless. They are subtended by five

Bothalia 30,1 (2000)

5

unequal sepals closely resembling the cylindrical leaves.

Flowers of F. pulchra are bright magenta with a white or

light yellow centre and are 25-35 mm in diam. Those of

F. humilis are white with a yellow centre, sometimes

tipped with pale pink, and are generally smaller (15-20

mm diam.). The petals number between 30 and 45 in F.

pulchra and tend to have blunt, rounded tips, whereas

those of F. humilis number between 20 and 30 and usu-

ally have acuminate tips (Figure 3 A, B).

Pollen

Pollen in both species of Frithia is yellow. The grains

are tricolpate and simplicolumellate in F. pulchra (Punt

et al. 1994) with a perforate surface and lumens of dif-

ferent sizes. Pollen in F. humilis has a perforate surface,

and is pluricolumellate, with lumens of more or less

equal size (Figure 5).

Fruit

Fruits are hygrochastic capsules, the shape resem-

bling a barrel. Thick tissue surrounds the capsules of F.

pulchra, whereas that of F. humilis is more fragile

(Figure 6C). However, this character is not constant for

the two species and seems to vary with environmental

conditions. Capsules of both species tend to break up

shortly after ripening. In both species, there are five or

six locules, no valve wings and no closing bodies.

Expanding keels are parallel with divergent tips (Figure

6), and are dark brown in F. pulchra and lighter brown in

F. humilis. Covering membranes are reduced to a ledge

in both cases.

Seeds

From Figure 7 it can be seen that the seed of F. pul-

chra is quite different from that of F. humilis. In F. pul-

chra (Figure 7A) the end where the seed has been

FIGURE 5. — Pollen of Frithia. SEM

micrographs of unacetolyzed

pollen grains: A, F. pulchra\

B, F. humilis. Detail of pol-

len surface: C, F. pulchra', D,

F. humilis. A, C, Burgoyne

6699c\ B, D, Burgoyne 6694b.

Scale bars: A, B, 1 pm; C, D,

0.5 pm.

attached to the funicle (hilar end) has a sharp point,

whereas in F. humilis it is more rounded. The length of

the micropylar regions appears to be similar in the two

species. The surface sculpturing is irregular in both

species of Frithia (Figure 7B, E) and although there are

no microbaculae present, the surface of the epidermal

cells differs markedly between the two species (Figure

7C, F), that of F. pulchra being rough-textured whereas

the cell surface of F. humilis is smoother.

NEAREST RELATIVES

The position of Frithia within the Stomatium Group

proposed by Hartmann (1998) has always been tentative

because of the outlier geographical distribution range

and unique leaf characters displayed by the genus.

Perhaps the most unusual feature is the spirally arranged

leaves, a unique character within the subfamily Ruschi-

oideae Schwantes in Ihlenf., Schwantes & Straka (1962)

emend. Bittrich & H.E.K. Hartmann. Superficially, plants

of Frithia resemble the genus Fenestraria N.E.Br., also

with windowed leaf tips. However, Fenestraria occurs in

northern Namaqualand and Namibia and no other char-

acters are shared by the two genera.

As stated by Hammer (1998), Frithia has characters

in common with Delosperma, but major differences still

separate the two genera. One of the similarities is the leaf

epidermis which is covered by opaque idioblasts.

Idioblasts of Delosperma deilanthoides S. A. Hammer

(1998) most closely resemble those of Frithia and are

also arranged in rows. Capsules in both genera lack cov-

ering membranes and closing bodies. Distributions of

these two genera overlap. The distribution of Delo-

sperma deilanthoides is centred in the Steenkampsberg,

Mpumalanga. It has similar habitat requirements (sandy,

well-drained soil with a high organic content and porous,

6

Bothalia 30,1 (2000)

FIGURE 6. — Fruit capsules of Frithia: A, F. pulchrcc, B, F. humilis', C,

closed capsule of F. humilis.', A, Burgoyne 6699c; B, C,

Burgoyne 6694b. Scale bars: 1 mm.

coarse lithology) to those of Frithia, but the Steen-

kampsberg receives a higher rainfall (± 1 200 mm per

annum). Flowers of both genera belong to the white/pink

colour range and open from mid-morning to mid-after-

noon (Smith et al. 1998). However both flower colour

and the presence or absence of epidermal idioblasts are

considered to be pleisiomorphic and are not suitable to

indicate relationships.

The spiral leaf arrangement, not opposite as in other

mesembs, gives rise to the interpretation that Frithia may

have retained this primitive character (spirally arranged

leaves are regarded as primitive) while developing

advanced states in other characters. A more likely expla-

nation may be that the spiral leaf arrangement found in

Frithia is a derived feature, since it is not present- in any

genus holding a more basal position within the Me-

sembryanthemaceae. Based on this data it could be

assumed that Frithia may be considered a highly spe-

cialised ‘Delosperma’ .

The leaves of Neohenricia sibbettii (L. Bolus) L. Bolus

most closely resemble those of Frithia (Figure 2C) in

shape, but the leaf surfaces of these two genera differ

markedly. Leaves of Neohenricia are covered by wart-

like crustose epidermal outgrowths, with opaque

idioblasts scattered among them. Leaves of Neohenricia

are opposite, whereas those of Frithia are arranged spi-

rally. Capsules of the two genera are similar except that

those of Neohenricia have four to six locules (four

locules being the norm) and are shallow; those of Frithia

are five- or six-locular and barrel-shaped. Moreover, the

capsules of Neohenricia are borne on a remarkably thin

pedicel and stand above the mass of leaves, whereas

those of Frithia are buried within the leaves on a short

pedicel and tend to be expelled via leaf pressure when

ripe. However, flower colour and morphology differ con-

siderably. Flowers of Neohenricia are pale yellow, borne

on long pedicels and have thin spiky petals, opening in

the mid-afternoon to evening. Nectaries are in the form

of a glandular ring in Neohenricia, whereas those of

Frithia are free. Flowers of both genera have very short

stigmas with the staminodes and stamens that are fused,

almost forming a hypanthium. The distribution ranges of

these two genera do not overlap.

Further investigation is being done to examine the

relationship (if any) between Conophytum limpidum

S. A. Hammer and Frithia (Burgoyne in prep.) as they

have some characters in common: a hypanthium is pre-

sent; the petals, petaloid staminodes and anthers are

comparable; windowed leaf tips present; fruit capsules

can be compared as closing bodies; covering membranes

are absent. Both are summer rainfall mesembs.

CONSERVATION STATUS

Use of the IUCN red list of categories (IUCN Species

Survival Commision 1994), indicate that both species of

Frithia should be regarded as Vulnerable, as the total

area that they occupy is less than 100 km2. Although the

areas where these species grow are not in any immediate

danger of being destroyed because they are too rocky, the

limited distribution poses the risk that human activity

could wipe out a large part of the populations should

their habitat be used and transformed in the future. One

locality of F. humilis is situated at the edge of an infor-

mal housing development, but the habitat is so unsuitable

for any utilisation by man that it has remained largely

undisturbed except for littering. The conservation status

of F. pulchra is more secure, as a large part of the popu-

lation is situated in the Rustenburg Nature Reserve. All

other areas where populations of both species of Frithia

Bothalia 30,1 (2000)

7

FIGURE 7. — Seed of Frithia: SEM micrographs: A-C, F. pulchra, Burgoyne 6699c\ D-F, F. humilis , Burgoyne 6694b. Scale bars: A, D, 100 |im;

B, E, 10 pm; C, F, 1 pm.

grow are in the hands of private land owners, many not

even aware of the presence of these tiny plants.

Unscrupulous succulent collectors may pose the greatest

threat to populations of Frithia. Further population stud-

ies of both species of Frithia will be undertaken and their

new conservation status will be determined (Burgoyne,

Krynauw & Smith in prep.).

ACKNOWLEDGEMENTS

Mrs A. Romanowski who assisted with the photogra-

phy and Mrs E. du Plessis who helped with the prepara-

tion of the manuscript are gratefully acknowledged. Dr

Otto Leistner is thanked for his assistance with the Latin

description. Ms R Chesselet and Prof. A.E. van Wyk are

thanked for useful discussions.

REFERENCES

BRINK, RD. 1985. Album. Frithia pulchra. Aloe 22: 43.

BROWN, N.E. 1925. Mesembryanthemum. The Gardeners’ Chronicle

ser. 3, 78: 433.

BROWN, N.E. 1926. Ficoidaceae. In J. Burtt Davy, A manual of the

flowering plants and ferns of the Transvaal with Swaziland,

South Africa 1,41: 162. Longmans, Green, London.

BURGOYNE, P.M. (in prep.) Similarities between Frithia pulchra,

Frithia humilis and Conophytum limpidum. Cactus and

Succulent Journal (US).

BURGOYNE, P.M., KRYNAUW, S. & SMITH, G.F. (in prep.) Popu-

lation studies and new conservation status of the genus Frithia

(Mesembryanthemaceae). Proceedings of the XVIth AETFAT

International Congress: special issue, systematics and geogra-

phy of plants.

DE BOER, H.W. 1968. Frithia pulchra var. minor de Boer, var. nov.

Succulenta 47: 147, 148.

GERMISHUIZEN, G. 1975. Some endangered succulents of the Trans-

vaal. Fauna & Flora 26: 5-7.

HAMMER, S. 1998. Frithia’s fairest friends. Mesemb Study Group

Bulletin 13,3: 64.

HARDY, D. & FABIAN, A. 1992. Succulents of the Tranvaal. Southern

Book Publishers, Halfway House.

HARTMANN, H.E.K. 1998. Groupings in Ruschioideae (Aizoaceae).

Mesemb Study Group Bulletin 2: 35, 36.

IHLENFELDT, HD., SCHWANTES, G. & STRAKA, H. 1962. Die

hoheren Taxa der Mesembryanthemaceae. Taxon 1 1 : 52-56.

IUCN SPECIES SURVIVAL COMMISSION. 1994. IUCN Red List

Categories. IUCN, Gland.

KROON, N. 1997. Tribute to two amateurs: William Nelson F.R.H.S.,

L.M.C.A., F.L.Sc. (1852-1922) and Frank Frith (1872-1954).

PlantLife 16: 15-17.

PUNT, W„ BLACKMORE, S„ NILSSON, S. & LE THOMAS, A.

1994. Glossary of pollen and spore terminology. L.P.P.

Contributions Series No. 1. Utrecht, The Netherlands.

SMIITH, G.F., CHESSELET, P., VAN JAARSVELD, E.J., HART-

MANN, H„ HAMMER, S„ VAN WYK, B-E., BURGOYNE,

P, KLAK, C. & KURZWEIL, H. 1998. Mesembs of the world.

Illustrated guide to a remarkable succulent group. Briza

Publications, Pretoria.

STEFFENS, F.E. 1988. Frithia pulchra. Aloe 25: 71, 72.

VENTER, F. 1979. Frithia pulchra, window plant of the Magaliesberg.

Aloe 17: 47-51.

VENTER, F. 1983. Frithia pulchra. Fauna & Flora 40: 19, 20.

ZIMMERMANN, N.F.A. 1996. Frithia pulchra N.E.Br. — Eine Reise

zu zwei Populationen im Transvaal mit Besprechung der succu-

lenten Begleitvegetation. Kakteen und Andere Sukkulenten 47:

146-152.

Bothalia 30,1 : 9-15 (2000)

A review of generic concepts in the Stilbaceae

J.P. ROURKE*

Keywords: Eurylobium Hochst., generic review, Kogelbergia Rourke, new combinations, new genus, Retzia Thunb., Stilbaceae

ABSTRACT

The generic concepts in the Cape endemic family Stilbaceae (± 14 species), are reviewed. Proposals by various authors

to include Retzia capensis Thunb. in the Stilbaceae are supported. Xeroplana Briq and Eurylobium Hochst. are shown to

be congeneric with Stilbe L., resulting in one new combination and one new name in Stilbe. Seven currently recognised

species of Stilbe are enumerated. A key to the genera of the Stilbaceae, as presently understood, is provided. A new genus

Kogelbergia is described to accommodate two species previously assigned to Stilbe sect. Amphistilbe and two new combi-

nations in Kogelbergia are proposed.

CONTENTS

Introduction 9

Inclusion of Retzia in the Stilbaceae 9

Stilbaceae subfamily Stilboideae 9

Current generic position 10

Synopsis of Stilbe 11

Key to genera of Stilbaceae 11

Kogelbergia, a new genus in the Stilbaceae 11

Acknowledgements 15

References 15

INTRODUCTION

The Stilbaceae Kunth, a small family in the Lamiales, is

endemic to the southern Western Cape. It is variously

regarded as a subfamily within the Verbenaceae by some

authors (Cronquist 1981) or as a family by others (Dyer

1975; Dahlgren 1980). The entire family consists of about

twelve currently recognised species arranged in six genera.

Four of these ( Eurylobium Hochst., Campylostachys

Kunth, Euthystachys A. DC. and Thesmophora Rourke)

are monotypic. While Campylostachys and Thesmophora

are defined on sound morphological characters, Eurylo-

bium and Xeroplana Briq. (two species) are here consid-

ered to be congeneric with Stilbe L. On present evidence

Euthystachys seems sufficiently distinct to be upheld as

a monotypic genus but further studies of fresh material

from a range of populations are required before an ade-

quate assessment can be made. Over and above these

genera there is material of at least one and possibly two

further undescribed taxa represented in local herbaria so

it is unlikely that the final species count for the family

will exceed fourteen.

INCLUSION OF RETZIA IN THE STILBACEAE

Before discussing these generic concepts any further,

the position of Retzia Thunb. must also be considered in

any review of the Stilbaceae and its relationships.

* Compton Herbarium, National Botanical Institute, Kirstenbosch,

Private Bag X7, 7735 Claremont, Cape Town.

MS. received: 1999-05-26.

Retzia capensis Thunb., a monospecific endemic Cape

genus, has been assigned to a wide range of families

since it was described by Thunberg in 1776 (see Dahl-

gren et al. 1979, for a detailed review of its past taxonomic

history). It has also been treated as a monospecific fami-

ly (Dyer 1975). Similarly, Cronquist (1981) upheld the

Retziaceae as a monospecific family, dismissing any

relationship with the Stilbaceae largely on account of the

placentation and ‘organisation of the gynoecium’ in

Retzia. In a similar view, Takhtajan (1997), while

acknowledging that Stilbaceae and Retziaceae are close-

ly related, continued to uphold both as separate families.

However, modern data from several different sources

now indicate conclusively that Retzia is most closely

allied to the Stilbaceae.

In a comparative study of iridoid glucosides in Stilbe

ericoides L. and Retzia capensis , biogenetically similar

compounds were found in both species (Dahlgren et al.

1979). These authors further pointed out that morpho-

logical and anatomical evidence also suggested a close

relationship between the two families. An extensive

investigation of the wood anatomy of several genera in

the Stilbaceae has further highlighted the similarities

between Retzia and the Stilbaceae (Carlquist 1986).

More recently, rbcL sequences for Euthystachys

(Stilbaceae) and Retzia were found to be very similar,

‘differing by only eight substitutions’, (Bremer et al.

1994), again confirming the close relationship between

the two families. Thus morphological, anatomical, phy-

tochemical and molecular evidence supports the inclu-

sion of Retzia in the Stilbaceae.

In an updated phylogenetic classification of the flow-

ering plants, Thorne (1992) incorporated Retzia into the

Stilbaceae, dividing the Stilbaceae in two subfamilies,

Retzioideae and Stilboideae. This view is followed in the

present treatment and is, similarly, the current opinion of

the Angiosperm Phylogeny Group (1998).

STILBACEAE SUBFAMILY STILBOIDEAE

All members of the the Stilbaceae subfamily Stil-

boideae are small, inconspicuous ericoid shrublets with

very small white or pinkish flowers. Although typical

components of the Cape fynbos they are never dominant

10

Bothalia 30,1 (2000)

and tend to be easily overlooked. Several taxa are rare

localised endemics, and some are possibly palaeoen-

demics. In almost all cases the flowers are extremely

small (shorter than 10 mm) which has made the dissect-

ing and interpretation of herbarium material very diffi-

cult, often leading to serious misinterpretations of the flo-

ral structures (Rourke 1977). During the course of revis-

ing the family I have been able to examine a wide range

of fresh living material collected in the field which has

greatly clarified an understanding of the floral structures.

This review of the generic characters within the fami-

ly has led to an expanded concept of Stilbe which will be

discussed here and also the exclusion from Stilbe of two

species previously known as Stilbe mucronata N.E.Br.

[more recently as Stilbe verticillata (Eckl. & Zeyh.)

Moldenke] and Stilbe phylicoides A. DC. It is proposed

that they be placed in a new genus, Kogelbergia Rourke.

As presently understood the Stilbaceae subfamily

Stilboideae consists of five genera, two of which are char-

acterised by having four-lobed corollas ( Campylostachys

and Thesmophora) and three which have five-lobed

corollas ( Stilbe , Euthystachys and Kogelbergia).

CURRENT GENERIC POSITION

The problem of finding a rational generic classifica-

tion of the Stilbaceae has been known for some time

(Dyer 1975; Rourke 1977). In this review an interim

solution is provided to satisfy the immediate needs of

two generic treatments of the family for ‘ Seed plants of

southern Africa: families and genera ’ and K. Kubitski’s

'The families and genera of vascular plants' , currently in

progress. A major taxonomic difficulty concerns the pre-

sent circumscription of Stilbe, Xeroplana, Eurylobium

and Euthystachys.

Stilbe L.

Two distinct groups can be recognised within Stilbe

based on the symmetry of the corolla and the presence or

absence of pubescence on the corolla lobes.

Stilbe verticillata and S. phylicoides A.DC. have actino-

morphic corollas and densely pubescent corolla lobes of

equal size [Stilbe sect. Amphistilbe (Pearson 1901)]. The

ovary in both species is single-chambered (but has mar-

ginal septum remnants), with two basal ovules.

The remaining species are characterised by promi-

nently bilabiate flowers with two broad, erect posterior

corolla lobes and three narrow anteriorly deflected corol-

la lobes, all of which are quite glabrous [Stilbe sect.

Eustilbe (Pearson 1901)]. The sepals are fused to form a

tubular, usually glabrous calyx. The ovary is two-cham-

bered with a single ovule in each chamber. At anthesis

one of the ovules and its locule aborts, leaving a single

ovule in a single chamber.

It is proposed that the present broad circumscription

of Stilbe be emended to encompass merely Stilbe sect.

Eustilbe and that the two species currently placed in

FIGURE 1 — Diagramatic representation of the differences in corolla

structure and pubescence between A, Stilbe and B, Kogel-

bergia.

Stilbe sect. Amphistilbe be removed to a new genus,

herein described as Kogelbergia (Figure 1).

Xeroplana Briq.

In Xeroplana the corolla is strongly bilabiate with two

broad, erect posterior petals, and three narrow anterior

petals. The corolla lobes are glabrous. Initially the ovary

is two-chambered with a single ovule in each chamber,

but the abaxial chamber and ovule soon abort, leaving a

single adaxial chamber with a solitary basal ovule. These

similarities with Stilbe sect. Eustilbe have already been

noted; moreover, it has previously been pointed out that

the generic status of Xeroplana is questionable (Rourke

1977). On the basis of gross morphological evidence

Xeroplana and Stilbe sect. Eustilbe are congeneric.

Consequently, as no rational justification can be found to

uphold Xeroplana, it is proposed that it be merged with

Stilbe sect. Eustilbe under a modified concept of Stilbe,

after having removed the two species in Stilbe sect.

Amphistilbe to a new genus, Kogelbergia.

Eurylobium Hochst.

Vegetative and florql characters in Eurylobium are the

same as in Stilbe sect. Eustilbe, namely, a tubular corol-

la, strongly bilabiate with two large, erect posterior lobes

and three narrower, anteriorly deflected lobes. The corol-

la lobes are glabrous. The sepals are fused for 3/4 of their

length to form a glabrous, actinomorphic calyx tube.

Initially the ovary is bilocular with an ovule in each

chamber but at an early stage one of the ovules aborts,

leaving the ovary to become unilocular with a single,

functional, basally attached ovule. As these characters

are the same as in Stilbe sect. Eustilbe, there is no justi-

fication for upholding Eurylobium as a monotypic genus.

Euthystachys A.DC.

The single species in this monotypic genus, E. abbre-

viata A.DC. is distinguished by the following characters:

soft, pubescent sepals fused for about ‘A of their length

at the base; an actinomorphic corolla of five, glabrous,

equal corolla lobes and a bilocular ovary with a single

functional ovule in each chamber.

Bothalia 30,1 (2000)

11

These characters, though slender, provide sufficient

justification for maintaining Euthystachys as a mono-

typic genus at this stage. Nevertheless, there are some

differences in the degree of fusion of the sepals in certain

collections. Further field work and the study of a range

of fresh material is required before the generic status of

Euthystachys can be adequately assessed.

Under the revised generic concept of Stilbe discussed

above, two genera in the family, Xeroplana Briq. and

Eurylobium Hochst. fall clearly within the emended cir-

cumscription of Stilbe and must be reduced to synonymy

as indicated below.

SYNOPSIS OF STILBE

Stilbe P.J.Bergius, Descriptiones plantarum ex Capite

Bonae Spei: 30, t. 4 (1767); H. Pearson: 183 (1901).

Type: Stilbe vestita P.J.Bergius.

Eurylobium Hochst.: 228 (1842); Walp.: 173 (1845); A.DC.: 608

(1848); H. Pearson: 187 (1901). Type: Eurylobium serrulatum Hochst.

Xeroplana Briq.: 336 (1896); H. Pearson: 188 (1901); Rourke: 1

(1977). Type: Xeroplana zeyheri Briq.

Ericoid shrublets, 0.5-1. 5 m tall, with a single main

stem, occasionally multistemmed and lignotuberous.

Leaves ericoid, linear-subulate to narrowly lanceolate,

ascending, in whorls of 3-5, margins revolute. Inflores-

cence a dense sessile spike, 10-40 mm long. Flowers

white or pink, sessile, subtended by single anterior bract.

Floral bracts 2, opposite, foliaceous. Calyx tubular, usu-

ally hard, with 5 free lobes occasionally slightly abaxial-

ly curved. Corolla tubular, 5-merous, bilabiate with two

erect glabrous posterior lobes and three narrower

glabrous anterior lobes; interior of throat ringed with

dense pubescence, only rarely glabrous. Stamens 4,

equal, posterior stamen absent; filaments glabrous, fili-

form; anthers basifixed. Ovary oblong, laterally com-

pressed, potentially two-chambered but abaxial chamber

abortive; adaxial chamber with a single basally attached

ovule; style filiform glabrous, slightly adaxially curved;

apex occasionally minutely bilobed. Fruit cylindric, lat-

erally compressed, brown, shiny, dehiscing into 2 valves

at apex. Seeds ovoid, straw-coloured.

An enumeration of the species of Stilbe currently

recognised within the new delimitation of the genus is

given here.

1. Stilbe vestita P.J.Bergius , Descriptiones plan-

tarum ex Capite Bonae Spei: 30, t. 4 fig. 6 (1767).

2. Stilbe ericoides L., Mantissa plantarum altera:

305 (1771).

3. Stilbe albiflora E.Mey., Commentariorum de

plantis Africae australioris 1: 279 (1836).

4. Stilbe rupestris Compton in Journal of South

African Botany 10: 127, 128 (1944).

5. Stilbe serrulata (Hochst.) Rourke, comb. nov.

Eurylobium serrulatum Hochst. in Nora 15: 229 (1842). Type: ‘Inter

rapes cacuminis montis prope Genadenthal in Colonia Capensis’ Dec.

1838, Krauss 1110. TUB6019 (TUB, holo.!).

6. Stilbe overbergensis Rourke , nom. nov.

Xeroplana zeyheri Briq. in Bulletin de l’Herbier Boissier, Ser. 1,4:

336 (1896). Type: Western Cape, on banks of Riviersonderend River at

Appelskraal and on adjacent mountains Ecklon & Zeyher sub Stilbe sp.

nov. no. 8 in herb. Delesserl (G, holo.!).

The epithet zeyheri cannot be used due to the prior

existence of Stilbe zeyheri Gand. (1913), a synonym of

Stilbe albiflora E.Mey.

7. Stilbe gymnopharyngia ( Rourke ) Rourke, comb.

nov.

Xeroplana gymnopharyngia Rourke in Journal of South African Botany

43: 6 (1977). Type: Western Cape, Langeberg Mountains, Riversdale

above the Farm Langkloof on ‘Annex Langekloof’, along summit ridge.

Rourke 1451 (NBG, holo.!, PRE, BOL, K, MO, G, S, iso.!).

Key to genera of Stilbaceae

la Corolla tube 45-55 mm long, pubescent externally, orange-

red with black tips (subfamily Retziaceae) Retzia

lb Corolla tube shorter than 12 mm, glabrous externally, uni-

formly mauve, pink or white (subfamily Stilboideae):

2a Corolla 4-lobed:

3a Sepals free; corolla lobes glabrous Campylostachys

3b Sepals fused forming tube; corolla lobes pubescent . . .

Thesmophora

2b Corolla 5-lobed:

4a Corolla actinomorphic; lobes equal or almost equal:

5a Corolla lobes pubescent Kogelbergia

5b Corolla lobes glabrous Euthystachys

4b Corolla prominently bilabiate with two larger erect poste-

rior petals and three narrow anterior petals Stilbe

KOGELBERGIA , A NEW GENUS IN THE STILBACEAE

As previously mentioned, it is proposed that the two

species in Stilbe sect. Amphistilbe be removed to a new

genus, Kogelbergia. Its distinguishing characters are dis-

cussed here.

Inflorescence structure

Superficially the inflorescence appears to be a sessile,

globose capitulum. Elowever, careful dissection of

K. verticillata indicates that it is paniculate in structure

with several highly condensed axillary branches bearing

3 or 4 flowers. The inflorescence structure in all other

Stilbaceae subfamily Stilboideae is spicate. It seems

probable that although much reduced and condensed, the

paniculate inflorescence of K. verticillata represents a

less specialised condition than the racemose pattern in

the rest of the Stilbaceae. In K. phylicoides the inflores-

cence is a condensed spike with no trace of a paniculate

structure.

Corolla

Although the corolla is actinomorphic at anthesis con-

sisting of 5 equal, linear corolla lobes, as the corolla

opens, two lobes briefly assume an erect (posterior) posi-

tion with the remaining three assuming a patent (anteri-

or) position. This condition can only be briefly detected

by observing developmental stages in living material and

12

Bothalia 30,1 (2000)

is soon lost as the flower matures. It seems probable that

actinomorphy in the corolla is a plesiomorphic condition

with bilateral symmetry and a bilabiate corolla repre-

senting the derived condition.

The presence of dense villous pubescence on the lin-

ear corolla lobes of Kogelbergia is a character highly

unusual in the family. All other Stilbaceae except

Thesmophora and Retzia, have completely glabrous

corolla lobes. This pubescence is formed by the presence

of 1-2 mm long white trichomes on the inner surface,

apex, and outer surface of each corolla lobe. Like almost

all other Stilbaceae, Kogelbergia has a distinctive ring of

pubescence in the throat of the corolla. Only three other

species in the entire family lack this feature. The dense

pubescence on the apices of the corolla lobes appears to

be unrelated to the throat pubescence and is regarded

here as a completely new character.

Ovary

Typically the ovary in Stilbaceae subfamily

Stilboideae is bilocular with a single ovule in each cham-

ber, or initially bilocular but with one abortive ovule in

an atrophied or abortive second locule leaving a single

functional ovule to develop. In Kogelbergia the ovary is

unilocular with two basal ovules. However, serial sec-

tioning of the ovary shows clear remnants of a septum in

the position where this would normally be expected to be

(i.e. at right angles to the axis). The septum breaks down

very early in the development of the ovary leaving a

unilocular condition which is clearly derived (Figure 2).

Pollen

A review of pollen structure in the Stilbaceae (exclud-

ing Retzia) showed that the pollen of the ‘Stilbe

mucronata type’ was quite distinct within the family (Raj

FIGURE 2. — Kogelbergia verticillata', cross section through base of

ovary showing single chambered ovary with two ovules, x 77.

Note remnant of ovary septum placed at right angles to inflor-

escence axis. Two anterior sepals at base of calyx tube slightly

larger than posterior sepals.

TABLE 1 . — Summary of major morphological differences between

Stilbe and Kogelbergia

1983). Instead of having a rugulose exine (as in Cam-

pylostachys) or tectate-perforate exine as in the remain-

ing species in the family, the exine of Kogelbergia (the

‘ Stilbe mucronata type’) has a striato-reticulate exine, set-

ting it apart from all its congeners (Raj 1983).

Conclusion

Fundamental morphological differences exist be-

tween the two species previously known as Stilbe verti-

cillata and Stilbe phylicoides and other members of the

genus Stilbe (Table 1). These are considered sufficient to

justify separate generic status. Accordingly, a new genus

Kogelbergia is described to accommodate these species.

Part of the distribution area of one of these species falls

within the Kogelberg Nature Reserve, one of the centres

of the highest endemism in the Cape Floristic Kingdom.

Kogelbergia Rourke, genus novum Stilbei affine,

a quo corolla actinomorpha 5 petalis aequalibus lobis

dense pubescentibus, et ovario uniloculari ovulis duobus

basilaribus, differt. Typus: Kogelbergia verticillata

(Eckl. & Zeyh.) Rourke.

The genus Kogelbergia is distinguished from all other

genera by its dense, globose, sometimes inconspicuously

branched (apparently paniculate) or spicate inflores-

cences, the actinomorphic corolla, 5 equal petals densely

pubescent at their apices, and the single-chambered ovary

with 2 basal ovules. Two species are currently recog-

nised.

Kogelbergia verticillata (Eckl. & Zeyh.) Rourke ,

comb. nov.

Trichocephalus verticillatus Eckl. & Zeyh., Enumeratio plantarum

africae australis 131 (1835). Stilbe verticillata (Eckl. & Zeyh.)

Moldenke: 474 (1948). Type: in lateralibus montium prope

Palmietrivier, supra Grietjiesgat, June, Alt. 4, Ecklon & Zeyher 1003

(SAM, iso.!).

Stilbe mucronata N.E.Br.: t 2526 (1897); H.Pearson: 184 (1901).

Type: in declivibus montium Houwhoek, April 1895, 1400 ped., Bolus

8409 (K, lecto.! here designated).

Stilbe mucronata N.E.Br. var. cuspidata H.Pearson: 184 (1901). S.

verticillata (Eckl. & Zeyh.) Moldenke var. cuspidata (H.Pearson)

Bothalia 30,1 (2000)

13

Moldenke: 474 (1948). Type: Zwartberg, Caledon, Dec., Bolus s.n. (K,

holo.!).

Stilbe chorisepala Suess.: 56 (1950). Type: Franschhoek Pass, 17-

11-1946, S. Rehm in herb. W. Giess 1377 (M, holo.!).

Small, rigid, well-branched shrublet up to 1 m high

with a single main trunk up to 50 mm diam. Flowering

branches 2-3 mm diam., densely lanate initially, later

becoming glabrous. Leaves in whorls of 5, patent to slight-

ly recurved, ericoid, narrowly lanceolate, 10-12 x 2 mm,

apices prominently mucronate, upper surface glabrous

when mature, under surface lanate, margins recurved.

Inflorescence a sessile, globose, much condensed panicle

15-22 mm diam; lateral axes much condensed, bearing 3

or 4 flowers; each flower subtended by a lanceolate-

acute, apically mucronate bract and two opposite floral

bracts. Floral bracts very narrowly oblanceolate,

glabrous basally but lanately crinite apically. Calyx

actinomorphic, tubular, 5 mm long, of 5 equal lobes (but

anterior sepals slightly thicker at base), fused into a

glabrous tube for lU of their length; free lobes very nar-

rowly oblanceolate, glabrous basally but lanately crinite

apically. Corolla 8-10 mm long, funnel-form, tubular,

tube region glabrous, hyaline, actinomorphic but the 2

posterior petals erect, three anterior petals patent at

anthesis; throat fringed with a dense ring of lanate pubes-

cence; lobes at anthesis linear-acuminate, densely

tomentose. Stamens 4, prominently exserted, 5 mm long,

inserted between corolla lobes; posterior stamen absent;

anthers versatile, dorsifixed, saggitate. Ovary ellipsoid,

purple in upper half, bilaterally flattened, single-cham-

bered with 2 basal ovules in chamber; style straight,

terete, 6-7 mm long. Seed cylindric-ovoid, 2 mm long,

surface colliculate, tending to tuberculate basally, pale

straw-coloured (Figure 3).

Distribution, habitat and biology

This species ranges in an arc from French Hoek Pass

southwards through the Hottentots Holland Mountains to

Kogelberg; thence eastwards to the Caledon Swartberg

and Genadendal in the Riviersonderend range. Outlying

populations occur in the mountains near Napier and Elim

with an isolated collection made on the Potteberg.

Kogelbergia is essentially a montane species occurring

mainly above 450 m, but may range from 250-1 150 m.

It generally favours moist south-facing habitats and is

often found in peaty soil overlying Table Mountain

Sandstone. Flowering time: May-November. Seeds are

shed in January (Figure 4).

Specimens examined

WESTERN CAPE. — 3318 (Cape Town): Jonkershoek, east of top

of zig-zag path above 2nd waterfall, (-DD), Jan., Rycroft 146 (PRE).

3319 (Worcester): Franschhoek Pass, (-CC), Nov., Rehm in herb.

W. Giess 1377 (M). 3418 (Simonstown): east of Kogelberg at summit

of peak at north end of Five Beacon Ridge, (-BB), June, Rourke 479

(NBG, PRE); mountains of Sir Lowry’s Pass, March, Burchell 8821 (K,

PRE); southeast slopes of Moordenaarskop, (-BD), July, Boucher 3056,

April, Boucher 1242 (NBG). 3419 (Caledon): Dwarsberg, Hottentots

Holland Mountains, (-AA), Jan., Rourke 1844 (NBG, PRE); Kathleen

Murray Nature Reserve, Nuweberg Forest Reserve, June, Rourke 1968

(NBG); top of ridge east of Viljoen’s Pass, June, Goldblatt 2049

(NBG); Houwhoek, April, Schlechter 7574 (K, PRE, SAM); in lateral-

ibus montium prope Palmietrivier supra Grietjiesgat, June, Ecklon &

Zeyher 1003 (SAM); Hottentots Holland Mts, July, Ecklon &

Zeyher 2215 (SAM); Lebanon Catchment, Grabouw, June, Kruger 76

(PRE); Swartberg, Caledon, (-AB), Oct., Bodkin in herb Bolus 6785

(NBG; PRE), Bolus s.n. (K); Highlands, (-AC), Aug., Compton 7348

(NBG); Genadendal, (-BA), Dec., Bodkin in herb. Guthrie 3623 (NBG);

Boskloof Farm, eastern ridges of Paardeberg, (-BC), Dec., Rourke 2053

(NBG); Groot Hagel Kraal, Haelkraal River area, NE of farmstead,

(-DA), April, Oliver 5884 (NBG); Elim, (-DB), April, Schlechter 7636

(K, PRE, NBG). 3420 (Bredasdorp): Potteberg, Albertsdal Farm,

(-BC), May, Compton 19526 (NBG).

Kogelbergia phylicoides (A. DC.) Rourke, comb.

nov.

Stilbe phylicoides A. DC., Prodromus systematis naturalis regni

vegetabilis 12: 606 (1848); H. Pearson: 183 (1901). Type: Voormans-

bosch near Swellendam (sub loc. no. 70), Zeyher 3589 in herb. Boiss.

(G, holo.; K!, SAM!).

Campylostachys phylicoides Sond.: 202 (1847), nom. nud.

Small, upright, multiple-stemmed, lignotuberous

shrublet, 0.5- 1.0 m tall. Flowering branches sparsely

sericeous initially, soon glabrous, 1-3 mm diam.; lower

branches covered in prominent leaf scars. Leaves in

pseudowhorls of 5-7, patent to slightly incurved, ericoid,

very narrowly lanceolate-linear, 5-12 x 1 mm, glabrous,

apices curved upwards, strongly mucronate. Inflores-

cence a sessile, globose to ovoid terminal spike, 20-25 x

15 mm, usually solitary, occasionally with up to 3 axil-

lary spikes clustered below; each flower subtended by a

lanceolate-cymbiform bract, sericeous proximally, apex

glabrous, mucronate. Floral bracts 2, narrowly lanceo-

late-acuminate, 5-7 x 1 mm, sparsely sericeous, hyaline.

Flowers sessile. Calyx with 5 sepals; sepals free, nar-

rowly lanceolate, 6-7 x 1 mm, hyaline to papyraceous,

upper half sericeous, apices prominently rostrate; poste-

rior sepal smaller than remainder. Corolla actinomor-

phic, 8-10 mm long; lobes narrowly lanceolate-acumi-

nate, patent, densely pubescent mainly on inner surface,

2-3 mm long; tube glabrous 4-5 mm long; throat fringed

with dense ring of pubescence. Stamens 4, exserted,

5-6 mm long, inserted between corolla lobes; posterior

stamen absent; anthers versatile, dorsifixed, saggitate.

Ovary 1 mm long, ovoid, bilaterally flattened, single-

chambered with two basal, erect ovules; style straight,

terete, 6 mm long. Seeds not seen.

Distribution, habitat and biology

Endemic to the Langeberg Range in the southern

Western Cape, Kogelbergia phylicoides occurs sporadi-

cally on the upper south slopes in mesic mountain fynbos

between the Clock Peaks at Swellendam and the Robin-

son Pass near Mossel Bay (Figure 4). Populations are

mostly small, usually consisting of less than 12 individ-

uals, and are generally found at elevations between 425

and 1 100 m. Flowering time: October and November.

Specimens examined

WESTERN CAPE. — 3320 (Montagu): first path below 10 O’clock

Peak, Swellendam, (-CD), Oct., Wurts 445 (NBG); Langeberg,

Swellendam, south slope of 12 O’clock Peak, Oct., Taylor 7203 (NBG,

K); Voormansbosch, (-DC), Oct., Zeyher 3589 (K, SAM); Zuurbraak

Peak, Oct., Barnard SAM37286 (SAM). 3321 (Ladismith): Garcia’s

Pass, Riversdale, (-CC), Oct., Thorne SAM38850: Garcia’s Pass, Oct.,

14

Bothalia 30,1 (2000)

FIGURE 3. — Kogelberf’ia verticillata, Raurke 1968. A, flowering shoot; B, mature seed, C, open flower and subtending bract; D, bract, abaxial

view; E, bract, adaxial view; F, corolla, opened; G, calyx and two floral bracts; H, leaf, adaxial view, 1, leaf, abaxial view, J, K, anthers

before dehiscence; L, anther after dehiscence; M, longitudinal section through ovary; N, gynoecium. Scale bars: A, 50 mm; B, J-L, 2 mm;

C-G, 10 mm; H, I, M, N, 5 mm.

Bothalia 30,1 (2000)

15

FIGURE 4. — Distribution of Kogel-

bergia verticillata , •; and

Kogelbergia phylicoides , A.

Galpin 4420 (K, PRE); summit of Kampscheberg, Riversdale, 9-12-

1814, Burchell 7127 (K, PRE); lower part of Kampscheberg, 1-12-

1814, Burchell 6937 (K); Langeberg Mountains, Riversdale, above

Farm Langkloof on ‘Annex Langekloof , Nov., Rourke 1446 (NBG).

3322 (Oudtshoom): Robinson’s Pass, Outeniqua’s, (-CC), Oct., Bond

7567 (NBG).

ACKNOWLEDGEMENTS

I am most grateful to the late Ellaphie Ward-Hilhorst

for preparing the drawing of Kogelbergia verticillata and

to Inge Oliver for the diagram in Figure 1.

REFERENCES

ANGIOSPERM PHYLOGENY GROUP (APG). 1998. An ordinal

classification for the families of flowering plants. Annals of the

Missouri Botanical Garden 85: 531-553.

BERGIUS, P.J. 1767. Descriptiones plantarum ex Capite Bonae Spei.

Laurent Salvii, Stockholm.

BREMER, B , OLMSTEAD, R.G., STRUWE, L„ & SWEERE, J.A.,

1 994. rbcL sequences support exclusion of Retzia, Desfontainia

and Nicodemia from the Gentianales. Plant Systematics and

Evolution 190: 213-230.

BRIQUET, J. 1896. Xeroplana zeyheri. Bulletin de VHerbier Boissier,

Ser. 1, 4: 336.

BROWN, N.E. 1897. Stilbe mucronata. Hooker’s leones plantarum 6:

t. 2526.

CARLQUIST, S. 1986. Wood anatomy of Stilbaceae and Retzliaceae:

ecological and systematic implications. Aliso 11: 299-316.

COMPTON, R.H. 1944. Stilbe rupestris. Journal of South African

Botany 10: 127, 128.

CRONQUIST, A. 198 1 . An integrated system of classification of flow-

ering plants. Columbia University Press, New York.

DAHLGREN, R. 1980. A revised classification of the Angiosperms.

Botanical Journal of the Linnean Society 80: 91-124.

DAHLGREN, R., NIELSEN, B.J., GOLDBLATT, P. & ROURKE, J.P.

1979. Further notes on Retziaceae: its chemical contents and

affinities. Annals of the Missouri Botanical Garden 66:

545-556.

DE CANDOLLE, A. 1848. Prodromus systematis naturalis regni veg-

etabilis. Masson, Paris.

DYER, R.A. 1975. The genera of southern African flowering plants 1 :

514-516. Department of Agricultural Technical Services, Pretoria.

ECKLON, C.F. & ZEYHER, K.L.P. 1835. Enumeratio plantarum

africae australis. Perthes & Besser, Hamburg.

GANDOGER, M. 1913. Manipulus plantarum novarum praecipue ame-

ricae australis. Bulletin de la Societe botanique de France 60: 25.

HOCHSTETTER, C.F. 1842. Nova genera plantarum africae turn aus-

tralis turn tropicae borealis. Flora 15: 228, 229.

LINNAEUS, C. 1771. Mantissa plantarum altera. Laurent Salvii,

Stockholm.

MEYER, E.H.F. 1836. Commentariorum de plantis Africae australioris

1 : 279. Voss, Leipzig.

MOLDENKE, H.N. 1948. Notes on new and noteworthy plants. V.

Phytologia 2: 474.

PEARSON, H.H.W. 1901. Verbenaceae. Flora capensis 5,1: 183-188.

Reeve, London.

RAJ, B. 1993. A contribution to the pollen morphology of Stilbaceae

Kunth. Pollen et spores 25: 395-408.

ROURKE, J.P. 1977. A revision of Xeroplana Briq. (Stilbaceae).

Journal of South African Botany 43: 1-8.

SONDER, O. 1847. Vergleichungen der von Ecklon und Zeyher und

von Drege gesammelten sudafrikanischen Pflanzen, mit den

Exemplaren von Zeyher’s neuesten Sammlungen, welche der-

selbe zum Verkauf stellt durch J.F. Drege in Borstel bei

Hamburg. Linnaea 20: 202.

SUESSENGUTH, K. 1950. Stilbe chorisepala. Mitteilungen der

Botanischen Staatssammlung, Miinchen, Heft 2: 56.

TAKHTAJAN, A. 1997. Diversity and classification of flowering

plants. Columbia University Press, New York.

THORNE, R.F. 1992. An updated phylogenetic classification of the

flowering plants. Aliso 13: 365—389.

WALPERS, W.G. 1845. Repertorium botanices systematicae 4: 173.

Hofmeister, Leipzig.

Bothalia 30,1 : 17-24(2000)

Studies in the Sphaerocarpales (Hepaticae) from southern Africa. 2.

The genus Sphaerocarpos and its only local species, S. stipitatus

S.M. PEROLD*

Keywords: Austrosphaerocarpos (subgenus), Hepaticae, Monocarpineae, Riellineae, southern Africa, Sphaerocarpaceae, Sphaerocarpales, Sphae-

rocarpineae, Sphaerocarpos stipitatus

ABSTRACT

A taxonomic account of the genus Sphaerocarpos, subgenus Austrosphaerocarpos, and its only local species, S. stipi-

tatus Bisch. ex Lindenb., is presented here. These taxa are classified in the suborder Sphaerocarpineae of the order

Sphaerocarpales, for which short notes and a key to all three constituent suborders, Monocarpineae, Riellineae and

Sphaerocarpineae, are provided.

INTRODUCTION

The genus Sphaerocarpos (Micheli) Boehmer was

first recorded from southern Africa by Proskauer (1954).

He received material from the algologist, Dr M.A.

Pocock, who had collected it in the Eastern Cape, on the

banks of pools in the Palmiet River, a tributary of the

New Years River, seven miles from Grahamstown, on the

Cradock Road. Miss Lilian Britten, a lecturer at Rhodes

University College, had originally discovered Sphaero-

carpos plants at this locality. A further collection from

Clanwilliam, by Prof. E.A.C.L.E. Schelpe, was also

recorded by Proskauer.

Amell (1963) did not report any new collections and the

only recent published South African record of the species

is by Long (1993) from a 1969 collection, H.R. Toelken

1978 (BM, BOL) from Paardeneiland, Cape Town.

Seven of the collections that I studied, have not been

reported in the literature before. Despite repeated visits

to the Northern, Western and Eastern Cape by various

collectors, myself included, such a small number of gath-

erings must surely reflect the rarity of these plants,

although Proskauer (1955) remarked that, ‘one does not

normally find members of the group in the field without

being familiar with them’. Their truncated life cycle,

from spore to spore, can take less than 45 days (Schuster

1992) , and dramatically reduces the time period during

which they are likely to be found.

It is, nonetheless, hoped that the present, fully illus-

trated paper, will familiarise more botanists with these

remarkable plants and that, in future, they will collect

them, thus enabling us to establish the local distribution

range more completely. Previously, it was thought that S.

stipitatus was restricted to the southern hemisphere, i.e.

Chile in South America (where the type specimen had

been collected by Bertero in 1829) and South Africa.

Recently, however, it was also collected as a presumed

adventive in Portugal (Sergio & Sim-Sim 1989), as well

as at a remote, high altitude locality in East Nepal (Long

1993) , where its accidental introduction is unlikely.

* National Botanical Institute, Private Bag X101, 0001 Pretoria.

MS. received: 1999-07-02.

This paper on Sphaerocarpos stipitatus, is the second

in a series of three on the Sphaerocarpales in southern

Africa.

MATERIAL AND METHODS

The same procedures as outlined in Perold (1999b)

were employed in the preparation of the material for

examination and photography by compound light micro-

scope and scanning electron microscope.

Order Sphaerocarpales Cavers in New Phyto-

logist 9: 81 (1910); Grolle: 55 (1983); Scott: 63 (1985).

[excluding the American Geothallus Campbell and the

Mesozoic fossil, Naiadita Brodie]

Plants ephemeral, very diverse, terrestrial; sometimes

minute, subspherical and pouch-like, occasionally with

branched ventral sprouts (Monocarpineae), or somewhat

larger, with prostrate, median stems, up to 4 (or more)

cell layers thick, pseudodichotomously branching, with

wings 2-ranked, lateral, incised, leafy, unistratose

(Sphaerocarpineae); otherwise aquatic, semifrondose,

stems ± erect, apical dominance seemingly persistent,

with broad, undulating wing on dorsal side, small leaf

scales at both sides and ventrally (Riellineae). Air cham-

bers and air pores (the latter in the carpocephalum wall)

present only in Monocarpineae. Cells all thin-walled,

large, delicate; smaller oil cells found only in Riellineae.

Rhizoids all smooth.

Asexual reproduction by gemmae, only in Riellineae.

Monoicous in Monocarpineae and rarely in Riella

species, otherwise dioicous, mostly heterothallic, with male

plants much or somewhat smaller. Antheridia ovoid to glo-

bose, in Monocarpineae located inside air chambers, pedi-

cel long, uniseriate; in Sphaerocarpineae individually sur-

rounded by involucres, crowded together on dorsal surface

of stems; in Riellineae enclosed in pockets along margin of

undulating wing, pedicel short, uniseriate. Archegonia with

6 rows of neck cells, in Monocarpineae borne on archego-

niophores, otherwise individually surrounded by large,

flask-shaped or pyriform involucres, dorsally on, or along

stems of female plants. Stalk variable in length, without rhi-

18

Bothalia 30,1 (2000)

zoid furrow. Capsule globose, cleistocarpous, wall unis-

tratose, cell walls delicate, in Monocarpineae only with

small, nodular to elongated thickenings, in other suborders

thickenings absent. Seta very short, becoming necrotic.

Spores medium-sized to very large, 40-140 pm diam., sin-

gle in Monocarpineae and Riellineae, mostly remaining in

tetrads in Sphaerocarpineae. Elaters absent. Nutritive cells

present in Sphaerocarpineae and Riellineae. Spore release

in all three suborders by dissolution or decay of capsule wall

and surrounding tissue.

Key to local suborders of Sphaerocarpales

la Plants terrestrial; distribution confined to winter rainfall regions of southern Africa; highly diverse in appear-

ance— subspherical and pouch-like or dorsiventrally flattened with lateral, leaf-like lobes; cells uniform, all

lacking oil bodies; spores single or in tetrads:

2a Thalli monoicous, very small, subspherical and pouch-like; outer protective layers with air chambers sur-

rounding carpocephalum, the wall in the latter interrupted by barrel air pores; spores single, hemispheri-

cal, ornamentation finely tuberculate suborder Monocarpineae sensu Carr

2b Plants dioicous, heterothallic, males very much smaller than females, stems dorsiventrally flattened, laterally

divided into unistratose, leaf-like lobes, without air chambers and air pores; gametangia surrounded by

small or large (depending on sex) flask-shaped involucres; spores in local species remaining in tetrads,

ornamentation ridged suborder Sphaerocarpineae Cavers

lb Plants aquatic; distribution in both winter and summer rainfall regions of southern Africa; uniform in appearance,

consisting of erect stem, dorsally bearing undulate wing and small leaf scales at sides and ventrally; cells

dimorphic, some with an oil body; spores single, ornamentation various suborder Riellineae R.M.Schust.

Sphaerocarpineae Cavers. R.M.Schust.: 807

(1992).

Plants terrestrial, dorsiventrally flattened, bilateral,

consisting of a stem (or axis), once or twice to repeated-

ly furcate, laterally invested with delicate, leaf-like

lobes. Ventral scales absent, instead invested with

ephemeral slime papillae. Cells thin-walled, lacking oil

bodies. Rhizoids all smooth, hyaline.

Asexual reproduction absent.

Dioicous, heterothallic, male gametophytes much

smaller than female ones. Antheridia individually devel-

oped inside small, unistratose, brownish purple, flask-

like involucres, in acropetal sequence, crowded along

stem dorsally. Archegonia also enclosed, almost always

individually, in much larger, green, flask-like involu-

cres, wall double-layered only in single South African

species, S. stipitatus', in other species unistratose. Calyp-

tra delicate, unistratose around cleistocarpous capsule,

3- or 4-stratose around short, necrotic seta and sub-

spherical foot, disintegrating when spores reach maturi-

ty, allowing their release, when flask wall also decays.

Spores in permanent tetrads in South African species,

large, external face covered with ridges; in some extra-

South African species separating at maturity, the spore

ornamentation areolate to lamellate. Nutritive cells pre-

sent, but elaters absent.

Sphaerocarpaceae (Dumort.) Heeg in Verhand-

lungen der kaiserlich-koniglichen zoologisch-botanischen

Gesellschaft in Wien 41: 573 (1891) (‘Sphaerocarpeae’);

Mull.Frib.: 310 (1951-1958); S.W.Arnell: 8 (1963);

R.M.Schust.: 810(1992).

Tribe Sphaerocarpeae Dumort. 13: 163 (1874).

Subfamily Sphaerocarpoideae Schiffn.: 50 (1893);

Engl.: 49 (1898).

The Sphaerocarpineae include only the single family,

Sphaerocarpaceae; hence, the details in the subordinal

description are not repeated here.

Sphaerocarpos (Micheli) Boehmer in C.G. Lud-

wig, Definitiones generum plantarum: 501 (1760);

Haynes: 219 (1910); Prosk.: 153 (1954); S.W.Arnell: 8

(1963); R.M.Schust.: 816 (1992). Type: S. michelii

Bellardi.

Sphaerocarpos Micheli: 4, t. 3 (1729).

Sphaerocarpus Adanson: 15 (1763) [orth. var.];

Steph.: 655 (1899); Mull.Frib.: 310 (1951-1958).

Plants ephemeral, gregarious to scattered; stems once

or twice (or more frequently) furcately branching, dor-

siventrally flattened, laterally invested with delicate, suc-

cubously inserted, leaf-like lobes; ventrally (and apical-

ly) with scattered slime papillae; ventrally with smooth,

hyaline rhizoids.

Dioicous, strongly heterothallic; gametangia individ-

ually enveloped by flask-shaped involucres, small in

male plants, but much larger in females, which are ses-

sile or stipitate, wall unistratose, only exceptionally bis-

tratose, eventually containing globose capsule and short

seta with foot. Spores remaining in tetrads or separating

at maturity, ornamentation reticulate or ridged.

Species in the genus fall into two, well-defined groups

that were recently segregated by Schuster (1992).

Key to subgenera of Sphaerocarpos

la Sporophytes delicate, sessile on stem or axis; wall of flasks

unistratose, without cells projecting as processes on their

inner surfaces subgenus Sphaerocarpos

lb Sporophytes firm, stipitate, but length of stalk variable; wall

of flasks bistratose, with inner surface bearing inward-

projecting, finger-like cells, rich in chloroplasts

subgenus Austrosphaerocarpos R.M.Schust.

Proskauer (1954) concluded that the specific epithet

stipitatus for this species was apt, since such a stalk did

not exist in any other species of the genus. The longest

stalk that he measured was ± 1.5 mm. Recently, howev-

er, Long (1993) reported that in the material from Nepal

and Portugal (Sergio & Sim-Sim 1989), ‘the female

Bothalia 30,1 (2000)

19

flasks have an extremely short basal stalk’. The stalks I

measured were also short, only 200-795 pm in length,

which would support Long’s observation that, ‘develop-

ment of the stipe must be considered variable’, hence the

minor alteration I made to Schuster’s (1992) key to the

subgenus Austrosphaerocarpus.

Sphaerocarpos stipitatus Bisch. ex Lindenb.,

Verhandlungen der kaiserlichen Leopoldinisch-carolinis-

chen deutschen Akademie der Naturforscher 10 (= Nova

acta Academiae Caesareae Leopoldino Carolinae ger-

manicae naturae curiosorum 18): 504 i (1837); Haynes:

228 (1910); Prosk.: 144 (1954); Prosk.: 63 (1955);

S.W.Arnell: 8 (1963); Sergio & Sim-Sim: 414 (1989);

Long: 77 (1993). Type: Chile, Quillota, Bertero 695 (PC,

holo.).

Sphaerocarpus berteri Mont.: 39 (1838); Nees: 369 (1838); Mont.:

50 (1839); Gottsche et al.: 595 (1846).

Sphaerocarpus berteroi Mont. orth. mut.; Steph.: 657 (1899); noin.

inval., orthographic variant for S. berteri.

Plants ephemeral to short-lived, gregarious or scat-

tered; stems once or twice, to repeatedly furcately

branching and forming partial rosettes; markedly het-

erothallic, male gametophytes tiny (Figure 1A, B),

antheridial flasks purplish brown, intimately associated

with much larger (x 10), light green, female gameto-

phytes. Stems somewhat indeterminate, in both sexes

prostrate, dorsiventrally flattened, in the middle up to

260-300 pm thick, consisting of (2-)3 or 4 layers of

thin-walled, 4- or 5-sided, elongated cells, 112.5-1 60.0 x

45-75 pm, ventral cell layer with scattered slime papil-

lae, 37.5-62.5 x 17.5-22.5 pm, raised on smaller or larg-

er basal cell; above mostly obscured by crowded

gametangia, each with a small or large (depending on

sex) flask-shaped involucre, produced in acropetal

sequence, usually in 2 series; laterally giving rise to deli-

cate, unistratose, translucent, alternate, succubously ori-

ented, leaf-like lobes; ventrally anchored to substrate by

rhizoids, all smooth, hyaline, 30-50 pm wide.

Dioicous. Male plants with branches (Figure 3A)

1250-1750 x 525-625 pm, tapering apically, close

together and almost parallel to widely divergent, lateral-

ly with mostly obscured, leaf-like lobes (Figure 1C, D),

fan-shaped, 650-1025 pm long, above 400—97 5 (— 1 1 00 )

pm wide, narrowing toward base, 100-650 pm wide;

upper marginal cells mostly subquadrate, 40-50 x

47.5- 70.0 pm, middle laminal cells long-rectangular,

132.5- 145.0 x 42.5-62.5 pm, basal cells 125.0-137.5 x

75.0- 87.5 pm. Antheridial involucres (Figures 2B; 3B,

C) generally in 2 or 3 series along segments, sessile,

flask-shaped, 250-280 pm high, apically contracted,

opening surrounded by 7 or 8 vertically oriented cells,

45.0- 62.5(-92.5) x 25-30 pm, their apices projecting

somewhat, median sector inflated, width across 180-200

pm, wall unistratose, cells generally 4-sided, sometimes

their walls bulging, mostly 50.0-67.5 x 27.5^45.0 pm.

Antheridia (Figure 2A) individually nearly completely

enclosed by flasks, only apical ones not discharged,

ovoid, ± 125 x 82.5 pm, raised on short pedicel, up to

57.5 x 17.5 pm, consisting of uniseriate column of 5

cells.

Female plants pseudodichotomously branching 1-3

times, forming partial rosettes, 4. 5-6. 5 mm across,

sometimes segments simple (Figure IE) or nearly so, ±

4.4 x 2.6 mm, lateral leaf-like lobes (Figure 1F-M)

mostly single, occasionally bilobed, obovate, size vari-

able, 1075-2500 x 625-3375 pm above, narrower below,

275-1125 pm wide; cells at upper margin (Figure IN)

usually isodiametric, 55.0-57.5 x 50.0-62.5 pm, thin-

walled, sometimes cell projecting at angle much larger, ±

52.5 x 145 pm, cells at lateral margin 87.5-125.0 x

30.0- 42.5 pm; upper laminal and median cells long-

hexagonal, 92.5-200.0 x 50-70 pm, basal cells

155.0- 177.5 x 55-105 pm. Archegonial involucres often

crowded together, obscuring leaf-like lobes (Figure 3D),

sessile when first initiated, later stalked (Figure 3E),

ellipsoidal or bluntly conical (Figure 2C-E), at maturity

1800-3000 pm high, width 750— J 125 pm across widest,

basal part, narrowing gradually upwards to mouth and

generally more abruptly below towards stalk, mouth

(Figure 3F) usually 250-300 pm wide, surrounded by

unistratose ring of 16 or more hyaline cells, 100-105 x

50-60 pm, rounded above and projecting somewhat

crenulately; rarely, larger involucre containing 2 sporo-

phytes (Figure 2F), 3625 pm high, at mouth ± 625 pm

wide, 2050 pm wide across basal part; below mouth rest

of involucral wall double-layered: outer cells hyaline,

chloroplasts absent, long-rectangular above (Figure 21),

125.0- 147.5 x 40.0^42.5 pm, lower down gradually

enlarging (Figure 2H), 5- or 6-sided, 150-300 x 75—125

pm; inner cells (Figure 2J-L) finger-like, much more

numerous, in clusters, rich in chloroplasts, with free,

downward curving, papilliform tips, up to 185 x 32.5 pm

in upper part of flask, smaller lower down and often

becoming partially destroyed. Stalk developing by inter-

calary growth, length variable, 200-795 pm, width

175-250 pm, in cross section (Figure 2G) mostly with 6

cells across, cells in outer row 30-40 x 40-60 pm, inner

cells in 4 rows, ± 75 x 75 pm. Calyptra remaining unrup-

tured for a long period enveloping mature capsule, seta

and foot (Figure 2M) hyaline, mostly unistratose, cells

4-6-sided, 65-145 x 65-100 pm, smaller above, where

crowned by persistent archegonial neck (Figure 2N) and

below, where closely investing seta and foot, here up to

4-stratose. Archegonial neck 75-85 x 30.0-37.5 pm,

formed by 6 rows of cells and reportedly (Schuster 1992)

with only 2 neck-canal cells. Capsule 650-825 pm

diam., wall (Figure 2Q) unistratose, lacking thickenings,

cells ± 5-sided, 37.5-65.0 x 37.5-50.0 pm, densely cov-

ered with starch granules, particularly when young; with

age their number apparently reduced, subround or oval,

5.0- 12.5 x 5-10 pm. Seta (Figure 20) with 4 cell rows,

very short, ± 100 x 40 pm, becoming necrotic. Foot part-

ly fused with calyptra below, bulbous, ± 270 x 220 pm,

resting on ‘platform’ of cells filling basal part of flask,

cells 100-170 x 75-125 pm; in cross section (Figure 2P)

centrally with several ‘haustorial’ cells, roughly triangu-

lar, corners rounded, ± 42.5 pm long, up to 47.5 pm wide

across widest part, surrounded by thin layer of non-cel-

lular tissue and then by 2 or 3 rows of hyaline cells, the

largest, outermost cells 30.0-37.5 x 32.5-37.5 pm.

Spores remaining in tetrads, 87.5-112.5 pm diam., yel-

low-brown to dark brown; 3 (Figure 4C) or all 4 spores

(Figure 4A, B, D, E) of tetrad visible at the same time,

joined at thin line in narrow groove (Figure 4F), ± 2.5

pm wide, ornamentation ridged, with 16-19 minutely

20

Bothalia 30,1 (2000)

FIGURE 1 — Sphaerocarpos stipitatus. A, B, dorsal view of male gametophytes with antheridial involucres; C, D, young male, and older, leaf-

like lobes, respectively; E, dorsal view of female gametophyte with stem (‘axis’) and lateral, leaf-like lobes; F-M, female leaf-like lobes;

N, detail of leaf-like lobe. A-N, Koekemoer 1425. Scale bars: A-M, 500 pm; N, 100 pm. Artist: Gillian Condy.

papillate ridges, sometimes branched and up to 10 pm

apart, running from centre of distal faces outward and

stopping short at groove, rarely forming incomplete are-

olae. Nutritive cells (Figure 2R1-3) generally ovoid,

42.5-50.0 x 30.0-37.5 pm, apparently with several

nuclei (4-8), resulting from mitoses and therefore

remaining diploid, sometimes 4-celled; although con-

taining starch granules and chloroplasts, their role in

nutrition of developing spores not proven. Release of

spores and nutritive cells by disintegration of the capsule

wall, then by the calyptra and finally by the wall of the

flask.

DISCUSSION

The genus Sphaerocarpos has aroused much interest,

because sex chromosomes in plants were first found in

S. donnellii by Allen (1919). Since then, it has been

Bothalia 30,1 (2000)

21

FIGURE 2. — Sphaerocarpos stipitatus. A, male involucre containing antheridium partly exposed; B, male involucre; C, archegonium and flask;

D, E, older flasks; F, large flask with 2 capsules; G, c/s stalk; H, cells in outer wall of flask (without chloroplasts). I-L, wall of flask: I,

mouth and upper part from outside; J, mouth and upper part with papilliform cells from inside; K, c/s with larger, clear outer cells and

smaller, inner cells containing numerous chloroplasts; L, inner cells more enlarged. M, younger capsule, with seta and foot enveloped by

calyptra; N, calyptra, above with persistent archegonial neck, below covering seta and foot; O, c/s seta; P, c/s foot; Q, capsule wall with-

out thickenings, but covered with starch grains; R1-R3, nutritive cells. A-R3, Koekemoer 1425. Scale bars: A-C, L, P, 50 pm; D-F, 500

pm; G-K, 100 pm; M, N, 200 pm; O, Q, R1-R3, 25 pm. Artist: Gillian Condy.

22

Bothalia 30,1 (2000)

FIGURE 3. — Sphaerocarpos stipitatus. A, male plant covered with numerous antheridial involucres; B, antheridial flasks, with arrow indicating

upper cell of slime papilla; C, mouth of antheridial flask; D, female plant with leaf-like lobes and 3 flasks, 4th flask removed, remainder

of its stalk indicated by arrow; at top, right comer tip of male branch covered in flasks, to compare difference in sizes of male and female

plants; E, female plant with single flask on longer stalk, partly hidden by leaf-like lobe; F, mouth of female flask. A-F, Koekemoer 1425.

A, x 45; B, x 135; C, x 254; D, x 8.5; E, x 14; F, x 83.

repeatedly studied with regard to its genetics and cytol-

ogy; furthermore, a number of cultures have been main-

tained for lengthy periods of time.

In several aspects, the gametophyte, with its prostrate

stem (or axis) and delicate, 2-ranked, succubously insert-

ed leaf-like lobes, resembles Fossombronia species, but

so-called ‘angle’ leaves at the fork of dichotomies, as

described by Proskauer (1954) and Schuster (1992), have

not been observed by me in the present study, although I.

am familiar with them from my work on the Fossom-

broniaceae (Perold 1999a).

FIGURE 4. — Sphaerocarpos stipitatus. A-E, spore tetrads; F, groove where 3 of tetrads are joined. A, E.A.C.L.E. Schelpe 4877 (BOL); B, Toelken

1978 (BOL); C, A.S.L. Schelpe 760 (BOL); D, S.M. Perold 1838; E, Koekemoer 1425; F, Oliver 8961. A, D, x 366; B, x 439; C, x 374; E,

x 347; F, x 682.

Bothalia 30,1 (2000)

23

The gametophyte in Sphaerocarpos species is delicate

and lacks physiological drought resistance (Schuster

1992); it is, nevertheless, adapted to warm and some-

times disturbed areas, with adequate moisture only sea-

sonally and with intervening long, dry periods. It is

regarded as an ecological specialist and is a temporary

component of vegetation that frequently endures drought

in summer. To cope with such an environment, its life

cycle is considerably shortened and development of the

sporophyte is accelerated; it survives the dry periods as

spores, which generally remain in tetrads, eventually

giving rise to two male plants in close proximity to two

female plants, thus ensuring that fertilisation takes place

and that a new generation will ensue.

In using a molecular approach by determining

sequences for the 18S-r RNA gene to construct phyloge-

netic trees of a number of bryophytes, Capesius & Bopp

(1997) concluded that the Sphaerocarpales diverged

early as a separate clade among the Marchantiopsida.

Ecology

In southern Africa, S. stipitatus is restricted to the

winter rainfall regions of the Northern, Western and

Eastern Cape, which experience hot, dry summers

(Figure 5). The plants grow at altitudes ranging from sea

level to ± 1 200 m, on moist, fine-grained, sandy to

clayey soils or on mud; at drainage ditches, water cours-

es or on earth banks; under fynbos, short karroid scrub or

grass, often together with mosses and Riccia species. The

localities fall in several different vegetation types, name-

ly Mountain Fynbos, Upland and Lowland Succulent

Karoo, as well as Grassy Fynbos (Low & Rebelo 1996).

Dates when sporulating material was collected, range

from the end of July to early October.

Specimens examined

Britten 58348 (BOL).

Koekemoer 1425 (PRE).

Magill & Schelpe 3961 A (PRE).

Oliver 8961 (PRE).

Perold 1838 (PRE).

A.S.L. Schelpe 760 (BOL58344)\ E.A.C.L.E. Schelpe 4877 ( BOL58346 ),

4918b ( BOL58345 ).

Toelken 1978 (BM) (BOL58347) (ridges rather thicker in the spores

from this specimen; see Figure 4B).

ACKNOWLEDGEMENTS

My sincere thanks are extended to Dr D.G. Long for

kindly refereeing this paper, to the Curator of BOL for

the loan of specimens; to Dr O.A. Leistner for his valu-

able advice and to my colleagues at NBI for kindly col-

lecting specimens of S. stipitatus, and particularly, to Ms

M. Koekemoer, Curator of PRE, for collecting and care-

fully tending live material; also to Ms G. Condy for the

drawings; Mrs A. Romanowski for developing and print-

ing the micrographs and to Ms D. Maree for typing the

manuscript.

12 14 16 18 20 22 24 26 26 30 32 34

FIGURE 5. — Distribution of Sphaerocarpos stipitatus in southern

Africa. Pocock specimen not indicated on map (grid reference

3326AD), because I did not examine it.

REFERENCES

ADANSON, M. 1763. Famille des Plantes 2: 1-640. Vincent, Paris.

ALLEN, C.E. 1919. The basis of sex inheritance in Sphaerocarpos.

Proceedings of the American Philosophical Society 58:

289-316.

ARNELL, S.W. 1963. Hepaticae of South Africa. Swedish Natural

Science Council, Stockholm.

CAPESIUS, I. & BOPP, M. 1997. New classification of liverworts

based on molecular and morphological data. Plant Systematics

and Evolution 207: 87-97.

CAVERS, F. 1910. The inter-relationships of the Bryophyta. New

Phytologist 9: 81-91.

DUMORTIER, B.C.J. 1874. Hepaticae Europae. Bulletin de la Societe

Royale de Botanique de Belgique 13: 1-203.

ENGLER, A. 1898. Syllabus der Pflanzenfamilien, edn 2: 49.

GOTTSCHE, C M., LINDENBERG, J.B.G. & NEES AB ESEN-

BECK, C.G. 1846. Synopsis hepaticarum. Meissner, Hamburg.

Reprinted 1967, Cramer, Lehre.

GROLLE, R. 1983. Nomina generica hepaticarum: references, types

and synonymies. Acta Botanica Fennica 121: 1-62.

HAYNES, C.C. 1910. Sphaerocarpos hians sp. nov., with a revision of

the genus and illustrations of the species. Bulletin of the Torrey

Botanical Club 37: 215-230.

HEEG, M. 1891. Verhandlungen der kaiserlich-kbniglichen zoologisch-

botanischen Gesellschaft in Wien 41: 573 (‘Sphaerocarpeae’).

LINDENBERG, J.B.W. 1837. Nachtrage zur Monographic der

Riccieen. Verhandlungen der kaiserlichen Leopoldinisch-car-

olinischen deutschen Akademie der Naturforscher 10 (= Nova

acta Academiae Caesareae Leopoldino Carolinae germanicae

naturae curiosorum 18): 504. i. t. 32.

LONG, D.G. 1993. Notes on Himalayan Hepaticae 1. Sphaerocarpos

subg. Austrosphaerocarpos Schust. in the Nepal Himalaya.

Journal of the Hattori Botanical Laboratory 74: 77-81 .

LOW, A.B. & REBELO, A.G. (eds). 1996. Vegetation of South Africa,

Lesotho and Swaziland. Department of Environmental Affairs

& Tourism, Pretoria.

LUDWIG, C.G. 1760. Definitiones generum plantarum. Leipzig.

MICHELI, PA. 1729. Nova plantarum genera. Firenze.

MONTAGNE, J.F.C. 1838. Centurie des plantes cellulaires exotiques

nouvelles. Annales des Sciences Naturelles, Bot. Ser. 2, 9:

38-57.

MONTAGNE, J.F.C. 1839. Florula boliviensis. In A. d’Orbigny,

Voyage dans I’Amerique meridionale 7, part 2. Paris.

MULLER, K. (Miill.Frib.) 1951-1958. Die Lebermoose Europas. Dr.

L. Rabenhorst's Kryptogamen-Flora 6, edn 3.

NEES AB ESENBECK, C.G. 1838. Naturgeschichte der europdis-

chen-Lebermoose 4. Breslau.

PEROLD, S.M. 1999a. Studies in the liverwort genus Fossombronia

(Metzgeriales) from southern Africa. 10. Three new species

from Northern and Western Cape. Bothalia 29: 83-93.

PEROLD, S.M. 1999b. Studies in the Sphaerocarpales (Hepaticae)

from southern Africa. 1 . The genus Monocarpus and its only

member, M. sphaerocarpus. Bothalia 29: 225-230.

24

Bothalia 30,1 (2000)

PROSKAUER, J. 1954. On Sphaerocarpos stipitatus and the genus

Sphaerocarpos . Journal of the Linnean Society of London,

Botany 55: 143-157.

PROSKAUER, J. 1955. The Sphaerocarpales of South Africa. The

Journal of South African Botany 21: 63-75.

SCHIFFNER, V. 1893. Hepaticae (Lebermoose). In H.G.A. Engler &

K.A.E. Prantl (1909). Die natiirlichen Pflanzenfamilien , edn 1,

1.3: 3-141 [publication date Sept. 1893 fide Stafleu & Cowan,

Taxonomic literature 5: 151 (1985)].

SCHUSTER, R.M. 1992. The Hepaticae and Anthocerotae of North

America 5: 798-827. Field Museum of Natural History, Chicago.

SCOTT, G.A.M. 1985. Southern Australian liverworts. Australian

Flora and Fauna Series, No. 2.

SERGIO, C. & SIM-SIM, M. 1989. Sphaerocarpos stipitatus Bisch. ex

Lindenb. na Europa. Especie introduzida em Portugal desde o

seculo passado. Notulae bryoflorae lusitanicae III. Portugaliae

Acta Biologica, Serie B, 15: 414—416.

STEPHANI, F. 1899. Species hepaticarum 1. Bulletin de I'Herbier

Boissier 7. Geneve.

Bothalia 30,1 : 25-30 (2000)

FSA contributions 15: Piperaceae

K.L. IMMELMAN*

Herbs, shrubs or rarely small trees, evergreen, often

succulent when herbaceous, often epiphytic or lithophyt-

ic, sometimes aromatic, erect or scandent; nodes often

swollen or jointed; stems with vascular strands distinct

and somewhat scattered as in monocots. Leaves usually

alternate, rarely opposite or verticillate, simple, entire;

stipules absent or adnate to petiole. Inflorescence a dense

cylindrical fleshy spike; spikes may be arranged in an

umbel. Flowers minute, bisexual or unisexual, without

perianth, each flower subtended by a peltate bract.

Stamens 1-10, hypogynous; filaments usually free;

anthers often articulated on the filaments; thecae 2, dis-

tinct or confluent. Ovary superior, unilocular, with a sin-

gle basal, erect ovule; stigmas 1-5, short, often brush-

like and lateral in Peperomia. Fruit baccate, with a suc-

culent, thin or dry pericarp, often sunken into succulent

rachis. Seeds globose, ovoid or oblong in outline, testa

membranous or rather fleshy, embryo minute.

A family of about eight genera and over 3 000

species; widespread in warm areas but especially com-

mon in South and Central America and in central Asia.

There are two genera represented in the FSA area, Piper

and Peperomia.

Various species are the source of pepper, cubebs and

various narcotics, and species of both genera are grown

as houseplants for their decorative foliage.

Shrubs; anther thecae usually distinct; stigmas 2 Piper

Herbs; anther thecae usually confluent; stigma 1 Peperomia

1862000 PIPER

Piper L., Species plantarum, edn 1: 28 (1753);

L.: 18 (1754); C.DC.: 240 (1869); Benth. & Hook.f.:

129 (1880); Baker & C.H.Wright: 144 (1902-1913);

C.H. Wright: 488 (1912); Keay: 84 (1954); Emboden: 91,

129 (1972); R.A.Dyer: 29 (1975); Tebbs: 518 (1993);

Verde.: 1 (1996); Diniz: 25 (1997); Mabb.: 560 (1997).

Type: P. nigrum L.

Erect or scandent herbs, shrubs or rarely trees;

branches terete, jointed at nodes. Leaves alternate, entire,

sometimes asymmetrical at base, penninerved; stipules

adnate to petiole or connate into a leaf-opposed sheath.

Inflorescence a dense, cylindrical spike, terminal on ends

of stems or on leaf-opposed short shoots; bracts sessile,

peltate. Flowers bisexual, sessile, without perianth.

Stamens 2; filaments very short; thecae distinct. Ovary

sessile; stigmas 2, distinct, recurved; ovule solitary,

basal. Fruit globose, compressed, sessile; pericarp thin

and dry.

* Range and Forage Institute, Agricultural Research Council, Private

Bag X05, 0039 East Lynne, Pretoria.

Characters not applicable in South Africa: Flowers

unisexual, may be pedicellate. Stamens 3 or 4, rarely

5-many; filaments occasionally longer than bracts.

Ovary rarely stipitate; stigmas up to 5, may be erect.

Fruit rarely stipitate; pericarp may be succulent.

A large genus of over 2 000 species; in Asia as far east

as Japan, in tropical America, with a few species in

Africa. A single species, P capense, occurs in southern

Africa.

As well as P. capense, a second species, P. borbo-

nense (Miq.) C.DC., was recorded from the FSA area by

Wright (1912). He cited specimens of Verraux and

Gueinzius, both without precise locality. Killick (1970)

considered that the Gueinzius specimen was probably

destroyed during World War II, while the Verraux speci-

men proved to be a true P borbonense. Killick suggest-

ed that the wrong label was placed on this specimen, and

certainly no other specimen of this species has been seen

from southern Africa by me. Possibly the Gueinzius

specimen was either mislabelled or misidentified, but

this is now impossible to determine. P borbonense is

quite different from P. capense, as it is parasitic and dioe-

cious.

Various species of Piper are the source of spices and

narcotics. P. nigrum, originally from India and brought

from there to Java, is the source of commercial pepper.

The ripened fruit with the pericarp removed gives white

pepper, while the more pungent black pepper comes

from the unripened fruit milled whole.

In S Asia and the islands of the Indian and Pacific

oceans, Piper betle leaves are rolled with a piece of betel

palm ( Areca catechu) and lime, and chewed as a mild

stimulant. P methysticum, which is closely related to

pepper, is cultivated in the Pacific Islands. The roots and

lower stem are chewed and used to make a beverage

called kava-kava, which has a relaxing effect (Emboden

1972).

Piper capense L.f, Supplementum plantarum 90

(1781); Thunb.: 443 (1823); C.DC.: 339 (1869); C.DC.:

224 (1894); Engl.: 146 (1913); Baker & C.H.Wright: 146

(1902-1913); C.H.Wright: 488 (1912); Eyles: 337

(1915); Keay: 84 (1954); Killick: t. 1583 (1970); Agnew:

88 (1974); Coates Palgrave: 90 (1977).

var. capense Verde.: 5 (1996); Diniz: 27 (1997).

Type: Western Cape, growing in forest in Outeniquas and

Grootvadersbos, near river, Thunberg s.n. (UPS, holo.;

microfiche in PRE, No. 713!).

Herb or soft-wooded shrub, rarely a small tree, some-

times scrambling, 0.5— 2.5(— 3.5) m high; stems swollen

at nodes. Leaves ovate to very broadly ovate, 70-130 x

26

Bothalia 30,1 (2000)

26— 95(— 1 15) mm, palmately 5-7-nerved, apex attenuate,

base truncate or cordate, occasionally cuneate, may be

symmetrical or asymmetrical, dark glossy green above,

paler below, with hairs when young, later only on under-

surface along veins, glabrescent above; petiole chan-

neled above, 10-50 mm long; stipules present. Inflores-

cence a slender cylindrical spike, terminating either the

main shoot or the short leaf-opposed side shoots, ±

30-50 mm long (in flower), up to 80 mm long (in fruit).

Flowers subtended by peltate bracts; stamens 2 or 3;

ovary ovoid; style short, with 2 stigmas. Fruits com-

pressed-ovoid, with a membranous reddish pericarp.

Figure 1.

Occurs in central, western, eastern and southern

Africa; in South Africa is found from Northern Province

to Eastern Cape, and then disjunctly from the Knysna-

Tsitsikamma area as far west as Western Cape: Swel-

lendam, Grootvadersbos (Figure 2). It grows in the un-

derstorey of densely shady, moist forests.

Vouchers: Botha & Van Wyk 1017 (PRE); Galpin 10084 (PRE);

Meeuse 10005 (PRE); Strey 9340 (NH, PRE); Watt & Brandwyk 1767

(PRE).

FIGURE 1. — Piper capense: A,,

habit, x 0.87 ; B, fruiting

inflorescence, x 1.75; C, part

of fruiting inflorescence, x 5;

D, flower with bract and two

stamens, x 13. A, D, Werger

1443 ; B, Story 4207 ; C, Strey

9014. Artist: G. Condy.

The main shoot often ends in an inflorescence or a

leaf and growth then continues from a shoot further back.

These side shoots may continue to grow but more usual-

F1GURE 2. — Distribution of Piper capense var. capense in southern

Africa.

Bothalia 30, 1 (2000)

27

ly end in a leaf or sometimes in an inflorescence. When

they do end in a leaf, the leaves of the plant may appear

superficially to be opposite rather than alternate, but

careful examination will show the presence of the stipule

or of the scar left where the stipule has abscised. The

stipules may either be deciduous or else clasp the petiole.

The fruits have a pleasant spicy smell rather like that

of cloves; this intensifies when they are boiled. They

have been recorded as having been used as a spice in the

past but, as far as is known, are no longer so used.

1866000 PEPEROMIA

Peperomia Ruiz & Pav., Flore peruvianae et

chilensis prodromus: 8 (1794); C.DC.: 391 (1869);

Benth. & Hook.f.: 132 (1880); Baker & C.H.Wright: 147

(1902-1913); C.H.Wright: 489 (1912); Metcalf & Chalk:

1120 (1950); Keay: 81 (1954); Dull: 56 (1973);

R. A. Dyer: 30 (1975); Van Jaarsv.: 67 (1992); Tebbs: 519

(1993); Verde.: 9 (1996): Diniz: 29 (1997); Mabb.: 540

(1997). Type: P. secunda Ruiz. & Pav.

Annual or perennial succulent herbs, ascending or

prostrate; branches jointed at nodes. Leaves alternate,

opposite or verticillate, entire, palmately or penninerved;

stipules absent. Inflorescence a cylindrical spike, termi-

nal or axillary, solitary or aggregated. Flowers bisexual,

sessile on thickened rachis, without perianth, with short-

stalked peltate bracts. Stamens 2; filaments short; anther

thecae 2, confluent. Ovary sessile; stigma undivided,

usually penicellate; ovule solitary, basal. Fruit minute,

globose or ellipsoid; pericarp thin and dry.

Characters not applicable in South Africa: species

may be a subshrub or climber.

A genus of about 1 000 species, cosmopolitan in

warm areas, especially America. In Africa there are 18

species. Five species are recorded from the FSA area by

Dull, but only four are accepted in this treatment.

The species excluded from this treatment is P. pellu-

cida (L.) Humb., Bonpl. & Kunth. Dull (1973) cited only

one specimen from our area, Bey rich 99 (B; Beleg durch

Bomben vernichtet), from Pondoland, which was proba-

bly destroyed in Berlin. The same specimen was quoted

by Wright (1909) under the name P. nana (a synonym of

P. pellucida ), but he did not indicate that he saw the speci-

men, as was his practice with all other specimens seen.

The description is a direct translation from De Candolle

(1869), where P. nana was originally described. Baker &

Wright (1909) did not record P. pellucida from further

south than the Zambesi in Mozambique, and it is possi-

ble that Beyrich 99 was misidentified. No specimens of

P. pellucida from within the FSA area have been seen by

me, and it has therefore been decided to exclude it from

this treatment. The species is annual, with the leaves

alternate, glabrous, cordate at the base, about as wide as

long, and widest below the middle.

Key to species

la Leaves ± ovate in whorls of 3 or 4 at nodes; stems succulent;

flowers densely crowded on rachis 1 . P. tetraphylla

lb Leaves 1 or 2 at nodes; flowers usually loosely arranged

along the axis or, if dense, then leaves orbicular and

steins not succulent:

2a Leaves opposite, pilose 4 P. blanda

2b Leaves mostly alternate, may be opposite below an inflo-

rescence, glabrous or pilose:

3a Leaves orbicular, occasionally broadly ovate, 4—10 mm

long; stems prostrate and non-succulent; fertile por-

tion of inflorescence 2-6 mm long 2. P. rotundifolia

3b Leaves elliptic, ovate or obovate, 8-50 mm long; stems

succulent, usually ascending; fertile portion of inflores-

cence 10-45 mm long 3. P retusa

1. Peperomia tetraphylla (G.Forst.) Hook. & Arn.,

The botany of Captain Beechey’s Voyage: 97 (1841);

Yuncker: 188 (1962); Dull: 72 (1973); Bond & Goldblatt:

355 (1984); Van Jaarsv.: 68 (1992); Verde.: 12 (1996);

Diniz: 29 (1997). Type: Society Islands, Forster s.n.

(GOET, holo.; K, iso.).

Piper tetraphyllum G.Forst.: 5 (1786).

Piper reflexum L.f.: 91 (1781); Thunb.: 443 (1823). Peperomia

reflexa (L.f.) A.Dietr.: 180 (1831) non Humb., Bonpl. & Kunth (1816);

C.DC.: 451 (1869); Hook.f.: 99 (1890); Baker & C.H.Wright: 155

(1902-1913); C.H.Wright: 490 (1912); Compton: 171 (1976). Type:

Western Cape, Outeniqua Mountains, in woods, among mosses on tree

trunks, Thunberg s.n. (UPS, holo.; microfiche in PRE No. 752!).

Peperomia reflexa forma capense Miq.: 169 (1843). P. reflexa var.

capense (Miq.) C.DC.: 451 (1869); C.H.Wright: 490 (1912); Batten &

Bokelmann: 57, t. 50,1 (1966). Syntypes: Western Cape, Cape

Peninsula, Drege s.n. (not found); Eastern Cape, near Uitenhage,

Verraux 1831 in hb. Delessert (G!); Western Cape, Hangklip, Mund &

Maire s.n. (not found); Bourbon Island, Richard in Paris Mus. hb. 707

(P!); Mauritius, Aub. du Petit Thours (not found).

Micropiper pusillum Miq.: 62, t. 5, fig. B (1839). Type: Java,

Tjirebon Province, Tjerimai Mountain, upper slopes on tree trunks,

Blume s.n. (not found).

Succulent herb; stems ascending to suberect (Figure

3E), rooting at nodes and forming a mat. Leaves in

whorls of 3 or 4, succulent, broadly rhomboid-ovate to

ovate, glabrous, 7—1 5(— 25) x 5— 1 1 (— 1 4) mm, base and

apex rounded, faintly 3-veined from base, may have 3

pale stripes along veins, dark green above and paler to

grey-green below, may be punctate below; petiole

1. 5-4.0 mm long. Inflorescence terminal only, fertile

portion 7-11 mm long (in flower), 1 3— 20(— 25) mm long

(in fruit). Flowers with bracts ± 0.3 mm diam. Fruit ± 1

mm long, glanduliferous, long-ellipsoid, with pseudo-

cupule at base.

Pantropical; in southern Africa is found in a number

of localities in Northern Province, North-West, Mpuma-

langa, Gauteng, Swaziland, KwaZulu-Natal, Lesotho

and in Eastern and Western Cape, from East Griqualand

to Bredasdorp (Figure 4). Epiphytic or lithophytic, often

growing among mosses.

Vouchers: Compton 26918 (PRE); Gibbs Russell 3824 (PRE);

Meeuse 10355 (PRE); Scheepers 69 (PRE); Van Wyk & Theron 4534

(PRE).

28

Bothalia 30,1 (2000)

2. Peperomia rotundifolia (L.) Humb., Bonpl. &

Kunth , Nova genera et species plantarum 1: 65 (1816);

Keay: 83 (1954); Dull: 85 (1973); Van Jaarsv.: 68 (1992);

Verde.: 12 (1996); Diniz: 33 (1997). Iconotype: America

Calidiore (Martinique), Plunder, Traitedes Fougeres de

l’Amerique 52, t. 69 (1693).

Piper rotundifolium L.: 30 (1753).

Herb; stems slender, trailing, not succulent, rooting at

nodes. Leaves alternate, succulent, orbicular or some-

times broadly obovate, glabrous or pilose, 4-10 x

3— 7(— 1 0) mm, apex rounded to emarginate, base round-

ed tq broadly cuneate; petiole 1-3 mm long.

Inflorescence axillary, short and dense, 2-6 mm long (in

flower). Fruit not seen.

Found in South America and the southern USA., in

Africa, the Comores and Madagascar; within South

Africa it is rare and restricted in distribution to southern

KwaZulu-Natal and Eastern Cape (Figure 5). It is epi-

phytic or lithophytic.

FIGURE 3. — A-D, Peperomia blan-

da , Hardy 399 1: A, habit, x

0.88; B, flowering inflores-

cence, x 13; C, fruiting inflo-

rescence, x 13; D, fruit, x

26.6. E, Peperomia tetra-

phylla, Venter 3786: habit, x

0.88. Artist: G. Condy.

Vouchers: Flanagan 1821 (PRE); Pegler 763 (PRE); Strey 5823,

6648 (PRE); Venter 883 (PRE).

FIGURE 4. — Distribution of Peperomia tetraphylla in southern

Africa.

Bothalia 30,1 (2000)

29

FIGURE 5. — Distribution of Peperomia rotundifolia in southern

Africa.

3. Peperomia retusa (L.f.) A.Dietr., Species plan-

tarum 1: 155 (1831); C.DC.: 446 (1869); C.H.Wright:

491 (1912); Marloth: 128 (1913); Keay: 82 (1954); Dull:

89 (1973); Compton: 171 (1976); Goldblatt in Bond &

Goldblatt: 355 (1984); Van Jaarsv.: 68 (1992); Verde.: 13

(1996); Diniz: 33 (1997). Type: Cape Province, in

woods, among mosses on tree branches and on rocks,

Thunberg s.n. (UPS, holo.; microfiche in PRE no. 754!).

Piper retusum L.f.: 91 (1781); Willd.: 165 (1797); Thunb : 443

(1813).

Peperomia retusa var. ciliolata C.DC.: 447 (1869); C.H.Wright: 491

(1912). Type: Diaboli Mountain, Fischer s.n. (not seen, LE vide Diill).

It is uncertain whether this locality is American or South African. There

is a Mount Diablo in California, on the mainland opposite San

Fransisco, or it may refer to Devils Peak, Cape Town. No collecter by

the name of Fischer is known to have been in South Africa, according

to Dr L.E.W. Codd (pers. comm.).

Peperomia retusa var. alternifolia C.DC.: 446 (1869). Type: Cape

Province, Ecklon & Zeyher s.n. (B, holo.!).

Peperomia bachmannii C.DC.: 227 (1894); C.H.Wright: 491 (1912).

P. retusa var. bachmannii (C.DC.) Diill: 90 (1973). Type: KwaZulu-

Natal. Pondoland, Bachmann 419 (B, holo.!).

Peperomia rehmannii C.DC.: 227 (1894). Syntypes: Northern

Province, Houtbosch, Rehmann 5969 (not found); Western Cape,

Knysna, Rehmann 489 (Z, holo.!).

Peperomia wilmsii C.DC.: 282 (1898). Type: Mpumalanga,

Lydenburg, Spitzkop Mountain, Wilms 1354 (BM!; G!; K!).

Mat-forming herb; stems rooting at nodes, succulent.

Leaves alternate, obovate or sometimes elliptic, 8-32 x

6-16 mm, succulent, glabrous, apex usually rounded to

emarginate, base broadly cuneate to cuneate, dark green,

may be slightly discolorous; petiole 3-7 (-13) mm long.

Inflorescence slender with flowers widely spaced, termi-

nal and axillary, 10-25 mm long (in flower), 25^15 mm

long (in fruit). Fruit ± 1 mm long.

Occurs in tropical and subtropical Africa and

Madagascar; in southern Africa it is the most widespread

species of the genus, being found from Northern

Province, Mpumalanga, Swaziland, KwaZulu-Natal,

Eastern and Western Cape in moist areas as far west as

FIGURE 6. — Distribution of Peperomia retusa in southern Africa.

the Cape Peninsula, and from sea level to the Dra-

kensberg in Lesotho (Figure 6). Epiphytic or lithophytic.

Vouchers: Bos 1052 (PRE); Mohle 262 (PRE); Venter & Vorster 139

(PRE); Wager s.n. ( PRE30235)\ Zeyde 131 (PRE).

Diill divides the species into three varieties, two of

which, var. retusa and var. bachmannii, occur within the

FSA region. The characters he gives as differing, howev-

er, show considerable intergradation, and it has therefore

been decided here not to keep the two varieties separate.

4. Peperomia blanda (Jacq.) Humb., Bonpl. &

Kunth , Nova genera et species plantarum 1: 67, t. 3'

(1816); Verde.: 18 (1996). Type: specimen grown at

Schoenbrunn, Vienna, from seed collected at Caracas,

Venezuela, Jacquin s.n. (W, holo.; K, photo.!).

Piper blandum Jacq.: 211(1791).

Peperomia leptostachya Hook. & Am.: 70 (1839-1841).

Peperomia arabica Decne. in Miq.: 121 (1843); C.DC.: 442 (1869);

Baker & C.H.Wright: 154 (1902-1913); C.H.Wright: 490 (1912);

Compton: 171 (1976). Syntypes: Arabie felice, PA. Botta s.n. (P!);

KwaZulu-Natal, Port Natal, Drege s.n. (G!).

Peperomia blanda var. leptostachya (Hook. & Am.) Diill: 110, Abb.

16 (1973); Van Jaarsv.: 69 (1992). Type: Hawaii, Oahu Island, Beechey

s.n. (G, holo.!; K!).

Semi-erect herb; stems succulent, rooting at lower

nodes. Leaves opposite, succulent, pilose, ovate, ovate-

rhomboid or slightly obovate, 28-50 x 12-31 mm, apex

acute to rounded, base cuneate, lower leaves smaller and

more rounded; petiole pilose, 6-11 mm long. Inflores-

cence terminal or axillary, slender, with widely-spaced

flowers, 20-45 mm (in flower), 22-145 mm (in fruit).

Bracts ± 0.6 mm diam. Fruit ± 1 mm long. Figure 3A-D.

Pantropical; is the most common Peperomia species

in Central and South America, but less common in

Africa. In southern Africa it occurs from Northern Pro-

vince, Mpumalanga, Swaziland, KwaZulu-Natal to

Eastern Cape (Figure 7). It is lithophytic, rarely epiphytic,

growing on cliffs and in pockets of soil among rocks.

30

Bothalia 30,1 (2000)

FIGURE 7, — Distribution of Peperomia blanda in southern Africa.

Voucher^: Buitendag 470 (PRE); Codd 7029 (PRE); Hemrn 539

(PRE); Schweickerdt 1425 (PRE); Venter 4577 (PRE); Ward 4110

(PRE).

REFERENCES

AGNEW, A.D.Q. 1974. Upland Kenya wild flowers. Oxford University

Press, London.

BAKER, J.G. & WRIGHT, C.H. 1902-1913. Piperaceae. In T.H.

Thiselton-Dyer, Flora of tropical Africa 6,1: 143-162. Reeve,

London.

BATTEN, A. & BOKELMAN, H. 1966. Wild flowers of the eastern

Cape province. Cape & Transvaal Printers, Cape Town.

BENTHAM, G. & HOOKER, J.D. 1880. Genera plantarum 3,1: 129.

Reeve, London.

BOND, P. & GOLDBLATT, P. 1984. Plants of the Cape Flora. Journal

of South African Botany. Suppl. Vol. 13.

COATES PALGRAVE, K. 1977. Trees of southern Africa. Struik, Cape

Town.

COMPTON, R.H 1976. The Flora of Swaziland. Journal of South

African Botany. Suppl. Vol 1 1 National Botanic Gardens, Cape

Town.

DE CANDOLLE, A. C.P. 1869. Piperaceae. In A.P de Candolle, Pro-

dromus systematis naturalis regni vegetabilis 16,1: 39 1 —47 1 .

Treuttel, Paris.

DE CANDOLLE, A. C P 1894 Piperaceae africanae et madagascarien-

sis. Botanische Jahrbiicher 19: 225-230.

DE CANDOLLE, A. C.P. 1898. Piperaceae nova. Annuaire Conserva-

toire et du Jardin botaniques de Geneve 2: 276-288.

DIETRICH, A. 1831. Species plantarum , edn 6, 1.1: 141-186. Nauk,

Berlin.

DINIZ, M.A. 1997. Piperaceae. In E. Launert, Flora zambesiaca 9,2:

24—37 Flora Zambesiaca Managing Committee, London.

DULL, R. 1973. Die Peperomia-Aften Afrikas. Botanische Jahr-

biicher 93: 56-119.

DYER, R.A. 1975. The genera of southern African flowering plants 1.

Department of Agricultural Technical Services, Pretoria.

EMBODEN, W. 1972. Narcotic plants. Studio Vista, London.

ENGLER, A. 1913. Die Pflanzenwelt Afrikas 6,1: 146. Engelmann,

Leipzig.

EYLES, F 1915. A record of plants collected in southern Rhodesia.

Transactions of the Royal Society of South Africa 5: 337.

FORSTER, J.G. A. 1786. Florulae insularum australium prodromus.

Joann Christian Dietrich, Goettingen.

HOOKER, J.D. 1890. The flora of British India 5: 99. Reeve, London.

HOOKER, W.J. & ARNOTT, G.A.W. 1839-1841. The botany of

Captain Beechey's Voyage, part 8-10: 70-97 . Bolin, London.

HUMBOLDT, F.W.H. VON, BONPLAND, A.J.A. & KUNTH, C.S.

1816. Nova genera et species plantarum I: 65. Lutetiae Pari-

siorum, Paris.

JACQUIN, N.J. VON (1789) 1791. Piper blandum. Collecteana Aus-

triaca ad botanicum, chemiam, et historiam naturalem spectan-

tia cumfiguris 3: 211.

KEAY, R.W.J. 1954. In Flora of West tropical Africa, edn 2, 1.1: 81-84.

Crown Agents, London.

KILLICK, D.J.B. 1970. Piper capense. The Flowering Plants of Africa

40: t. 1583.

LINNAEUS, C. 1753. Species plantarum, edn 1 Laurentius Salvius,

Stockholm.

LINNAEUS, C. 1754. Genera plantarum, edn 5: 48. David, Paris.

LINNAEUS, C. fil. 1782 (‘1781’). Supplementum plantarum. Orphano-

trophi, Brunsvigae.

MABBERLEY, D.J. 1997. The plant book. A portable dictionary of the

higher plants, edn 2. Cambridge University Press, Cambridge.

MARLOTH, R. 1913. Piperaceae. The flora of South Africa 1. Darter,

Cape Town.

METCALF, C.R. & CHALK, L. 1950. Anatomy of the Dicotyledons 2:

1120-1127. Clarendon Press, Oxford.

M1QUEL, F.A.W. 1839. Commentarii phytographici 62, t. 5, fig. B, S

& J. Luchtmans, Leiden.

MIQUEL, F.A.W. 1843. Systema piperacearum, fasc. 1: 120-175.

Kramers, Rotterdam.

MIQUEL, F.A.W. 1844. Systema piperacearum, fasc. 2: 303, 343.

Kramers, Rotterdam.

RUIZ, H. & PAVON, J.A. 1794. Flore peruvianae & chilensis prodro-

mus. Cramer, Lehre.

TEBBS, M.C. 1993. Piperaceae. In K. Kubitzki, J.G. Rohwer & V.

Bittrich, The families and genera of vascular plants. 2. dicotyle-

dons: 516-520. Springer- Verlag, Berlin.

THUNBERG, C.P 1823. Flora capensis 1: 433. Cottae, Stuttgart.

VAN JAARSVELD, E.J. 1992. Peperomia species of South Africa.

Aloe 29: 67-69.

VERDCOURT, B. 1996. Piperaceae. Flora of tropical East Africa:

1-21. Balkema, Rotterdam.

WILLDENOW, C.L. VON 1797. Species plantarum I: 165. Berlin.

WRIGHT, C.H. 1912. Piperaceae. In W.H. Thiselton-Dyer, Flora

capensis 5,1: 487^)92. Reeve, London.

YUNCKER, T.G. 1962. Nomenclatural notes on Piperaceae. Brittonia

14: 188.

Bothalia 30,1: 31-33 (2000)

FSA contributions 16: Sphenocleaceae

W.G. WELMAN*

Annual glabrous hygrophytic herbs. Stems erect or

decumbent, somewhat succulent and spongy, usually

swollen with aerenchymatous tissue at base, lower sub-

merged part with a pithy covering. Leaves alternate, sim-

ple, entire, venation pinnate, particularly visible on

lower surface; stipules absent. Inflorescences terminal,

densely spicate, acropetal. Flowers small, bisexual, regu-

lar, subtended by 1 bract and 2 bracteoles. Calyx persis-

tent; tube hemispherical, adnate to ovary; lobes 5, round-

ed, shortly connate, imbricate. Corolla campanulate-

urceolate, perigynous, caducous; lobes 5, imbricate.

Stamens 5, inserted on corolla tube, alternating with

petals; filaments very short, dilated at base; anthers

short, free, suborbicular, 2-thecous, dehiscing longitudi-

nally. Ovary semi-inferior, 2-locular; style 1, very short,

glabrous; stigma capitate, slightly 2-lobed; ovules very

numerous, anatropous, attached to a large spongy stipi-

tate axile placenta. Fruit a membranous, 2-locular,

depressed-globose, circumscissile capsule (pyxidium);

operculum carrying away calyx lobes. Seeds very

numerous, minute, oblong; testa irregularly plicate-

costate; endosperm scanty or absent; embryo axile,

straight, subterete.

A close relationship between Sphenoclea and the

Campanulaceae has been suggested for a long time.

Most systematists have treated the genus either as a

member of the Campanulaceae (Dahlgren 1983), or as a

closely related monogeneric family (Monod 1980;

Cronquist 1988). Sphenoclea is separable from the

Campanulaceae by the imbricate aestivation of the corol-

la lobes, the circumscissile capsule, the absence of a sec-

ondary pollen presentation mechanism, the glabrous

style, and by the apparent absence of latex canals.

Airy Shaw (1948, 1968) regarded this family as an

isolated group, probably marginally related to the Cen-

trospermae, e.g. Phytolacca (habit, anatomy) and per-

haps also to the Primulaceae (circumscissile capsules).

However, Subramanyam (1950) had shown that embry-

ological and anatomical features reveal important differ-

ences between Sphenoclea , Phytolaccaceae and Primu-

laceae. Cosner et al. (1994) suggested a position for

Sphenoclea near the Hydrophyllaceae in the Solanales-

Boraginales, but that family has a late sympetaly. Erbar

(1995) did SEM-investigations on the floral develop-

ment of S. zeylanica and concluded that the early sym-

petalous corolla justifies a position of the Spheno-

cleaceae near the Campanulales-Asterales, to which the

family is usually aligned, whereas it is separable from

these orders by the absence of a secondary pollen pre-

sentation mechanism. Secondary pollen presentation is

one of the characters that help to define the Cam-

panulales-Asterales complex. The family may be more

* National Botanical Institute, Private Bag X101, 0001 Pretoria.

MS. received: 1998-12-15.

primitive than the Campanulaceae and Gustafsson &

Bremer (1995) suggested that Sphenoclea does not

belong in or near the Asterales, but rather in the other

main branch of the Asteridae. The Angiosperm Phylo-

geny Group (led by K. Bremer, M.W. Chase & P.F. Ste-

vens) concluded in 1998 that the Sphenocleaceae,

together with the Convolvulaceae, Hydroleaceae, Montin-

iaceae and Solanaceae form the Solanales which belong

to the Euasterids I; this seems to be strongly supported

by molecular data.

8680000 SPHENOCLEA

Sphenoclea Gaertn., De fructibus et seminibus

plantarum 1: 113, t. 24/5 (1788) nom. cons.; Hemsl.: 480

(1877); Hepper: 307 (1963); Airy Shaw: 1 (1968);

Thulin: 1 (1973); C.D.K.Cook et al.: 533 (1974);

R. A. Dyer: 643 (1975); Thulin: 116 (1983); C.D.K.Cook:

217 (1990); Relief & P.RJ.Herman: 624 (1997). Type:

Sphenoclea zeylanica Gaertn.

Characters of the family.

Two species have been described, namely S. zeylani-

ca Gaertn., which is almost pantropical but probably

introduced and naturalised in the Americas and southeast

Asia, and S. dalzielii N.E.Br. (1912) which is endemic to

West Africa from Senegal to the Central African

Republic. This species, like S. zeylanica, grows in wet

places, but is a straggly herb with sessile obovate-ellip-

tic leaves. The generic name Sphenoclea (Greek) is

derived from spheno (wedge-shaped, sphenoid) and kies

(abundance of), presumably referring to the flowers,

while the species name zeylanica refers to Ceylon (Sri

Lanka) where the type was collected. Common name:

soapweed; Clark 482 (PRE).

Sphenoclea zeylanica Gaertn., De fructibus et

seminibus plantarum 1: 113, t. 24/5 (1788); Hemsl.: 481

(1877); Hepper: 307, t. 272 (1963); Roessler: 138

(1966); Airy Shaw: 1, t. 1 (1968); Thulin: 2, t.l (1973);

Thulin: 116, t. 20 (1983); C.D.K.Cook: 363, t. 367

(1996); Retief & P.P.J. Herman: 624 (1997). Type:

Ceylon, collector unknown (L, holo.).

Roots numerous, long, cord-like, tenuous, orange.

Stems stout, hollow, erect, up to 1.5 m tall, often much

branched. Leaves pale grey-green, linear-lanceolate to

oblong-lanceolate, base attenuate, apex acute to obtuse,

up to 50 x 150 mm; petiole up to 30 mm long. Spikes

cylindrical-conical, ± 10 mm diam. and up to 120 mm

long, but usually much shorter, narrowed at apex; pedun-

cle up to 80 mm long. Bracts and bracteoles spathulate

to oblanceolate-spathulate, tips arched over flowers

except at anthesis. Flowers densely crowded, though

usually only a few open at a time, rhomboid or hexago-

32

Bothalia 30,1 (2000)

FIGURE 1. — Sphenoclea zeylanica

Gaertn. A, habit; B, part of

flowering branch; C, rachis of

inflorescence, showing scars

left by fallen capsules; D,

bract; E, bracteole; F, flower

bud, apical view; G, bud

beginning to open, side view;

H, flower, showing opening

corolla, oblique view; I, gyno-

ecium and calyx, with two

sepals removed, showing cu-

neate base; J, corolla opened

out; K, stamen; L, fruit partly

dehisced; M, t/s fmit; N, seed.

B, N, Milne -Redhead. & Taylor

7463; F, M, Jones FH1 18808;

K, Deighton 132a. A, x 0.2;

B, x 0.9; C-J, L, M, x 5.2; K,

X 10.4; N, x 52. Artist: Olive

Milne -Redhead. Reproduced

with permission of the Royal

Botanic Gardens, Kew.

nal by compression, sessile, wedge-shaped below,

attached longitudinally to rachis by a linear base. Calyx

lobes grey, broadly triangular, apex obtuse, 1 .0-1.5 mm

long, ultimately slightly accrescent and connivent.

Corolla green, tinged white, yellow, mauve or occasion-

ally purple, ± 4 mm across; lobes 2-4 mm long, ovate-

triangular, obtuse or acute, united about half-way, con-

nivent. Filaments shorter than anthers; anthers ±0.5 mm

long. Ovary obovoid, 1 .5-2.5 mm long, apex broad, free,

truncate. Capsule 4-5 mm diam., dehiscing below calyx

lobes, leaving scarious base persistent on rachis. Seeds

yellowish brown, ± 0.5 mm long. Figure 1.

In Africa S. zeylanica is widespread south of the

Sahara (excluding the NE Horn), extending south to

northern Namibia, northern Botswana, Swaziland and

also eastern Mpumalanga and northern KwaZulu-Natal in

South Africa. Also recorded from Madagascar. Figure 2.

It is autogamous but perhaps also sometimes pollinat-

ed by insects. The seeds are dispersed in muddy water

and probably also by other means such as in the mud

stuck to the feet of birds. In southern Africa, S. zeylani-

ca has been recorded on black clay soil, greyish brown

sandy clay-loam and also orange-grey sandy loam. It is

common in swampy areas, in wet or dry mud on the mar-

gins of periodically inundated depressions and flood

plains; it is often found along the banks of water courses

such as tidal creeks and irrigation channels, also in still,

shallow pools. It can be seasonally submerged, emergent

or temporarily terrestrial and can tolerate brackish water.

S. zeylanica has been recorded from grassland and also

Bothalia 30,1 (2000)

33

among Acacia trees; it can be browsed by animals e.g.

zebra. It is often gregarious, becoming a troublesome

weed in rice fields, but it is also cultivated; young plants

are eaten as a vegetable in Java (Indonesia). It grows

from about sea level to 1 500 m altitude. Flowering and

fruiting time: Dec. to May in southern Africa.

Vouchers: Clark 482 (PRE); Giess 15101 (PRE, WIND); Pooley

1609 (NU, PRE); Smith 3710 (PRE, SRGH); Van der Schijff 2593

(KNP, PRE).

ACKNOWLEDGEMENTS

The referees are thanked for valuable comments. Ms

H. Snyman is thanked for providing the distribution map,

through the MAPPIT programme of the NBI. The

Director, Royal Botanic Gardens, Kew is thanked for

permission to reproduce Figure 1.

REFERENCES

FIGURE 2. — Distribution of Sphenoclea zeylanica Gaertn. in South

Africa.

AIRY SHAW, H.K. 1948. Sphenocleaceae. Flora malesiana, ser.1,4:

27, 28.

AIRY SHAW, H.K. 1968. Sphenocleaceae. Flora of tropical East Afri-

ca: 1-3.

ANGIOSPERM PHYLOGENY GROUP. 1998. An ordinal classifica-

tion for the families of flowering plants. Annals of the Missouri

botanical garden 85: 531-553.

BROWN, N.E. 1912. Diagnoses africanae XLIX. Kew Bulletin 1912:

270-283.

COOK, C.D.K., GUT, B.J., RIX, E.M., SCHNELLER, J. & SEITZ, M.

1974. Water plants of the world: 533, 534. Junk, The Hague.

COOK, C.D.K. 1990. Aquatic plant book : 216, 217. SPB Academic

Publishing, The Hague.

COOK, C.D.K. 1996. Aquatic and wetland plants of India: 363. Oxford

University Press, Oxford.

COSNER, M E., JANSEN, R.K. & LAMMERS, T.G. 1994. Phylo-

genetic relationships in the Campanulales based on rbch

sequences. Plant systematics and evolution 190: 79-95.

CRONQUIST, A. 1988. The evolution and classification of flowering

plants, edn 2. New York Botanical Garden, New York.

DAHLGREN, R. 1983. General aspects of angiosperm evolution and

macrosystematics. Nordic Journal of Botany 3: 119-149.

DYER, R.A. 1975. The genera of southern African flowering plants 1 :

643. Department of Agricultural Technical Services, Pretoria.

ERBAR, C. 1995. On the floral development of Sphenoclea zeylanica

(Sphenocleaceae, Campanulales) — SEM investigations on her-

barium material. Botanische Jahrbiicher 1 17: 469^483.

GAERTNER, J. 1788. De fructibus et seminibus plantarum 1: 113, t.

24/5. Academia Carolinae, Stuttgart.

GUSTAFSSON, M.H.G. & BREMER, K. 1995. Morphology and phy-

logenetic interrelationships of the Asteraceae, Calyceraceae,

Campanulaceae, Goodeniaceae and related families (Asterales).

American Journal of Botany 82: 250-265.

HEMSLEY, W.B. 1877. Sphenoclea. Flora of tropical Africa 3: 480,

481.

HEPPER, F.N. 1963. Sphenocleaceae. Flora of West tropical Africa ,

edn 2,2: 307-309.

MONOD, T.H. 1980. A propos du Sphenoclea zeylanica (Spheno-

cleaceae). Adansonia, ser. 2,20: 147-164.

RETIEF, E. & HERMAN, P.P.J. 1997. Plants of the northern provinces

of South Africa: key and diagnostic characters. Strelitzia 6: 624.

ROESSLER, H. 1966. Sphenocleaceae. Prodromus einer Flora von

Siidwestafrika 138: 1.

SUBRAMANYAM, K. 1950. A contribution to our knowledge of the

systematic position of the Sphenocleaceae. Proceedings of the

Indian Academy of Science B3 1,1: 60-65.

THULIN, M. 1973. Sphenocleaceae. Flore d'Afrique centrale: 1-5.

THUL1N, M. 1983. Sphenocleaceae. Flora zambesiaca 7,1: 114—117.

Bothalia 30,1: 35-42 (2000)

Taxonomic studies in the Aizoaceae from South Africa: three new

species and some new combinations

C. KLAK*

Keywords: Aizoaceae, new species, new combinations. South Africa, taxonomy

ABSTRACT

Two new species of Brownanthus, B. glareicola Klak and B. fraternus Klak and one new species of Scopelogena , S.

bruynsii Klak are described. S. gracilis L.Bolus is reduced to synonymy under S. verruculata (L.) L.Bolus. Three new com-

binations are made: Antimima excedens (L.Bolus) Klak, Erepsia dunensis (Sond.) Klak and Hammeria meleagris (L. Bolus)

Klak and full synonomy is given. Lampranthus maximilianii (Schltr. & A. Berger) L.Bolus is transferred back to Braunsia

maximilianii (Schltr. & A. Berger) Schwantes and the identity of Ruschia polita L.Bolus is discussed. The taxonomic posi-

tion of Mesembryanthemum purpureostylum L.Bolus is clarified.

Antimima N.E.Br.

This genus was described by N.E. Brown (1930) and

later amended by Dehn (1989). Recently about 100 spe-

cies were moved from Ruschia Schwantes to Antimima

(Hartmann 1998a). However, no revision of the species

is available. Although species of Ruschia and Antimima

were for many years included in one genus, they are

thought to be unrelated, due to differences in the mor-

phology of their respective capsules. Antimima has cap-

sules of the Leipoldtia type. These have relatively large,

stalked closing bodies, which is considered to be an apo-

morphy for the Leipoldtia group (Hartmann 1991).

Whereas all other genera with a Leipoldtia type of fruit

are multilocular, capsules of Antimima are always 5-loc-

ular. As a consequence of this circumscription of

Antimima , a further species is now moved from Ruschia

to Antimima.

Antimima excedens ( L.Bolus ) Klak, comb. nov.

Ruschia excedens L.Bolus, Notes on Mesembryanthemum and allied

genera part 3: 278 (1954). Type: South Africa, Vanrhynsdorp Div.,

Knersvlakte, Sept. 1950, Meyer SUG12505, sheet 3 (BOL, lecto , here

designated).

Braunsia Schwantes

Braunsia is a small genus of about four or five species

which is thought to be related to the Lampranthus group

(Hartmann 1991). The echinate seeds and leaves that are

fused for a quarter to half of their length, with white, car-

tilagenous margins, were so far considered to be the main

characteristics of the genus.

The lack of echinate seeds in one of the species,

Braunsia maximilianii, prompted L. Bolus (1965) to

place it rather in Lampranthus N.E.Br., which never has

echinate seeds. This view stood in contrast to the earlier

opinion of N.E. Brown (1929). He considered the pres-

* Bolus Herbarium, University of Cape Town, 7701 Rondebosch, Cape

Town, e-mail: klak@botzoo.uct.ac.za

MS. received: 1999-04-08.

ence or absence of echinate seeds to be unimportant in

terms of the genus boundaries, since he had noticed that

this character is found in completely unrelated taxa.

Recent investigations now confirm the views of

Brown. Apart from the morphology of the flowers and

leaves, which are typical for Braunsia, the morphology

of the capsules excludes this species from Lampranthus.

In contrast to Lampranthus, where the expanding keels

always diverge from the base, those of Braunsia are par-

allel for most of their length and only diverge towards

the tips. In addition, the covering membranes in

Lampranthus are always complete and more or less firm,

often with additional closing devices below the covering

membranes. In Braunsia the covering membranes only

cover the locules partly, are fairly flexible and do not

have any additional closing devices. The capsules of

Lampranthus maximilianii were found to be typical of

species of Braunsia in all these details. Therefore the

older name, Braunsia maximilianii is re-instated here.

Braunsia maximilianii (Schltr. & A. Berger) Schwant.

in Gartenwelt 32: 644 (1928a).

Mesembryanthemum maximiliani Schltr. & A. Berger: 633 (1922).

Echinus maximilianii (Schltr. & A. Berger) N.E.Br.: 57 (1929).

Lampranthus maximilianii (Schltr. & A. Berger) L.Bolus: 172 (1965).

Type: Pakhuisberg, 12 Aug. 1897, Schlechter 10817 (B, holo.; BOL!,

GRA!).

Mesembryanthemum apiculatum var. mutica L.Bolus: 149 (1913).

Type: Gifberg, Sept. 1911, Phillips 7664 (BOL, holo.!).

M. phillipsii L.Bolus: 5 (1925). Type: Schlechter 10817 (B; BOL!,

holo.; GRA!, iso.).

M. binum L.Bolus: t. 263 (1927), non N.E.Br. Type: Vanrhynsdorp,

Rood NBG1 249/23 (BOL, holo.).

Specimens examined

WESTERN CAPE. — 3118 (Vanrhynsdorp): Gifberg Plateau,

(-DC), Acocks 14897 (BOL); plateau between Matsikamma and

Gifberg, (-DC), Esterhuysen 30750 (BOL); Nardouw, (-DD), L. Bolus

s.n. (BOL); Cederberg, Kraaibosberg, (-DD), Klak 437 (BOL). 3119

(Calvinia): top of Vanrhyn’s Pass, (-AC), Esterhuysen 7773 (BOL);

entrance to Oorlog’s Kloof, (-AC), L. Bolus BH 19232 (BOL);

36

Bothalia 30,1 (2000)

FIGURE 1. — Epidermal surface of Brownanthus. A-C: stems; E-G, leaves. A, E, B. glareicola, Klak 457 ; B, F, B. corallinus , Klak 64, C, G, B.

fraternus, Klak 171 . D, H, seeds of Brownanthus. D, B. glareicola, Klak 457\ H, B. fraternus, Klak 171. Scale bars: B, F, 200 pm; A, C,

E, G, H, 250 pm; D, 500 pm.

Bothalia 30,1 (2000)

37

Nieuwoudtville, (-AC), E.E. Galpin s.n., (BOL). 3218 (Clanwilliam):

northeast of Pakhuis Pass, (-BB), Esterhuysen 32202 (BOL).

Note : the name ‘maximilianii’ is spelt here with ‘ii’

since ‘Maximilian’ is assumed not to be a latinized name

and therefore ‘maximiliani’ needs to be corrected by

adding an ‘i’ (ICBN Art. 60.11).

Brownanthus Schwantes

Broxvnanthus is one of 11 genera placed in the sub-

family Mesembryanthemoideae. Apart from Psilocaulon

N.E.Br., Aptenia N.E.Br. and Aspazoma N.E.Br., it is one

of the few genera in the Aizoaceae in which stem succu-

lence has developed (Bittrich 1986). At present ten

species are recognised (Pierce & Gerbaulet 1997). The

main characteristics of the genus are that the sepals re-

main upright throughout anthesis, the white or cream-

coloured flowers never possess any filamentous stami-

nodes and the lower part of the capsule is shallowly

bowl-shaped (Ihlenfeldt & Bittrich 1985). Recent explora-

tion in Western Cape has brought two new species to

light both of which can unambiguously be placed in

Brownanthus.

Brownanthus fraternus Klak, sp. nov., a B. coral-

lino foliis minus papillosis manifeste, seminibus grandi-

oribus cum tumore hilari differt.

TYPE.- — Western Cape, 3420 (Bredasdorp): west of

Swellendam, (-AA), Klak 171 (BOL, holo.; K).

Dwarf shrub, decumbent to erect, 150-400 mm high,

500 mm wide. Branches articulated, green, succulent,

becoming woody with age towards bases; internodes

cylindrical, 2. 5-8.0 x 3. 5-4.0 mm, epidermal bladder

cells xeromorphic, ± isodiametric (Figure 1C). Leaves

deciduous, up to 9 mm long, flattened above, convex

below, free towards bases, with mesomorphic epidermal

bladder cells without hair-like extensions (Figure 1G).

Calyx with 5 lobes; lobes shortly connate and erect dur-

ing anthesis. Flowers solitary, borne at tip of stem, 24

mm diam. Staminodes: petaloid staminodes white to

cream-coloured, free to bases, recurving over calyx

lobes; filamentous staminodes absent. Ovary semi-inferi-

or; placentation axile. Fruit a hygrochastic capsule, 5-

locular, 5-7 mm diam., valve wings inflexed' over

expanded valves, seed bags absent. Seeds dark brown, D-

shaped, 1.4- 1.5 mm long, without crest, testa with raised

central papillae, with hilar bulge (Figure 1H). Flowering

time: December.

Distribution and ecology: Bredasdorp and Swellendam

Districts (Figure 2); on quartz patches overlaying clay or

on gravelly shale slopes; winter rainfall up to 400 mm.

Specimens examined

WESTERN CAPE.— 3419 (Caledon): 10 km north of Napier,

(-BD), Bruyns 6844 (BOL), Klak 275 (NBG); 2 km north of Napier,

(-BD), Bruyns & Klak 7807 (K). 3420 (Bredasdorp): west of

Swellendam, (-AA), Klak 171 (BOL, holo.; K).

The habit, floral morphology and, in particular, the

epidermal features of the stem of B. fraternus (Figure

1C) are very similar to those of B. corallinus (Thunb.)

Ihlenf. & Bittrich (Figure IB), but the former may be dis-

tinguished by its less papillate leaves (Figure 1G). The

most conspicuous difference lies in the seeds where there

is a hilar bulge in B. fraternus (Figure 1H), which is

absent in B. corallinus.

B. fraternus is known from three localities between

Swellendam and Bredasdorp, where it appears to be

under threat from agricultural activities. This is much

further south than any other Brownanthus and it is the

only species of Brownanthus which grows on the coastal

plains south of the Langeberg. The nearest locality of its

closest relative, B. corallinus, is over 300 km to the

northwest at the southern end of Namaqualand. The

widespread, but less closely related B. ciliatus, is found

no nearer than 100 km away from B. fraternus.

Brownanthus glareicola Klak, sp. nov., a B.

corallino habitu parviore compactiore, pagina caulium

hirta, foliis minus papillosis manifeste supra concavis

infra convexis foliis junioribus imbricatis, seminibus

cristatis distinguenda est.

TYPE. — Western Cape, 3118 (Vanrhynsdorp): south-

west of Vanrhynsdorp, (-DA), Klak 457 (BOL, holo.; K).

Dwarf, erect shrub up to 170 mm high, 250 mm wide.

Stems articulated, green, succulent, becoming slightly

woody with age towards bases; internodes cylindrical,

3. 5-5. 5 x 3. 5-4. 5 mm, epidermal bladder cells xero-

morphic with hair-like extensions (Figure 1A). Leaves

deciduous, up to 3-5 mm long, concave above, convex

below, overlapping at bases, epidermal bladder cells

mesomorphic without hair-like extensions (Figure IE).

Calyx with 5 lobes; lobes shortly connate and erect

FIGURE 2. — Distribution of Brownanthus glareicola, ▼; B. coralli-

nus, ■; B. fraternus, •.

38

Bothalia 30,1 (2000)

during anthesis. Flowers solitary at tips of stems, 10-15

mm diam. Staminodes : petaloid staminodes white to

cream-coloured, free to bases, recurving over calyx

lobes; filamentous staminodes absent. Ovary semi-infe-

rior; placentation axile. Fruit a hygrochastic capsule, 5-

locular, (4-)5-6 mm diam., valve wings inflexed over

expanded valves, without seed bags. Seeds brown, D-

shaped, 1.3-1. 5 mm long, with a crest, hilar bulge

absent, testa cells slightly raised (Figure ID). Flowering

time : October.

Distribution and ecology : Knersvlakte, southwest of

Vanrhynsdorp (Figure 2); flat to gently sloping patches

of quartz gravel on clay; winter rainfall, 100-200 mm

per year.

Etymology, glareicola = gravel dweller.

Specimens examined

WESTERN CAPE. — 3118 (Vanrhynsdorp): southwest of

Vanrhynsdorp, (-DA), Klak 457 (BOL, holo.; K); gypsum mine,

(-DA), Klak 458 (BOL).

The material upon which the description is based,

comes from a locality south of Vanrhynsdorp, where the

plants were first noticed by A. Ellis and P. Desmet in

1996. The species had previously been misidentified as

Brownanthus corallinus and grows together with it in

one of the known localities. The new species is distinct-

ly smaller and more compact in growth than B. coralli-

nus and the older stems are only slightly woody towards

their bases. It also differs from B. corallinus (Figure IB)

by the hair-like epidermal cells which cover the stems

(Figure 1A). Furthermore, the bladder cell idioblasts of

the leaves are noticeably more reduced in size compared

to the ones in B. corallinus and the bases of young leaves

overlap, which is never found in B. corallinus (Figure

IE, F). In addition, the leaves in B. corallinus are sub-

cylindrical, whereas in B. glareicola they are convex

below and concave above. Further differences may be

found in the seeds. These have a conspicuous crest in B.

glareicola, which is not known for any other species of

Brownanthus (Figure ID).

Whereas B. corallinus is known from numerous local-

ities over a distance of about 250 km in Namaqualand, B.

glareicola appears to be much rarer and is known from

only two localities in the southern Knersvlakte (Figure

2). In all of these it grows on patches of flat to gently

sloping quartz gravel.

Erepsia N.E.Br.

Erepsia was revised by Liede (1989). Characteristics

separating it from the closely allied genus Lampranthus

N.E.Br., are the triquetrous leaves with tough epidermis

and the presence of a hypanthium. Liede (1989) subdi-

vided the genus into four sections, three of which share

flowers with vertical hypanthium walls and leaves usual-

ly less than 5 mm diam. The remaining section, Crassi-

foliae, consists of seven species in which the walls of the

hypanthium are sloped outwards and the leaves are most-

ly fairly thick (5-10 mm diam.).

Recently, plants of the very rare Lampranthus dunen-

sis were rediscovered on the Cape Flats. This species was

described as Mesembryanthemum dunense by Sonder

(1862). In Jacobsen (1960), L. Bolus moved it to Lam-

pranthus and placed it in the informal section Reptantes.

An examination of the flowers of this material showed

that they differ from those found in all species of

Lampranthus, in that they possess a conspicuous, out-

wardly sloped hypanthium. In addition, the broad, tri-

quetrous leaves with tough epidermis are typical of

species placed in Erepsia section Crassifoliae. This fea-

ture, together with the noticable triquetrous leaves suggest

that L. dunensis belongs in Erepsia. The leaves are 4-6

mm broad and fall within the limits included by Liede in

Sect. Crassifoliae. It is now transferred to Erepsia.

Erepsia dunensis (Sond.) Klak, comb. nov.

Mesembryanthemum dunense Sond., Flora capensis 2 : 411 (1862).

Lampranthus dunensis (Sond.) L. Bolus in Jacobsen: 1197 (1960).

Syntypes: Western Cape, seashore near Cape Town, Ecklon s.n.. Pappe

s.n. (TCD, lecto.!, here designated; S).

Mesembryanthemum macrocalyx Kensit: 153 (1909). Type: Western

Cape, Skurfkop near Somerset West, Nov. 1907, Pillans 1423 (BOL,

holo.).

Specimens examined

WESTERN CAPE. — 3418 (Cape Town): Sandvlei, (-AB), L. Bolus

18590 (BOL); Klipfontein Road, (-AB), Maytham s.n. (BOL); north-

west of Fish Hoek, (-AB), Pillans 3626 (BOL); Zeekoevlei, (-AB),

Starke s.n. (BOL); Wolfgat Nature Reserve, (-BB), Klak 453 (BOL).

Hammeria P.M.Burgoyne

Hammeria was recently established for two species

from the Ceres Karoo. The type of the genus is Ham-

meria salteri (L.Bolus) RM.Burgoyne, with Ruschia salteri

L. Bolus as its basionym (Burgoyne et al. 1998). Even

more recently Hartmann (1998b) published a new name,

Lampranthus tanquanus H.E.K. Hartmann, based on the

same type, without reference to the earlier publication of

Burgoyne et al. (1998).

The absence of closing bodies and the presence of

valve wings were the reasons for excluding Ruschia

salteri from Ruschia. Burgoyne et al. (1998) suggested

morphological similarities to Antimima, Cheiridopsis

and Chasmatophyllum, but were unable to place R.

salteri into any of the existing genera. Hartmann (1998b)

noted correctly that the fruits resembled those of the

Titanopsis type, with thin, flexible, covering membranes

bending down into the empty locules, and almost com-

plete covering membranes. In addition, the locules are

shallow, so that the lower part of the capsule is bowl-

shaped. In contrast, fruits of Lampranthus are always

deep, ± funnel-shaped, with firm, complete covering

membranes and often additional closing devices below

the covering membranes. It is therefore not possible for

this species to remain in Lampranthus. From the capsule

morphology, it appears more likely that closer relation-

ships may be found with species placed in the Titanopsis

or the Stomatium group. A closer study of this complex

is needed before it can be decided whether Hammeria

Bothalia 30,1 (2000)

39

may be included in one of the existing genera or indeed

deserves generic status.

However, it has so far been overlooked that the same

species was named several times by L. Bolus. Since one

of these names predates R. salteri , a new combination

and synonomy are necessary.

Hammeria meleagris ( L.Bolus ) Klak, comb. nov.

Mesembryanthemum meleagris L.Bolus, in Notes on Mesem-

brianthemum and allied genera, part 2: 17 (1928). Lampranthus melea-

gris (L.Bolus) L.Bolus: 169 (1939). Type: South Africa, Western Cape,

Ceres Div., near Karoopoort, Sept. 1928, Pillans 6130 (BOL, holo. !).

Lampranthus longisepalus L.Bolus: 169 (1939). Mesembryanthemum

longisepalum L.Bolus: 198 (1930). Type: Western Cape, between

Karoopoort and Calvinia, Jan. 1930, Leipoldt BOL19135 (BOL, holo.!).

Ruschia salteri L.Bolus: 371 (1932b). Hammeria salteri (L.Bolus)

P.M.Burgoyne et al.: 206 (1998). Lampranthus tanquanus H.E.K. Hart-

mann: 70 (1998b). Type: South Africa, Western Cape, beyond Karoo-

poort near the road to Sutherland, 3 Aug. 1932, Salter 2668 (BOL,

lecto.!).

Lampranthus stoloniferus L.Bolus: 307 (1965). Type: South Africa,

Western Cape, Sutherland Div., 55 miles east of Karoopoort, June

1965, F.J. Stayner KG214/65 (BOL, holo.!).

Scopelogena L.Bolus

Mesembryanthemum verruculatum was already

known to Linnaeus and had been introduced to England

in the early 1730’s (Dillenius 1732). L. Bolus later

placed this species in Lampranthus (Bolus 1950). The

plant, however, could be separated from Lampranthus by

the apparent lack of valve wings and the fact that the cap-

sules do not close completely once they have opened. On

the other hand it differs from Ruschia by the absence of

closing bodies and by the yellow colour of the petals.

These reasons prompted L. Bolus to establish a new

genus, Scopelogena, in 1962 for this species. At the same

time she described a second species of Scopelogena , S.

gracilis L.Bolus (1962).

S. verruculata is known only from the Cape Peninsula,

whereas S. gracilis was described from Grootvadersbosch,

east of Swellendam. The latter was said to differ from S.

verruculata by the more slender, sometimes obtusely

keeled, often laterally compressed leaves, with the capsule

obconical inside and not angled. Newly collected material

from west and south of Swellendam has shown that none

of these characters consistently separate S. verruculata

from S. gracilis and that they are conspecific. S. verrucu-

lata is distributed from Cape Town sporadically eastwards

to Herbertsdale, usually occurring on exposed, locally

arid, sandstone outcrops.

These investigations have also shown that, contrary to

previous descriptions (Bolus 1962), newly ripened cap-

sules which were opened for the first time possessed very

narrow, ‘seam-like’ valve wings. -In most of the old cap-

sules this character is no longer visible and this might

explain the fact that it was not mentioned in previous

descriptions.

Recent exploration has brought to light a second

species, which occurs further north of the distribution

area of S. verruculata. This is now described. Despite the

large size of the plant, this species has so far been over-

looked by most collectors. It is named after P.V. Bruyns,

who was the first to notice it in Namaqualand in 1992.

Scopelogena bruynsii Klak, sp. nov., a S. verrucu-

lata floribus parvioribus luteis ad roseis salmoneisve,

praesentia staminodiorum filamentosorum, fructibus re-

petite claudentibus discedit.

TYPE. — South Africa, Western Cape, Namaqualand,

3118 (Vanrhynsdorp): 10 km north of Nuwerus, (-AB),

Klak & Bruyns 462, (BOL, holo.; K).

Woody shrub up to 0.3 m high, 1 m wide, with stout,

erect or spreading grey to brown stems. Leaves crowded,

shortly connate, incurved, erect, 3-angled to cylindrical,

up to 45 x 5-8 mm, bluntish, shortly mucronate, soft,

whitish grey to slightly reddish. Calyx with 5 subequal

lobes. Flowers in much-branched inflorescence; pedicels

4-11 mm long; bracts up to 15 x 11 mm. Staminodes :

petaloid staminodes in 1 series; filamentous staminodes

present, conically collected, papillate at bases; yellow,

salmon or pale pink. Stamens erect, outer stamens papil-

late up to middle; anthers yellow; filaments same colour

as staminodes. Nectaries in crenulate ring. Ovary semi-

inferior; stigmas filiform, shorter than tallest stamens;

placentation parietal. Fruit a hygrochastic capsule,

ochre, relatively soft and not woody, 5-locular, 3. 0-4. 5

mm diam., top convex, lower part deep, funnel-shaped,

with very narrow valve wings when young, covering

membranes present, almost completely covering locules,

with fine ridge below covering membrane, expanding

keels diverging, without closing body. Seeds dark brown,

obovate, tuberculate, 1.0-1. 1 mm long. Flowering time:

September and October.

Distribution and ecology : Namaqualand to Clan-

william and Ceres Karoo (Figure 3); on low, sandstone

cliffs; in areas receiving winter rainfall of 100-200 mm.

Specimens examined

NORTHERN CAPE. — 3018 (Kamiesberg): near Leeukuil, (-DC),

Bruyns 5267a (BOL).

WESTERN CAPE. — 3 1 1 8 (Vanrhynsdorp): 1 0 km north of Nuwe-

rus, (-AB), Klak & Bruyns 462, (BOL, holo.; K); Matsikamma, (-DB),

Helm 1656 (BOL). 3119 (Calvinia): Soutpan, (-CD), Klak 568 (BOL).

3219 (Wuppertal): Dassiekloof, (-BA), Klak 424 (BOL); south of

Elandsvlei, (-DA), Klak 565 (BOL), Van Jaarsveld 13579 (BOL).

Scopelogena verruculata (L.) L.Bolus in Journal

of South African Botany 28: 9 (1962).

Mesembryanthemum verruculatum L.: 486 (1753). Lampranthus ver-

ruculatus (L.) L.Bolus: 385 (1950). Ruschia verruculata (L.)

G.D. Rowley: 9 (1978). Iconotype: Dill., Hort. Eltham. t. 203, fig. 259

(1732).

Scopelogena gracilis L.Bolus: 10 (1962). Ruschia scopelogena

G. D. Rowley: 9 (1978). Type: Western Cape, Grootvadersbos, Dec. 1958,

H. Hall 1506 (BOL, holo.!).

40

Bothalia 30,1 (2000)

FIGURE 3. — Distribution of Scopelogena verruculata , •; and

Scopelogena bruynsii, Y.

Woody shrub up to 0.3 m high, 2 m wide, with stout,

erect or spreading, grey to brown stems. Leaves crowd-

ed, shortly connate, incurved, erect, 3-angled to cylindri-

cal, obtuse, up to 16-50 x 3-5 mm, shortly mucronate,

very soft, grass green to grey or sometimes reddish.

Calyx of 5 subequal lobes. Flowers in much-branched

inflorescence; pedicels 4-15 mm long; bracts up to 15 x

15 mm. Staminod.es : petaloid staminodes in 1 or 2 series,

yellow rarely white; no filamentous staminodes observ-

ed. Stamens erect, outer stamens dorsally papillate up to

middle, inner ventrally papillate at middle; anthers and

filaments yellow. Nectaries in crenulate ring. Ovary

semi-inferior; stigmas filiform, longer than stamens; pla-

centation parietal. Fruit a hygrochastic capsule, not fully

closing again once opened, ochre, relatively soft, not

woody, 5-locular, 5-9 mm diam., top convex, lower part

deep, funnel-shaped, with very narrow valve wings when

young, covering membranes incomplete to almost com-

plete, with inconspicuous ridge below covering mem-

brane, expanding keels diverging, without closing body.

Seeds dark brown, obovate, tuberculate, 1. 1-1.2 mm

long. Flowering time : September-October.

Distribution and ecology: Cape Peninsula to

Riversdale (Figure 3); on sandstone cliffs, with winter

rainfall of 300-600 mm annually.

Specimens examined

WESTERN CAPE. — 3318 (Cape Town): rocks west of Lion’s

Head, (-CD), Hall s.n. (BOL); Table Mountain, (-CD), Klak 318

(BOL), Marloth 2851 (BOL); Lion’s Head, (-CD), Wolley Dod 2420

(BOL). 3419 (Caledon): 10 km N of Napier, Karsrivier, (-BD), Klak

278 (BOL). 3420 (Bredasdorp): 7 km W of Swellendam, (-AB), Klak

177 (BOL); Grootvadersbos, (-BB), Hall 1506 (BOL); Boskloof,

Potberg, (-BC), Burgers 1627 (NBG). 3421 (Riversdale): Glen Leith,

(-AA), Muir 4320 (BOL).

S. verruculata and S. bruynsii are morphologically

very similar and there appears to be no doubt that this

new species belongs in Scopelogena. This is dispite the

fact that in the new species the capsules open and close

repeatedly. Consequently one of the main distinctions of

Scopelogena (that the capsules remain open) falls away.

Similar variability in this feature is also found in species

of Aridaria (Gerbaulet 1996). The flowers of S. bruynsii

may be yellow, pink or salmon-pink. S. verruculata has

predominantly yellow or rarely white flowers.

S. bruynsii is found over a wide area, from Nuwerus

to Clanwilliam and eastwards to the Ceres Karoo. The

plants are always found growing on low sandstone cliffs.

The character combinations present in Scopelogena

do not suggest clear affinities to any other genus in the

Ruschiodeae and, consequently the relationships of Sco-

pelogena need further study.

The correct identity of Mesembryanthemum purpureo-

stylum L. Bolus

In recent literature, much confusion has arisen over

the correct identity of Mesembryanthemum purpureosty-

lum L. Bolus. The type was based on a collection from

Bonnievale. L. Bolus later included this name in the syn-

onymy of Ruschia forficata (L.) L. Bolus. Bruyns (1997)

was the first to notice that in fact two different species

were involved: R. forficata (L.) L. Bolus was found to be

a synonym of Erepsia forficata (L.) Schwantes. Based on

the presence of closing bodies and the absence of valve

wings, Bruyns (1997) transferred M. purpureostylum to

Ruschia , but, shortly afterwards, it was moved to Acro-

don by Burgoyne (1998). In the same year, Hartmann

(1998b) published a new combination in Cerochlamys

based on the type of M. purpureostylum L. Bolus, how-

ever, without reference to any of these recent publica-

tions.

The absence of a capsule on the type poses some dif-

ficulties for the correct position of this species. However,

a watercolour painting of the type specimen exists in the

Bolus Herbarium (by M. Page) and this shows the habit,

as well as the leaves and flowers. From this it is obvious

that the flowers have their filamentous staminodes and

stamens collected into a central cone, with the filamen-

tous staminodes overtopping the stamens and partly con-

cealing them. This arrangement is found in numerous

other genera such as Ruschia and Antimima, but in none

of the other species of Cerochlamys. In Cerochlamys the

filamentous staminodes are loosely arranged in a cylin-

der around the stamens, sometimes with the outer ones

spreading horizontally (Hartmann 1998b: 51, t. 16, 18,

20). The flower of Acrodon purpureostylus shown by

Burgoyne matches that of the type. Hartmann, on the

other hand, makes no mention of this typical cone flower.

In addition, Hartmann did not find any closing bodies

in the capsules of the two collections which she cited.

Both Bruyns and Burgoyne mention the presence of clos-

ing bodies in their collections. Both Bruyns’ and

Burgoyne’s collections agree with the illustration of the

type in all details of habit, leaves and flowers. It can

therefore only be deduced that Hartmann must have

based her conclusions on misidentified material.

Bothalia 30,1 (2000)

41

From the characteristics found to be typical of

Acrodon (Hartmann 1996), Burgoyne (1998) showed con-

clusively that the species should be placed in Acrodon

and no reasons have been put forward to alter this. A

complete list of synonyms is given below.

Acrodon purpureostylus ( L.Bolus ) P.M. Burgoyne

in Aloe 35: 60 (1998).

Mesembryanthemum purpureostylum L.Bolus: 6 ( 1920). Erepsia pur-

pureostyla (L.Bolus) Schwantes: 68 (1928b). Ruschia purpureostyla

(L.Bolus) P.V.Bruyns: 41 (1997). Cerochlamys purpureostyla (L.Bolus)

H.E.K. Hartmann: 56 (1998b). Type: South Africa, Bonnievale,

Mathews sub NBG3426/15 , sheet 2 (BOL, lecto., here designated).

The identity of Ruschia polita L.Bolus

Until now much uncertainty has existed about the

identity of Ruschia polita L.Bolus, which was described

in 1932. The type specimen was collected by G. Nel at

Touws River in April 1930. Bolus was uncertain whether

this species belonged in Ruschia, as she placed a ques-

tion mark next to the genus in her text. The absence of

fruits on the type sheet may have added to the confusion.

Annotations on the type sheet suggest that it belongs

either in Corpuscularia or Antimima.

However, species placed in Corpuscularia are charac-

terised by a peculiar epidermis, which is made up of

tightly packed, dome-shaped epidermal bladder-cells.

The surface of R. polita, on the other hand, consists of

flattened cells and is completely smooth. In addition,

species of Corpuscularia have so far only been recorded

from a fairly small area around Port Elizabeth, in the

Eastern Cape. Thus, the difference in epidermal mor-

phology and the very disjunct distribution would suggest

that the type specimen of Ruschia polita is unlikely to

belong to Corpuscularia.

Exclusion of this species from Antimima and Ruschia

is more difficult since the type lacks any fruits.

Nevertheless, since Bolus did not mention the clustering

of filamentous staminodes into a dense cone, it seems

unlikely that it is either a species of Antimima or of

Ruschia.

A notable feature of the type specimen is, however, its

sharply keeled, thick leaves with cartilaginous margins.

This is typical of species placed in Braunsia. The

description of Braunsia geminata matches very closely

that of Ruschia polita, both with respect to the morphol-

ogy of the leaves as well as the flowers. In addition, the

distribution of B. geminata extends from the Ceres

Karoo to Prince Albert and the Little Karoo and so the

type locality of R. polita falls within this area. Thus it is

reasonable to conclude that R. polita is conspecific with

B. geminata.

Braunsia geminata (Haw.) L.Bolus in Journal of

South African Botany 33: 306 (1967).

Ruschia polita L.Bolus: 332 (1932a); L.Bolus: 260 (1954). Type:

Laingsburg Dist.; Touw’s River, Apr. 1930, G. Nel SUG9095 (BOL,

holo.!).

ACKNOWLEDGEMENTS

This research was funded by the FRD, the Mesemb

Study Group and the American Cactus and Succulent

Society of America.

REFERENCES

BITTRICH, V. 1986. Untersuchungen zum Merkmalsbestand, Glie-

derung und Abgrenzung der Unterfamilie Mesembryanthe-

moideae (Mesembryanthemaceae Fenzl.). Mitteilungen aus

dem Institut fiir Allgemeine Botanik Hamburg 21: 5-116.

BOLUS, H.M.L. 1913. Mesembrianthemurn. Annals of the South Afri-

can Museum 9: 140-154.

BOLUS, H.M.L. 1920. Novitates africanae. Annals of the Bolus Her-

barium 3: 6.

BOLUS, H.M.L. 1925. Novitates africanae. Annals of the Bolus Her-

barium 4: 5.

BOLUS, H.M.L. 1927. Mesembrianthemurn binum. The Flowering

Plants of South Africa 7: t. 263.

BOLUS, H.M.L. 1928. Notes on Mesembrianthemurn and allied gen-

era, part 2: 17-32. Bolus Herbarium, University of Cape Town.

BOLUS, H.M.L. 1930. Notes on Mesembrianthemurn and allied gen-

era, part 2: 193-208. Bolus Herbarium, University of Cape

Town.

BOLUS, H.M.L. 1932a. Notes on Mesembrianthemurn and allied gen-

era, part 2: 309-336. Bolus Herbarium, University of Cape

Town.

BOLUS, H.M.L 1932b. Notes on Mesembrianthemurn and allied gen-

era, part 2: 357-376. Bolus Herbarium, University of Cape

Town.

BOLUS, H.M.L. 1939. Notes on Mesembryanthemum and allied gen-

era, part 3: 139-188. Bolus Herbarium, University of Cape

Town

BOLUS, H.M.L. 1950. Lampranthus verruculatus (L.) L.Bolus. In R.S.

Adamson & T.M. Salter, Flora of the Cape Peninsula. Juta,

Cape Town.

BOLUS, H.M.L. 1954. Notes on Mesembryanthemum and allied gen-

era, part 3: 237-288.

BOLUS, H.M.L. 1962. Notes on Mesembryanthemum and allied gen-

era. Journal of South African Botany 28: 9-11.

BOLUS, H.M.L. 1965. Notes on Mesembryanthemum and allied gen-

era. Journal of South African Botany 31 : 172, 307.

BOLUS, H.M.L. 1967. Notes on Mesembryanthemum and allied gen-

era. Journal of South African Botany 33: 308, 309.

BROWN, N.E. 1929. New species and critical notes on Mesem-

bryanthemum and allied genera. Kew Bulletin 1929: 56-62.

BROWN, N.E. 1930. Mesembryanthemum. The Gardeners' Chronicle,

ser. 3, 87: 13-211.

BRUYNS, RV. 1997. A new combination in the genus Ruschia and the

correct name for ‘ Cynanchum capense' . South African Journal

of Botany 63: 241

BURGOYNE, RM. 1998. Finding a place in the sun: where does

Ruschia purpureostyla belong? Aloe 35: 60, 61.

BURGOYNE, P.M , SMITH, G.F. & CHESSELET, P. 1998. Ham-

meria, a new genus of Aizoaceae from South Africa. Cactus

and Succulent Journal (US) 70: 203-208.

DEHN, M. 1989. Untersuchungen zum Merkmalsbestand und zur Stel-

lung der Gattung Antimima N.E.Br. emend. Dehn (Mesem-

bryanthemaceae Fenzl.) Mitteilungen aus dem Institut fiir

Allgemeine Botanik Hamburg 22: 189-215.

DILLENIUS, J.J. 1732. Hortus elthamensis: 268, t. 203, fig. 259.

Londini, sumptibus auctoris.

GERBAULET, M. 1996. Revision of the genus Aridaria N.E.Br.

(Aizoaceae). Botanische Jahrbiicher 118: 41-58.

GREUTER, W„ BARRIE, F.R., BURDET, H.M., CHALONER, W.G.,

DESMOULIN, V„ HAWKSWORTH, D.L., JORGENSEN,

PM., NICOLSON, D.H., SILVA, PC. TREHANE, P. &

MCNEILL, J. (eds). 1994: International Code of Botanical

Nomenclature, adopted by the Fifteenth International Botanical

Congress, Yokohama, August-September 1993. Konigstein.

HARTMANN, H.E.K. 1991 . Mesembryanthema. In H P Linder & A.V.

Hall, Systematics, biology and evolution of some South African

taxa. Contributions from the Bolus Herbarium 13: 75-157.

HARTMANN, H.E.K. 1996. Miscellaneous taxonomic notes on

Aizoaceae. Bradleya 14: 29-56.

HARTMANN, H.E.K. 1998a. New combinations in Antimima (Ruschi-

oideae, Aizoaceae) from southern Africa. Bothalia 28: 67-82.

42

Bothalia 30,1 (2000)

HARTMANN, H.E.K. 1998b. New combinations in Ruschioideae,

based on studies in Ruschia (Aizoaceae). Bradleya 16: 44-91.

IHLENFELDT, H.-D. & BITTR1CH, V. 1985. Morphologic, Glieder-

ung und Abgrenzung der Gattung Psilocaulon N.E.Br. s.l. (Me-

sembryanthemaceae). Botanische Jahrbiicher 105: 289-322.

JACOBSEN, H. 1960. A handbook of succulent plants 3. London.

KENS1T, H.M.L. 1909. Contributions to the African flora. Transac-

tions of the Royal Society of South Africa 1: 153.

LIEDE, S. 1989. Untersuchungen zum Merkmalsbestand und zur

Taxonomie der Erepsiinae (Mesembryanthemaceae). Beitrdge

zur Biologie der Pflanzen 64: 391-479.

LINNAEUS, C. 1753. Species plantarum, edn 1: 486. Salvius, Stock-

holm.

PIERCE, S.M. & GERBAULET, M. 1997. Brownanthus Schwantes

(Mesembryanthemoideae, Aizoaceae): two new species and a

new combination from the Richtersveld and southwestern

Namibia. Aloe 37: 42^44.

ROWLEY, G. 1978. Reunion of some genera of Mesembryanthemaceae.

The National Cactus & Succulent Journal (UK) 33: 6-9.

SCHLECHTER, ERR. & BERGER, A. 1922. Mehrere neue Mesem-

brianthemum und eine Aloe. Botanische Jahrbiicher 57: 626-644.

SCHWANTES, G. 1928a. Braunsia. Gartenwelt 32: 644.

SCHWANTES, G. 1928b. Mesembriaceen unserer Kulturen in neuer

Benennung. Gartenflora 77: 68, 69.

SONDER, O.W. 1862. Mesembryanthemum. In W.H. Harvey & O.W.

Sonder, Flora capensis 2: 395^-60.

Bothalia 30,1: 43-55 (2000)

Notes on African plants

VARIOUS AUTHORS

AGAVACEAE

AGAVE VIVIPARA: A NATURALISED ALIEN IN SOUTHERN AFRICA

.invasive species are a kind of habitat destruction. When invasive

species take over a habitat, they erase the native richness and diversi-

ty of species. What we really need is an everyday concern among ordi-

nary people about biodiversity and the issue of invasiveness. '

— Dr Vandana Shiva in conversation with

IUCN’s Ricardo Bayon (1997-1998)

The indigenous succulent flora of southern Africa rep-

resents 55 mostly unrelated plant families and includes

all possible types of life forms. A small minority of exotic

succulents (24 species according to Smith et al. 1997)

can be regarded as naturalised in southern Africa. These

species are mostly representative of the Cactaceae (19

species) and, to a much lesser degree, the Basellaceae (1

species), Agavaceae (2 species) and Portulacaceae (2

species) (Smith et al. 1997). Some of the species of

Cactaceae, e.g. Opuntia ficus-indica (L.) Mill, (prickly

pear) and O. aurantiaca Lindl. (jointed cactus) are

aggressive noxious weeds that have invaded and trans-

formed certain parts of the southern African landscape.

In recent years, at least among urban gardeners, the

popularity of agavaceous taxa is seemingly on the

increase. The plants are being used as inexpensive, per-

manent barriers to ward off intruders. With their formid-

able spines, large and compact habit, drought resistance

and aggressive growing capability, they are ideally suit-

ed for this purpose. With the abandonment of habitations,

these hardy agaves can persist without human interven-

tion and subsequently become adventive aliens (Kloot

1987) in the local flora. One such species, Agave decipi-

ens Baker was recently recorded in South Africa (Smith

& Steyn 1999a). This paper deals with a comparable

alien, namely A. vivipara L. var. vivipara.

Agave vivipara. often known by one of its synonyms,

A. angustifolia Haw., is a variable species native to cen-

tral America (Gentry 1982; Forster 1992). It is thought

that the species was the wild ancestor of henequen (A.

fourcroydes Lem.), a cultivated species known world-

wide for the high quality of its fibres (Colunga-

GarclaMarin & May-Pat 1993; Colunga-GarcfaMarin et

al. 1999). Mr Bernard Ulrich (Pforzheim, Germany) has

in turn suggested to the second author that A. vivipara is

possibly a selected form of the widely cultivated A.

sisalana Perrine, source of sisal hemp. Currently, six

varieties, including the typical variety, and two cultivars

are recognised in A. vivipara (Forster 1992). According

to Gentry (1982), this species complex has the most

wide-ranging distribution of agaves in North America

(for a list of exsiccatae see Gentry 1982: 586-590). The

plants commonly occur in tropical savanna, thorn forest

and drought-deciduous tropical forests. However, the

species also survives in extreme habitats such as the arid

Sonoran Desert with about 250 mm rain per annum and

in montane pine-oak forests with a yearly precipitation of

1 680 mm. In these natural habitats, wild populations of

A. vivipara exhibit a gradient in morphological variation,

with characters such as plant size, length of leaves, dis-

tribution of marginal teeth and mass of leaf fibres show-

ing a high degree of plasticity. Also, improved growth

conditions in gardens result in an increase in leaf length

and fibre content and a decrease in thorniness (Colunga-

GarcfaMarin & May-Pat 1997).

Since pre-Hispanic times, wild populations of A.

vivipara have been prized by the inhabitants of central

America not only for their fibres — which are used for

hammocks, bags and fabrics — but also, the peduncles,

leaves, stems and roots are used for building material,

utensils, tools, food, fermented beverages and medicine

(Cruz-Ramos et al. 1985; Colunga-GarcfaMarin & May-

Pat 1993; Nobel 1994). With increasing attention being

paid to the utilisation of invasive aliens (Zimmermann &

Zimmermann 1987; Anon. 1988; Howell & Schnell

1991; Turksvykwekersvereniging 1997), southern Afri-

can environmentalists should take cognisance of the

Mexican ethnobotanical uses of A. vivipara with a view

to duplicating some of these practices locally.

The typical variety of A. vivipara is easy to distin-

guish from A. decipiens , the only species with which it

can be confused in southern Africa. Important morpho-

logical distinctions between the two taxa are given below

(Table 1).

TABLE 1. — Main morphological distinctions between Agave

decipiens and A. vivipara

44

Bothalia 30,1 (2000)

FIGURE 1. — Agave vivipara var.

vivipara: A, habit, rosette 1.5

m tall; B, panicle 2 m tall.

C-F, protandrous flower: C,

bud; D, male phase; E,

female phase; F, longitudinal

section. G, distal part of leaf

showing straight margin and

hooked teeth. C-G, x 0.65.

Scale bar: 20 mm. All draw-

ings made from live material

collected by E.M.A. Steyn

and deposited under G.F.

Smith & E.M.A. Steyn 2

(PRE). Artist: G. Condy.

Agave vivipara var. vivipara is widely cultivated in

Australia and is listed as an established alien in the Flora

of Australia (Forster 1996). In South Africa this variety,

as well as the attractive A. vivipara cv. Marginata with its

yellow-edged leaves, is frequently cultivated (Smith &

Steyn 1999b). In urban Pretoria the typical variety has

recently been found as a garden escape in a few loca-

tions, for example on an open stand (Smith & Steyn

1999b) where the plants had possibly spread from garden

refuse dumped earlier, and along a roadside in eastern

Pretoria. In the latter location the colony was freely suck-

ering and spreading onto the broad shoulder of the road.

Some of the suckering plants were in flower and bulbils

were found on an older inflorescence which bore no fruit

(Figures 1 & 2A). However, a dried-out infructescence

with numerous seed-filled capsules on a dead plant in the

same colony showed that the plants are also capable of

sexual reproduction (Figure 2B). During germination

tests conducted by the first author, ± 80% of the large,

black seeds were viable. Similar patterns of distribution

and establishment of the species have been observed near

Cape Town and Port Elizabeth and Uitenhage, Eastern

Cape (Figure 3). A. vivipara var. vivipara has therefore

apparently progressed beyond the adventive stage of nat-

uralisation in South Africa and is currently becoming an

established alien (Kloot 1987) like A. americana L. and

A. sisalana (Smith & Mossmer 1996) in the local flora.

1219000-00014 Agave vivipara L. var. vivipara.

Species plantarum 1: 323 (1753). Type: Commelin, Prae-

ludia botanica 65, t. 15(1 703) [lectotypified by Wijnands

(1983)].

Agave angustifolia Haw. var. angustifolia: 72 (1812); Gentry: 559

(1982); Serna & Lopez-Ferrari: 9 (1993); Colunga-Marin & May-Pat:

1455 (1997). Type: St. Helena Island, cultivated. [Neotype: Tozzeti: 2,

t. 6 (1810), designated by Gentry: 560, t. 20.6 (1982)].

A. pacifica Trel. in Standley: 118 (1 920). Type: Sinaloa, Isla Creston

cerca de Mazatlan, Trelea.ie s.n. (MO).

Bothalia 30,1 (2000)

45

FIGURE 2. — Parts of older inflorescences of Agave vivipara var.

vivipara : A, developing bulbils, x 0.9; B, seeding capsule,

x 0.9. Scale bar; 20 mm. Material collected by E M. A. Steyn

and deposited under G.F. Smith & E.M.A. Steyn 2 (PRE). Artist:

G. Condy.

A. yaquiana Trel. in Standley: 120 (1920). Type: Sonora, from rocky

hillsides between Hermosillo & Ures, Trelease 391 (MO).

A. owenii I.M.Johnst.: 999, 1000 (1924). Type: Sonora, on an islet in

Guaymas Harbor, Johnston 3085 (CAS).

(For a more complete list of synonyms, see Serna & Lopez-Ferrari: 9

(1993).

FIGURE 3. — Distribution of Agave vivipara var. vivipara in southern

Africa.

Surculose, caulescent, succulent perennial with radia-

ting rosettes ± 1.5 m diam.; trunk up to 1 m. Leaves light

green, margins straight, surfaces smooth, hard-fleshy,

very fibrous, rigidly spreading, ascending to descending,

0.75-0.9 m long, ± 90 mm wide at midblade, lanceolate,

flat in transverse section, narrowed and thickened

towards the base; leaf bases broadened, overlapping,

stem clasping; teeth easily detachable, usually prominent

with low, broad bases and slender cusps, 2-3 mm long,

evenly spaced, strongly decurved (apparently releasing

an allergen causing painful local swellings when punc-

turing the skin), commonly reddish brown to dark brown,

interstitial teeth usually absent; terminal spine 20-25

mm long, decurrent, conical, reddish brown to dark

brown, greying with age, pungent. Inflorescence panicu-

late, 3. 5-4.0 m long, bulbiferous after flowering; shaft as

long as or slightly longer than panicle; umbels varying in

number from 24-36; branches horizontally spreading.

Flowers yellowish green, about 55 mm long, short-

stalked, scented. Perianth with tepals unequal, cucullate

and pubescent at tip, margins involute, 21-22 mm long,

outer segments overlapping the inner; inner segments

prominently keeled, wilting before flower reaches

female phase, crimping outward and downward; perianth

tube succulent, cup-shaped, 11 mm diam. at level of fil-

ament insertion, 12 mm deep. Stamens exserted 20 mm

beyond tepals in male phase flower, yellowish with

brownish red speckles; filaments stout, tapered towards

apex, inserted in single series in throat of perianth tube,

45 mm long; anthers 22 mm long before dehiscence,

cylindrical, yellowish with conspicuous brownish red

speckles, versatile. Ovary small, 25 x 10 mm diam.,

cylindrical, terete, indistinctly grooved in distal region;

neck short, slightly constricted; style stout, terete, 72 mm

long, light green with brownish red speckles. Fruit large,

broadly ellipsoid, 30-35 x 20-25 mm, short-beaked,

freely seeding. Seeds large, 7-12 x 6-8 mm, dull black,

flattened, D-shaped with complete marginal wing, hilar

notch shallow. Chromosome number unknown. Flower-

ing time : February and March.

GAUTENG. — 2528 (Pretoria): Meyers Park area; along Simon

Vermooten Road, 0. 9 km north of entrance to German School on west-

ern shoulder of road, (-CB), G.F. Smith & E.M.A. Steyn 2 (PRE).

Illustrations: Gentry: t. 20.6-20.9 (1982); Smith & Steyn: t. 1

(1999b); Cruz-Ramos et al.: t.l, 2 (1985).

Common names: kleingaringboom (Afrikaans). Wild

populations in the Yucatan (Mexico) are referred to as

chelem (Colunga-Garci'aMarfn & May-Pat 1997).

REFERENCES

ANONYMOUS 1988. Red dye from an invader cactus weed. Plant

Protection News 13: 1,2.

BAYON, R 1997-1998. Face to face on invasive species: the end of

the biodiversity economy? World Conservation 4/1997-1/1998:

45—48.

COLUNGA-GARCIAMARIN, P. & MAY-PAT, F. 1993. Agave studies

in Yucatan, Mexico. I. Past and present germplasm diversity

and uses. Economic Botany 47: 312-327.

COLUNGA-GARCIAMARIN, P. & MAY-PAT, F. 1997. Morphologi-

cal variation of henequen ( Agave fourcroydes, Agavaceae)

germplasm and its wild ancestor (A. angustifolia) under uni-

form growth conditions: diversity and domestication. American

Journal of Botany 84: 1449-1465.

46

Bothalia 30,1 (2000)

COLUNGA-GARCIAMARIN, P„ COELLO-COELLO, J., EGUIAR-

TE, L.E. & PINERO, D. 1999. Isozymatic variation and phylo-

genetic relationships between henequen (Agave fourcroydes)

and its wild ancestor A. angustifolia (Agavaceae). American

Journal af Botany 86: 115-123.

CRUZ-RAMOS, C.A., ORELLANA, R. & ROBERT, M.L. 1985. Agave

research progress in Yucatan. Desert Plants 7: 71-92.

FORSTER, PL 1992. New varietal combinations in Agave vivipara

(Agavaceae). Brittonia 44: 74, 75.

FORSTER, P.I. 1996. Naturalized succulents in the Australian flora.

Haseltonia 4: 57-65.

GENTRY, H.S. 1982. Agaves of continental North America. University

of Arizona Press, Tucson.

HAWORTH, A H. 1812. Synopsis plantarum succulentarum , edn 1.

Richard Taylor, London.

HOWELL, R. & SCHNELL, A. 1991. Beproefde turksvy resepte. Publ-

isher and place of publication unknown.

JOHNSTON, I.M. 1924. Expedition of the California Academy of

Sciences to the Gulf of California in 1921. Proceedings of the

Californian Academy of Sciences, ser. 4, 12: 999, 1000.

KLOOT, PM 1987. The naturalised flora of South Australia. I. The

documentation of its development. Journal of the Adelaide

Botanical Garden 10: 81-90.

LINNAEUS, C. 1753. Species plantarum , edn 1. Salvius, Stockholm.

NOBEL, PS. 1994. Remarkable agaves and cacti. Oxford University

Press, New York.

SERNA, A.E. & LOPEZ-FERRARI, A. R. 1993. Las monocotileddneas

Mexicanas una sinopsis floristica, part 1. Consejo Nacional de

la Flora de Mexico, Mexico, D.F.

SMITH, G.F. & MOSSMER, M. 1996. FSA contributions 4: Agava-

ceae. Bothalia 26: 31-35.

SMITH, G.F. & STEYN, E.M.A. 1999a. A first record of Agave decip-

iens naturalised in southern Africa. South African Journal of

Botany 65: 249-252.

SMITH, G.F. & STEYN, E.M.A. 1999b. Agave vivipara: the correct

name for A. angustifolia (Agavaceae). Bothalia 29: 100.

SMITH, G.F., VAN JAARSVELD, E.J., ARNOLD, T.H., STEFFENS,

F.E., DIXON, R.D. & RETIEF, J.A. (eds). 1997. List of south-

ern African succulent plants. Umdaus Press, Pretoria.

STANDLEY, PC. 1920. Trees and shrubs of Mexico. Contributions

from the United States National Herbarium 23. Government

Printing Office, Washington.

TOZZETI, T. 1810. Annuli del museo imperiale di ftsica e storia natu-

rale di Firenze: 2, t. 6.

TURKSVYKWEKERSVERENIGING 1997. Riglyne vir die verbou-

ing van doringlose turksvye vir vrugteproduksie, edn 2. Groep

7 Trust Uitgewers, Sinoville.

WIJNANDS, D O. 1983. The botany of the Commelins. Balkema, Rot-

terdam.

ZIMMERMANN, H.G. & ZIMMERMANN, H. 1987. Aanwending van

jong turksvyblare vir menslike voeding. Boerdery in Suid-

Afrika: Onkruidreeks B. 1.2/1987.

E.M.A. STEYN* and G.F. SMITH*

* National Botanical Institute, Private Bag X101, 0001 Pretoria.

MS. received: 1999-03-31.

HYACINTHACEAE

CORRECTION OF A HISTORICAL ERROR IN THE TAXONOMIC DESCRIPTION OF URGINEA CILIATA

INTRODUCTION

The current interpretation of Urginea ciliata (L.f.)

Baker is complicated by an unexplained error in Baker’s

(1875) description of the species. Despite having seen

the Thunberg type specimen ( UPS-THUNB8281 ), Baker

clearly states that the distinct cilia on the leaf margins are

black. However, on examination of the type specimen,

we found that the cilia are hyaline and not black. This

was confirmed by herbarium specimens and living plants

in the field and in cultivation. The type description is fur-

ther inadequate due to the absence of a bulb on the speci-

men and the incorrect assumption that the two leaves

preserved on the specimen represent the entire plant.

Subsequent treatments (Schonland 1919; Fourcade 1941;

Jessop 1977) place the species in nomina dubia and

result in confusion in herbaria with Urginea marginata

(Thunb.) Baker.

MATERIALS AND METHODS

The present taxonomic correction is based on a study

of herbarium specimens of Urginea ciliata and related

taxa from relevant herbaria (BOL, GRA, NBG, PRE), of

living plants in the field and in cultivation and of archival

records housed in the Museum of Evolution, Uppsala

University. Representative specimens were compared

with a JEOL-JSM 840 scanning electron microscope.

Particular attention was paid to the diagnostic cilia found

on the leaf margins.

Urginea ciliata (L.f.) Baker in Journal of the

Linnean Society 13: 218 (1875); Baker: 464 (1897);

Schonland: 41 (1919); Fourc.: 103 (1941); Jessop: 315

(1977). Type: Caput Bonae Spei, Thunberg s.n. ( UPS -

THUNB8281 , holo.!).

Ornithogalum ciliatum L.f.: 199 (1782); Thunb.: 62 (1794); Willd.:

117 (1799); Roem. & Schult.: 528 (1817-1819); Thunb.: 316 (1823);

Kunth: 359 (1843).

Plant (100-)250(-300) mm tall. Bulb globose, slight-

ly flattened at top, hypogeal, 20-25 mm diam., white

turning greenish at apex, scales thick, fleshy, tightly

packed; outer tunic transparent on bulb, becoming loose

at apices, pale brown, dry, leathery, not neck-forming.

Leaves 10, the lower six in two whorls of three, ovate,

(20— )25(— 30) x (15— )17(— 19) mm, the upper four in two

opposite pairs, three being ovate elliptic, ( 1 5— )20(— 25) x

(7—) 10(— 1 5) mm and the uppermost lorate (8—) 1 5(— 20) x

(3-)4(-5) mm, dark glossy green, firm, leathery, surface

smooth, margins densely ciliate, hysteranthous, prostrate

(Figure 4). Desiccated leaves persistent during flowering

and dormancy period, brownish, transparent, leathery,

marginal cilia obvious (Figure 5). Inflorescence erect,

200-260 mm tall, racemose; peduncle and inflorescence

axis slender, 1.0-1. 5 mm thick, thickest at base, maroon,

glossy; fertile part (25— )30— 35(— 60) x (15-)20 mm with

10-15 flowers; sterile tuft minute, 1 mm long, held at an

angle away from the peduncle axis; bracts deltoid,

cupped, acute, held at 45° below pedicel, green tinged

purplish, 1 x 0.5 mm, basal spur dark purple, 0.5 mm

long, apex acute, facing downwards (Figure 6A).

Flowers scentless; pedicels held at 90° to peduncle in

bud, 45° at an thesis and 20° after an thesis, (4— )6 x 0.4

mm, 4-5 mm apart, reddish brown; buds pendulous;

perianth patent at anthesis becoming erect after pollina-

tion, 12 mm diam., translucent white with light brown

keels becoming pale towards apices; lobes spreading,

Bothalia 30,1 (2000)

47

FIGURE 4. — Leaf arrangement of Urginea ciliata, Dold 2383. Scale

bar: 10 mm.

equal in length, recurved on either side of keel, fused at

base by 0.4 mm, apices with minute stalked papillae,

inner lobes broadly ovate with obtuse apices, 5.6 x 2.6

mm, outer lobes ovate with acute apices, 5.6 x 1.9 mm

(Figure 6B). Stamens spreading; filaments white, 3 mm

long; anthers bilobed, 1 x 0.2 mm, green, basifixed.

Ovary obovoid, 2 x 1.5 mm, tri-locular, yellow-green;

style tubular, longitudinally grooved into three sections,

2 mm long, receptive surface glossy, not swollen.

Capsule ovoid, erect, 3-7 x 2. 5-5.0 mm, papery, pale

yellow green turning brown (Figure 6C). Seed irregular-

ly pyramidal, 2. 0-2. 5 x 2 mm at base, black, shiny,

winged; wings thin, papery; testa reticulate.

Urginea ciliata and U. marginata specimens are often

confused in herbaria as a result of Baker’s incorrect and

inadequate type description. Complicating matters fur-

FIGURE 5. — Marginal cilia of Urginea ciliata, Dold 2383. Scale bar:

100 pm.

ther, Baker (1875) describes the leaf margin of U. mar-

ginata as obscurely ciliated, however on examination of

the type specimen ( UPS-THUNB8393 ), herbarium spec-

imens (BOL, GRA, NBG, PRE) and living specimens we

observe that the leaf margin is without cilia at all but is

minutely verrucose. The two species are easily distin-

guished by means of the hyaline marginal cilia present on

short rigid prostrate leaves of the latter and the thick car-

tilaginous margins on spreading to erect leaves of the for-

mer. In addition U. ciliata has a racemose inflorescence

with basifixed anthers while U. marginata has an umbel-

late inflorescence and dorsifixed anthers.

Flowering period is from January to February.

Flowers open in pairs in the late afternoon and close

before dark — they do not open again. The vegetative

growing period is from February to October thereafter

dormant to January. The plant is leafless at anthesis with

basal leaves reappearing at the time of fruit development

or soon afterwards. Thunberg’s type sheet is a mixed

gathering representing both leaves and inflorescence.

Although the bulbs are hypogeal, the well preserved but

FIGURE 6. — Urginea ciliata, Dold 2383. A, inflorescence; B, perianth; C, capsules. Scale bars: A, C, 10 mm; B, 5 mm.

48

Bothalia 30,1 (2000)

desiccated leaves are persistent and the diagnostic cilia

can be seen on specimens collected over the dormant

period, enabling identification.

Distribution and habitat

Urginea ciliata is essentially an Eastern Cape species

with the majority of collections from the Port Elizabeth

and Joubertina Districts. One isolated record from

Riversdale reflects a disjunction in distribution (Figure

7). This may be a result of sporadic collections due to its

cryptic habit and localised habitat. U. ciliata appears to

be uncommon and restricted to low grassy fynbos where

it is often associated with quartzite outcrops of the Table

Mountain Group (Cape Supergroup Peninsula Forma-

tion). Populations of up to 40 plants are densely congest-

ed between rocks in very shallow sand and gravel soils

and are extremely localised. Altitudes from 110 m to 670

m have been recorded.

It has been established that there can be no question as

to the validity of the Thunberg type material, as the

younger Linnaeus had access to the Thunberg herbarium

and he and Thunberg worked in close collaboration on his

material in Uppsala (fide UPS-THUNB archival records).

Furthermore, there is no material of Omithogalum cilia-

tum L.f. in the Linnaeus herbarium that may have result-

ed in a mistaken type designation. It is clearly Baker’s

error, although it is not known why it was caused.

Baker repeated the error of his earlier type description

in Flora capensis (1897) and applied it as a diagnostic

key character. U. ciliata was subsequently interpreted as

having only two leaves with black cilia on the margins.

As a result later collections were erroneously placed in U.

marginata or not placed at all. Schonland (1919) attrib-

uted to the specimen Cruden 355 from Redhouse, the

manuscript name U. crudeni (ined.) and placed the speci-

men in a type cover (GRA herbarium practice). He does,

however, note that the specimen is closely allied to U. cil-

iata, but was probably misled by Baker’s description.

In 1927, Marloth received a specimen from Riversdale,

Muir 4126, with a note from the collector suggesting that it

may be a new species of Urginea closely related to U. cil-

iata, but there is no indication of Marloth’s opinion. Jessop

( 1 977) discussed Cruden 355 and Muir 4126 under Drimia

marginata (Thunb.) Jessop, comb. nov. and remarked that

the specimens differ from that species by virtue of the cili-

ate margins. He was undecided as to the status of the two

specimens. He refered to Omithogalum ciliatum L.f. and

Urginea ciliata under nomina dubia despite having seen

the Thunberg type ( UPS-THUNB8281 ).

Vouchers: Cruden 355 (GRA, PRE); Dold 2383 (GRA, UPS);

Fourcade 2930 (BOL), 3569 (BOL); Muir 4126 (PRE).

Specimens examined

WESTERN CAPE. — 3421 (Riversdale): flats around Riversdale,

(-AB), Muir 4126 (PRE).

EASTERN CAPE. — 3323 (Willowmore): rocky hills north of

Joubertina, (-DD), Fourcade 2930 (BOL); Wagenbooms River at

Joubertina, (-DD), Fourcade 3569 (BOL). 3324 (Steytlerville): between

Kareedouw and Assegai Bosch, (-CD), Fourcade s.n. (BOL). 3325

(Port Elizabeth): flats around Baakens River Valley, (-DC), Cruden 355

(GRA, PRE), Paterson 976. (GRA); Bridgemead, Parsons Vlei, (-DC),

Dold 2383 (GRA, UPS), Olivier 3489b (GRA), Yates 400 (GRA).

Without precise locality: Caput Bonae Spei, Thunberg s.n. (UPS).

ACKNOWLEDGEMENTS

We would like to thank the curators of BOL, PRE and

NBG for loans of their material and Dr M.C. Olivier and

Mrs M. Yates for locating Urginea ciliata populations.

Thanks to Mr E. Kruger for photographs.

REFERENCES

BAKER, J.G. 1875. Revision of the genera and species of Scilleae and

Chlorogaleae. Journal of the Linnean Society 13: 209-292.

BAKER, J.G. 1897. Liliaceae. In W.T. Thistelton-Dyer, Flora capensis

6: 464, 465. Reeve, London.

FOURCADE, H.G. 1941. Checklist of the flowering plants of the divi-

sions of George, Knysna, Humansdorp and Uniondale.

Memoirs of the Botanical Survey of South Africa No. 20: 103.

JESSOP, J.P. 1977. Studies in the bulbous Liliaceae in South Africa: 7.

The taxonomy of Drimia and certain allied genera. Journal of

South African Botany 43: 265-319.

KUNTH, C.S. 1843. Enumeratio plantarum 4: 359. Cotta, Stuttgart.

LINNAEUS, C. filius. 1782 (‘1781’)- Supplementum plantarum.

Orphanotropheus, Braunschweig.

ROEMER, J.J. & SCHULTES, J.A. 1817-1819. Systema vegetabilium

7: 528. Cotta, Stuttgart.

SCHONLAND, S. 1919. Phanerogamic flora of the divisions of

Uitenhage and Port Elizabeth. Memoirs of the Botanical Survey

of South Africa No. 1:41.

THUNBERG, C.P. 1794. Prodromus plantarum capensium. Edman,

Uppsala.

THUNBERG, C.P. 1823. Flora capensis, edn Schultes. Cotta, Stuttgart.

WILLDENOW, C.L. 1799. Species plantarum 2: 117. Nauk, Berlin.

A.P. DOLD* and R. MOBERG**

* Selmar Schonland Herbarium, Department of Botany, Rhodes

University, P.O.Box 94, 6140 Grahamstown, South Africa, e-mail:

botd@rhobot.ru.ac.za

** Museum of Evolution, Botany Section (Fytoteket), Uppsala Univer-

sity, Villavagen 6, SE-752 36, Uppsala, Sweden,

e-mail: Roland.Moberg@fyto.uu.se

MS. received: 1999-05-17.

Bothalia 30,1 (2000)

49

ERICACEAE

TWO NEW SPECIES OF ERICA FROM WESTERN CAPE, SOUTH AFRICA

Erica orthiocola E.G.H.Oliv., sp. nov. ( %Melaste -

mon), inflorescentibus floribus duobus in ramulis brevis-

simis et in synflorescentibus densis aggregatis, corolla

breviter infundibuliformi, sepalis pedicello base adnatis,

antheris muticis projectione apicale, foliis ciliatis pilis

brevibus validis plumosis distinguitur.

TYPE. — Western Cape, 3419 (Caledon): Rivier-

sonderend Mtns, Pilaarkop, ridge WNW of peak, 1 540

m, (-BB), 26-02-1999, E.G.H. & I.M. Oliver 11230

(NBG, holo.; K, NY, PRE).

Compact rounded woody shrub ± 300( — 400) x 300

mm, single-stemmed reseeder. Branches', numerous

main and secondary with continuous apical growth;

internodes short, less than leaf length, no infrafoliar

ridges, puberulous with simple spreading to reflexed

white hairs. Leaves 3-nate, erect to spreading, imbricate,

± 5.5 x 1.5 mm, oblong-elliptic, adaxially flattened,

abaxially rounded, hard and stiff, glabrous, margins with

short stubby plumose hairs, apex shortly cuspidate, sul-

cus narrow and closed at base; petiole 1 mm long,

appressed, adaxially glabrous, abaxially puberulous.

Inflorescence : flowers 2(1) terminal on highly reduced

lateral branchlets, appearing axillary on main branches,

these branchlets with only 2 bract-like prophylls; pedicel

6 mm long, with very short dense substellate hairs, white

to pinkish; bract about V5 up pedicel, 1.5 x 0.5 mm,

oblong, very slightly apiculate, white-pink, semitrans-

parent, glabrous, sparsely ciliolate, sulcus very small to

absent; bracteoles 2, 2/3 up pedicel, otherwise like bract.

Calyx 4-lobed, 2.5 x 2.0 mm, appressed to corolla; lobes

slightly fused at base and to apex of pedicel, ovate, apic-

ulate, the outer two slightly overlapping inner two, stout

and hard, scarious with semitransparent marginal zone,

pink and green, glabrous, margins with short stout

plumose hairs and finer simple hairs, sulcus narrow, ± '/3

length of lobe. Corolla 4-lobed, 4 x 3.5 mm, shortly and

broadly funnel-shaped, pink turning brownish, glabrous;

lobes erect, 1. 5-2.0 x 1.5 mm, ovate, subacute, keeled

with a dark longitudinal stripe, margins occasionally

toothed to slightly erose. Stamens 8, free, manifest; fila-

ments 2 x 0.4 mm, linear-oblong, slightly narrowed at

base, straight, glabrous, whitish; anthers dorsifixed near

base, bilobed, V-shaped, muticous; thecae 1.5 x 0.5 mm.

oblong with narrowed slightly elongated tip, papillate,

medium brown; pore relatively large, 2/3 length of theca;

pollen in tetrads. Ovary 4-locular, ±1x1 mm, broadly

ellipsoid to obovoid, slightly emarginate, glabrous, dull

pink, with no basal nectaries; ovules 6-10 per locule

pendulous from placenta in upper 2/3; style far exserted,

6 mm long, filiform, glabrous, white to reddish at apex;

stigma small narrowly cyathiform, red. Fruit a dehiscent

capsule, 2x2 mm, valves splitting for 2/3 their length to

angle of 30°, apices not incurved, septa equal on valves

and columella. Seeds 0.7 x 0.5 mm, ellipsoid, yellow to

dark brown, testa reticulate, with thick radial cell walls,

internally slightly wavy, externally ± straight. Figure 8.

This new species is characterised by the 2-flowered

inflorescence borne on very short lateral branchlets

arranged in a rather dense synflorescence near the ends

of the branches, the shortly funnel-shaped corolla, sepals

fused to the upper end of the pedicel which is covered

with short substellate hairs, and the muticous anthers

with an apical projection. The leaves are stiff and hard

and edged with short stout plumose hairs.

There are several species similar to E. orthiocola

based on the pedicel being longer than the corolla, the

corolla shape, and anther type with extended portion

above the pore, but it differs in a number of respects, the

main one being the short stout plumose hairs on the mar-

gins of the leaf. E. pillarkopensis , which is sympatric

with E. orthiocola , is a (arge woody shrub to small tree,

with the flowers numerous, bright pink and arranged all

over the branches, the leaves and sepals hairy with sim-

ple short hairs, and the calyx differently shaped and not

fused. E. seriphiifolia and E. cubica both have a simple

many-flowered inflorescence on the ends of the main

branches, the bract axial, long scattered distinctly

plumose hairs on the pedicel, the stem with distinct

infrafoliar ridges and the leaf margins with a few short

simple hairs. They also have a similar fusion of the sep-

als with the upper portion of the pedicel. E. obconica is

similar in the mucronate leaf and obconic stigma, but the

bract and bracteoles are in a higher position on the pedi-

cel and the inflorescence is simple and terminal on the

main branches. E. roseolobci has the similar fusion of

sepals but differs by the included style and simple hairs

on the pedicel.

The flowers of E. orthiocola lack nectaries around the

base of the ovary and have a slightly expanded stigma.

This would indicate that the pollination is probably

effected by wind. This could not be confirmed in the

field due to the slightly late stage of flowering we

encountered when collecting the type material.

The new species is very localised on a single slope in

the Riviersonderend Mountains just west of Pilaarkop

(Figure 9) where it grows on very steep south-facing

rock ledges or small outcrops just below the summit

ridge, hence the name orthiocola ( orthius = high, lofty;

colus = inhabiting). Several visits to the area provided

only a few scattered small groups of plants, the largest

with about five plants, which had escaped the major fire

seven years ago. The plants appear to be very slow-grow-

ing since no young plants were observed. Flowering

time'. February and March.

This locality is peculiar in having several species of

plants confined to the upper south slopes of the ridge.

There are several undescribed Erica species including

the following one, and the unusual and very rare

Lonchostoma esterhuyseniae Strid (Bruniaceae). This is

probably in part due to the habitat conditions which are

cool and moist, with an accumulation of cloudy mist

from the southeast winds caused by the leeward position

of the ridge in relation to the peak of Pilaarkop.

50

Bothalia 30,1 (2000)

FIGURE 8. — Erica orthiocola. A, flowering branch; B, branch; C, leaf, D, inflorescence on very short lateral branchlet with subtending leaf

removed; E, flower; F, corolla; G, corolla opened to show androecium; H, stamen, front, side and back views; I, ovary; J, ovary opened

laterally; K, capsule; L, seed; M, testa cells. All drawn from the type, Oliver & Oliver 11230. A, x 1; B, D, 4 mm; C, E-K, 2 mm; L, 1

mm; M, 100 pm. Artist: Inge Oliver.

E. pillarkopensis appears to have been very abundant

on the southern slopes of the ridges just west and east of

Pilaarkop as evidenced by the numerous silvery grey

skeletons remaining after the extensive fire in 1994. This

is particularly so in the upper portion of the main kloof

of Olifantsbosch. At present there are very few old plants

of the species remaining and these are only in rocky

areas where they escaped the fire. Some are in the region

of 3 m high with a trunk of 200 mm diam. Surprisingly

there are very few young plants visible on any of the

slopes.

E. seriphiifolia , E. cubica and E. obconica all occur

on the Langeberg range with E. cubica extending as far

away as the South Coast of KwaZulu-Natal. E. roseolo-

ba is restricted to a single peak in the Klein Swartberg

near Seweweekspoort and was only recently described

(Oliver & Oliver 1996).

Paratype material

WESTERN CAPE. — 3419 (Caledon); Riviersonderend Mtns,

Pilaarkop, 5400 ft, (-BB), 8-03-1970, Esterhuysen 32416 (BOL); ibid.,

4500-4800 ft, 7-03-1971, Esterhuysen 32579 (BOL, K); ibid., 9-04-

197 1, Esterhuysen BOL55092 (BOL); ibid., 1 520m, 9-10-1998, Oliver

11177 (NBG); ibid., 04-1940, Stokoe 7877 (BOL); ibid., 06-1949,

Stokoe SAM62327 (BOL, SAM).

Erica columnaris E.G.H.Oliv., sp. nov. (§ Hermes /

Chlorocodon), synflorescentibus densis columnaribus,

foliis floribus superantibus, bracteolis plerumque plene

recaulescentibus, calcaribus antherarum brevibus ex

parte decurrentibus, nectariis nullis distinguitur.

TYPE. — Western Cape, 3419 (Caledon): Rivier-

sonderend Mtns, Pilaarkop, moist southern slopes below

ridge WNW of peak, 1 500 m, (-BB), October 1998,

Bothalia 30,1 (2000)

51

FIGURE 9. — Known distribution of both Erica columnaris and E.

orthiocola.

E.G.H. & I.M. Oliver 11177 (NBG, holo.; BM, BOL, E,

K, MO, NY, P, PRE, S, W).

Stoutish, erect, virgate shrub 0.5(— 1 ) m high, single-

stemmed reseeder. Branches : main branches of previous

season often devoid of leaves and terminating in a com-

pact innovation zone of 7-15 fastigiate leafy secondary

branches, these terminating in synflorescences 15-40

mm long, branches glabrous with infrafoliar ridges,

tertiary branchlets very reduced, 0.5 mm long terminat-

ing in a florescence. Leaves 4-6-nate, subspreading-

incurved, imbricate, 7-8 x 1 mm, lanceolate, flat adaxi-

ally and slightly rounded abaxially, glabrous, ciliolate,

sulcus narrow and closed at base; those on tertiary

reduced branchlets in 2 or 3 whorls, very small, bract-

like and whitish; petiole 0.8 mm long, appressed,

glabrous, ciliolate. Inflorescence'. 1(2) flowers subtermi-

nally on very reduced lateral branchlets appearing axil-

lary to the longer leaves on secondary branches, these

arranged in column-like synflorescences 15-40 mm

long at ends of secondary branches; pedicel 2 mm long,

glabrous; bract partially recaulescent xf up the pedicel,

1 x 0.7 mm, ovate, acute, esulcate, glabrous, ciliolate,

light green; bracteoles 2, fully recaulescent as lateral

lobes of calyx, occasionally 2 partially recaulescent and

appressed to calyx, rarely in middle position, when par-

tially recaulescent then like the bract in shape. Calyx 4-

partite; lobes appressed to corolla, occasionally lateral

two imbricating at base, 1.5 x 1 mm, ovate-lanceolate,

glabrous, ciliolate, green often tinged red, sulcus nar-

row, ‘/ 3 length of sepal. Corolla 4-lobed, 2.5 x 2. 5-3.0

mm, urceolate, glabrous, wine-red with whitish base;

lobes recurved, 0.8 x 1 mm, broadly triangular, margins

entire. Stamens 8, free, included; filaments 1 mm long,

broadly oblong, flat, straight in upper half, glabrous,

white; anthers bilobed, ovate in front view, awned; the-

cae erect, appressed, glabrous, dark brown; awns decur-

rent on filament, ± 0.1 mm long, linear, simple, pendu-

lous, white; pore '/2 length of theca; pollen in tetrads.

Ovary 4-locular, 8-lobed, 0.5 x 0.8 mm, oblate obovoid,

emarginate, glabrous, reddish green, nectaries absent;

ovules 4 per locule, spreading from placenta in central

position on columella; style exserted, 2.5 mm long, nar-

rowly cylindric, sometimes slightly curved apically,

glabrous, greenish white base and dark red apex; stigma

peltate-capitate, sometimes reflexed, dark red. Fruit a

dehiscent capsule, 1.2 x 2 mm, valves splitting to ± 45°

and nearly to the base, septa ± 30% and very thin on col-

umella and 70% on valve, placenta very convoluted,

columella easily shed from capsule. Seeds 0.5 x 0.4 mm,

ellipsoid, circular in cross section, yellow brown, testa

reticulate, cells 60 x 90 |im, radial walls narrowly wavy.

Figure 10.

This new species is distinguished by the dense com-

pound synflorescence, the leaves of which are longer

than the lateral flowers, the bracteoles often fully

recaulescent as the lateral segments of the calyx, the

short anther spurs which are partially decurrent along the

apex of the filament, and the absence of nectaries. The

pseudospicate, columnar synflorescences give the species

its name, columnaris = pillar- or column-like.

Several species show alliances with E. columnaris

based on the long erect tightly packed synflorescences

with small coralline flowers and stems with infrafoliar

ridges. E. regerminans with a very similar habit and

method of branching, differs in its slightly larger, more

conspicuous dark pink flowers that are longer than the

leaves, its corolla being ellipsoid with a longer pedicel

and in the long anther spurs which are not decurrent. E.

dodii differs in having a soft delicate growth form, corol-

la more open cyathiform to campanulate, anthers small

and delicate with spurs much longer but slightly decur-

rent, style included, flowers pink not dark purplish pink

and larger and more visible. E. coarctata Wendl. differs

in the flowers being slightly smaller and more cyathi-

form and greenish cream-coloured sometimes tinged

pink, the anthers lacking spurs, the stigma peltate-cyathi-

form and far exserted, and the ovary not emarginate.

E. columnaris exhibits clearly the transition from par-

tially to fully recaulescent bracteoles in the Ericeae

which condition was noted in the problems with the

delimitation of the genera Philippia Klotzsch (Oliver

1988) and Ericinella Klotzsch (Oliver 1994). In ± 70%

of the flowers, it would appear that the bracteoles have

been lost, whereas they have recaulesced fully with the

pedicel and become incorporated in the calyx as the

apparent lateral sepals. This will be addressed in a forth-

coming publication dealing with the relationship of the

rest of the minor genera of the Ericeae to Erica.

The lack of nectaries and the expanded stigma com-

plex would indicate that the pollination syndrome for the

species is anemophily. On walking through a large dense

population, occasional puffs of pollen were noted com-

ing from disturbed plants. However, there were numbers

of small beetles visiting the inflorescences of some

plants. This could suggest a case of evolutionary change

from entomophily to anemophily taking place in a

species. We postulate that this is the situation in a num-

ber of other species of Erica. The occurrence of fully

recaulescent bract and bracteoles is always accompanied

by loss of nectaries, expansion of the stigma and anemo-

phily in the Ericeae.

52

Bothalia 30,1 (2000)

FIGURE 10. — Erica columnaris. A, flowering branch; B, flowering branch, close-up; C, stem with leaves removed; D, leaf; E, flower; F, bract;

G, sepal; H, stamen, front, side and back views; I, gynoecium; J, ovary opened laterally; K, capsule; L, seed; M, testa cells. A-J, drawn

from type, Oliver & Oliver 11117; K-M, drawn from Oliver 10940. A, x 1; B, 4 mm; C, D, 2 mm; E-G, 2 mm, H-K, 1 mm; L, 0.5 mm.

Artist: Inge Oliver.

The new species is confined to the steep southern

slopes of the ridge, just west of the main peak of Pilaar-

kop in the Riviersonderend range (Figure 9), which is the

only known locality. It forms dense stands of plants in

some areas, mostly on the loamy open slopes away from

the rocks inhabited by E. orthiocola. Flowering time :

September and October.

E. regerminans is confined to moist south-facing

slopes in the Langeberg range, whereas E. coarctata is

widespread from the Cape Peninsula to the George area

where it grows on dry lower slopes of sometimes coastal

flats associated with calcareous deposits. The latter spe-

cies has been noted by us to be a resprouter in several

well-separated localities.

E. dodii occurs from the Cape Peninsula to the Rivier-

sonderend Mountains where it grows sympatrically with

E. columnaris, sometimes less than a metre apart, but

then up against or under overhangs of large rocks and

rock faces with a soft delicate habit no more than 200

mm high.

Paratype material

WESTERN CAPE. — 3419 (Caledon): Riviersonderend Mtns near

Lindeshof, Pilaarkop, 4000 ft, (-BB), 17-11-1965, fruiting, Ester-

huysen 31400 (BOL, K); ibid., 4500 ft, 24-10-1971, Esterhuysen 32718

(BOL, NBG, PRE); ibid., 4750 ft, 28-10-1997, fruiting, Oliver 10940

(K, NBG, NY, PRE).

Bothalia 30,1 (2000)

53

REFERENCES

OLIVER, E.G.H. 1988. Studies in the Ericoideae (Ericaceae). VI. The

generic relationship between Erica and Philippia in southern

Africa. Bothalia 18: 1-10.

OLIVER, E.G.H. 1994. Studies in the Ericoideae (Ericaceae). XV. The

generic relationship between Erica and Ericinella. Bothalia 24:

121-126.

OLIVER, E.G.H & OLIVER, I.M 1996. Studies in the Ericaceae

(Ericoideae) XIX. Two new species of Erica from southern

Africa. Feddes Repertorium 107: 305-310.

E.G.H. OLIVER* and I.M. OLIVER*

* Compton Herbarium, National Botanical Institute, Private Bag X7,

7735 Claremont, Cape Town.

MS. received: 1999-05-17.

ASPHODELACEAE: ALOOIDEAE

THE GENUS POELLNITZIA INCLUDED IN ASTROLOBA

Considered by various authors to be a species of Aloe

L., Astroloba Uitewaal or Haworthia Duv., the monotyp-

ic Poellnitzia rubriflora (L. Bolus) Uitewaal has had a

particularly chequered taxonomic history since its origi-

nal description in the genus Apicra Willd. (Smith 1994).

Vegetatively the species shows close affinities to some

species of Astroloba and accords completely with the

genus in its tubular, actinomorphic flowers with includ-

ed stamens. A close relationship between the two genera

is also supported by the results of a preliminary survey of

lipophilic anthranoid aglycones in the roots of subfamily

Alooideae (B-E. van Wyk pers. comm.). The genus

Poellnitzia was distinguished from Astroloba on the

basis of the red flowers with connivent, reduplicate-val-

vate tepals (Smith 1995; Smith et al. 1995). Careful

examination of the flowers, however, shows that the aes-

tivation is in fact imbricate and the species thus differs

from Astroloba only in the more or less horizontal

racemes bearing secund, orange-red flowers with con-

nivent tepals. These floral adaptations are now recog-

nised as part of the syndrome of sunbird pollination

which is widespread in southern Africa (Goldblatt &

Manning 1999) and the resemblance between the flowers

of Poellnitzia and the bird-pollinated genus Microloma

R.Br. (Apocynaceae: Asclepiadoideae) is particularly

striking. In the wild the species is visited by Lesser dou-

ble-collared sunbirds, which insert just the tip of their

beak into the flowers before extending the tongue into

the tube to extract the nectar. Species of Astroloba are

pollinated by anthophorine bees and have smaller, dull-

coloured flowers. The sugar composition of the nectar of

Poellnitzia also reflects a shift from insect- to bird-polli-

nation and is hexose-dominant whereas the nectar of

Astroloba is sucrose-dominant (Van Wyk et al. 1993).

Adaptations for specialised pollination strategies alone

are insufficient grounds for the recognition of genera and

we believe that Poellnitzia is best treated as a species of

Astroloba adapted to pollination by sunbirds.

Astroloba rubriflora (L.Bolus) G.F.Sm. & J.C. Man-

ning, comb. nov.

Apicra rubriflora L.Bolus, The Annals of the Bolus Herbarium 3: 13,

t. 2D (1923). Type: South Africa, Western Cape, Bonnievale, Smith s.n.

(BOL, holo.!).

REFERENCES

BOLUS, H.M.L. 1923. Novitates africanae. The Annals of the Bolus

Herbarium 3: 13, t.2D.

GOLDBLATT, P. & MANNING, J.C. 1999. Evidence of bird pollina-

tion in Iridaceae of southern Africa. Adansonia 21: 25-40.

SMITH, G.F. 1994. Taxonomic history of Poellnitzia Uitewaal, a unis-

pecific genus of Alooideae (Asphodelaceae). Haseltonia 2: 74—

78.

SMITH, G.F. 1995. FSA contributions 3: Asphodelaceae/Aloaceae,

1028010 Poellnitzia. Bothalia 25: 35, 36.

SMITH, G.F., VAN WYK, B-E. & CONDY, G. 1995. Poellnitzia rubri-

flora. Flowering Plants of Africa 54: 94—98.

VAN WYK, B-E., WHITEHEAD, C.S., GLEN, H.F., HARDY, D.S.,

VAN JAARSVELD, E.J. & SMITH, G.F. 1993. Nectar sugar

composition in the subfamily Alooideae (Asphodelaceae).

Biochemical Systematics and Ecology 21: 249-253.

J.C. MANNING* and G.F. SMITH**

* Compton Herbarium, National Botanical Institute, Private Bag X7,

7735 Claremont, Cape Town.

** National Botanical Institute, Private Bag XI 01, 0001 Pretoria.

MS. received: 1998-12-14.

ALOE DELPHINENSIS IN ALOE SECT. LOMATOPHYLLUM

The genus Lomatophyllum Willd. has been disting-

uished from Aloe L. primarily by the former having

fleshy, indehiscent fruit (berries) with unwinged seeds

and the latter having dehiscent fruit (capsules) with

winged seeds. Smith & Van Wyk (1991) following a

cladistic analysis of the subfamily Alooideae (As-

phodelaceae), concluded that Lomatophyllum could be

easily included in Aloe but that a comprehensive taxon-

omic revision of the species concerned was required.

Evidence from chromosome morphology, leaf surface

anatomy, pollen morphology (Schill 1973) and leaf sap

flavanoid chemistry (Viljoen et al. 1998) all lend support

to this move to reduce Lomatophyllum to the synonymy

of Aloe.

Rowley (1996) has proposed that species included in the

genus Lomatophyllum be transferred to Aloe as members

of Aloe sect. Lomatophyllum G.D. Rowley. Various new

combinations and new names have been proposed to vali-

date this proposal (Rowley 1996; Newton & Rowley

1998). These automatic transfers have done little to resolve

the species relationships in the group. However, Rauh

(1998) has recently provided an identification key to the 18

species he included in his concept of Lomatophyllum.

54

Bothalia 30,1 (2000)

A comprehensive, phylogenetically based, infragener-

ic classification of Aloe is not currently available.

Reynolds (1966, 1982) chose to recognise 10 sections in

the genus Aloe s.s. However, the continued use of his

classification has been questioned in recent times on the

basis of chemical investigations (Viljoen et al. 1998).

Van Wyk & Smith (1996) did not mention these sections

at all and used informal groups to arrange the South

African species of Aloe. The recognition of Lomato-

phyllum at sectional rank in Aloe is probably not equiva-

lent to that used by Reynolds (1966, 1982) in defining

the various sections in the genus. Hence the infrageneric

classification of Aloe remains to be resolved and a dif-

ferent rank such as subgenus may need to be considered

for the taxa of Lomatophyllum, assuming monophylly of

this infrageneric unit.

Nevertheless the characters used to define the section

Lomatophyllum are distinctive and easily ascertained if

fresh fruit are available. There are a number of species of

Madagascan Aloe for which the fruit are unknown (cf.

Rauh 1995, 1998) and hence their sectional classification

is unknown. One of these is Aloe delphinensis Rauh

(1990), based on his collection Rauh 68629a from Pic St.

Louis, near Fort Dauphin in southwest Madagascar.

Rauh (1990) allied the new species to A. bakeri Scott

Elliot but subsequently considered that a further new

species A. lucile-allorgeae Rauh was the closest relative

of A. delphinensis (Rauh 1998).

Fruit set was stimulated in a cultivated specimen of

Aloe delphinensis using pollen from A. dinteri Berger, a

‘true’ Aloe from Namibia (Reynolds 1982). These fruits

were fleshy and indehiscent, and contained wingless

seeds (Figure 11).

Aloe delphinensis can be included in Aloe sect.

Lomatophyllum thus bringing the number of included

species in that section to nineteen. This species may be

distinguished from the other Madagascan taxa in the

group by inserting an additional couplet into the key of

Rauh (1998).

FIGURE 11. — Aloe delphinensis. A-C, fruit: A, side view, x 2; B, api-

cal view, x 2; C, cross section, x 2. D, E, seed, x 10. All from

fresh material of P.I. Forster PIF24630 [voucher at Queensland

Herbarium (BRI)] prepared from a plant purchased by N. Carr

in a Madagascan nursery as this species and in close agreement

with the protologue and accompanying dlustrations of Rauh

(1990, 1995, 1998).

Aloe L. sect. Lomatophyllum G.D. Rowley in

Excelsa 17: 59 (1996).

Aloe delphinensis Rauh in Cactus & Succulent

Journal (US) 62: 230 (1990). Type: Madagascar, on

granitic rocks on the Pic St. Louis, near Fort Dauphin,

100 m, Rauh 68629a (holo., HEID).

Fruiting pedicel 14-15 x ± 0.5 mm, glabrous. Fruit

oblong, triquetrous, indehiscent and fleshy, 18-20 x

9-10 mm diam. Seed ± globose, ± 2 mm diam., with a

minute ring or ridge around ± 2F of the circumference

(Figure 11).

Key to Madagascan species

(based mainly on Rauh 1998, but with nomenclature updated for inclusion within Aloe )

la Plants with bulbils in the inflorescence 2

lb Plants without bulbils in the inflorescence 3

2a Rosettes small, with leaves spreading, green with white bands and spots; bulbils 1 or 2, in the floral region

A. propagulifera (Rauh & Raz.) L.E.Newton & G.D. Rowley

2b Rosettes larger, with leaves erect, uniformly green; bulbils in the axils of bracts up the scape

A. schilliana L.E.Newton & G.D. Rowley

3a Stems decumbent, short, richly branched, forming dense mats A. socialis (H. Perrier) L.E.Newton & G.D. Rowley

3b Stems erect, single or branched from the base, but not forming dense mats 4

4a Plants with thick stems to 100 mm diam. and 2-3 m tall with a large terminal leaf rosette A. peyrierasii Cremers

4b Plants with thinner and shorter stems 5

5a Stems short, to 0.5 m and 10-20 mm in diam 11

5b Stemless rosette plants or with very short stems 6

6a Rosettes small, to 0.5 m in diameter; leaves spreading or erect 7

6b Rosettes larger, stemless or with a short stem; leaves 0. 1-1.0 x 0.1-0.12 m 8

7a Leaves narrow linear, erect, densely dentate along the margins; in cultivation forming dense clumps; inflorescence much longer

than leaves A. belavenokensis (Rauh & Gerold) L.E.Newton & G.D. Rowley

7b Leaves triangular, not densely dentate 9

Bothalia 30,1 (2000)

55

8a Leaves up to 1 x 0.1-0.12 m, acute; inflorescences much exceeding the leaves

A. occidental^ (H. Perrier) L.E.Newton & G.D. Rowley

8b Leaves shorter, with rounded apices; inflorescences shorter than the leaves A. orientalis (H Perrier) L.E.Newton & G.D. Rowley

9a Leaf blades with large deltoid marginal teeth, often chocolate brown; raceme few-flowered, globular

A. prostrata (H. Perrier) L.E.Newton & G.D. Rowley

9b Leaf blades with small marginal teeth, not chocolate brown; raceme elongated 10

10a Flowers lemon yellow with green tips A. citrea (Guillaumin) L.E.Newton & G.D. Rowley

10b Flowers rose pink A. rosea (H. Perrier) L.E.Newton & G.D. Rowley

1 la Inflorescences shorter than the leaves 12

1 lb Inflorescences longer than the leaves 13

12a Stems not creeping at the base; leaves very long and narrow, at least over 0.3 m A. oligophylla Baker

12b Stems very thin and stolon-like at the base, creeping and producing adventitious roots; leaves shorter, 0.2-0. 3 m long

A. antsingyensis (Leandri) L.E.Newton & G.D. Rowley

13a Stems not branching from base, leaf blades green, straight; raceme dense with hanging flowers

A. namorokaensis (Rauh) L.E.Newton & G.D. Rowley

13b Stems branching from base, leaf blades bluish green or brownish green, spirally rolled in upper part or decurved; raceme lax ... . 14

14a Leaf blades bluish green; marginal teeth 10-15 mm apart; floral tube 25-30 mm long; fruit globose, ± 12 mm diam.

A. anivorartonensis (Rauh & Hebding) L.E.Newton & G.D. Rowley

14b Leaf blades brownish green; marginal teeth 2-8 mm apart; floral tube 20-22 mm long; fruit oblong, 9-10 mm diam

A. delphinensis Rauh

ACKNOWLEDGEMENTS

Thanks to N. Carr for plant material and W. Smith,

Queensland Herbarium (BRI) for Figure 11.

REFERENCES

NEWTON, L.E. & ROWLEY, G.D. 1998. New transfers from Lomato-

phyllum to Aloe (Aloaceae). Bradleya 16: 114.

RAUH, W. 1990. Two new aloes from Madagascar. Cactus & Succu-

lent Journal (US) 62: 25-31.

RAUH, W. 1995. Succulent and xerophytic plants of Madagascar. Vol.

1 . Strawberry Press, Mill Valley.

RAUH, W. 1998. Three new species of Lomatophyllum and one new

Aloe from Madagascar. Bradleya 16: 92-100.

REYNOLDS, G.W. 1966. The aloes of tropical Africa and Madagas-

car. The Trustees of the Aloes Book Fund, Mbabane, Swazi-

land.

REYNOLDS, G.W. 1982. The aloes of South Africa, edn 4. Balkema,

Cape Town.

ROWLEY, G.D. 1996. The berried aloes: Aloe Section Lomatophyllum.

Excelsa 17: 59-62.

SCHILL, R 1973. Studien zur systematischen Stellung der Gattung

Lomatophyllum Willd., ein Vergleich mit Aloe Tourn. (Lilia-

ceae). Beitrdge zur Biologie der Pflanzen 49: 273-289.

SMITH, G.F. & VAN WYK, B-E. 1991. Generic relationships in the

Alooideae (Asphodelaceae). Taxon 40: 557-581

VAN WYK, B-E. & SMITH, G.F. 1996. Guide to the aloes of South

Africa. Briza Publications, Pretoria.

VILJOEN, A M., VAN WYK, B-E. & VAN HEERDEN, F.R. 1998.

Distribution and chemotaxonomic significance of flavanoids in

Aloe (Asphodelaceae). Plant Systematics & Evolution 211:

31-42.

PI. FORSTER*

* Queensland Herbarium, Brisbane Botanic Gardens, Mt Coot-tha

Road, Toowong, Queensland 4066, Australia.

MS. received: 1999-06-21.

Bothalia 30,1: 57-68 (2000)

Morphology and anatomy of the rhizome and frond in the African

species of Polystichum (Pteropsida: Dryopteridaceae)

J.P. ROUX* and A.E. VAN WYK**

Keywords: anatomy, lamina, morphology, Polystichum , Pteridophyta, rhizome

ABSTRACT

The generic circumscription of the polystichoid ferns within the Dryopteridaceae and their relationships has been and

still is much debated. Although floristic accounts for Polystichum are available for many parts of the world, its morpholo-

gy and evolutionary trends within the genus are still poorly understood. In this study, primarily based on the Polystichum

species from Africa and neighbouring islands, the morphology and anatomy of the rhizome and frond are addressed. Several

species from other genera within the Dryopteridaceae are used for comparison, to gain a better understanding of generic

affinities and evolution in this family.

INTRODUCTION

Polystichum Roth is a cosmopolitan genus comprising

180 to 200 species. Although floristic accounts of the

genus exist for many parts of the world, a comprehensive

modern taxonomic account for the genus as a whole has

never been undertaken. The circumscription and definition

of species within the genus are still weak. Reasons for the

group’s inadequate taxonomic status include the occur-

rence of common and widespread allopolyploid taxa

(D.H. Wagner 1979), sterile FI hybrids, apomictic taxa,

substantial phenotypic variation within populations and in

most cases, a lack of critical morphological studies.

The morphology and anatomy of Polystichum repre-

sentatives from the Indian subcontinent was studied by

Chandra & Nayar (1982) and those of western North

America by D.H. Wagner (1979). Polystichum, Arach-

niodes Blume, Cyrtomium C.Presl, Dryopteris Adans.

and Phanerophlebia C.Presl form a closely related group

of genera within the tribe Dryopterideae and are here

referred to as the polystichoid ferns. Cyrtomium and/or

Phanerophlebia are often included in Polystichum

(Tryon & Tryon 1982; Kramer 1990). Yatskievych

(1996), however, treated Cyrtomium and Phanerophle-

bia as distinct genera and showed that both have a clos-

er affinity with Polystichum than with each other, a

hypotheses first proposed by Christensen (1930). The

genera Arachniodes, Dryopteris and Polystichum are not

always clearly separable, as each include species that

display characters generally associated with the other

genera. The proposed close affinity of these genera is

also supported by the occurrence of a natural hybrid

between Dryopteris and Polystichum (W.H. Wagner

1985). Widen et al. (1981) showed that phloroglucinols

widespread in Dryopteris also occur in Arachniodes but

are rare in Polystichum. Indusia, when present, are

peltate in Polystichum and reniform in Arachniodes and

* National Botanical Institute, Compton Herbarium, Private Bag X7,

7735 Claremont, Cape Town.

** H.G.W.J. Schweickerdt Herbarium, Department of Botany, Univer-

sity of Pretoria, 0002 Pretoria.

* To whom correspondence should be addressed.

MS. received: 1998-10-20.

Dryopteris. All these genera, however, also contain species

that are exindusiate.

This study is primarily based on Polystichum species

from Africa, Macaronesia, the Madagascan region and

the Marion Island groups, but observations on species

from outside of this region are also considered. For com-

parative purposes several species from other genera

within the Dryopterideae were included in the study. The

aim is to gain a better understanding of generic affinities

within the Dryopterideae and to speculate on possible

evolutionary trends within the group.

MATERIALS AND METHODS

Material used in the anatomical study was obtained

from the wild and from cultivated plants (Table 1). Voucher

specimens are housed at the Compton Herbarium (NBG).

Fresh material was fixed in FAA for at least 24 hours. For

rhizomes, serial sections were cut by hand, stained with

alcian blue for one minute, rinsed in water and tem-

porarily mounted in glycerine. Camera lucida drawings

were prepared at x 7.5, x 15 or x 31.25 magnifications

using a Wild ‘Heerbrugg’ microscope. The stelar struc-

ture of the rhizome was reconstructed using these sec-

tions.

Roots, stipes, rachises, secondary rachises and lami-

nae were sectioned 45 pm thick using a Reichert-Jung

Hn 40 freeze microtome. Sections were stained with

safranin and alcian blue for one minute, rinsed in water

and dehydrated in a series of 50% ETOH, 40%

ETOH:60% toluene, 20% ETOH:80% toluene and 100%

toluene, after which they were permanently mounted in

Entellan (Art. 7961, E. Merck, Darmstadt).

For describing the vasculature of the stipe and lamina

fresh material was used; freehand serial sections were

made at 2.5 or 5 mm intervals along the entire length of

the axis, noting where bundles divide or fuse and where

pinnae and pinnules originate.

Epidermal features were studied using pinna or pin-

nule fragments removed from selected specimens, and

58

Bothalia 30,1 (2000)

TABLE 1. — Material examined in the study of the rhizome and frond stelar structure

Taxon

Voucher

Origin

Rhizome

pungens (Kaulf.) C.Presl

setiferum (Forssk.) T.Moore ex Woyn.

transkeiense W.B.G. Jacobsen

transvaalen.se N.C. Anthony

NORTHERN PROVINCE. — 2329 (Pietersburg): Louis Trichardt,

Hanglip Forest Station, (-BB).

KWAZULU-NATAL. — 2828 (Bethlehem): Royal Natal National Park,

Tugela Gorge, (-DB).

EASTERN CAPE. — 3227 (Stutterheim): Amabele, (-DA).

EASTERN CAPE. — 3226 (Fort Beaufort): Hogsback Forest Station,

(-DB).

WESTERN CAPE. — 3322 (Oudtshoom): Knysna, Touws River, (-DC).

MPUMALANGA. — 2530 (Lydenburg): Sabie, (-BB).

KWAZULU-NATAL. — 2828 (Bethlehem): Royal Natal National Park,

(-DB).

KWAZULU-NATAL — 2731 (Louwsburg): Ngome Forest Reserve.

KWAZULU-NATAL — 2828 (Bethlehem): Royal Natal National Park,

(-DB).

FREE STATE. — 2829 (Harrismith): Harrismith, Platberg, (-AC).

EASTERN CAPE.— 3325 (Port Elizabeth): Settler’s Park, (-DC).

WESTERN CAPE. — 3322 (Oudtshoom): Swartberg Pass, (-AC).

GAUTENG. — 2627 (Potchefstroom): Roodepoort, Strubens Valley,

(-BB).

LESOTHO. — 2928 (Marakabeis): Semonkong, LeBihan, (-CC).

EASTERN CAPE. — 3129 (Port St Johns): Port St Johns, (-DA).

KWAZULU-NATAL.— 2831 (Nkandla): Nkandla Forest, (-CA).

EASTERN CAPE. — 3227 (Stutterheim): Isidinge Forest, (-CA).

EASTERN CAPE. — 3227 (Stutterheim): Evelyn Valley Forest, (-CA).

FREE STATE. — 2829 (Harrismith): Farm Metz, (-CA).

KWAZULU-NATAL.— 2828 (Bethlehem): Royal Natal National Park,

(-DB).

EASTERN CAPE. — 3227 (Stutterheim): Amabele, (-DA).

EASTERN CAPE. — 3227 (Stutterheim): Pirie Forest, (-CB).

MPUMALANGA.— 2430 (Pilgrim’s Rest): Graskop, (-DD).

KWAZULU-NATAL. — 2829 (Harrismith): forest south of Seheletwane,

(-AC).

WESTERN CAPE. — 3318 (Cape Town): Table Mountain, (-CD).

Madeira.

EASTERN CAPE.— 3029 (Kokstad): Weza Forest, (-DA).

KWAZULU-NATAL. — 2829 (Harrismith): forest south of Seheletwane,

(-AC).

Roux 2370

Sine coll. s.n.

Roux 2498

Roux 2516

cleared using household bleach. Once cleared the mate-

rial was repeatedly rinsed in clean water and semi-

permanently mounted in either glycerine or glycerine

jelly. Cover slips were sealed with Entellan.

RESULTS AND DISCUSSION

MORPHOLOGY

Rhizome

Detailed studies of the rhizome of Polystichum are

scant and its morphology remains poorly known. The rhi-

zome anatomy of Cyrtomium falcatum was studied by

Gibson et al. (1984) and found to be basically identical to

that of Polystichum. Rhizomes are rarely represented in

herbarium collections, hence in modern floristic accounts

they are often not mentioned. In many species the rhi-

zome is bulky and this may explain why most collectors

are deterred from collecting them. The Polystichum

species studied by Chandra & Nayar (1982) in India, all

have short, stout, and erect rhizomes that are mostly

unbranched. In African species two distinct rhizome

types, namely, erect and decumbent occur. The erect rhi-

zome type is short and generally unbranched with a large

number of fronds crowded in the apical region (Figure

1A). Up to seven leaf gaps overlap at any given time. This

rhizome type is sheathed by a mass of wiry roots, with

FIGURE 1 — Rhizome and lamina morphology. A-D, Polystichum-. A, erect rhizome type in P. transvaalense, Roux 2585. B, C, P. incongruum :

B, decumbent rhizome type, Roux 2377 \ C, acroscopic proximal pinnules, basiscopic pinnule shows free venarion, Roux 2385. D, P.

macleae. Roux 2242, acroscopic pinna, section shows free venation. E, L, M, Arachniodes webbiana subsp . foliosa, Roux 1001 : E, basi-

scopic proximal pinnule; L, rachis section showing confluent nature of axes sulci; M, pinnule segment showing venation. F, Dryopteris

inaequalis , Taylor 4264, basiscopic proximal pinnule and acroscopic pinnule with free venation; G, Cyrtomium caryotideum var.

micropterum, Roux 1913, section of pinna outline with reticulate venation; H, Phanerophlebia auriculata, pinna outline with section show-

ing venation, after Yatskievych (1996). I, J, P. transkeiense, Roux 2493: I, proximal part of rachis showing non-confluent nature of axes

sulci; J, distal part of rachis showing confluent nature of axes sulci; K, C. falcatum ex hort., non-confluent nature of sulci. All vouchers

housed in NBG. Scale bars: A, B, 2 mm; C, I-M, 5 mm; D-F, 10 mm; G, H, 20 mm.

Bothalia 30,1 (2000)

59

60

Bothalia 30,1 (2000)

helically arranged persistent stipe bases several layers

deep and, at least in the younger parts, also with paleae,

giving greater bulk to the rhizome; the highly dissected

stellar structure suggests that it is derived. The decumbent

rhizome type is usually long and branched, terete to

slightly laterally or dorsoventrally flattened; up to five

leaf gaps overlap at a time; the fronds are generally more

widely spaced, often exposing the rhizome intemodes

(Figure IB). In both rhizome types paleae form a dense

covering especially over the apical region.

Branching of the rhizome takes place through the for-

mation of lateral buds at regular or irregular intervals

along the main stem. Branching in Polystichum

transkeiense W.B.G.Jacobsen is regular. In certain habi-

tats, P. incongruum J.P.Roux, P. monticola N.C. Anthony

& Schelpe and P. pungens (Kaulf.) C.Presl also show

regular branching. In P. dracomontanum Schelpe &

N.C. Anthony and P. marionense Alston & Schelpe, lat-

eral buds initiate the formation of long stoloniferous out-

growths. As a result, large clonal stands are formed by

these species. Fronds are initially widely spaced along

these stolons but marked primary thickening takes place

distally, and the fronds become more closely spaced.

Roots are irregularly formed over the entire length of the

stolons. Since the thinner stoloniferous branches can rot

away more readily, individual rhizomes may become iso-

lated thus serving as a means of clonal propagation.

Branching and stolon formation is here viewed as an

adaptation for plants occurring in environments not

always conducive to sexual reproduction. Species with

this type of rhizome often occur in more exposed, often

dryish habitats that may also be subjected to periodic

fires. It is therefore also possible that the clonal habit

might be an adaptation to periodic burning. The short,

suberect to erect rhizome type is, however, much more

common in the genus. In contrast to the decumbent rhi-

zome type, the erect rhizome of the African species

mostly remains unbranched resulting in these species

occurring as individuals rather than as clonal clusters.

Plants with this rhizome type are also adapted to a wide

range of habitat conditions, but tend to be more site-spe-

cific. The erect rhizome is generally well seated in the

substrate, whereas the decumbent rhizome type mostly

grows epigeally, although the stolons of P. dracomon-

tanum are subterranean. Occasionally the rhizome in P.

transkeiense may also be subterranean.

Species in the genera Arachniodes, Cyrtomium,

Dryopteris and Phanerophlebia have either erect or

decumbent rhizomes. If the cyatheoid ferns are accepted

as an ancestor of the dryopteroid ferns, as suggested by

Bower (1928); Ching (1940); Copeland (1947); Nayar

(1970, 1979) and Holttum (1973), then the short erect rhi-

zome should be viewed as the primitive state and the

decumbent rhizome as derived. However, a dennstaedtioid

origin as proposed by Holttum (1947) and Pichi Sermolli

(1977) would suggest the reverse. Hasebe etal. (1995) and

Pryer et al. (1995) presented strong morphological and

molecular evidence that Dennstaedtiaceae branched off

the tree below the Dryopteridaceae, with Cyatheaceae

considerably below that. The evolutionary transformation

of the rhizome habit thus remains open to debate.

Roots

An abundance of wiry adventitious roots occurs irregu-

larly over the entire surface of the rhizome, often form-

ing a dense fibrous mass. Roots appear to be long-lived

but they are, however, only initiated at or near the grow-

ing apex of the rhizome. Young roots are whitish to pale

brown and mostly simple, whereas older roots are dark

brown to black and frequently monopodially branched.

Golden brown root hairs form a dense covering a short

distance behind the root apex. They do, however, lose

their function on the older parts of the roots and become

abraded. In the erect rhizome type, roots also provide

good anchorage.

Fronds

All the Polystichum species within the study area are

evergreen; the fronds remain alive for at least two annual

cycles. The fronds are persistent and are shed through

decay or mechanical damage only. A type of frond re-ori-

entation as described by Nooden & W.H. Wagner (1997),

occurs in P. wilsonii H. Christ. In this high altitude

species, which frequently grows in exposed sites, a large

percentage of the fronds may collapse during the cold

winter months. The plants are, however, never entirely

deciduous.' Frond length varies considerably among

species and in P. incongruum it can attain a length of 1.8

m. There is a strong correlation between the rhizome type

and the number of fronds borne by a plant. Species with

an erect rhizome always bear more fronds than those with

a decumbent rhizome. In P. transvaalense N.C. Anthony,

up to 27 fronds are borne by each plant, whereas in

species with a decumbent rhizome, the number of fronds

rarely exceeds seven. Fronds in the Polystichum species

studied are monomorphic, with no evident differentiation

between the sterile and fertile fronds. The fronds are dif-

ferentiated into a distinct stipe and lamina. Vernation of

the fronds is initially circinate, but as the stipe and lami-

na elongate, the juvenile fronds become hook-shaped in

species such as P. dracomontanum, P. luctuosum (Kunze)

T.Moore, P. transvaalense and P. wilsonii. Monomorphic

fronds are most common in the Dryopterideae, but in

Dryopteris, e.g. D. dracomontanum Schelpe & N.C. An-

thony, they are dimorphic.

Stipe

The use of stipe characters in fern taxonomy has been

studied by Lin & De Vol (1977, 1978). They clearly illu-

strate that stipe characters, especially anatomy, are more

useful at the generic and family level than at the species

level. The stipe in all the studied species of Polystichum

is firm. In species with erect rhizomes the stipes grow

directly upwards, whereas in species with a decumbent

rhizome, they are initially strongly curved upwards.

Stipe length and diameter are variable but may be up to

930 mm long in P. incongruum and up to 10 mm diam.

in P. volkensii (Hieron.) C.Chr. In most species the stipe

base is variously rounded adaxially and abaxially, but

distally they all become shallowly to deeply sulcate

adaxially. In most taxa the basal part of the stipe is cas-

taneous (brownish) to ebeneous (black) and often lus-

trous (nitid), becoming paler distally in dried material. In

Bothalia 30,1 (2000)

61

live plants, however, the distal part of the stipe generally

remains green. The stipe bases of the African

Polystichum species appear not to be modified into dis-

tinct trophopods as described by W.H. Wagner &

Johnson (1983). Trophopods have been reported for all

the Phanerophlebia species with the exception of P.

macrosora (Baker) Underw. (Yatskievych 1996). In most

species the dorsolateral aerophore line is conspicuous

throughout the length of the stipe, generally being some-

what paler in colour than the surrounding tissue. In P.

luctuosum the aerophore line is often green, thus con-

trasting strongly with the generally castaneous to ebe-

neous stipe. The stipe is always clothed with paleae, the

density of which shows considerable variation amongst

the species.

Lamina

Most Polystichum species in the study area have 2-

pinnate pinnatifid to 3-pinnate laminae, with 1-pinnate

laminae recorded in P. falcinellum (Sw.) C.Presl, P.

kalambatitrense Tardieu, P. macleae (Baker) Diels and

P. maevaranense Tardieu only. P. macleae has laminae

ranging from 1-pinnate to 2-pinnate. Lamina length

varies considerably between species with the longest

laminae, up to 925 mm, having been recorded in P.

volkensii (Hieron.) C.Chr. Lamina outline also shows

variation between species, but within a species the

degree of variation is fairly restricted.

Pinnae and pinnules are borne subopposite to alter-

nate on the rachis and secondary rachis. Proximally the

pinnae and pinnules are usually short-stalked but distally

they become sessile and eventually adnate. They are

mostly herbaceous in texture, but in P. dracomontanum

and P. marionense, both of which grow in harsh condi-

tions, the pinnules are usually coriaceous. The lamina is

always discolorous with the adaxial surface darker in

colour. Adaxially the pinnules are pale to dark green in

colour, but in P. dracomontanum and P. incongruum at

Hogsback in the Eastern Cape, exposed and older fronds

generally turn bronze.

Polystichum is characterised by a largely acroscopic

pinna development (Figure 1C, D), but basiscopically

developed laminae are often present in P. vestitum

(G.Forst.) C.Presl and P. whiteleggei Watts. The degree

to which the proximal acroscopic pinnules are developed

shows considerable inter- and infraspecific variation.

Interspecific variation is exemplified especially in P.

drepanum (Sw.) C.Presl, P. incongruum and P. macleae.

In Arachniodes and Dryopteris the laminae are most-

ly basiscopically developed (Figure IE, F). In Arach-

niodes and in Cyrtomium and Phanerophlebia, where

most species have 1-pinnate laminae, the pinnae are also

acroscopically developed and often conspicuously auri-

cled (Figure 1G, H, M).

In Polystichum the pinnae are mostly oblong-acumi-

nate to narrowly oblong-acuminate in outline, but in

some species they may be ovate, narrowly triangular, or

in P. marionense often deltoid. In some species the proxi-

mal pinnae are reduced towards the base of the frond,

and often also strongly conduplicate and deflexed. In

most species the proximal pinnae are usually widely

spaced with no or little overlap with the more distal ones.

Towards the lamina apex, however, the pinnae frequent-

ly become imbricate. The number of stalked pinnae per

lamina ranges from 12 to 15 in P. maevaranense, but up

to 45 in P. setiferum (Forssk.) T.Moore ex Woyn. Pinna

length in most species falls within the 100 to 200 mm

range, but in P. marionense the pinnae may reach a maxi-

mum length of only 36 mm, whereas in P. drepanum the

proximal pinnae may be up to 450 mm long.

Pinnules are always inaequilateral, asymmetric and

variously ovate, trullate or rhomboid in outline, with an

acroscopic auricle. The base is mostly truncate acroscopi-

cally whilst basiscopically it is mostly narrowly to broad-

ly cuneate. Margins may be lobate, serrate, doubly ser-

rate or dentate with the teeth being obtuse, pungent or

aristate. Pichi Sermolli (1972), in describing P. kiliman-

jaricum Pic.Serm., emphasized the number and direction

of the pinnule awns, but we found them to be extremely

variable and of no taxonomic value. In P. marionense the

pinnule margins of plants growing in exposed conditions

are often revolute. The number of short-stalked pinnules

per pinna ranges from five in P. marionense to 27 in P.

zambesiacum Schelpe. Pinnules are mostly small, but in

P. drepanum the proximal acroscopic pinnule can be up

to 83 x 15 mm. Indumenta occur on both the adaxial and

abaxial lamina surfaces of most species with the abaxial

surface generally more densely set.

Rachis

Chromatically the rachis does not differ significantly

from the stipe, being green throughout, but in older

fronds it may become dark brown proximally. The

aerophore line that extends from the stipe is generally

visible with the naked eye and may be paler or darker in

colour than the surrounding tissue. The rachis is mostly

straight throughout its length but distally it may become

slightly flexuose in some species. Adaxially the rachis

forms a V-shaped sulcus along its entire length. In P. luc-

tuosum the sulcus is shallow and not very prominent.

Holttum (1959) stressed the importance of the external

shape of the rachis in defining related groups: in the

Dryopteris- group of ferns, to which Polystichum be-

longs, the rachis has a median sulcus that opens adaxial-

ly to admit the sulcus of the secondary rachises; the

raised edges of the secondary rachis sulcus join the sides

of the costa sulcus with the edge of the pinnule-lamina

decurrent on the secondary rachis as a lateral wing.

Our observations on live Polystichum, Cyrtomium

and Dryopteris material show that the rachis sulcus does

not always open to admit the secondary rachis sulcus, as

the pinna stalk is often terete, especially in the lower part

of the lamina (Figure II, J, K). In Arachniodes, however,

the sulci of the rachis and that of the lower order axes are

confluent (Figure 1L). Paleae similar to those occurring

on the stipe extend to the rachis. Palea density on the

rachis is, however, determined by the species and to a

lesser degree also by the prevailing environmental con-

ditions.

62

Bothalia 30,1 (2000)

FIGURE 2. — Rhizome, root and frond anatomy in Polystichum. A, B, t/s rhizomes showing distribution of vascular tissue:

A, P. transvaalense , Roux 2510A (NBG); B, P. dracomontanum. C, t/s root; D, t/s root showing vascular and non-

vascular tissue. E, t/s rachis; F, t/s dorsal vascular bundle of rachis. G, P. crinulosum , sine coll. s.n. (PJ, adaxial epi-

dermis; H, I, P. aculeatum , Moire s.n. (RAB), adaxial and abaxial epidermal cells; J, P. transkeiense, Roux 2498

(NBG), schematic presentation of frond vasculature showing branching and fusion of vascular bundles along axes.

B-D, Roux 2715 (NBG); E, F, Roux 2377 (NBG). Scale bars: A, B, 5 mm; C, 800 pm; D, F, 100 pm; E, 2 mm; G-I,

140 pm; J, 50 mm.

Bothalia 30,1 (2000)

63

Bulbils

Bulbils are here defined as buds or outgrowths capa-

ble of developing into an independent plant. Although

external stimuli may contribute to the formation of bul-

bils, this ability largely appears to be fixed genetically. In

the polystichoid ferns the ability to produce gemmae

appears to be restricted to Polystichum , Dryopteris and

Phanerophlebia juglandifolia (Willd.) J.Sm. Within

Polystichum a diverse group of species is capable of pro-

ducing bulbils. They have been recorded in sections

Lasiopolystichum Daigobo, Metapoly stichum Tagawa,

Macropoly stichum Daigobo, Cyrtomiopsis Tagawa, Ste-

nopolystichum Daigobo and Haplopolystichum Tagawa

emend. Daigobo. Bulbils in Polystichum are borne either

at the retuse apex of a terminal pinna, at the apex of an

extended glabrous rachis or, as in the case of the species

within the study area, adaxially on the rachis near the

frond apex in or near the pinna ‘axils’. One to three bul-

bils per frond appear to be the norm, but in P. pauciac-

uleatum Bonap. up to five bulbils are borne by a frond.

Paleated bulbils may also occur near the pinna apices in

P. tsaratananense Tardieu and occasionally on the stipe

of P. setiferum (Moore 1855). African and Madagascan

species bearing gemmae include P. crinulosum , P. kili-

manjaricum, P. maevaranense, P. magnificum F.Ballard,

P. pauciaculeatum , P. tsaratananense and P. volkensii

and total 27% of the species within the defined area.

Within the Madagascan region 62% of the species are

bulbiliferous compared with the 64% of species occur-

ring in the West Indies (Mickel 1997).

Most taxa which produce bulbils seem to occur in

moist or tropical conditions where this form of vegeta-

tive reproduction may contribute to more rapid colonisa-

tion in areas of fierce competition. Bulbil formation in

the ferns, its distribution and ecological implications,

requires further study. The fact that bulbil formation

occurs in clearly unrelated taxa suggests that it originat-

ed independently in these groups.

Venation

Venation patterns in the genera of the Dryopterideae

are diverse. This is the principle character commonly

used to segregate Cyrtomium and Phanerophlebia from

Polystichum. In Polystichum the veins of all the species

in the study area are always free and anadromous (Figure

1C, D), but in the circumboreal P. braunii (Spenn.) Fee,

at least some of the veins are catadromous (Kramer

1987). Most veins run into the teeth where they always

terminate near the lamina margin. In P. volkensii the

veins show a slight enlargement towards their apices.

In the fertile pinnules of 2-pinnate species of

Polystichum, the veins mostly terminate in the soral

receptacle midway between the costa and the margin.

This feature is also found in several Arachniod.es and

Dryopteris species. Variations, however, do occur with

the fertile vein often extending for a short distance

beyond the sorus. However, in 1-pinnate Polystichum

and free-veined Phanerophlebia species, the venation of

the fertile pinnae shows no or little differentiation from

that of the sterile pinnae, in that the veins bearing sori are

not conspicuously shortened, thus extending well beyond

the sorus (Figure 1H). Also in most Dryopteris species

the sori-bearing veins are not shortened and extend to the

margin (Figure IF).

Cyrtomium and some Phanerophlebia species have

reticulate venation (Figure 1G). Studies showed that

reticulation in these genera have different origins

(Mitsuta 1977). In Cyrtomium the reticulations have

either a costal or a discal origin, whereas in Phanero-

phlebia they are exclusively marginal (W.H. Wagner

1979). The areolae formed by the reticulations in

Cyrtomium all have one or two (often three) free excur-

rent veinlets on which the sori are borne. In Phanero-

phlebia the areolae have no included veinlets.

ANATOMY

Rhizome: stelar structure

The stelar structure of the Polystichum rhizome can

best be classified as a dictyostele as defined by Schmid

( 1 982). By this definition, two or more perforations (leaf,

root or branch gaps) overlap along the vascular cylinder.

In those Polystichum species with the erect rhizome type,

the number of vascular bundles or meristeles (leaf, root

and branch traces excluded) may number as many as

eight (Figure 2A). In the decumbent rhizome type, how-

ever, the number is typically lower at four or five (Figure

2B), but in P. monticola up to seven vascular bundles

may occur. These bundles are variable in size and shape

and are situated approximately in a medial ring.

The rhizome of Cyrtomium falcatum is a radially

symmetrical dictyostele but here only three traces vascu-

larise each frond; the lowest one diverging into the abax-

ial region of the frond axis and the other two traces into

the adaxial region of the frond axis (Gibson et al. 1984).

The rhizome branch trace in P. transkeiense is a

strand that originates laterally from one of the larger vas-

cular bundles of the main axis. This trace is dorsoventral

and haplostelic as defined by Schmid (1982). The first

frond borne by this rhizome branch is not associated with

a leaf gap. However, when the second and third fronds

develop, non-overlapping leaf gaps are formed. At this

stage the stele is still dorsoventral. A true dictyostele is

established later.

Root traces branch at irregular distances off the dorsal

or lateral surfaces of the main axis vascular bundles and

are not associated with root gaps. Chandra & Nayar

(1982) claim that they originate from the base of the leaf

trace as well, but we were unable to confirm this. The

stele in the roots of most species is circular in cross sec-

tion, whereas in P. luctuosum it is often conspicuously

elliptic with the protoxylem centres forming the distal

poles. In all the Polystichum species studied the roots

were found to be consistently diarch and exarch (Figure

2C, D). In Cyrtomium caryotideum var. micropterum, a

taxon often included in Polystichum (Kramer 1990), the

roots were found to be triarch.

64

Bothalia 30,1 (2000)

Frond: stelar structure of the axes

The stelar structure of the stipe base is characterised

by four to five (rarely seven) vascular bundles arranged

in a U-shape towards the adaxial surface. Khullar &

Gupta (1979) reported up to nine vascular bundles in P.

biaristatum (Blume) T.Moore, whereas smaller species

like P. obliquum (D.Don) T.Moore merely have two. In

Polystichum the arc is a broken line with the larger vas-

cular bundles located in a dorsolateral position at either

end and the smaller ones between (Figure 2E). The

smaller vascular bundles are the first to branch from the

lateral ends of two adjacent rhizome meristeles and the

larger ones are the last to branch from the rhizome meris-

teles. A reduction in the number of vascular bundles usu-

ally takes place along the length of the stipe as some of

the smaller bundles merge. An increase often occurs as a

bundle may also split to fuse again after a short distance.

The xylem strand in the larger meristeles is curved and

hooked adaxially, the so-called hippocampus shape when

viewed in cross section (Lin & De Vol 1977) (Figure 2F).

Xylem bundles in the smaller meristeles are circular to

broadly elliptic in cross section. In Dryopteris we

observed a consistently larger number of vascular bun-

dles (up to 12) in the stipe.

Xylem tissue in the vascular bundles is composed of

helical and scalariform tracheids with thin plates of

xylem parenchyma irregularly dispersed between them.

Protoxylem of the smaller and larger vascular bundles

has an endarch arrangement. The phloem, which forms a

sheath around the xylem, is more prominent in the outer

dorsolateral and ventral sides of the xylem. The pericy-

cle consists of large parenchymatous cells two to four

cell layers deep. The endodermis is characterised by

inner and radial walls that are secondarily thickened and

in which numerous simple and branched pits of variable

sizes are scattered. Pit apertures are always elliptic in

outline.

In all the species of the Dryopterideae studied the

arrangement of the vascular tissue in the rachis is essen-

tially the same as that of the stipe. In the rachis, howev-

er, there is a further reduction in the number of vascular

bundles towards the lamina apex. Also here the smaller

bundles initially merge with one another at irregular

intervals and often also divide as in the stipe. The final

small bundle eventually merges with one of the larger

dorsolateral bundles some distance from the lamina

apex. The two dorsolateral meristeles eventually fuse

and continue as a single vascular bundle to the lamina

apex (Figure 2 J ).

The vascular tissue serving each pinna, branches as a

single dorsolateral trace from one of the two larger vas-

cular bundles. This trace soon divides and the two (rarely

three) bundles that are formed run parallel to one anoth-

er for most of the pinna length before they finally merge

near the pinna apex. This condition holds true for 1 -pin-

nate species and those with more compound laminae

suggesting the costa of 1 -pinnate species to be analogous

with the secondary rachis of species with more com-

pound laminae. Costules and veins branch from one of

the two pinna meristeles or the terminal bundle of the

pinna apex and remain single-stranded.

Rhizome: non-vascular tissue

In most of the species investigated the non-vascular

tissue of the rhizome is distinctly differentiated into

parenchymatous ground tissue and sclerenchyma.

Histologically the cortex and pith are identical, consist-

ing of small isodiametric parenchyma cells. Towards the

epidermis, however, the cells become compressed with

the anticlinal walls somewhat sinuate. A narrow sheath

of sclerenchyma several cell layers deep is situated

beneath the epidermis. In the outer cortical layer this

sclerenchyma usually extends to the root trace. In P.

transkeiense a small sclerotic cap is formed on the outer

periphery of the root trace only, whereas in P. dra-

comontanum and P. incongruum a sclerotic sheath often

forms around the entire root trace. Chandra & Nayar

(1982) also found sclerenchyma bundles associated with

the departing leaf trace bundles. Small to large groups of

partially to moderately thick-walled sclerenchymatous

cells occur at random throughout the ground tissue of

most species and often extend to the stipe bases. Their

abundance appears to be determined by age and the envi-

ronment with more cell aggregates present in older parts

of the rhizome as well as in plants occurring in exposed

habitats. In older rhizomes they appear to be concentrat-

ed near the larger meristeles. Considering the abundance

and distribution of sclereid clusters in the ground tissue

of the rhizome, Chandra & Nayar (1982) suggested some

evolutionary trends and species relationships within

Polystichum. D.H. Wagner (1979) also found the pres-

ence of sclereid clusters in the rhizome of P. munition

(Kaulf.) C.Presl diagnostic in separating it from P. imbri-

cans (D.C.Eaton) D.H. Wagner where they are mostly

absent. However, because of their variability, we consid-

er them to be of no taxonomic value for the species we

studied. The deposition of phenolic substances in the

sclerenchymatic clusters in especially the older parts of

the rhizome may act as a preservative against decay, thus

promoting longevity.

The cortex of Polystichum roots is composed of a

dense inner sclerenchymatic sheath and parenchymatous

outer cortex. In old roots the outer non-sclerotic cortical

layer decomposes. Sclerenchyma cells opposite the pro-

toxylem poles are generally not as strongly lignified as

the rest of the sheath (Figure 2D). Schneider (1996)

terms these non-sclerotic cells opposite the protoxylem

poles passage cells, a feature that characterises the

Davallia type of root.

Frond: non-vascular tissue

In cross section the stipe base is usually circular,

tetragonal or transversely broadly ovate in outline. In the

upper part of the stipe, however, the adaxial surface is

usually variously sulcate, as is the rachis. In cross section

the epidermal cells are small and isodiametric in outline

and their walls generally densely lignified, similar to the

underlying band of sclerenchyma. In young material this

layer is interrupted dorsolaterally by the aerophore lines

studied by Davies (1991). In younger fronds the lateral

aerophore lines are composed of parenchymatous cells

with large intercellular spaces and are usually not very

rich in phenolic substances. With age, however, this tis-

Bothalia 30,1 (2000)

65

sue also becomes sclerified. Stomata occurring on the

stipe and rachis are confined to the aerophore line and

may be raised but are mostly flush with the epidermis. A

tanniferous sclerenchyma layer is deposited on the inner

tangential and radial walls of the cell layer adjoining the

endodermis. The density of this layer is determined by

age with that in older fronds being more prominent.

The lamina anatomy of species with 1 -pinnate fronds

differs slightly from those species with more compound

laminae. In cross section the pinnules are flat with a shal-

low V-shaped sulcus adaxially along the costa that has

slightly raised margins. The costa and veins that are situ-

ated equidistant between the adaxial and abaxial surfaces

are enveloped by a narrow layer of parenchymatous tis-

sue, a bundle sheath, that stretches from the adaxial to

the abaxial epidermis. This tissue becomes lignified as

the frond ages. In P. macleae, a 1 -pinnate species, the

pinna structure is similar except that the costa is signifi-

cantly enlarged abaxially. This enlarged costa is com-

posed of parenchymatous mesophyll in which the two

vascular strands that run parallel to each other occur.

Epidermal cells

Epidermal cells can provide many characters of taxo-

nomic value (Stace 1984), such as cell size, shape, ori-

entation and anticlinal wall undulation. Cell size, howev-

er, may be influenced by ploidy level and environmental

factors (Metcalfe & Chalk 1950).

Epidermal cell size in Polystichum has been studied

by Chandra (1977), who concluded that Polystichum has

a closer affinity with Cyrtomium than it has with

Arachniodes and Dryopteris. Epidermal cells in all the

species of this investigation are elongated parallel to the

veins and are irregular in shape. Cell size of the adaxial

and abaxial epidermis differ, those of the abaxial epi-

dermis usually being larger. The anticlinal walls of the

epidermal cells are generally sinuous to deeply lobed

(Figure 2H), but in P. crinulosum (Desv.) J.P.Roux the

cells walls are almost straight (Figure 2G). Epidermal

cells occurring abaxially along the veins are generally

narrow and their anticlinal walls are not as deeply sinu-

ate as those of the intercostal cells. Cell size varies con-

siderably between species, with the smallest cells in P.

ammifolium and the largest in P. wilsonii (Table 2).

Stomata

Stomatal ontogeny in pteridophytes has been studied

by Mickel & Lersten (1967), Pant & Khare (1969), Sen

& Hennipman (1981) and Mehra & Soni (1983) to name

a few. The most comprehensive studies on the morphol-

ogy of the mature stomatal complex are those of Van

Cotthem (1970), Sen & Hennipman (1981) and Mehra &

Soni (1983).

Stomata in Polystichum are confined to the abaxial

surface of the lamina and are positioned in the same plane

as the epidermal cells, oriented with their longitudinal

axis parallel to the lateral veins (Figure 21). Anomocytic,

eupolocytic, copolocytic and staurocytic stomata have

been reported for Polystichum (Van Cotthem 1970;

Chandra 1977; Mehra & Soni 1983). In the species dealt

with here, we found eupolocytic stomata to be the most

common type, followed by the anomocytic state. A few

copolocytic stomata were noted in P. aculeatum (L.)

Roth, P. dracomontanum, P. falcinellum, P. incongruum

and P. kalambatitrense. Guard cell length is fairly vari-

able: the smallest stomata were recorded in P. crinulosum

and the largest in P . falcinellum (Table 3). In P. setiferum

stoma initials often abort during early development.

Stoma size appears to be related to ploidy level (Table 3).

TABLE 2. — Adaxial epidermal cell length in Polystichum

Taxon Range and mean (pm) SD N

aculeatum (L.) Roth

ammifolium (Poir.) C.Chr.

crinulosum (Desv.) J P.Roux

dracomontanum Schelpe & N.C. Anthony

drepanum (Sw.) C.Presl

falcinellum (Sw.) C.Presl

incongruum J.P.Roux

kalambatitrense Tardieu

kilimanjaricum Pic.Serm.

luctuosum (Kunze) T.Moore

macleae (Baker) Diels

maevaranense Tardieu

magnificum F.Ballard

marionense Alston & Schelpe

monticola N.C. Anthony & Schelpe

pauciaculeatum Bonap.

pungens (Kaulf.) C.Presl

setiferum (Forssk.) T.Moore ex Woyn.

transkeiense W.B.G. Jacobsen

transvaalense N.C. Anthony

tsaratananense Tardieu

volkensii (Hieron.) C.Chr.

wilsonii H. Christ

zambesiacum Schelpe

x maderense J.Y.Johnson

x saltum J.P.Roux

SD, standard deviation; N, number of observations.

66

Bothalia 30,1 (2000)

TABLE 3. — Guard cell length in Polystichum

SD, standard deviation; N, number of observations.

CONCLUSIONS

This study illustrates our poor understanding of the

genus. Polystichum, like other polystichoid ferns, is

characterised by a set of features, most of which are also

shared by other members within the assemblage, rather

than a set of unique characters (Table 4). The genus can,

however, be readily distinguished by an adaxially sulcate

rachis with four to five (rarely seven) vascular bundles

arranged in a U-shape, axes that are adaxially shallowly

sulcate, the sulcus of the secondary axes confluent or not

with that of the main axis, 1 -pinnate to 3-pinnate fronds

with (mostly) acroscopically developed pinnae and pin-

nules, anadromous free venation, indumentum composed

of paleae (rarely also clavate glandular hairs), sori occur-

ring dorsally on the veins or at a vein apex, and a peltate

indusium (secondarily lost in a number of species).

In Polystichum the short, erect rhizome type is the

most common, occurring in a wide range of taxa of

which many are not considered related and is also wide-

spread in the other polystichoid genera. Species belong-

ing to this group are mostly forest dwelling and occur as

individuals. Species in the second rhizome type can

reproduce vegetatively and have the ability to grow in a

wider range of habitats. This clone-forming group can

consequently be found growing in forests, forest margins

or open habitats. The sharing of resources by clone-form-

ing plants (Hutchings 1997) may compensate for the

fewer fronds produced by each rhizome branch.

Chandra & Nayar (1982) and D.H. Wagner (1979) con-

sider the presence and distribution of sclerenchymatic cell

aggregates in the cortex and pith of the rhizome of taxo-

nomic value. Our observations, however, suggest their

presence and number is influenced by age and habitat and

we consequently consider them of no taxonomic value.

Root morphology and anatomy also appear to have

little taxonomic value at the species or genus level. In

Polystichum both the Dennstaedtia type and Davallia

type of cortex occur (Schneider 1996). These root types

are widespread within the Pteropsida occurring in sever-

al groups generally not considered related. The latter

type, characterised by the presence of passage cells

opposite the protoxylem poles, we observed in P. luctu-

osum only. All species had diarch roots apart from

Cyrtomium caryotideum var. micropterum, where triarch

roots were observed, a feature that requires further study.

Frond morphology in African Polystichum, as in most

polystichoid ferns, is of the monomorphic type. Only in

Dryopteris, however, do several species have dimorphic

fronds. This derived feature is widespread and within the

Pteropsida must have arisen independently on numerous

occasions. The morphology and anatomy of the axes in

polystichoid ferns also exhibit some important evolu-

tionary trends. Holttum (1959) considered the external

shape of the rachis and the way in which the shape is

modified when a secondary rachis is attached to it, of sig-

nificant evolutionary and taxonomic importance. He

used this feature in separating fern groups with a similar

habit and sori. Our observations in Polystichum, howev-

er, show that this feature exhibits intra- and interspecific

variation in the pinnae stalks often being terete rather

than sulcate. The pinna sulcus is thus not confluent with

that of the rachis. The large number of small vascular

bundles in the stipe and rachis of Dryopteris taxa we

studied appears to be unique within the Dryopterideae,

suggesting it to be of taxonomic value, but the evolu-

tionary significance of this is not understood. The sec-

ondary rachis being homologous with the costa in 1-pin-

nate species of polystichoid ferns suggests frond simpli-

fication to be derived (Mitsuta 1977; Yatskievych 1989,

1996).

Bothalia 30,1 (2000)

67

TABLE 4. — Summary of some morphological and anatomical character trends in the rhizome and frond of polystichoid ferns based on species

studied and a literature survey

* Depending on the circumscription of the genus, for example P. dubium (Karst.) Diels has reticulations.

Within the polystichoid ferns as defined earlier, only

certain Polystichum and Dryopteris species have the

ability of producing bulbils on the laminae. In these gen-

era one or more bulbils are mostly borne on the rachis

behind the apical pinna. In the West Indies, however,

most of the Polystichum species bear bulbils, either at a

retuse apex of the terminal pinna or at the apex of an

extended rachis. The ability to produce bulbils is here

viewed as a derived feature. The distinct way in which

the bulbils are borne suggest that this feature has also

arisen independently on more than one occasion in

Polystichum.

Ontogenetic studies show that monopodially branch-

ed free veins are ancestral and that reticulate venation

within the polystichoid ferns is derived (Mitsuta 1977). It

is widely accepted that reticulate veins are derived inde-

pendently in Cyrtomium and Phanerophlebia. Yatskie-

vych (1996) is also of the opinion that within Phanero-

phlebia it arose independently on two occasions. Sorus-

bearing veins terminating near the lamina margin are

here considered ancestral, whereas modified veins termi-

nating in the sorus or extending for a short distance

beyond the sorus are considered derived.

The epidermal cell anticlinal walls show considerable

variation in the degree to which they are undulated and

are here not considered to be of any taxonomic value.

Polystichoid ferns are all hypostomatic, and the stomata

are mostly of the anomocytic type. Our observations

confirm that within closely related species ploidy level is

reflected in the guard cell length as P. pungens, an octo-

ploid (pers. obs.) has larger stomata than P. incongruum

which is tetraploid (Table 3).

Judging by the information provided, more detailed

studies on a wider range of species are required if a bet-

ter understanding of Polystichum , and its affinity with

the other genera in the Dryopterideae is to be formed.

Further studies may reject or support some of the specu-

lative evolutionary trends proposed here. The view of

Yatskievych (1989) who suggested Polystichum to be

polyphyletic, is supported here.

ACKNOWLEDGEMENTS

We wish to express our thanks to Yvonne Reynolds,

librarian at the Harry Molteno Library, Kirstenbosch, and

the various South African Botanical Liaison Officers that

served at Kew since 1987 for obtaining literature. Our

thanks also go to the anonymous reviewers for their valid

criticism on an earlier draft of this paper.

REFERENCES

BOWER, F.O. 1928. The ferns. Vol. 3. Cambridge University Press,

England.

CHANDRA, P. 1977. Epidermal studies in some species of Poly-

stichum Roth. New Botanist 4: 101-107.

CHANDRA, P. & NAYAR, B.K. 1982. Morphology of some polysti-

choid ferns. II. The sporophyte of Arachniodes , Cyrtomium and

Polystichum. Journal of the Indian Botanical Society 61:

391-403.

CHING, R.C. 1940. On the classification of the family Polypodiaceae.

Sunyatsenia 5: 201-268.

CHRISTENSEN, C. 1930. The genus Cyrtomium. American Fern Jour-

nal 20: 41-52.

COPELAND, E.B. 1947. Genera filicum. Choranica Botanica,

Waltham, Massachusetts.

DAVIES, K.L. 1991. A brief comparative survey of aerophore structure

within the Filicopsida. Botanical Journal of the Linnean Society

107: 115-137.

GIBSON, A.C., CALKIN, H.W. & NOBEL, P.S. 1984. Xylem anato-

my, water flow, and hydraulic conductance in the fern

Cyrtomium falcatum. American Journal of Botany 1 1 : 564—574.

HASEBE, M., WOLF, P.G., PRYER, K.M., UEDA, K., ITO, M„

SANO, R„ GASTONY, G.J., YOKOYAMA, J„ MANHART,

J.R., MURAKAMI, N„ CRANE, E.H., HAUFLER, C.H. &

HAUK, W.D. 1995. Fern phylogeny based on rbch nucleotide

sequences. American Fern Journal 85: 134—181.

HOLTTUM, R.E. 1947. A revised classification of the leptosporangiate

ferns. Botanical Journal of the Linnean Society 53: 123-158.

HOLTTUM, R.E. 1959. Vegetative characters distinguishing the vari-

ous groups of ferns in Dryopteris of Christensen's Index fil-

icum, and other ferns of similar habit and sori. Gardens Bul-

letin, Singapore 17: 361-367.

HOLTTUM, R.E. 1973. Posing the problems. Botanical Journal of the

Linnean Society 61 , Suppl. Vol. 1: 173-189.

HUTCHINGS, M.J. 1997. Resource allocation in clonal herbs and their

consequence for growth. In F.A. Bazzaz & J. Grace, Plant

resource allocation'. 161-189. Academic Press, San Diego.

KHULLAR, S.P. & GUPTA, S.C. 1979. Morphological studies of some

Himalayan polystichums. Nova Hedwigia Beiheft 63,2: 41-58.

KRAMER, K.U. 1987. A brief survey of the dromy in fern leaves, with

an expanded terminology. Botanica Helvetica 92: 219-228.

KRAMER, K.U. 1990. Dryopteridaceae. In K. Kubitzki, The families

and genera of vascular plants 1. Pteridophytes and Gymno-

sperms: 101-144. (K.U. Kramer & P.S. Green, volume editors).

Springer- Verlag, Berlin.

LIN, B.L. & DE VOL, C.E. 1977. The use of stipe characters in fem

taxonomy I. Taiwania 22: 91-99.

LIN, B.L. & DE VOL, C.E. 1978. The use of stipe characters in fem

taxonomy II. Taiwania 23: 77-95.

MEHRA, P.N. & SONI, S.L. 1983. Stomatal patterns in pterido-

phytes— an evolutionary approach. Proceedings of the Indian

National Science Academy, Part B, 49: 155-203.

68

Bothalia 30,1 (2000)

METCALFE, C.R. & CHALK, L. 1950. Anatomy of the dicotyledons.

Claredon Press, Oxford.

MICKEL, J.T. 1997. A review of the West Indian species of Poly-

stichum. In R.J. Johns, Holttum Memorial Volume : 119- 143.

Royal Botanic Gardens, Kew.

MICKEL, J.T. & LERSTEN, N.R. 1967. Floating stomates (adenosto-

my) in ferns: distribution and ontogeny. American Journal of

Botany 54: 1181-1185.

MITSUTA, S. 1977. Evolution of simple fronds in Cyrtomium — a pat-

tern morphological study. Acta Phytotaxonomica Geobotanica

28: 131-142.

MOORE, T. 1855. A popular history of British ferns and allied plants.

Reeve, London.

NAYAR, B.K. 1970. A phylogenetic classification of the homosporous

ferns. Taxon 19: 229-236.

NAYAR, B.K. 1979. The classification of ferns: its present status and

problems. In S.S. Bir, Aspects of Plant Sciences 3: 1-38.

NOODEN, L.D. & WAGNER, W.H. 1997. Photosynthetic capacity and

leaf re-orientation in two wintergreen ferns, Polystichum acros-

tichoides and Dryopteris intermedia. American Fern Journal

87: 143-149.

PANT, D.D. & KHARE, P.K. 1969. Epidermal structure and stomatal

ontogeny in some eusporangiate ferns. Annals of Botany 33:

795-805.

PICHI SERMOLLI, R.E.G. 1972. Fragmenta pteridologiae III. Webbia

27: 389-460.

PICHI SERMOLLI, R.E.G. 1977. Fragmenta pteridologiae VII. Web-

bia 32: 69-93.

PRYER, K.M., SMITH, A.R. & SKOG, J.E. 1995. Phylogenetic rela-

tionships of extant ferns based on evidence from morphological

and rbcL sequences. American Fern Journal 85: 205-282.

SCHMID, R. 1982. The terminology and classification of steles: his-

torical perspective and the outlines of a system. Botanical

Review 48: 817-931.

SCHNEIDER, H. 1996. The root anatomy of ferns: a comparative

study. In J.M. Camus, M. Gibby & R.J. Johns, Pteridology in

perspective : 271-283. Royal Botanic Gardens, Kew.

SEN, U. & HENNIPMAN, E. 1981. Structure and ontogeny of stoma-

ta in Polypodiaceae. Blumea 27: 175-201.

STACE, C.A. 1984. The taxonomic importance of the leaf surface. In

V.H. Heywood & D.M. Moore, Current concepts in plant tax-

onomy. 67-94. The Systematics Association Special Volume

No. 25. Academic Press, London/Orlando.

TRYON, R.M. & TRYON, A.F. 1982. Ferns and allied plants with

special reference to tropical America. Springer- Verlag, New

York.

VAN COTTHEM, W.R.J. 1970. Comparative morphological study of

the fem stomata in the Filicopsida. Bulletin du Jardin Botanique

National de Belgique 40: 81-239.

WAGNER, D.H. 1979. Systematics of Polystichum in western North

America north of Mexico. Pteridologia 1: 1-64.

WAGNER, W.H. 1979. Reticulate veins in the systematics of modem

ferns. Taxon 28: 87-95.

WAGNER, W.H. 1985. Morphological variation and evolution in

Polystichum. In D.S. Barrington, The present evolutionary and

taxonomic status of the fem genus Polystichum: The 1984

Botanical Society of America Pteridophyte Section Sympo-

sium. American Fern Journal 75: 22-28.

WAGNER, W.H. & JOHNSON, D.M. 1983. Trophopod, a commonly

overlooked storage structure of potential systematic value. Taxon

32: 268, 269.

WIDEN, C.J., MITSUTA, S. & IWATSUKI, K. 198 1 . Chemotaxonomic

studies on Arachniodes (Dryopteridaceae) III. Phloroglucinol

derivatives of putative hybrids. Botanical Magazine (Tokyo)

94: 127-139.

YATSKIEVYCH, G. 1989. A new combination in South American Poly-

stichum. American Fern Journal 79: 26, 27.

YATSKIEVYCH, G. 1996. A revision of the fem genus Phanerophle-

bia (Dryopteridaceae). Annals of the Missouri Botanical Gar-

den 83: 168-199.

Bothalia 30,1:69-86 (2000)

The epidermis in Passerina (Thymelaeaceae): structure, function

and taxonomic significance

C.L. BREDENKAMP* and A.E. VAN WYK**

Keywords: anatomy, cuticle, epicuticular waxes, epidermis, Passerina, southern Africa, stomata, taxonomy, Thymelaeaceae

ABSTRACT

Epidermal features were studied in all 1 7 species of Passerina , a genus endemic to southern Africa. Leaves in Passerina

are inversely ericoid, the adaxial surface concave and the abaxial surface convex. Leaves are inversely dorsiventral and

epistomatic. The adaxial epidermis is villous, with unicellular, uniseriate trichomes and relatively small thin-walled cells,

promoting flexibility of leaf margins owing to turgor changes. In common with many other Thymelaeaceae, abaxial epi-

dermal cells are large and tanniniferous with mucilaginous cell walls. The cuticle is adaxially thin, but abaxially well devel-

oped, probably enabling the leaf to restrict water loss and to tolerate high light intensity and UV-B radiation. Epicuticular

waxes, present in all species, comprise both soft and plate waxes. Epidermal structure proves to be taxonomically impor-

tant at family, genus and species levels. Interspecific differences include arrangement of stomata and presence or absence

of abaxial epidermal hair. Other diagnostic characters of the abaxial epidermal cells are arrangement, size and shape, cutic-

ular ornamentation and presence or absence of wax platelets. Two groups of species on the basis of abaxial epidermal cell

orientation are recognised. Many leaf epidermal features in Passerina are interpreted as structural adaptations to the

Mediterranean climate of the Cape.

CONTENTS

Introduction 69

Material and methods 71

Light microscope studies (LM) 71

Scanning electron microscope studies (SEM) .... 72

Transmission electron microscope studies (TEM) . 72

Terminology 72

Trichome structure 72

Cuticle 72

Cuticular ornamentation 72

Epicuticular wax 72

Results 72

Macromorphology of the leaf 72

Anatomy of the leaf 73

Adaxial (dorsal) epidermis 73

Cuticle 73

Epidermal cells 73

Stomatal complex 73

Trichomes 75

Abaxial (ventral) epidermis 75

Trichomes 75

Epidermal cells 75

Cuticle 77

Cuticular ornamentation 79

Epicuticular waxes 79

Discussion 79

Adaxial epidermis 79

Stomatal complex 79

Trichomes 81

Cuticular ornamentation 81

Epicuticular waxes 83

Functions of epicuticular waxes 84

Systematic value 84

Ecological aspects of leaf epidermis 84

* National Botanical Institute, Private Bag X101, 0001 Pretoria.

** H.G.W.J, Schweickerdt Herbarium, Department of Botany, University

of Pretoria, 0002 Pretoria.

MS. received: 1999-06-07.

Conclusions 85

Acknowledgements 85

References 85

INTRODUCTION

The genus Passerina L. comprises about 17 species, all

endemic to southern Africa (Thoday 1924; Bond &

Goldblatt 1984). Despite the now outdated revision by

Thoday (1924), taxonomic boundaries in Passerina remain

a problem, mainly owing to the apparent lack of marked

morphological differences between the species. The pre-

sent paper emanates from a comparative leaf-anatomical

survey of the genus, undertaken as part of a monographic

study of the group. This survey highlighted the importance

of the epidermis as a source of taxonomic evidence.

The combined distribution of all the Passerina species

is shown in Figure 1. Most species of Passerina are

FIGURE 1 — Number of species per grid in the distribution of Passerina.

Lines PQ and RS: boundaries between summer (A), intermediate

(B) and winter (C) rainfall areas. Line XY shows northern bound-

ary of Cape Supergroup rock outcrops.

70

Bothalia 30,1 (2000)

TABLE 1 . — Passerina specimens examined and housed at PRE

Material used for the SEM study of the ad- and abaxial epidermises. +Fresh material collected for the TEM study.

Bothalia 30,1 (2000)

71

endemic to the Cape Floristic Region. From here the dis-

tribution of P.filiformis and P. montana extends east and

north along the eastern mountains and Great Escarpment

of southern Africa. In the Cape the climate is for the most

part Mediterranean or semi-Mediterranean. In the west, it

rains in winter; along the south coast, winter rainfall is

complemented by some summer rain which increases

eastwards. The western Karoo and Namaqualand (Suc-

culent Karoo Biome) are characterised by winter precipi-

tation and summer drought. KwaZulu-Natal and the east-

ern mountains of southern Africa are predominantly sum-

mer rainfall areas. Distribution of the species of

Passerina coincides with the geography and climate

along the whole distribution area. P. ericoides, P.

paleacea, P. paludosa, P. galpinii and P. burchellii are

endemic to Western Cape. The first three species are

found along beaches and salt marshes only, P. galpinii

grows mainly on calcrete in the Agulhas Plain area

(Goldblatt & Manning in press) and P. burchellii is found

on the high mountains at Genadendal and Villiersdorp.

P. comosa grows on mountain slopes and summits in the

Kamiesberg, Great Winterhoek and Klein Swartberg

Ranges. P. glomerata is found from Worcester to Tul-

bagh, in the Clanwilliam area and extends to the

Witteberg south of Matjiesfontein. P. obtusifolia is ubiqui-

tous in the Cape, distributed from Worcester in Western

Cape to Alice in Eastern Cape and on some of the moun-

tain ranges in and around the Little Karoo. A new species,

of which the plants are often buried under snow during

winter, grows at high altitudes in the Ceres Karoo. P. vul-

garis is a pioneer with a wide distribution from Western

Cape to East London in Eastern Cape. P. falcifolia is

found on mountain ranges between George and Uiten-

hage and P. pendula is endemic to the KwaZunga

Catchment Basin and the Zwartkops River area of Eastern

Cape. P. rubra is common in the Port Elizabeth to

Uitenhage area, with outliers in the Swellendam and

Bredasdorp Districts. P. drakensbergensis is endemic to

the high Drakensberg in the Bergville District of

KwaZulu-Natal and P. rigida is distributed all along the

coast, from northern KwaZulu-Natal to the Cape

Peninsula. Passerina sp. nov. 2 is found on the northern

Cederberg Mountains, P. sp. nov. 3 at mountain tops in the

Uitenhage area and the Swartberg Pass and P. sp. nov. 4

on the Kouga Mountains and the Montagu Pass.

The most important studies applying the ‘anatomical

method’ for the delimitation of the Thymelaeaceae were

published by Van Tieghem (1893) and Supprian (1894).

The presence of mucilaginous epidermal cells in P. eri-

coides (= Chymococca empetroides Meisn.) as opposed

to the total lack thereof in the other species, was also

mentioned by Supprian (1894). Subsequently, Gilg

(1894) critically discussed the ‘anatomical method’ as

applied by Van Tieghem (1893) and Supprian (1894) for

the delimitation of the Thymelaeaceae and concluded that

certain characters would not uphold criticism. He regard-

ed former systems based on floral morphology as more

suitable for a taxonomic grouping of the Thymelaeaceae.

The twentieth century yielded very little anatomical

work on the Thymelaeaceae. Standard works were those

of Solereder (1908) and Metcalfe & Chalk (1950, 1979).

Thoday (1921) described the structure and behaviour in

drought of the ericoid leaves of P. filiformis and P. cf.

falcifolia ; he also supplied some notes on their anatomy.

In a discussion of inversely dorsiventral leaves, Kugler

(1928) included a description of the leaves of P. fili-

formis (= P. pectinata Hort.). More recently, leaf anato-

my of the genera Lachnaea L. and Cryptadenia Meisn.

was treated by Beyers (1992) and leaf and involucral

bract characters of systematic use in Gnidia L. were

studied by Beaumont et al. (1994). The scanty informa-

tion on leaf anatomy in Thymelaeaceae calls for further

research in this field, especially in the genus Passerina.

Previous leaf anatomical studies identified mucilagi-

nation of the epidermal cells as being of possible taxo-

nomic importance. Recently Bredenkamp & Van Wyk

(1999) clarified the structure of the epidermal cells and

origin of the mucilage, concluding that mucilagination of

epidermal cells is of taxonomic importance mainly at the

family level.

The wide distribution of Passerina in the Cape Floris-

tic Region, along the southern and eastern coastline and

along the Great Escarpment of southern Africa as far north

as Zimbabwe, illustrates the adaptation of these plants to a

wide range of habitats, including Mediterranean and sum-

mer rainfall regimes. Decreasing rainfall from the eastern

Escarpment to the northwestern Cape is reflected by adap-

tive changes in the leaf structure of the group. The present

paper provides a description of epidermal characters in

Passerina as well as an assessment of their taxonomic sig-

nificance. It also speculates on the possible adaptive value

of the observed structural features of the leaf.

MATERIAL AND METHODS

Fresh leaf material of 17 species of Passerina (Table 1)

was collected, fixed and stored in a 0.1 M phosphate-

buffered solution at pH 7.4, containing 2.5% formalde-

hyde, 0.1% glutaraldehyde and 0.5% caffeine [modified

Karnovsky fixative; Karnovsky (1965)]. Whenever possi-

ble, material from at least five different localities was

collected, fixed and air-dried for each species and

herbarium specimens were made.

Light microscope (LM) studies

The LM was used for general leaf anatomy as well as

epidermal studies. Unless stated otherwise, the tenth leaf

from the growing point of a twig was used in all com-

parative studies. To prepare transverse sections of the

main vein as well as both leaf margins, a 1 mm wide seg-

ment of leaf material was cut from the centre of each

leaf. Samples were dehydrated, embedded in glycol

methacrylate (GMA) and sectioned according to the

methods of Feder & O’Brien (1968). Sections were

stained with the periodic acid/Schiff’s reaction and in

toluidine blue ‘O’, then mounted in Entellan (Art. 7961,

E. Merck, Darmstadt).

The following three methods were followed in the

study of the cuticles:

1. GMA transverse sections of leaves were stained for

10 minutes in 1% Sudan Black B dissolved in 70%

ethanol. Sections were rinsed twice in 70% ethanol for a

few seconds and mounted in glycerine jelly.

72

Bothalia 30,1 (2000)

2. Cuticular mounts were obtained by maceration

according to the method of Kiger (1971). Specimens

were slightly over-stained in a 1% acqueous safranin

solution, dehydrated in methyl cellusolve and mounted

in Entellan.

3. Epidermal mounts were obtained by removing

small pieces of ad- and abaxial epidermis manually and

by paradermal hand sections. Epidermises were stained

in 1% safranin dissolved in 50% ethanol, dehydrated in a

graded ethanol series and mounted in Entellan.

Scanning electron microscope (SEM) studies

The SEM was used to study the epidermal surface

features (including epicuticular waxes), as well as to ver-

ify the structure of the cuticle. Leaves from air-dried

material were used for all species. Whole leaves were

used as they are small and ericoid, but trichomes were

manually removed adaxially to reveal the stomata.

Leaves were mounted onto aluminium stubs with silver

paint, exposing the ad- and abaxial surfaces separately

and sputter-coated with gold. For the purpose of studying

epicuticular waxes, the sputter-coating process was mod-

ified to prevent high temperatures from changing the

wax surfaces. Specimens were sputter-coated for 30 sec-

onds and left to retain their normal temperature for one

minute. This was repeated five times after which the

specimens were viewed with a Jeol 840 SEM.

For the verification of the authenticity of epicuticular

wax droplets and small round protrusions observed in

certain species of Passerina, leaves were washed in chlo-

roform for one minute, before they were pasted onto alu-

minium stubs. The procedure described above was used

for sputter-coating and viewing.

Transmission electron microscope (TEM) studies

The TEM was used for the study of the structure of

mucilaginous epidermal cell walls in Passerina. The sec-

ond, fifth and tenth leaf from the growing points of

P. ericoides, P. falcifolia and P. paleacea were used to

study the structure of the cell wall. Leaf segments of

± 1 mm2 were fixed in a modified Kamovsky fixative

(Karnovsky 1965). Fixed material was rinsed in 0.075 M

phosphate buffer, pH 7.4-7. 5, post-fixed for one hour in

0.25% aqueous OsCL, washed in three changes of water,

dehydrated in a graded acetone series and embedded in

Quetol 651 resin (Van der Merwe & Coetzee 1992).

Ultrathin sections were contrasted in 4% aqueous uranyl

acetate for 10 minutes and rapidly rinsed in water three

times. The sections were then contrasted with lead citrate

(Reynolds 1963), rinsed in water and examined with a

Phillips 301 TEM.

For the verification of wettability and possible

absorption of water by laminar epidermal hairs, we fol-

low Alvin (1987). He proposed a mechanism through

which water is absorbed by the specialised abaxial epi-

dermal trichomes of Androstachys johnsonii Prain. This

process involves the wettability of the hairs which he

investigated by spraying the glabrous adaxial surfaces of

the leaves with water. Water seeped round the leaf mar-

gins to the abaxial surface, wetting approximately 50%

of the abaxial surface within 5 minutes. In the present

study, the glabrous abaxial surfaces of five cymbiform

leaves (from dried herbarium specimens) were pasted

onto a sticky surface, exposing the villous concave adax-

ial surface. A drop of water was placed in the adaxial

groove at the base of each leaf (average leaf size 2.5 x

4.0 mm) and left overnight. This experiment was repeat-

ed using 0.5% aqueous safranin, followed after 20 min-

utes by a rinse with water.

Terminology

Trichome structure

We have followed the terminology of Stace (1965)

and Theobald et al. (1979).

Cuticle

Although the interpretation proposed by Martin &

Juniper (1970) for the cuticle of plants has been widely

followed by many workers, Holloway (1982) reviewed

the historical perspective of the plant cuticle and attempt-

ed to adopt the most workable interpretation of the cutic-

ular membrane (CM) in practice. In response, we follow

Jeffree (1986), whose uncomplicated and pragmatic inter-

pretation distinguishes three main zones, namely the cuti-

cle proper, the cuticular layer and the cell wall. The cutic-

ular membrane comprises the cuticle proper plus the

cuticular layer and is bonded to the outer periclinal walls

of the epidermal cells by a pectin-rich layer, which is

equivalent to the continuous middle lamella. A layer of

epicuticular wax generally coats the cuticle proper.

Cuticular ornamentation ( LM and SEM )

We follow Wilkinson (1979) in our choice of termi-

nology to describe cuticular ornamentation.

Epicuticular wax

The recognition of soft waxes in the present study is

based on the criteria proposed by Amelunxen et al.

(1967), Wilkinson (1979) and Barthlott et al. (1998).

RESULTS

Macromorphology of the leaf

Leaf arrangement decussate, sometimes imbricate,

closely adherent to stem or spreading at angle of

5°-20°(-60°); spreading of leaves often prominent in juve-

nile plants. Lamina inversely ericoid; adaxial surface con-

cave, often forming a groove facing stem and lined with

woolly hairs; abaxial surface convex, orientated more or

less acroscopically, thus exposing a large surface area to

the environment; cuticle often amber-coloured (in herbari-

um material) and outline of epidermal cells often macro-

scopically visible. Leaf shape cymbiform (boat-shaped),

falcate or cigar-shaped; plane shape linear, oblong, lanceo-

late, ovate or trullate. Leaf base sessile or cuneate. Leaf

apex truncate and hump-backed, obtuse, rounded, acumi-

Bothalia 30,1 (2000)

73

nate or acute to almost spine-tipped. Margins sometimes

ciliate. Size (1.5-)2.5-4.0(-8.0) x (0.8-)1.2-2.0(-3.0) mm.

Anatomy of the leaf

Transverse section (LM): leaves epistomatic. Adaxial

epidermis concave, villous, with unicellular, uniseriate tri-

chomes; cuticle relatively thin, 2-5 pm; epidermal cells

uniserial, relatively small (10—) 1 5— 25(— 35) x 1 0— 1 7(— 20)

pm; vacuoles large with tanniniferous substances, cell

walls thin; stomata present, with guard cells at same level,

sunken below, or raised above adjacent epidermal cells.

Abaxial epidermis convex, glabrous or sparsely hairy; cuti-

cle relatively thick ( 1 0— )20 — 40(— 70) pm; epidermal cells

relatively large, periclinal diam. of cells (20-)30-60(-65)

pm, anticlinal diam. (25-)30-75 (-105) pm (Table 2), tan-

niniferous, often with mucilaginous cell walls. Mesophyll

inversely dorsiventral (Kugler 1928); spongy parenchyma

situated adaxially and palisade parenchyma abaxially.

Main vascular bundle collateral, surrounded by parenchy-

matous bundle sheath with ample amounts of tanniniferous

substances; bundle sheath adaxially irregularly biseriate,

abaxially strengthened by sclerenchyma. Secondary vascu-

lar bundles ± 6; bundle sheaths irregular, parenchymatous

and tanniniferous. Figure 2A, B.

Adaxial (dorsal) epidermis

Cuticle

Transverse section (LM): cuticular membrane 2-5

pm thick, smooth, ridged along boundaries of guard cells

(Figure 2G), gradually thickening close to leaf margins,

equalling abaxial cuticle in thickness and sculpturing at

margins.

Surface view (LM and SEM): smooth (Figure 2C),

except in Passerina sp. nov. 1, where markings on epi-

cuticular wax are most probably caused by snow (Figure

3D, E).

Epidermal cells

Transverse section (LM): cells uniserial, irregularly

shaped,- relatively small with periclinal diam. (10-)

1 5— 25(— 35) pm, anticlinal diam. 10—1 7(— 20) pm; cell

walls thin, outer periclinal wall convex; vacuoles large,

containing tanniniferous substances (Figure 2A, F-H).

Margin formed by a few rows of conically shaped or

anticlinally elongated cells.

Surface view (LM and SEM): cells polygonal,

4-many-sided, walls usually undulate with loose, wide

u-shaped curves of shallow amplitude (Figures 2D, K;

3C), arranged in rows and straight-walled in Passerina

sp. nov. 1 (Figure 3D, E). Nodular walls observed in P.

falcifolia (Figure 2D). Vacuoles with ample tanninifer-

ous substances.

Stomatal complex

Transverse section (LM): lamina epistomatic; stoma-

ta dispersed randomly over adaxial surface, but absent

from edges of leaf margin; raised or at same level as

other epidermal cells (Figure 2E-H); dispersed in two

columns in adaxial epidermal folds, with ± 3-5 rows of

epidermal cells in between; raised, sunken or arranged in

stomatal crypts in Passerina sp. nov. 1 (Figure 3F).

Guard cell outline in all species varying between widely

obtrullate, very widely obtrullate or widely depressed

obtrullate, with angles slightly rounded; cell walls thick-

TABLE 2. — Dimensions of abaxial epidermal cells and cuticular membrane (CM) in Passerina. Measurements in pm in cross section

and surface view

74

Bothalia 30,1 (2000)

' ♦

Bothalia 30,1 (2000)

75

ened (Figure 2F, H); periclinal and anticlinal dimensions

for individual guard cells 1 0.0— 1 2.5(— 1 5.0) x ( 1 0.0—)

12.5-1 5. 0(-20.0) (irn. Cuticular membrane (Figures 2G;

3B) covering outer periclinal walls of epidermal and

guard cells, as well as poral epidermal walls of guard

cells, smooth or slightly crenate when lining the pore

(Figure 3B), contracted into a pair of ± continuous outer

stomatal ledges above guard cells, thus forming an entire

outer cavity (not divided into compartments); inner

stomatal ledges and inner cavity present. Epidermal cells

surrounding guard cells not different from other epider-

mal cells in size, shape or staining properties (Figure 2F).

Peristomatal cuticular rims conspicuous on epidermal

cells bordering guard cells (Figure 2G).

Surface view (LM and SEM): stomata anomocytic;

outline elliptic to circular, dimensions (20-)26-30(-35)

x (15-)24-30(-35) pm, circular in Passerina sp. nov. 1,

dimensions 27.5 x 27.5 pm. Epidermal cells surrounding

guard cells 3-5(6), irregularly shaped with sinuate walls

and long axis parallel to guard cells, corresponding in

orientation, size, shape and staining properties to other ±

elongated epidermal cells (Figures 2D, K; 3C); pentago-

nal to heptagonal epidermal cells in Passerina sp. nov. 1,

with walls slightly sinuate to straight, possibly nodular

(Figure 3E). Stomata raised above or at same level as

other epidermal cells in all species (Figure 21, J, L); dis-

persed in two columns in adaxial epidermal folds, with

± 3-5 rows of epidermal cells in between, sunken or

arranged in stomatal crypts in Passerina sp. nov. 1

(Figure 3D). Guard cells often conspicuously raised

(Figure 21, J). Peristomatal cuticular rims conspicuous

on epidermal cells bordering guard cells (Figures 21, L;

3A), rims also visible as 1-4 small semilunar protrusions

bordering guard cells in cuticular preparations and epi-

dermal peels (Figure 3C) (rims should not be confused

with small subsidiary cells, an interpretation which could

result in stomata being erroneously classified as paracy-

tic or cyclocytic). Outer stomatal ledges ± continuous,

present above guard cells (Figures 2I-L; 3A, C).

Stomatal poles (where guard cells meet) retuse; T-pieces

(cuticular thickenings of common walls between guard

cells) well developed (Figures 21, J, L; 3C).

Trichomes

LM and SEM: adaxial surface of leaf villous. Tri-

chomes nonglandular, unbranched, devoid of surface fea-

tures or constrictions, mostly strongly spiralled (Figure

3G, H), terete, with central lumen, covered by cuticle

(Figures 2G; 31). Hair bases with pore, poral rim some-

what thickened (Figures 2C; 3C, G); hair base cells most-

ly 4-6 and slightly radially elongated (Figure 3C, G).

Trichomes bordering leaf margin conspicuous in

P. burchellii , P. paludosa and P. pendula (Figure 31, J).

Trichome foot scarcely modified, inserted between epi-

dermal cells (Figure 31), usually straight, but with indi-

vidual trichomes strongly spiralled (Figure 3J) in P. pen-

dula (brown in dried material).

Wettability and the possible absorption of water by

the laminar epidermal hairs in Passerina were assessed

by means of laboratory tests. We found that water had

formed a film over the felty layer of hair at the leaf base,

whereas the adaxial surface had remained dry. A treat-

ment with 0.5% aqueous safranin revealed that only the

exposed parts of the spiralised hairs in the felty indu-

mentum at the leaf bases stained pinkish. Although the

longer hairs at the leaf margins were stained, those on the

rest of the adaxial surface remained unstained.

Abaxial (ventral) epidermis

Trichomes

Abaxial surfaces of bracts and young leaves in P.

comosa , P. sp. nov. 3 and P. sp. nov. 4 tomentose to

sparsely hairy (Figure 4B), older leaves often glabrous.

Description of trichomes as described under adaxial epi-

dermis.

Epidermal cells

Transverse section (LM and TEM) (Figures 2A, B, E;

3K-L): epidermis uniserial. Stomata absent. Epidermal

cells more or less oblong in outline; outer periclinal walls

straight or convex, inner periclinal walls straight, convex

or bulging towards mesophyll, often mucilaginous and

then superficially resembling a multiple epidermis; peri-

clinal diam. of cells (20-)30-60(-65) pm, anticlinal

dianr. (25— )30— 75(— 1 05) pm (Table 2). Mucilaginous

cell walls increasing progressively from leaf margin to

midrib (Figure 2B), affecting mainly inner periclinal but

also anticlinal cell walls (Figure 3K, L); mucilage with a

layered appearance (Figures 2E; 3K), eventually occupy-

ing about two-thirds of epidermal cell and separated

from cytoplasm by innermost cellulose layer of inner

periclinal cell wall (Figure 3L). Cytoplasm compressed

by mucilage, remaining as a thin layer appressed to large,

usually tanniniferous vacuole. Anticlinal layer of inner

periclinal cell wall often plicate but gradually straighten-

ing and often disappearing as mucilagination increases,

eventually breaking under pressure of accumulating

FIGURE 2. — LM photographs and SEM micrographs of epidermis of inversely ericoid leaf in Passerina. A, P. falcifolia , Bredenkamp 91 7, ad-

and abaxial epidermis with mucilage accumulating abaxially; B, P. galpinii, Bredenkamp 946, mucilaginous abaxial epidermal cells; C,

P. filiformis, Bredenkamp 1016, smooth adaxial cuticle, stomata and poral rims of hair bases; D, P. falcifolia, Bredenkamp 915, adaxial

epidermal walls undulate, nodular; E, P. ericoides, Taylor 4042, stomata at different levels in relation to adaxial epidermis; F, P. comosa,

Bredenkamp 1034, PAS staining of guard cell walls and surrounding epidermal cells, showing width; G, P pendula, Bredenkamp 909, t/s

adaxial epidermis stained with Sudan Black B, showing cuticular membrane; H, P. pendula, Bredenkamp 909, raised stomata stained with

toluidine blue; I, P. paleacea, Bredenkamp 961, with peristomatal rim, raised epidermal and poral walls of guard cells, conspicuous outer

stomatal ledges; J, P. galpinii, Bredenkamp 946, with distinct outer stomatal ledge aperture; K, P. filiformis, Bredenkamp 1016, and L,

P. pendula, Bredenkamp 909, with T-pieces at stomatal poles. Abbreviations: ad, adaxial epidermis; ab, abaxial epidermis; CM, cuticular

membrane; e, epidermal cell; egc, epidermal wall of guard cell; es, epidermal cell surrounding guard cell; gc, guard cell; ic, inner cavity;

il, inner stomatal ledge; 1, lumen of trichome; m, mucilage; oc, outer stomatal cavity; ol, outer stomatal ledge; ola, outer stomatal ledge

aperture; p, pore; pgc, poral wall of guard cell; pr, trichome poral rim; ps, peristomatal rim; s, stomata; sc, stomatal crypt; t, trichome; T,

T-piece at stomatal pole. Scale bars: A, B, D, E, F, K, 100 pm; C, G, H, I, J, L, 10 pm.

76

Bothalia 30,1 (2000)

Bothalia 30,1 (2000)

77

TABLE 3. — Abaxial epidermal characters in Passerina

Abax-

ial

hair

pre-

sent

Epicuticular

remains of epidermal cells and adjacent mesophyll

(Figure 2 A) (Bredenkamp & Van Wyk 1999).

Surface view (SEM micrographs and cuticular prepara-

tions): shape pentagonal to heptagonal, cells mostly isodia-

metric or transversely oblong in P. glomerata, P. ericoides

(Figure 4D, E) and P. obtusifolia (Figure 4G, H), but oblong

in P. burchellii (Figure 4J, K); cells mostly slightly oblong

or oblong in all other species of Passerina (Figure 5; Table

2). Arrangement random in P. glomerata, P. ericoides , P.

obtusifolia and P. burchellii (Figure 4D-K), in rows in all

other species of Passerina (Figure 5; Table 3).

Cuticle

Transverse section (LM): epicuticular wax absent

owing to chemical treatment during fixation, embedding

and staining. Cuticular membrane (CM) well devel-

oped, ( 1 0— )20— 30(— 70) pm thick (Table 2); cuticle prop-

er delineated by narrow, lightly stained outer zone and

cuticular layer by wider, darker stained zone; cuticular

pegs present, formed by layer projecting into grooves

between anticlinal walls of adjacent epidermal cells.

Outer periclinal cell walls not staining with Sudan

Black B (Figures 41; 5C, I, L).

TEM: cuticle structure corresponding to the cuticular

structural type 3, described by Holloway (1982). Cuticle

proper and CM not distinguishable. Cuticular membrane

(Figure 4A) comprising a wide, mainly amorphous outer

zone and narrow faintly reticulate inner zone; osmio-

philic granules aligned on border of clearly defined cell

wall; cuticular pegs with unknown (possibly pectina-

ceous) substance (stained light grey) between cell wall

and peg, forming part of middle lamella.

FIGURE 3. — LM photographs and SEM micrographs in Passerina. A-F, structure of stomatal complex. A-C, P. rigida , Bredenkamp 1013, Ward

7211\ A, surface view of stomata showing peristomatal rims, raised guard cells and pronounced outer stomatal ledges; B, t/s adaxial epider-

mis stained with Sudan Black B, with crenate surface of cuticular membrane lining poral walls of guard cells; C, epidermal maceration

stained with safranin, showing structure of epidermal cells surrounding guard cells, peristomatal rims. D-F, Passerina sp. nov. 1,

Bredenkamp 1046: D, sunken stomata in cavity of cymbiform leaf; E, epidermal maceration stained with safranin, with structure of epider-

mal cells and sunken stomata; F, t/s leaf, with raised stomata as well as stomatal crypts. G-J, structure of trichomes. G, P. rubra , Bredenkamp

905 , with poral rims in relation to adaxial epidermal cells. H, P . falcifolia, Bredenkamp 915, with unicellular, long, spiralised, pointed tri-

chomes; I, P. paludosa , Bredenkamp 1035, with trichome foot and conspicuous lumen; J, P. pendula, Bredenkamp 909, trichomes strongly

spiralised. K, L, TEM micrographs of abaxial leaf epidermal cells of P . falcifolia, Bredenkamp 917, in cross section: K, mucilage accumu-

lated between innermost and outermost cellulose layers of inner periclinal cell wall; L, innermost cellulose layer of inner periclinal cell wall.

Abbreviations: aw, anticlinal cell wall; cy, cytoplasm; iip, innermost layer of inner periclinal cell wall; oip, outer layer of inner periclinal cell

wall; m, mucilage; op, outer periclinal cell wall; v, vacuole. Scale bars: K, L 5 pm; A, B, H, 10 pm; C-F, G, I, J, 100 pm.

78

Bothalia 30,1 (2000)

Bothalia 30, 1 (2000)

79

Cuticular ornamentation

In transections and surface view of leaves, LM and

SEM studies showed that two groups of species, hence-

forth called Groups A, Intermediate and B (Table 3), can

be distinguished on the basis of the arrangement and

shape of epidermal cells as well as cuticular ornamenta-

tion.

Group A

Epidermal cells mostly isodiametric or transversely

oblong in surface view; arranged randomly; cuticle most-

ly papillate; outer per iclinal walls of cells convex in all

species. Cuticular membrane (CM) smooth in P. eri-

coides and P. glomerata (Figures 4D-F; 6C); papillate in

P. obtusifolia, with one dome per cell, situated ± central-

ly on outer periclinal wall of pentagonal or heptagonal

cells (Figures 4G-I; 6E); with several domes per cell in

P. burchellii (Figures 4J-L; 6F).

Group B

Epidermal cells mostly oblong in surface view,

arranged in rows; concavities (depressions in centre

region of cell) and convexities (roundish cells forming a

low dome) more or less alternating (Figure 5G, J); cuti-

cle with ridges at junction of epidermal cell walls most-

ly conspicuously raised, exhibiting a definite striate pat-

tern (Figure 5D, G, J), otherwise ± plane.

Cuticular membrane pronounced at junctions of epi-

dermal cell walls and grooved between anticlinal walls

of adjacent cells (Figure 51), more or less smooth in P.

vulgaris , P.filifonnis, P.falcifolia, P. pendula, P. rigida,

and P. galpinii , except in Passerina sp. nov. 1, in which

the presence of snow, at the time of collecting, seemed to

have caused markings on the cuticular wax (Figure 5D,

E). Small globular papillae visible between cuticular

ridges in Passerina sp. nov. 1 (Figure 5D-E), P. rubra ,

P. paleacea and P. paludosa (Figure 5J-L).

Intermediate

Epidermal cells arranged in rows but CM less pro-

nounced at junctions of epidermal cell walls and cuticu-

lar ridges less conspicuous, were recorded in P. comosa

(Figures 4B; 6A, B), P. drakensbergensis (Figure 5A, B),

P. montana, P. sp. nov. 3 and P. sp. nov. 4. CM smooth

or with small globular papillae in P. montana and P. sp.

nov. 4; domed with a ‘molar’ -like crown in P. comosa

(Figure 4B, C), with several domes per cell in P. sp. nov.

3 and with 9 or 10 globular papillae per cell in P. drak-

ensbergensis (Figures 5A-C; 6G).

Epicuticular waxes

Soft waxes present, coating entire abaxial surface:

wax protruding through amorphous layer of CM in a

variety of configurations: droplets conspicuous in P.

comosa , P. ericoides and P. burchellii (Figure 6A, D, F);

droplets and small round protrusions forming flat, shape-

less lumps in P. paleacea (Figure 6L). Crystalloids: wax

platelets and plates present or absent (Table 3); thin wax

platelets, with margins entire or non-entire, flaking from

wax surface in P. comosa and P. rigida (Figure 6A, J)

and changing to plates as margins become distinctly

edged. Upright plates separating from surrounding wax

in P.filiformis (Figure 6H). Platelets and plates varying

from sparse to abundant; platelets ± square to irregularly

shaped, plates ± square to oblong and usually arranged

perpendicular to cell rows.

The authenticity of epicuticular wax droplets and

small round protrusions, observed in P. ericoides,

P. obtusifolia and P. paleacea (Figure 7), was verified by

washing leaves in chloroform for one minute and com-

paring them to unwashed specimens under SEM.

Epicuticular wax droplets were clearly discernible in

unwashed P. paleacea (Figure 7A), while small pores

appeared in the cleaned, de-waxed cuticle after washing

(Figure 7B-E). Similar pores were also present in P. eri-

coides (Figure 7F). No pores were present in the papil-

late CM of P. obtusifolia, but the corroded apices of the

papillae clearly showed an accumulation of epicuticular

waxes at these points (Figure 7G-I).

DISCUSSION

Adaxial epidermis

Plants of high mountains in the tropics usually have

straight to curved anticlinal epidermal cell walls, the per-

centage of species with undulated walls increasing as

altitude decreases (Wilkinson 1979). The straight-walled

arrangement of the cells in Passerina sp. nov. 1 (Figure

3D, E), a high-altitude montane species, seems to com-

ply with this pattern.

Stomata! complex

In all but one species of Passerina the stomata are usu-

ally raised or at the same level as other epidermal cells

(Figure 2E, G, H), indicating that this character is of lim-

FIGURE 4. — A, TEM micrograph of cuticular membrane in Passerina paleacea, Bredenkamp 961 , with wide, amorphous outside zone, narrow

faintly reticulate inner zone, osmiophilic granules at border of cell wall and cuticular peg. B-L, LM photographs and SEM micrographs

of abaxial leaf epidermis in Passerina. Epidermal macerations stained with safranin and t/s of epidermis stained with Sudan Black B. B,

C, P. comosa, MacDonald 2125, Andraea 1288: B, trichomes present; C, CM domed, with ‘molar-like’ crown to each dome. D-F, P. eri-

coides, Bredenkamp 956, 962, Taylor 4042: D, CM smooth, epidermal cells randomly arranged, ± isodiametric, outer periclinal cell walls

convex; E, cells randomly arranged, ± isodiametric; F, convex outer periclinal walls and smooth CM. G-I, P. obtusifolia, Bredenkamp

1034: G, CM with one dome per cell; H, epidermal cells randomly arranged, transversely oblong with one dome per cell; I, convex outer

periclinal cell wall and CM with one dome per cell. J-L, P. burchellii. Bolus 687 , Bredenkamp 1545 ; J, CM with several domes per cell;

K, randomly arranged cells, transversely oblong with rounded angles, several domes per cell; L, t/s epidermis in polarised light showing

CM with several domes per cell. Abbreviations: CM, cuticular membrane; cw, outer periclinal cell wall; d, dome; me, molar-like crown;

og, osmiophilic granules; oz, amorphous outside zone; pe, cuticular peg; rz, narrow faintly reticulate inner zone; t, trichome. Scale bars:

A, 5 pm; C, F, I, 10 pm; B, D, E, G, H, J-L, 100 pm.

80

Bothalia 30,1 (2000)

Bothalia 30,1 (2000)

81

ited taxonomic value at species level, except in Passerina

sp. nov. 1, which has stomatal crypts or sunken stomata.

Classification of the stomatal complex into stomatal types

is often a problem owing to the subtle distinction of sub-

sidiary cells (Wilkinson 1979; Van Wyk et al. 1982).

Patel (1978) considers subsidiary cells as morphologi-

cally and physiologically different from other epidermal

cells and proposes a number of criteria to distinguish

subsidiary cells in mature epidermis. Of these criteria we

used the following in the distinction of subsidiary cells:

size, shape, contents and position of cells. We found that

the cells adjacent to the guard cells did not differ from

other epidermal cells, except that they might be raised or

sunken (Figures 2K; 3C). Furthermore, when stained

with PAS, periclinal walls of subsidiary cells should be

lightly stained compared with other epidermal cells,

owing to less carbohydrates in these cell walls according

to Patel (1978). In Passerina the periclinal walls of the

cells adjacent to the guard cells stained homogenously

with other cells in the stomatal complex (Figure 2F) and

the anticlinal walls are not comparatively thinner than

those of other epidermal cells, thus the cells adjacent to

the guard cells cannot be considered subsidiary cells

(Figure 2F, H). Stained with Sudan Black B, the contents

of the cells surrounding the guard cells do not differ from

those of other epidermal cells and no lipid bodies are pre-

sent (Figure 2G).

We therefore conclude that the epidermal cells sur-

rounding the guard cells in Passerina are not differentia-

ted as subsidiary cells and we classify the stomatal appa-

ratus in Passerina as anomocytic. This corresponds to the

prevailing state in the Thymelaeaceae (Solereder 1908;

Metcalfe & Chalk 1979). However, although we prefer to

regard the epidermal cells surrounding the guard cells as

similar to other epidermal cells, the presence of conspic-

uous peristomatal cuticular rims on the outer periclinal

cell walls of epidermal cells around the guard cells may

be used in support of a view that these cells are subsidiary

cells. The stomatal apparatus could then be classified as

staurocytic (Wilkinson 1979) or anomotetracytic (Dilcher

1974). As the number of epidermal cells surrounding the

guard cells varies from 3-5(6), it would seem appropriate

to classify the stomatal apparatus as anomostaurocytic

(Van Wyk et al. 1982).

Trichomes

Passerina leaves are often cymbiform with spiralised

trichomes densely arranged in the concave ventral

space. This indumentum is likely to play an important

role in the water relations of the plant. Water droplets

precipitating from the atmosphere, or running down

from leaves directly above, would accumulate in the

concave leaf space. Droplets would be repelled by the

hydrophobic cuticle of the trichomes and owing to cohe-

sion forces cause a moisture layer in the upper part of

the dense trichomes. One may speculate that water

vapour escaping through the stomata would not be

drawn outwards by capillary forces because of the

water-repelling nature of the cuticle surrounding the tri-

chomes, thus retaining a high concentration of moisture

in the vicinty of the stomata. The overall high concen-

tration of water vapour over the adaxial surface of the

leaf is likely to decrease the transpiration rate.

Laboratory tests to assess the wettability and the possi-

ble absorption of water by the laminar epidermal hairs in

Passerina, suggest that the wettability of the spiralised

hairs is quite low and that absorption of water by these

trichomes is highly improbable. However, our sugges-

tion of an overall high concentration of water in the

adaxial cavity of the leaf, which serves to decrease the

transpiration rate, is supported by these tests.

Cuticular ornamentation

Cuticular thickness may be affected by light, temper-

ature, soil, atmospheric moisture and altitude (Wilkinson

1979). In Passerina, with many species adapted to the

Cape Mediterranean climate, all members have a rela-

tively thick cuticle, but it was the thickest in P. comosa,

P. glomerata, P. burchellii, P. galpinii and P. paleacea

(Table 2). The first two species grow in the northwestern

parts of the Western Cape and on the mountains in and

around the Little Karoo (= Karoo Mountain Centre sensu

Weimarck 1941), areas with high light intensity, high

temperature and low atmospheric moisture. P. burchellii,

growing on high mountains at Villiersdorp and

Genadendal, is exposed to high light intensity as well as

high and low critical temperatures. P. galpinii grows on

calcrete and P. paleacea is exposed to salt spray and

wind. In P. drakensbergensis, P. falcifolia, P. paludosa

and P. sp. nov. 1, the thickness of the CM is ± 20 pm. Of

these species, P. falcifolia, from the mountains between

George and Uitenhage, and P. drakensbergensis, from

high altitudes in the Bergville District of KwaZulu-

Natal, are exposed to relatively high atmospheric mois-

ture. However, it is difficult to speculate on the function-

al significance of the relatively thin cuticles in P. palu-

dosa, from salt marshes in the Cape Peninsula, and P. sp.

nov. 1, a species from Waboomberg, one of the highest

points in the Western Cape and often covered by snow in

winter.

FIGURE 5.— Abaxial leaf epidermis and structure of CM in Passerina. Epidermal macerations stained with safranin and t/s of epidermis stained

with Sudan Black B. A-C, P. drakensbergensis, Bredenkamp 1018, 1019'. A, cells arranged in rows with 9 or 10 globular papillae per cell;

B, inner surface facing upwards, cells oblong in shape with 9 or 10 papillae per cell; C, CM layered, with cuticular layer and cuticle prop-

er, also globular papillae. D-F, Passerina sp. nov. 1, Bredenkamp 1044, 1046: D, several domes per cell, CM irregularly marked by ice

crystals; E, cells arranged in rows, oblong in shape with CM irregularly marked by ice crystals; F, geometrical plates, flat or slightly raised.

G-I, P. rigida, Bredenkamp 1013, Ward 7211: G, cells arranged in rows, plates abundant; H, cells arranged in rows, isodiametric to slight-

ly oblong; I, CM pronounced at junctions of epidermal cell walls, grooved in midline of joining walls, concavities and convexities not con-

spicuous. J-L, P. paludosa, Bredenkamp 1035, Thoday 100: J, cells arranged in rows, CM pronounced at junctions of epidermal cell walls,

grooved in midline of joining walls, concavities and convexities conspicuous; K, cells arranged in rows, cells oblong; L, CM pronounced

at junctions of epidermal cell walls, grooved in midline of joining walls. Abbreviations: cl, cuticular layer; co, concavity; cp, cuticle prop-

er; cw, outer periclinal cell wall; cx, convexity; gr, groove in CM; pa, papillae; pe, cuticular peg; pi, plates. Scale bars: A, D, E, G, H, J-L,

100 pm; B, C, F, I, 10 pm.

82

Bothalia 30,1 (2000)

FIGURE 6. — SEM micrographs of abaxial leaf surfaces, the CM and epicuticular waxes in Passerina. A, B, P. comosa, MacDonald 2125: A,

droplets present in epicuticular wax, platelets flaking from smooth wax coating; B, wax platelets flaking from smooth wax coating, plates

present C, P. glomerata, Bredenkamp 973 , outer periclinal wall convex, plates scarce, square to oblong, raised 30°-90°; D, P. ericoides ,

Bredcnkamp 956, droplets present in epicuticular wax; E, P. obtusifolia , Bredenkamp 929 , smooth wax coating also covering domes. F, G,

P burchellii, Stokoe 2542: F, droplets at apices of domes; G, small round protrusions at apices of papillae. H, P. filiformis, Bredenkamp

K)16, upright plates separate from surrounding wax, orientated at an angle to cell rows; I, P. pendula , Bredenkamp 908 , plates frequent,

perpendicular to cell rows, square to oblong, flat or raised; J, P. rigida, Bredenkamp 1013 , platelets and plates ; K, L, P. paleacea,

Bredenkamp 961 , wax droplets, protrusions and flat shapeless lumps contributing towards soft wax coating or smooth layer. Abbreviations:

dr, droplets in epicuticular wax; pi, plates; pr, small round protrusions of epicuticular wax; ps, platelets. Scale bars: A-K, 10 pm; L, 1 pm.

Bothalia 30,1 (2000)

83

FIGURE 7. — SEM micrographs of abaxial leaf surfaces of Passerina washed in chloroform for one minute, compared to unwashed specimens.

A-E, P. paleacea, Bredenkamp 96P. A, unwashed leaf showing droplets in smooth wax coating; B, low magnification of washed leaf,

showing CM devoid of epicuticular wax; C-E, higher magnifications showing pores in CM. F, P. ericoides, Bredenkamp 956, washed

specimen showing pores in CM. G-I, P. obtusifolia, Bredenkamp 929: G, unwashed specimen; H, I, washed specimens showing corroded

apices of papillae. Abbreviations: dr, droplets in epicuticular wax; p, pore; pa, papillae. Scale bars: A, C-F, I, 1 pm; B, G, H, 100 pm.

Haberlandt (1914), following a study of plants in tropical

rain forests, considered the function of papillose epidermal

cells as concentrating limited light by acting as lenses.

Bredenkamp & Van Wyk (1999) speculate that, in

Passerina, the convex outer periclinal epidermal cell

wall may well focus light rays onto the mesophyll,

whereas large vacuoles filled with phenols and the

mucilage formed by the cellulose slimes (inner periclinal

walls) protect the mesophyll from potentially dangerous

UV-B radiation. According to Wilkinson (1979) the pres-

ence and prominence of papillae are diagnostically unre-

liable because they vary with the climate or distribution

of the species; only morphologically distinct types can be

used for diagnostic purposes. However, distinct epider-

mal cell papillae characterise P. comosa, P. obtusifolia

P. burchellii, P. drakensbergensis and P. sp. nov. 2

(Figures 4B-C, G-L; 5A-C). The presence of these

papillae could have been induced by the high light inten-

sity of the areas in which these plants grow.

Epicuticular waxes

In their study of the epicuticular waxes in the families

of the Dilleniidae and Rosidae, Ditsch & Barthlott

(1997) documented the numbers of genera, species and

hybrids in which different wax types occur, without iden-

tifying the various taxa. The epicuticular waxes of 12

genera, 31 species and two hybrids were studied in the

Thymelaeaceae. Of these, nine genera and 26 species

have wax flakes, one species has angled platelets and

four genera and five species have no crystalloids. Our

84

Bothalia 30,1 (2000)

observations indicate that the simple plate-type waxes

found in Passerina correspond well to those described by

Ditsch & Barthlott (1997) in the Thymelaeaceae. Of the

17 species in Passerina , two have wax flakes, eight have

platelets or angled plates and seven are devoid of crys-

talloids (Figure 6, Table 3).

The mechanism of wax extrusion through the cuticle is

highly controversial (Baker 1974; Jeffree et al. 1975;

Hallam 1982). Baker (1982) discusses the extrusion of

wax by means of ‘pores and channels, the liquid extrusion

theory, polymerization theory and the crystallization the-

ory’. Hallam (1982) proposes that wax or wax precursors

in their protein or glycoprotein ‘shells’ move through the

cuticle and burst on the surface, liberating the wax from

the ‘package’; on crystallization, the protein coats stick to

the surface as the wax crystals develop.

Our results indicate small pores in the cleaned, de-

waxed cuticle of P. paleacea and P. ericoides (Figure

7B-F), after washing leaves in chloroform. Both Baker

(1982) and Hallam (1982) are convinced that detailed

investigations by many investigators have failed to con-

firm the presence of pores or microchannels in certain

plant cuticles and that pores have not been shown to con-

nect with the plasmalemma of the epidermal cytoplasm

below. Although the presence of pores has been con-

firmed by our study, further research on the ultrastructure

of the CM in Passerina could be most informative.

Freeman et al. (1979), working on Citrus, found

amorphous wax layers on immature leaves and fruit,

with small protrusions and isolated regions of upright

platelets developing, eventually followed by cracks and

irregular plates. Similarly in Passerina, wax droplets,

protrusions and flat, shapeless lumps contribute towards

a soft wax coating or a smooth layer. Species of Pas-

serina with soft wax coatings, without platelets or plates,

are summarised in Table 3. In P. comosa, P. filiformis

and P. rigida (Figure 6B, H, J) platelets and plates are

formed as a result of cracks developing on the outer wax

surface, crystallising into irregularly shaped flakes,

which gradually become square or oblong with ‘entire’

or ‘non-entire’ margins, often becoming distinctly edged.

In P. filiformis (Figure 6H) upright plates separate from

the surrounding wax, orientating themselves at an angle

to the cell rows, eventually resulting in most plates being

arranged more or less perpendicularly to the cell rows.

Wax type, as well as the presence or absence of plates

and platelets, is apparently genetically determined

(Baker 1982). For example, P. ericoides, P. rigida and P.

paleacea (Figure 6D, J, K) all grow along the sea shore,

where they are subjected to wind, salt spray and high

light intensity, and yet, P. ericoides and P. paleacea have

coverings of soft waxes only, whereas platelets and

plates are abundantly present in P. rigida. However, in

plate waxes the number of platelets and plates, size, con-

figuration and distribution of the surface wax structures

can be considered as environmentally induced (Baker

1974, 1982).

Functions of epicuticular waxes

Possible functions of epicuticular waxes are discussed

by Jeffree (1986). In Passerina, large areas of the abaxi-

al epidermis are exposed to the atmosphere because the

inverse-ericoid leaves are usually closely appressed to

the stem. In response to the warm, dry summers of the

Mediterranean climate of the Cape, it is proposed that the

CM, including the abaxial epicuticular waxes, has a

water-proofing function, protecting the leaves against

desiccation and limiting transpiration to the adaxial epi-

dermis only. As the leaves are decussately arranged, the

water-repelling function of the waxes would cause

droplets of water to run off the abaxial epidermis, into

the concave, hairy adaxial surface of the lower leaf,

resulting in a decreased transpiration rate owing to the

higher adaxial water concentration. According to Jeffree

(1986) the wettability of the plant surface is determined

by its microroughness. The presence of crystalloid

platelets and plates, and especially their arrangement

perpendicular to cell rows, may facilitate the retention of

moisture.

Systematic value

Epicuticular waxes have been proven taxonomically

valuable, among others in the study of the Centro-

spermae (Engel & Barthlott 1988), Dilleniidae and

Rosidae, including the Thymelaeaceae (Ditsch &

Barthlott 1997), at sectional level in Eucalyptus L’Her.

(Hallam & Chambers 1970) and at species level in

Hordeum L. (Baum et al. 1989). In Passerina the pres-

ence or absence of crystalloid platelets or plates com-

bined with characteristics of the CM and the outer peri-

clinal cell walls of the abaxial epidermis, makes it possi-

ble to distinguish between two groups in the genus. This

distinction is species-specific for most of the 17 species

examined (Table 3).

Ecological aspects of leaf epidermis

The structure and function of the epidermis should be

considered in context with gross leaf morphology and

arrangement. Leaf arrangement is of vital importance to

the physiology of the plant. The epidermis serves as an

envelope, physically protecting the mesophyll, the

largest part of the abaxial epidermis forming a multi-

functional barrier to the environment. The thin adaxial

epidermis is concealed in the groove of the cymbiform

leaf in most cases; it is almost covered by dense, long,

spiralised uniserial trichomes and contains the stomata,

which are often raised. This arrangement is likely to

reduce the rate of transpiration, especially if moisture

can be retained by the indumentum. The abaxial epider-

mis is probably multifunctional. The whole of the CM

has a waterproofing function and the epicuticular waxes

also have a water-repelling function. At the same time

the CM may play a major part in focusing light rays onto

the palisade parenchyma. Large tanniniferous vacuoles

may play a role in the possible absorption of UV-B radia-

tion, and mucilage formed by the cellulose slimes (inner

periclinal walls) possibly protects the mesophyll from

desiccation (Bredenkamp & Van Wyk 1999).

The expansion and inrolling of the leaf margins in

Passerina, as a result of changing turgor pressure in the

epidermal cells, were described by Thoday (1921). He

regards the main mechanism involved as the co-ordina-

Bothalia 30,1 (2000)

85

tion between the turgor pressure and the difference in

size and thickness of cell walls of the ad- and abaxial epi-

dermis, whereas the plicate anticlinal cell walls of the

abaxial epidermis protect the cells against bending stress.

Stomata (or at least the indumentum) are exposed when

the leaf margins expand and are protected in a villous

groove when the leaf margins are rolled inwards, thus

regulating the rate of transpiration.

CONCLUSIONS

Leaf shape and structure in Thymelaeaceae exhibit a

transformation series from mainly dorsiventral, the pre-

vailing family feature, to isobilateral or centric in

Diarthron Turcz., Pimelea Banks & Soland. and Thyme-

laea Juss. (Metcalfe & Chalk 1950). All the mentioned

states are present in Lachnaea and Cryptadenia (Beyers

1992) and, as the most advanced state, inversely dor-

siventral leaves in Passerina. A transformation series can

also be illustrated by the presence of amphistomatic,

hypostomatic and epistomatic leaves in the Thyme-

laeaceae (Metcalfe & Chalk 1950), the epistomatic state

in Passerina considered to be the most advanced (the

collateral vascular bundles of the leaves, with xylem

arranged adaxially and phloem abaxially, rule out the

possibility of resupination of the leaves).

The most pronounced epidermal characters of the

Thymelaeaceae are anomocytic stomata (Metcalfe &

Chalk 1950), unicellular trichomes and mucilagination

of epidermal cells. In the present study the presence or

absence, distribution of or changes in the above-men-

tioned structures, were used as distinguishing characters

at both generic and species levels. Mucilagination of epi-

dermal cells is often found both ad- and abaxially in the

leaves of Thymelaeaceae. In Passerina, mucilagination

takes place in the abaxial epidermis only. At species level

the sunken stomata and stomatal crypts of Passerina sp.

nov. 1 are used in the delineation of the new taxon and

P. comosa is distinguished by the presence of unicellular

trichomes on the abaxial surface of the leaves.

On the basis of abaxial cuticular characters, it has

been possible to distinguish two groups of species in the

genus. Group A comprises P. burchellii Thoday, P.

comosa C.H. Wright, P. ericoides L., P. glomerata

Thunb. and P. obtusifolia Thoday. Group B comprises P.

drakensbergensis Hilliard & B.L.Burtt, P. falcifolia

C.H. Wright. P.filiformis L., P. galpinii C.H. Wright, P.

montana Thoday, P. paleacea Wikstr., P. paludosa

Thoday, P. pendula Eckl. & Zeyh., P. rigida Wikstr., P.

rubra C.H.Wright, P. vulgaris Thoday, P. sp. nov. 1, P.

sp. nov. 2, P. sp. nov. 3 and P. sp. nov. 4. Certain species

in each of the two groups seem to be naturally allied.

Distribution patterns of P. obtusifolia and P. glomerata

coincide at Worcester and transitional types can be clear-

ly distinguished. Transitional types are similarly present

in P.fdiformis and P. vulgaris in the Cape Peninsula and

in P.fdiformis and P. falcifolia near Knysna.

Hence it can be concluded that the conspicuous dif-

ferences as well as the concise characters of the ad- and

abaxial epidermis, critically described and discussed in

this paper, can be used as taxonomic tools at the family,

genus and species levels. Furthermore, the leaf epidermis

in Passerina is probably most valuable to the plant in

terms of ecological adaptation, considering the wide dis-

tribution of the genus in southern Africa as well as the

accompanying geographical and climatic variation. The

gross leaf morphology and the ad- and abaxial epidermal

characters have been most useful in the interpretation of

the possible functioning of the leaves and are of vital

importance in the survival strategies of the plant.

ACKNOWLEDGEMENTS

The authors wish to thank Mmes H. du Plessis and C.

Steyn and Dr E. Steyn for assistance with the LM, Mrs

A. Romanowski for developing and printing many excel-

lent photographs and Prof. J. Coetzee and Mr C. van der

Merwe, both of the University of Pretoria, for assistance

with the SEM and TEM.

REFERENCES

ALVIN, K.L. 1987. Leaf anatomy of Androstachys johnsonii Prain and

its functional significance. Annals of Botany 59: 579-591.

AMELUNXEN. F„ MORGENROTH. K. & PICKSAK, T. 1967. Unter-

suchungen an der Epidermis mit dem Stereoscan-Elektronen-

mikroskop. Zeitschrift fiir Pflanzenphysiologie 57: 79-95.

BAKER, E.A. 1974. The influence of environment on leaf wax devel-

opment in Brassica oleracea var. gemmifera. New Phytologist

73: 955-966.

BAKER, E.A. 1982. Chemistry and morphology of plant epicuticular

waxes. In D.F. Cutler, K.L. Alvin & C.E. Price, The plant cuti-

cle: 139-165. Academic Press, London.

BARTHLOTT, W„ NEINHUIS, C„ CUTLER, D„ DITSCH, F„ MEU-

SEL, I., THEISEN, 1. & WILHELMI, H. 1998. Classification

and terminology of plant epicuticular waxes. Botanical Journal

of the Linnean Society 126: 237-260.

BAUM, B R., TULLOCH, A.P. & BAILEY, G.L. 1989. Epicuticular

waxes of the genus Hordeum: a survey of their chemical com-

position and ultrastructure. Canadian Journal of Botany 67:

3219-3226.

BEAUMONT, A.J., EDWARDS, T.J. & VAN STADEN, J. 1994. Leaf

and involucral bract characters of systematic use in Gnidia

(Thymelaeaceae). Proceedings of the Electron Microscopy Society

of South A frica 24: 47.

BEYERS, J.B.P. 1992. The generic delimitation of Lachnaea and

Cryptadenia (Thymelaeaceae). M.Sc. thesis, University of Stellen-

bosch, Stellenbosch.

BOND, P. & GOLDBLATT, P. 1984. Plants of the Cape flora: a de-

scriptive catalogue. Journal of South African Botany 13: 1-455.

BREDENKAMP, C L. & VAN WYK, A.E. 1999. Structure of mucila-

ginous epidermal cell walls in Passerina (Thymelaeaceae).

Botanical Journal of the Linnean Society 129: 223-238.

DILCHER, D.L. 1974. Approaches to the identificaton of angiosperm

leaf remains. Botanical Review 40: 1-157.

DITSCH, F. & BARTHLOTT, W. 1997. Mikromorphologie der Epi-

cuticularwachse und das System der Dileniidae und Rosidae.

Akademie der Wissenschaften und der Literatur, Mainz.

ENGEL, T.E. & BARTHLOTT, W. 1988. Micromorphology of epicu-

ticular waxes in Centrosperms. Plant Systematics and Evolution

161: 71-85.

FEDER, N. & O’BRIEN, TP 1968. Plant microtechnique: some prin-

ciples and new methods. American Journal of Botany 55: 1 23—

142.

FREEMAN, B„ ALBRIGO, L.G. & BIGGS, R.H. 1979. Ultrastructure

and chemistry of cuticular waxes of developing Citrus leaves

and fruits. Journal of the American Society for Horticultural

Science 104: 801-808.

GILG, E. 1894. Studien iiber die Verwandtschaftsverhaltnisse der

Thymelaeales und iiber die ‘anatomische Methode’. Botanische

Jahrbiicher 18: 489-574.

GOLDBLATT, P & MANNING, J.C. in press. Cape plants: conspec-

tus of the Cape flora of South Africa. Strelitzia.

86

Bothalia 30,1 (2000)

HABERLANDT, G. 1914. Physiological plant anatomy. Reprint edn

1965. Today & Tomorrow’s Book Agency, New Delhi.

HALLAM, N.D. 1982. Fine structure of the leaf cuticle and the origin

of leaf waxes. In D.F. Cutler, K.L. Alvin & C.E. Price, The plant

cuticle: 197-214. Academic Press, London.

HALLAM, N.D. & CHAMBERS, T.C. 1970. The leaf waxes of the

genus Eucalyptus L’Heritier. Australian Journal of Botany 18:

335-386.

HOLLOWAY, P.J. 1982. Structure and histochemistry of plant cuticu-

lar membranes: an overview. In D.F. Cutler, K.L. Alvin & C.E.

Price, The plant cuticle : 1-32. Academic Press, London.

JEFFREE, C.E. 1986. The cuticle, epicuticular waxes and trichomes of

plants, with reference to their structure, functions and evolu-

tion. In B. Juniper & R. Southwood, Insects and the plant sur-

face: 23-64. Edward Arnold, London.

JEFFREE, C.E., BAKER, E.A. & HOLLOWAY, P.J. 1975. Ultra-

structure and recrystallization of plant epicuticular waxes. New

Phytologist 75: 539-549.

KARNOVSKY, M.J. 1965. A formaldehyde-glutaraldehyde fixative of

high osmolality for use in electron microscopy. Journal of Cell

Biology 27: 137a.

KIGER, R.W. 1971. Epidermal and cuticular mounts of plant material

obtained by maceration. Stain Technology 46: 71-75.

KUGLER, H. 1928. fiber inverse-dorsiventrale Blatter. Planta 5: 89-

134.

MARTIN, J.T. & JUNIPER, B.E. 1970. The cuticles of plants. Edward

Arnold, London.

METCALFE, C.R. & CHALK, L. 1950. Anatomy of the dicotyledons

2. Clarendon Press, Oxford.

METCALFE, C.R. & CHALK, L. 1979. Anatomy of the dicotyledons

1, edn 2. Clarendon Press, Oxford.

PATEL, J.D. 1978. How should we interpret and distinguish subsidiary

cells? Botanical Journal of the Linnean Society 77: 65-72.

REYNOLDS, E.S. 1963. The use of lead citrate at high pH as an electron-

opaque stain in electron microscopy. Journal of Cell Biology

17:208-212.

SOLEREDER, H. 1908. Systematic anatomy of the dicotyledons 2.

Clarendon Press, Oxford.

STACE, C.A. 1965. Cuticular studies as an aid to plant taxonomy. The

Bulletin of the British Museum (Natural History), Botany, ser.

4: 1-78.

SUPPRIAN, K. 1894. Beitrage zur Kenntnis der Thymelaeaceae und

Penaeaceae. Botanische Jahrbiicher 18: 306-353.

THEOBALD, W.L., KRAHULIK, J.L. & ROLLINS, R.C. 1979.

Trichome description and classification. In C.R. Metcalfe & L.

Chalk, Anatomy of the dicotyledons 1 : 40-53.

THODAY, D. 1921. On the behaviour during drought of leaves of two

Cape species of Passerina , with some notes on their anatomy.

Annals of Botany 35: 585-601.

THODAY, D. 1924. XVIII. A revision of Passerina. Kew Bulletin 4:

146-168.

VAN DER MERWE, C.F. & COETZEE, J. 1992. Quetol 651 for gen-

eral use: a revised formulation. Proceedings of the Electron

Microscopy Society of Southern Africa 22: 31, 32.

VAN TIEGHEM, PH. 1893. Recherches sur la structure et les affinites

des Thymeleacees et des Peneacees. Annates des Sciences

Naturelles Botanique. 7,17: 185-294.

VAN WYK, A.E., ROBBERTSE, P.J. & KOK, PDF. 1982. The genus

Eugenia L. (Myrtaceae) in southern Africa: the structure and

taxonomic value of stomata. Botanical Journal of the Linnean

Society 84: 41-56.

WEIMARCK, H. 1941. Phyto geo graphical groups, centres and inter-

vals within the Cape Flora. Lunds Universitets Arsskrift. N.F.

Avd. 2, Bd. 37, Nr. 5. Otto Harrassowitz, Leipzig.

WILKINSON, H.P. 1979. The plant surface. In C.R. Metcalfe & L.

Chalk, Anatomy of the dicotyledons, edn 2, 1: 97-165. Claren-

don Press, Oxford.

Bothalia 30,1: 87-96(2000)

Vegetation of the coastal fynbos and rocky headlands south of George,

South Africa

D.B. HOARE*f, J.E. VICTOR**, R.A. LUBKE* and L. MUCINA*

Keywords: gradient analysis, numerical syntaxonomy, ordination, phytogeography, phytosociology

ABSTRACT

Community structure and composition of the coastal fynbos and rocky headland plant communities south of George,

southern Cape, were studied. Vegetation was analysed using standard sampling procedures of the floristic-sociological

approach of Braun-Blanquet. The releve data were subject to TWINSPAN-based divisive classification, and ordinated by

Principal Coordinates Analysis with the aim to identify vegetation coenocline subsequently interpreted in terms of under-

lying environmental gradients. Most of the sampled vegetation was classified as coastal fynbos. The Leucadendron

salignum-Tetraria cuspidata Fynbos Community was found to occupy sheltered habitats, whereas the Relhania calyci-

na-Passerina vulgaris Fynbos Community was found in exposed habitats. The other two communities characterise strong-

ly exposed rocky headlands. The Pterocelastrus tricuspidatus-Ruschia tenella Community is wind-sheared scrub, and the

Gazania rigens- Limonium scabrum Rocky Headland Community is a loose-canopy, low-grown herbland, characterised by

the occurrence of partly salt-tolerant and succulent herbs. The ordination of the fynbos communities revealed a horseshoe

structure allowing a direct recognition of a coenocline spanning two fynbos communities along the Axis 1 interpreted in

terms of exposure to wind and salt spray. A considerable amount of alien plant infestation was also present. This appears to

be the largest threat to the continued existence of this coastal fynbos.

INTRODUCTION

Fynbos occurring in close vicinity of the coast has been

studied in detail in many regions of South Africa (Boucher

1977; Van der Merwe 1979; Cowling 1984; Taylor 1985;

Hellstrom 1990; Taylor & Boucher 1993; Hoare 1994).

There are, however, still many portions of the coastline,

especially outside formal reserves, which remain only

poorly known. Only 79 km of the central south coast falls

into existing protected areas, namely the Goukamma

Nature Reserve, the Robberg Nature Reserve and the

Tsitsikamma Coastal National Park (Jarman 1986). The

promulgation of the Agulhas National Park may alter these

statistics (World Wide Fund for Nature 1999).

The area along the coast south of George, the present

study area, is of interest because of its scenic beauty and its

location in the centre of the popular Garden Route. Interest

in coastal development for recreational purposes through-

out the Garden Route is likely to impact on areas which

have not yet been encroached upon. The extent of vegeta-

tion cover outside formally protected areas has been con-

siderably transformed, and of the remaining natural areas

large portions have been invaded by alien species. It was

therefore suggested by Cape Nature Conservation (George)

that a detailed study be undertaken along the coast south of

George to provide information on the local vegetation and

flora and possible importance of the study area for conser-

vation (G. Hellstrom pers. comm.). This particular study

features descriptions of the plant communities of habitats

close to the coastline.

* Department of Botany, Rhodes University, 6140 Grahamstown,

South Africa.

t Present address: Agricultural Research Council, Range & Forage

Institute, Private Bag X05, 0039 Lynn East, Pretoria.

** National Botanical Institute, Private Bag X101, 0001 Pretoria.

+ Department of Botany, University of Stellenbosch, Private Bag XI,

7602 Stellenbosch, South Africa.

MS. received: 1998-07-24.

STUDY AREA

The study area extended from Rooiklip, SE of

Pacaltsdorp, to Ghwanobaai, 3 km E of Glentana (Figure

1), and covered ± 190 hectares. Glentana, and the area west

of it, has been extensively developed and was therefore

found unsuitable for the intended study. The fynbos vegeta-

tion in the study area has been classified as Asteraceous

Fynbos by Cowling & Holmes (1992) and it was broadly

classified by Acocks (1988) as ‘cultivated land, plantations,

dense alien communities and open sandy areas’, a descrip-

tion which gives a clear indication of the transformed state

of the vegetation. The vegetation of the coastline area is

classified as Dune Thicket (Low & Rebelo 1996), which

forms a mosaic with Dune Fynbos (Low & Rebelo 1996)

in the region including the study area.

The coastline consists of steep coastal cliffs ranging in

height from 50 to 70 metres, forming rocky headlands.

The plenitude of alternating bays and headlands has

resulted in microhabitats with varying degrees of expo-

sure to the prevailing winds, salt spray and sun. Coastal

soils vary markedly according to substrate, but are often

calcareous and coarse-grained. Topography is the domi-

nant factor affecting soil formation and the removal of

the products of weathering may exceed their formation,

especially on slopes. The coastal cliffs of the study area

are Rooiklip Granite-Gneiss of the Kaaimans Group and

are pre-Cape intrusive granite rocks (South African

Committee for Stratigraphy 1980). These rocks are

important because, upon weathering, they form base-rich

substrates containing exchangeable cations that are

important for soil formation and plant nutrient cycling

(Deacon et al. 1992).

The Koppen’s climate classification code for the

George coastal area is Cfb, which indicates warm, tem-

perate climate (Schulze & McGee 1978). The mean annu-

al temperature is ± 17°C, with a mean temperature range

Bothalia 30,1 (2000)

FIGURE 1. — Map of the study area and climate diagram for George.

of 8°C near the coast. The area has precipitation during all

months of the year, with three prominent peaks in spring,

summer and autumn, although there is no precipitation

deficit at any time of the year (Figure 1 ). Sea mists may

provide additional moisture on seaward-facing slopes.

Pronounced strong winds blow along the entire coastal

belt, varying in direction according to location and season.

In the study area, wind from the southwest and west were

found to be more important, since their frequency was

higher (Weather Bureau 1994). Wind velocity has been

found to be important in coastal plant communities, since

the greater the velocity of the wind, the higher the salt load

of the moving air, which may have a serious detrimental

effect on the growth of plants (Avis & Lubke 1985).

Importance values (Table 1) were determined which took

into account the velocity and frequency of winds from dif-

ferent directions (IV = mean monthly direction frequency

multiplied by mean monthly velocity expressed as per-

centage of total of all values for year). Westerly winds had

the highest importance values throughout the year but

were of greatest importance during the winter months.

Southwesterlies showed a similar pattern. Easterlies and

southeasterlies had the second highest importance values,

but these were most prevalent during spring and summer

(September to March).

MATERIAL AND METHODS

Aerial photographs were used to stratify the region

into broad vegetation complexes based on vegetation

structure and releves (25 m2 each) were made within

these zones. The quadrat size was determined from

species/area curves drawn from data collected in the

study area, but also conformed to a scale-related

approach to vegetation sampling, i.e. the size of the

releve was related to the scale at which the vegetation

was studied (Rutherford & Westfall 1994). Riparian

thicket, dune thicket, dense alien stands and agricultural

lands were not sampled. Standard field techniques and

the 7-grade sampling scale of Braun-Blanquet (Westhoff

& Van der Maarel 1973; Werger 1974) were used to

record the cover/abundance values for each vascular

species encountered in the releves. Topographic informa-

tion, including slope, aspect and altitude were also

TABLE 1. — Importance values of winds from different directions in

the months of the year. Bold text shows cells with importance

values greater than 1. (IV’s calculated from data from Weather

Bureau. See text for method)

Bothalia 30,1 (2000)

89

recorded. Physical soil properties, such as pH, conduc-

tivity, organic matter content, water-holding capacity, as

well as the relative fraction of coarse, medium and fine

sand, silt and clay, were ascertained in a subset of 20

releves. The general methodology used by the US

Department of Agriculture (USDA 1972) was followed

for analysing soils.

The vegetation data were initially classified using

TWINSPAN (Hill 1979), producing a rough species-by-

releve matrix, which was further rearranged in order to

finely tune the releve/species coincidence patterns sup-

posed to carry meaningful ecological information. The

data from only the fynbos communities were subject to

ordination by Principal Coordinates Analysis based on

similarity ratio as resemblance measure with no a priori

data transformation using the programme package SYN-

TAX-5 (Podani 1993, 1994).

For the descriptions of vegetation communities, three

informal ranks of vegetation units are recognised: com-

munity, subcommunity, and facies. Facies (Braun-

Blanquet 1964) represents the lowest-ranked unit and

corresponds to vegetation stands dominated by a single

species, mostly an alien element.

RESULTS AND DISCUSSION

Classification of plant communities

The rearranged species-quadrat matrix (Table 2)

revealed two groups of plant communities: the majority

of the releves were classified as coastal fynbos (Com-

munities A & B) and the remainder of the releves as

(non-fynbos) rocky headland communities (Communities

C & D). A summary of the community environmental and

floristic relationships is given in Figure 2. The following

communities, subcommunities and facies were identified:

A. Leucadendron salignum-Syncarpha paniculata Fyn-

bos Community

This community occurs in areas that are protected

from coastal winds either by headlands or the proximity

of coastal thicket (Figure 3) and has a different species

composition to the fynbos on the exposed summit of the

coastal cliffs. Diagnostic species: Leucadendron salig-

num, Aspalathus asparagoides, Hermannia angularis ,

Metalasia acuta , Helichrysum cymosum and Bobartia

aphylla. Dominant species: Syncarpha paniculata and

Passerina vulgaris. Common species: Tetraria cuspidata,

Cliffortia falcata, Metalasia pungens, Lobelia tomen-

tosa. Erica discolor and Phylica confusa. Many stands of

this community have been seriously invaded by the alien

species Acacia cyclops, Leptospermum laevigatum and

Pinus spp.

Slopes vary from moderate to steep and soils are gen-

erally deeper and finer-grained than on cliff summits.

There is high species richness (19 species on average) in

this community. The habitat is more variable than on the

cliff summits, thus leading to a greater species turnover

between localities. The geographic distribution of this

community beyond the present study area is unknown.

Four subcommunities are recognised and described.

Aa. Thamnochortus cinereus Subcommunity

Diagnostic species: Thamnochortus cinereus,

Polygala microlopha and Cliffortia sp. ( Victor 313). It is

situated below a housing development at Herold’s Bay

Extension; there was no evidence of recent fire and the

subcommunity is possibly one form of a fire-climax veg-

etation in protected areas.

Ab. Protea neriifolia Subcommunity

Diagnostic species: Protea neriifolia. It has developed

in the absence of fire and is possibly a fire-climax vege-

tation of protected areas. Bobartia aphylla and Pelar-

gonium fruticosum serve as common linking species to

subcommunites Aa and Ab.

Ac. Typical Subcommunity

Lacks the diagnostic species of the other three sub-

communities, but contains the diagnostic species of the

Leucadendron salignum-Syncarpha paniculata Fynbos

Community. There was evidence of recent fires in a num-

ber of the releves.

Ad. Hermannia althaeifolia Subcommunity

Diagnostic species: Hermannia althaeifolia, Hibiscus

aethiopicus, Ficinia albicans and Ursinia saxatilis as

well as a number of infrequently occurring species

(Table 2). Most of the releves had evidence of being

burnt recently, suggesting that it is an early post-fire suc-

cessional stage.

The vegetation height varies from 1 m, where dwarf

shrubs are dominant, to over 2.5 m where shrubs of the

exotic Leptospermum laevigatum are found. The mean

species richness of this subcommunity is 20 species on

average, but this may be reduced to only eight species

where invasion by exotic shrubs has taken place, as is the

case of the Acacia meamsii facies (Table 2, rel. 23).

B. Relhania calycina-Phylica confusa Fynbos Commu-

nity

This community makes up the greatest proportion of

the fynbos in the study area and also extends beyond the

boundaries of the present study area along the summit of

the coastal cliffs towards Knysna (Figure 3). Diagnostic

species: Relhania calycina subsp. calycina and Viscum

capensis. Dominant species: Erica discolor, Phylica con-

fusa, Passerina vulgaris and Tetraria cuspidata. Common

species: Thesium virgatum, Agathosma ovata, Erica pelta-

ta, Syncarpha paniculata and Cliffortia serpyllifolia.

The dwarf shrub layer of this community is usually

about 1 m in height and the herb layer about 0.4 m. The

total cover of the vegetation is slightly lower than for

90

Bothalia 30,1 (2000)

TABLE 2. — Classification of the vegetation of the coastal cliff habitats south of George, South Africa

Key to communities and subcommunities:

A: Leucadendron salignum-Syncarpha paniculata Fynbos Community

Aa: Thamnochortus cinereus Subcommunity

Ab: Protea neriifolia Subcommunity

Ac: Typical Subcommunity

Ad: Hermarmia althaeifolia Subcommunity

F: Acacia mearnsii facies

B: Relhania calycina-Phylica confusa Fynbos Community

Ba: Tetraria cuspidata Subcommunity

Bb: Eriocephalus africanus Subcommunity

C: Pterocelastrus tricuspidatus-Ruschia tenella Coastal Scrub Community

S: Sporobolus virginicus facies

D: Gazania rigens-Limonium scabrum Rocky Headland Community

G: Gazania rigens facies

Legend to vegetation layers: hi, herb layer; jl, juvenile woody species; si, upper shrub layer; s2, low shrub layer; t3, low tree layer. Taxonomic

notes, Gazania rigens = Gazania rigens var. uniflora\ Limonium scabrum agg. is a new taxon pending formal description (L. Mucina, in prep.) and

is closely related to L. scabrum (Thunb.) Kuntze; Drosanthemum marinum agg. is a complex of D. marinum and D. delicatulum and might repre-

sent one taxon after a revision (P. Burgoyne pers. comm ).

Coastal Fynbos

Tetraria cuspidata

Thesium virgatum

Syncarpha paniculata

Erica formosa

Cliffortia falcata

Metalasia pungens

Erica pel tat a

Schizaea pectinata

Restio triticeus

Lobelia tomentosa

Centella virgata

Pentaschistis eriostoma

Ficinia nigrescens

Carissa edulis

Aspalathus alopecurus

Anthospermum prostratum

Cullumia bisulca

Aspalathus florifera

Falkia repens

Crassula subulata

Disparago kraussii

Erica ericoides

Rhus lucida

Hahlenbergia desman tha

Leucadendron salignum-Syncarpha paniculata Fynbos Co

Leucadendron salignum

Aspalathus asparagoides

Hermannia angularis

Metalasia acuta

Helichrysum cymosum

Bobartia aphylla

Pelargonium fruticosum

mmunity

Bothalia 30,1 (2000)

91

Tliamnocliortus cinereus Subcommunity

Thamnochortus cinereus -hi

Polygala microlopha -s2

Cliffortia sp. (Victor 313) -s2

Helichrysum felinum -hi

Fuirena hirsuta -hi

Tetraria microstachys -hi

Printzia poli folia -hi

Protea neriifolia Subcommunity

Protea neriifolia

Hermannia althaeifolia Subcommunity

Hermannia althaeifolia -hi

Hibiscus aethiopicus -hi

Ursinia saxatilis -hi

Ficinia albicans -hi

Clutia alaternoides -s2

Relhania pungens -s2

Chrysanthemoides monilifera -s2

Diospyros lycioides -s2

Aspalathus kougaensis -s2

Tritoniopsis antholyza -hi

Gerbera serrata -hi

Ficinia quinquangularis -hi

Ischyrolepis triflora -hi

Themeda triandra -hi

Acacia mearnsii facies

Acacia mearnsii -t3

Leptospermum laevigatum -t3

32 |. . |br3.b| I | . | . + r | | 1 . | I ■ |

Relhania calycina-Phylica confusa Fvnbos Community

Relhania calycina -hi

Vise um capense -hi

Helichrysum tereti folium -hi

Cassytha ciliolata -hi

Carpobrotus edulis -hi

Stoebe microphylla -hi

Secamone alpini -hi

Tetraria robusta -hi

Tetraria compressa -hi

Eriocephalus africanus Subcommunity

Eriocephalus africanus -s2

Ficinia repens -hi

b+b+a+aaal

+++ .

b.3. .r.3. I . I | .

I . I .

Ruschia teneila-Gazania rigens Coastal Rocky Headland Community

Ruschia tenella -hi

Limonium scabrum agg. -hi

Gazania rigens -hi

Drosanthemum marinum agg. -hi

Tetragonia decumbens -hi

Sporobolus virginicus -hi

Sarcocaulon natalense -hi

Carpobrotus deliciosus -hi

Thesidium fragile -hi

Pterocelastrus tricuspidatus-Ruschia tenella Coastal Scrub Community

Pterocelastrus tricuspidatus -s2

Sideroxylon inerme -s2

Delosperma edwardsiae -hi

Cineraria britteniae -hi

Ir.r. I

I I

I - - r . |

I I

bb45 . . .

. . .r . . .b.

. . a . b . . . .

Other native spp. common to coastal fynbos

Passerina vulgaris

Erica discolor

Phylica confusa

Cliffortia serpylli folia

Agathosma ovata

Colpoon compressum

Agathosma apiculata

Indigofera heterophylla

Lampranthus sociorum

Thesidium podocarpum

Chironia baccifera

Crassula sp.

Oxalis sp.

Invasive alien spp.

Acacia cy clops

Acacia cy clops

Pinus pinaster

Pinus pinaster

Leptospermum laevigatum

Casuarina sp.

Leptospermum laevigatum

Acacia mearnsii

Hakea sericea

92

Bothalia 30,1 (2000)

TABLE 2. — Classification of the vegetation of the coastal cliff habitats south of George, South Africa (cont.)

other fynbos communities in the study area, mostly due

to an absence of a distinct restioid stratum. The commu-

nity is exposed to sun and wind and is consequently hot

and dry. Compounding this dryness, the soils are gener-

ally shallow and stony due to natural surface erosion at

the summit of these cliffs. Slopes vary from flat to mod-

erate. The community tends to grade into thicket inland,

often with a tall Passerina belt before the true thicket.

There is a high degree of invasion by Acacia cyclops

(present in 78% of releves), possibly causing habitat

modification, which could ultimately lead to irreversible

changes in species composition and vegetation struc-

ture.

Ba. Tetraria cuspidata Subcommunity

This subcommunity has the same dominant and com-

mon species as the Community itself, but with Viscum

capense as a common species. It is usually found more

inland of the other subcommunity suggesting that the

two subcommunities form a gradient from lower altitude

to wind-protected inland plant communities.

Bb. Eriocephalus africanus Subcommunity

Transition to lower altitude, steep-slope communities.

It has slightly steeper slopes than the typical cliff summit

community and is often moderately exposed to the influ-

ence of salt spray. There are also higher soil conductivi-

ty levels, which can be attributed to higher levels of

wind-home salt effects (see below). Pterocelastrus tricus-

pidatus and Lampranthus sociorum are occasionally pre-

sent. This community is further marked by the absence

of: Tetraria cuspidata, Viscum capense and Syncarpha

paniculata.

C. Pterocelastrus tricuspidatus-Ruschia tenella Coastal

Scrub Community

The woody component of this community reflects the

species composition of the coastal scrub in the deeply

incised valleys along this coastline and there is possibly a

floristic gradient from this community into thicket on

other steep slopes. Diagnostic species: Pterocelastrus tri-

cuspidatus as well as infrequent occurrence of Sideroxylon

inerme, Delosperma edwardsiae and Cineraria britteniae.

Bothalia 30,1 (2000)

93

BP

o

<D

O

O.

£

94

Bothalia 30,1 (2000)

FIGURE 3. — Generalised profile

diagram of coastal area south

of George indicating the spa-

tial relationship of communi-

ties to one another and the

sea.

Common or dominant species: Passerina vulgaris,

Phylica confusa, Lampranthus sociorum, Agathosma ovata

and Eriocephalus africanus, as well as the shrub Colpoon

compressum. Shrubs are wind-cropped to a maximum of

1 m, and the total aerial cover of the vegetation averages

60%. A facies, dominated by Sporobolus virginicus (Table

2, rel. 62), was distinguished within this community.

D. Gazania rigens-Limonium scabrum Rocky Headland

Community

Salt-tolerant asteraceous Gazania rigens var. uniflora

and a taxon from the Limonium scabrum complex ac-

companied by mesembs such as Drosanthemum mari-

num and Ruschia tenella, as well as other leaf succulents

such as Crassula sp. and Sarcocaulon natalense, form an

assemblage typical of the exposed rocky headlands in the

southern Cape.

Community C and the rocky headland ‘herbland’

Community D have many species in common, which can

be ascribed to a so-called neighbourhood effect (or mass

effect; Shmida & Wilson 1985). Naturally, the exposed

Community D is virtually lacking (except of Sarco-

caulon natalense) in its ‘own’ diagnostic species because

of the environmental stress in the form of deposition of

wind-borne salt that poses a major selective pressure on

the potential species pool.

Due to the low altitude this community is greatly

exposed to the influence of the wind and especially to

salt spray (Figure 3). The vegetation has a very low cover

and is wind-cropped. The average height of the vegeta-

tion is between 0.13 and 0.33 m. Of all the community

types, this one is lowest in species richness with a mean

of 10 species on average. Shallow soil covering granite

rocks is characteristic, and is often less than 5 cm deep

and very stony. Surface boulders are often present and

modify the microhabitat to some degree. A facies with

Gazania rigens var. uniflora was distinguished within

this community (Table 2, rel. 68).

This species composition corresponds with the

description of similar rocky headland communities

described for other parts of the coastline at Plettenberg

Bay (Hellstrom 1990) and Port Alfred (Lubke 1983).

Gradient analysis

The ordination of the fynbos communities (Figure 4)

revealed a horseshoe structure allowing the direct recog-

nition of a coenocline spanning two fynbos communities

AXIS 1

FIGURE 4. — Ordination of fynbos

vegetation using Principal

Coordinates Analysis. +, re-

leves of the Relhania calyci-

na-Phylica confusa Fynbos

Community; O, releves of

the Leucadendron salignum-

Syncarpha paniculata Fynbos

Community.

Bothalia 30,1 (2000)

95

TABLE 3. — Soil properties for communities (mean with standard deviation in brackets)

A, Leitcadendron salignum-Syncarpha paniculata Fynbos Community; B, Relhania calycina-Phylica confusa Fynbos Community; C,

Pterocelastrus tricuspidatus-Ruschia tenella Coastal Scrub Community; D, Gazania rigens-Limonium scabrum Rocky Headland Community.

along Axis 1 of the scatter diagram. There was a clear

separation along this coenocline between the Relhania

calycina-Phylica confusa Fynbos Community and the

Leucadendron salignum-Tetraria cuspidata Fynbos

Community. This pattern is interpreted in terms of expo-

sure to wind and salt spray as reflected in topographical

features and soil properties.

A classification scheme accompanied by associated

environmental relationships (Figure 2) suggests that the

Leucadendron salignum-Tetraria cuspidata Fynbos

Community occurs on deep soils inland of the cliff sum-

mits where the vegetation is protected from wind and salt

spray. The Relhania calycina-Passerina vulgaris Fynbos

Community occurs on shallow, stony soils on the cliff

summits, where there is a higher exposure to wind and

salt spray.

Physical soil properties varied from community to

community (Table 3). Cliff summits had the highest pro-

portion of coarse sand, slopes above the cliff summits

had the highest proportion of silt, and steep tall us slopes

and rocky headland slopes had the highest proportions of

medium sand, fine sand and clay. This is a trend in

decreasing sand particle size away from the cliff summit

(below and above) and larger proportion of clay to silt

below the cliff summits (steep tallus slopes and rocky

headland slopes).

Soil conductivity levels showed a decreasing trend as

exposure to wind from the sea decreased. High levels

were recorded on the steep tallus slopes, the cliff sum-

mits and the rocky headland slopes. The two cliff summit

communities showed a difference in conductivity levels

with the Eriocephalus africanus Subcommunity having

higher conductivity measures than the Tetraria cuspida-

ta Subcommunity. These higher conductivity levels were

attributed to higher salinity levels from wind-borne salt

spray effects.

Fire appeared to play an important role in the nature

of the communities especially in the Leucadendron

salignum-Syncarpha paniculata Fynbos Community

where the subcommunities Aa and Ab could be separat-

ed from the subcommunities Ac and Ad based on evi-

dence of recent burning. We suggest that the subcommu-

nities are distributed along a fire-induced succession gra-

dient, but with site-specific effects also coming into play.

ACKNOWLEDGEMENTS

We are indebted to Gavin Hellstrom, previously of

Cape Nature Conservation, George, for initiating this

study and for the provision of maps, ortho-photographs

and aerial photographs.

REFERENCES

ACOCKS, J PH. 19S8. Veld types of South Africa, edn 3. Memoirs of

the Botanical Survey of South Africa No. 57.

AVIS, A M. & LUBKE. R.A. 1985. The effect of wind-borne sand and

salt spray on the growth of Scirpus nodosus in a mobile dune

system. South African Journal of Botany 51:1 00-1 1 0.

BRAUN-BLANQUET, J. 1964. Pflanzensoziologie. Grundzuege der

Vegetationskunde, edn 3. Springer- Verlag, Wien.

BOUCHER, C. 1977. A provisional checklist of flowering plants and

ferns in the Cape Hangklip area. Journal of South African

Botany 43: 57-150.

COWLING, R.M. 1984. A syntaxonomic and synecological study in

the Humansdorp region of the Fynbos Biome. Bothalia 15:

175-227.

COWLING, R.M. & HOLMES, RM. 1992. Flora and vegetation. In

R.M. Cowling, The ecology of fynbos: nutrients, fire and diver-

sity: 23-61. Oxford University Press, Cape Town.

DEACON, H.J., JURY, M R. & ELLIS, F. 1992. Selective regime and

time. In R.M. Cowling, The ecology of fynbos: nutrients, fire

and diversity: 6-22. Oxford University Press, Cape Town.

HELLSTROM, G. 1990. A phytosociological classification, with man-

agement proposals, of the Robberg Nature Reserve, Plettenberg

Bay. B.Sc. (Hons) thesis, Rhodes University, Grahamstown.

HILL, M.O. 1979. TWINSPAN — a FORTRAN program for arranging

multivariate data in an ordered two-way table by classification

of the individuals and attributes. Ecology & Systematics,

Cornell University, Ithaca, N Y.

HOARE, D.B. 1994. Assessing successional effects on plant diversity

in the Goukamma Nature Reserve, southern Cape. B.Sc. (Hons)

thesis, Rhodes University, Grahamstown.

JARMAN, M.L. 1986. Conservation priorities in lowland regions of

the Fynbos Biome. South African National Scientific Program-

mes Report No. 87. CSIR, Pretoria.

LOW, A.B & REBELO, A.G. (eds). 1996. Vegetation of South Africa,

Lesotho and Swaziland. Companion to the vegetation map of

South Africa, Lesotho and Swaziland. Department of Environ-

mental Affairs and Tourism, Pretoria.

LUBKE, R.A. 1983. A survey of the coastal vegetation near Port

Alfred, Eastern Cape. Bothalia 14: 725-738.

PODANI, J. 1993. SYN-TAX-pc. Computer programs for multivariate

data analysis in ecology and systematics. Version 5.0. User's

Guide. Scientia Publishing, Budapest.

PODANI, J 1994. Multivariate analysis in ecology and systematics.

SPB Academic Publishing, The Hague.

RUTHERFORD, M.C. & WESTFALL, R.H. 1994. Biomes of southern

Africa — an objective categorization, edn 2. Memoirs of the

Botanical Survey of South Africa No. 63.

96

Bothalia 30,1 (2000)

SOUTH AFRICAN COMMITTEE FOR STRATIGRAPHY (SACS)

1980 Stratigraphy of South Africa. Part 1. Lithostratigraphy of

the Republic of South Africa, South West Africa/Namibia, and

the Republics of Bophuthatswana, Transkei and Venda. Geo-

logical Survey of South Africa Handbook No. 8. Government

Printer, Pretoria.

SCHULZE, R E. & MCGEE, O S. 1978. Climatic indices and classifi-

cations in relation to the biography of southern Africa. In

M.J.A. Werger, Biogeography and ecology of southern Africa.

Junk, The Hague.

SHMIDA, A & WILSON, M.V. 1985. Biological determinants of

species diversity. Journal of Biogeography 12: 1-20.

TAYLOR, H.C. 1985. An analysis of the flowering plants and ferns of

the Cape of Good Hope Nature Reserve. South African Journal

of Botany 51: 1-13.

TAYLOR, H.C. & BOUCHER, C. 1993. Dry coastal ecosystems of the

South African south coast. In E. Van der Maarel, Ecosystems of

the World 2B. Dry coastal ecosystems: Africa, America, Asia

and Oceania. Elsevier, Amsterdam.

UNITED STATES DEPARTMENT OF AGRICULTURE (USDA)

1972. Soil survey laboratory methods .and procedures for col-

lecting soil samples. Soil Survey Report 1. US Government

Printing Office, Washington, DC.

VAN DER MERWE, C.V. 1979. Plantekologiese aspekte en bestuurs-

probleme van die Goukamma-natuurreservaat. M.Sc. thesis,

University of Pretoria.

WEATHER BUREAU 1994. Monthly weather report for the months

January to December 1994. Weather Bureau, Pretoria.

WERGER, M.J.A. 1974. On concepts and techniques applied in the

Ziirich-Montpellier method of vegetation survey. Bothalia 1 1 :

309-323.

WESTHOFF, V. & VAN DER MAAREL, E. 1973. The Braun-Blanquet

approach. In R.H. Whittaker, Ordination and classification of

plant communities. Junk, The Hague.

WORLD WIDE FUND FOR NATURE (WWF) 1999. Southernmost

point of Africa: a gift to the earth. Africa — Environment and

Wildlife Suppl. to 7,3: 9.

Bothalia 30,1: 97-101 (2000)

Checklist of plant species of the coastal fynbos and rocky head-

lands, south of George, South Africa

J.E. VICTOR**, D.B. HOARE* *+ and R.A. LUBKE*

Keywords: checklist, coastal fynbos, endemics, phytogeography, rocky headlands. South Africa, Western Cape

ABSTRACT

A checklist of vascular plants and cryptograms was compiled for the fynbos and rocky headland communities of the

coastal region south of George. The area studied is a 12 km stretch of steep sandstone cliffs forming alternating bays and

headlands situated between Glentana and Wilderness. The plant communities of the natural vegetation inhabiting the coast-

line are a mixture of coastal thicket, riparian thicket, fynbos and rocky headland types. The extent of natural vegetation has

been reduced by the spread of agricultural land and urban development and is under further threat from the spread of natu-

ralised alien invader species, particularly Acacia cyclops. The checklist records the occurrence of 271 taxa including 16

alien species (6% of taxa). Of the flowering plant species recorded, 6% were regional or local endemics.

INTRODUCTION

The study area is a 12 km section of coastline south of

George extending from Rooiklip, southeast of Pacalts-

dorp, to Ghwanobaai, 3 km east of Glentana (see Hoare

et al. 2000 for details). It includes a band of vegetation

within 500 m of the high tide mark on steep sandstone

cliffs which form alternating bays and headlands. The

plant communities of the natural vegetation along the

coastline are a mixture of coastal thicket, riparian thick-

et, fynbos and rocky headland types. The study was con-

fined to the fynbos and rocky headland vegetation.

Riparian thicket, dune thicket, dense alien stands and

agricultural lands were not sampled. Rainfall along this

section of coastline occurs throughout the year, but with

three distinct peaks in spring, summer and autumn

(Hoare et al. 2000). Because of its close proximity to the

sea, vegetation structure and composition are greatly

influenced by oceanic winds. The extent of the natural

vegetation has been reduced by the spread of agricultur-

al land and urban activities and is under further threat

from the spread of naturalised alien species, particularly

Acacia cyclops.

This study area falls into that part of the Fynbos

Biome called Limestone Fynbos of the Mossel Bay

District (Low & Rebelo 1996) of which ± 14% is con-

served and 40% transformed, although the recent procla-

mation of the Agulhas National Park may affect these

statistics. The particular study area is of interest because

of its scenic beauty and its location in the centre of the

Garden Route— a popular tourist attraction. Most of the

study area is owned by private landowners who have

used the land mostly for agriculture (in places agricultur-

al lands extend to within a few metres of the summit of

the coastal cliffs), but also for plantation forestry on a

** National Botanical Institute, Private Bag X101, 0001 Pretoria.

* Department of Botany, Rhodes University, 6140 Grahamstown,

South Africa.

t Present address: Agricultural Research Council, Range & Forage

Institute, Private Bag X05, 0039 Lynn East, Pretoria.

MS. received: 1999-11-29.

small scale. Urban development occurs in nodes, e.g.

Herold’s Bay and Victoria Bay. Recreational activities,

e.g. fishing, occur on a small scale along the coast.

Disturbances related to all these activities have resulted

in invasion by alien trees and shrubs, particularly Acacia

cyclops. Pinus species have spread from the plantations

and are also a potentially serious threat.

The aim of this work was to provide a checklist of the

plant species along this stretch of coastline that can act as

a basic reference for floristic and ecological work and as

a baseline for future development in the region. This

checklist forms a link between checklists and floristic

studies done for the Western Cape coastal region

(Boucher 1977; Taylor 1985), southern Cape coast

(Hellstrom 1990; Hoare 1994; Van der Merwe 1979) and

the Eastern Cape coast (Lubke 1983; Lubke et al. 1988;

Cloete & Lubke 1999). This provides the opportunity for

comprehensive examination of floristic gradients along

the coastal region of the Fynbos Biome.

METHODS

Fieldwork was done in all four seasons of the year to

cover as many flowering times as possible. Voucher

specimens of most taxa were deposited in the National

Herbarium, Pretoria, and additional taxa listed were

obtained from sight records made during the course of

fieldwork. The herbarium collection at PRE was consult-

ed to obtain records of plant species previously collected

in the study area, and these were added to the checklist.

RESULTS

The checklist lists 271 taxa comprising nine lichen

species, three bryophytes, three pteridophytes, two gymno-

sperms, 56 monocotyledons and 198 dicotyledons (Table

1). The most commonly represented angiosperm families

(Table 2) are Asteraceae (15% of species), Poaceae (7%),

Cyperaceae (6%), Fabaceae (6%), Mesembryanthema-

ceae (5%), Ericaceae (5%) and Crassulaceae (4%). The

genera with the most species are Erica (11), Crassula

(11), Aspalathus (8) and Helichrysum (6). Ficinia, Lam-

98

Bothalia 30,1 (2000)

TABLE 1 . — Number of families, genera and species recorded in the

vegetation of the coast south of George

pranthus, Phylica and Hermannia were each represented

by five species.

All of the 20 largest genera listed for the Cape flora by

Bond & Goldblatt (1984) are represented in this coastal

area, as are 13 of the 15 largest families. Of the 16 alien

species recorded, Acacia Cyclops was by far the most

abundant.

Some 230 flowering plant species were classified

according to phytogeographical range and affinity (Table

3). It was found that 35% are endemic to the Fynbos

Biome and 7% are regional (southern Cape) endemics

and one was a local endemic — Silene vlokii , which has a

restricted range from Herold’s Bay to Glentana.

DISCUSSION

Asteraceous Coastal Fynbos is defined as having high

asteraceous and non-ericaceous ericoid cover and often

high grass cover (Cowling 1992). Phylica , Passerina,

Agatliosma (and other Diosminae), Aspalathus , Restio and

Cliffortia are listed as dominant genera in this vegetation

type (Cowling 1992), a view which is consistent with what

was found in the study area (Floare et al. 2000).

A comparison of the flora of the study area with those

of the Goukamma Nature Reserve (Table 4) shows that

the number of species and genera in the present study

area is comparatively high in relation to its size, espe-

cially considering that not all vegetation types were sam-

pled. Goukamma Nature Reserve is a larger area but with

TABLE 2. — Families of angiosperms in the study with the highest

number of genera and species

TABLE 3. — Distribution of angiosperms found in the study area

TABLE 4. — Comparison of the indigenous angiosperm flora of the

study area with nearby Goukamma Nature Reserve (Van der

Merwe 1979; Hoare 1994)

Goukamma

Study area Nature

Reserve

fewer species, indicating that there is lower diversity in

the Dune Fynbos and thicket vegetation of that region

compared with the communities along the rocky shore

and promontories of this study area. This coastal region

therefore has a surprisingly high diversity for such a

small area, probably due to its abundance of microhabi-

tats. Further studies encompassing the non-fynbos vege-

tation types should be carried out to contribute to the

knowledge of this relatively understudied and poorly

conserved region.

REFERENCES

BOND, P. & GOLDBLATT, P. 1984. Plants of the Cape Flora: a

descriptive catalogue. Journal of South African Botany Suppl.

Vol. 13: 455.

BOUCHER, C. 1977. A provisional checklist of flowering plants and

ferns in the Cape Hangklip area. Journal of South African

Botany 43: 57-80.

BRUM MITT R.K. & POWELL, C.E. 1992. Authors of plant names.

Royal Botanic Gardens, Kew, London.

CLOETE, E.C. & LUBKE, R.A. 1999. Flora of the Kap River Reserve,

Eastern Cape, South Africa. Bothalia 29: 139-149.

COWLING, R.M. 1992. The ecology of fynbos: nutrients , fire and

diversity. Oxford University Press, Cape Town.

HELLSTROM, G. 1990. A phytosociological classification, with man-

agement proposals , of the Robberg Nature Reserve, Plettenberg

Bay. M:Sc. thesis, Rhodes University, Grahamstown.

HOARE, D.B. 1994. Assessing successional effects on plant diversity

in the Goukamma Nature Reserve, southern Cape. B.Sc. (Hons)

thesis, Rhodes University, Grahamstown.

HOARE, D.B., VICTOR, J.E., LUBKE, R.A. & MUCINA, L. 2000.

Vegetation of the coastal fynbos and rocky headlands south of

George, South Africa. Bothalia 30: 87-96.

LOW, A.B. & REBELO, A.G. 1996. Vegetation of South Africa,

Lesotho and Swaziland. Department of Environmental Afrairs

& Tourism, Pretoria.

LUBKE, R.A. 1983. A survey of the coastal vegetation near Port

Alfred, eastern Cape. Bothalia 14: 725-738.

LUBKE, R.A., GESS, F.W. & BRUTON, M.N. 1988. A field guide to

the eastern Cape coast. Wildlife Society, Grahamstown.

TAYLOR, H.C. 1985. An analysis of the flowering plants and ferns of

the Cape of Good Hope Nature Reserve. South African Journal

of Botany 51: I -1 3.

VAN DER MERWE. C.V. 1979. Plantekologiese aspekte en bestuurs-

probleme van die Goukamma-natuurreseivaat. M.Sc. thesis,

University of Pretoria.

Bothalia 30,1 (2000)

99

CHECKLIST

Taxa are arranged alphabetically, and author citations follow Brummitt & Powell (1992). Except for site records, col-

lectors’ names and numbers follow the author citation; specimens are housed at PRE. Naturalised alien species are

marked with an asterisk*. Abbreviations for collectors’ names: Bo, P. Bohnen; Da, G. Davidse; Th, M.F. Thompson;

V, J.E. Victor; VFC, C.M. Van Wyk, A. Fellingham & M. O’Callaghan; V&H, J.E. Victor & D.B. Hoare; Wi, I.J.M.

Williams.

LICHENS

Cladia aggregata (5w.) Nyl., V 327

Cladonia

Chasmariae sp., V 328

chlorphaea (Floerke) Spreng., V 325

Cocciferae sp., V 326

confusa R.Sant., V 324

coniocraea (Floerke) Spreng., V 328b

Pycnoporus sp., V 545

Teloschistes flavicans (Sw.) Norm., V 546

Usnea rubicunda Stirt., V 547

BRYOPHYTES

FUNAR1ACEAE

Funaria hygrometrica Hedw., V 290

PH YLLOGON I ACE AE

Catagonium nitens (Brid.) Card, subsp. maritimum (Hook.) S-H.Lin, V 352

POTTIACEAE

Tortella xanthocarpa (C.Muell.) Broth., V 303

PTERIDOPHYTES

ADIANTACEAE

Cheilanthes hirta Sw. van hirta, V 367

ASPLEN1ACEAE

Asplenium rutifolium (P.J.Bergius) Kunze, V 351, 366

SCHIZAEACEAE

Schizaea pectinata (L.) Sw., V 317

GYMNOSPERMS

PINACEAE

Pinus

^pinaster Aiton

*radiata D.Don

ANGIOSPERMS: MONOCOTYLEDONS

ASPHODELACEAE

Anthericum cooperi Baker, V 561; V&H 43

CYPERACEAE

Ficinia

albicans Nees, V&H 83

cf. gracilis (Pair.) Schrad., V 259, 321 , 329, 359

laciniata (Thunb.) Nees, Da 33719

nigrescens (Schrad.) J.Raynal, V 261, 285 ; V&H 68

repens (Nees) Kunth, V 292

Fuirena hirsuta (P.J.Bergius) P.L. Forbes, V 306

Isolepis tenuissima (Nees) Kunth, V 288

Mariscus

congestus (Vahl) C.B. Clarke, V 557

thunbergii (Vahl) Schrad., V 282

Pycreus polystachyos (Rottb.) Beauv. var. polystachyos, V 287

Schoenoxiphium sparteum (Wahlenb.) C.B. Clarke, Da 33726

Tetraria

bolusii C.B. Clarke, VFC 234

compressa Turrill, V 246; V&H 40

cuspidata (Rottb.) C.B. Clarke, V 242; V&H 78

microstachys (Vahl) Pfeiffer, V 309, 316, 333

Trianoptiles capensis (Steud.) Harv., V 289

HYACINTHACEAE

Lachenalia bulbifera ( Cyr.) Engl., VFC 1 70

Omithogalum sp.

HYPOXIDACEAE

Empodium sp., V 207

Spiloxene trifurcillata (Nel) Fourc., V 335

1RIDACEAE

Babiana fourcadei G.J. Lewis, V 349

Bobartia aphylla (L.f.) Ker Gawl., V 307, 540; V&H 47

Chasmanthe aethiopica (L.) N.E.Br., VFC 171

Freesia

alba ( G.L.Mey.) Gumbleton

leichtlinii Klatt, V 296

Gladiolus floribundus Jacq. subsp. floribundus, Th 609

Hesperantha falcata (L.f.) Ker Gawl., V 213

Micranthus alopecuroides (L.) Rothm., V 558

Tritoniopsis antholyza (Pair.) Goldblatt, V 495

JUNCACEAE

Juncus

acutus L. subsp. leopoldii (Pari.) Snog., V 345

dregeanus Kunth, V 286, 534

ORCHIDACEAE

Disperis capensis (L.) Sw. var. capensis, V 238

Herschelianthe hians (L.f) Rauschert, V 529, 533

POACEAE

Cynodon dactylon (L.) Pers., V 516

Eragrostis

capensis (Thunb.) Trin., V 503

plana Nees, V 511

Ehrharta

calycina J.E.Sm., V 513

capensis Thunb., V 513b; V&H 80

erecta Lam. var. erecta

*Lolium perenne L„ V 524

Paspalum

*dilatatum Pair., V 522

distichum L„ V 517

Pentaschistis eriostoma (Nees) Stapf

Polypogon strictus Nees, V 350

Setaria sphacelata (Schum.) Moss var. sphacelata, V 523

Sporobolus

africanus (Pair.) Robyns & Tournay, V 518

virginicus (L.) Kunth

Stenotaphrum secundatum (Walt.) Kuntze

Themeda triandra Forssk., V 510

Tribolium uniolae (L.f.) Renvoize, V&H 81

PONTEDERIACEAE

*Eichhomia crassipes (Mart.) Solms-Laub ., V 562

RESTIONACEAE

Hypodiscus willdenowia (Nees) Mast., Da 33722

Ischyrolepis tnflora (Rottb.) Linder

Restio triticeus Rottb., V 253, 254, 320; V&H 36

Thamnochortus cinereus H P. Linder, V 308; V&H 45

DICOTYLEDONS

ACANTHACEAE

Hypoestes aristata (Vahl) Roem. & Schult. var. thiniorum Balkwill, V 341

AIZOACEAE

Tetragonia

fruticosa L„ V 273

decumbens Mill.

spicata L.f. var. spicata, V 338

virgata Schltr., V&H 42

ANACARDIACEAE

Rhus

crenata Thunb., V 302

glauca Thunb., VFC 177

lucida L. forma lucida, V 208

APIACEAE

Centella virgata (L.f.) Drude, V 319

APOCYNACEAE

Astephanus marginatus Decne.

Sarcostemma viminale (L.) R.Br.

100

Bothalia 30,1 (2000)

ASTERACEAE

Arctotheca prostrata (Salisb.) Britten , VFC 185

Athanasia trifurcata (L.) L., V 509

Athrixia capensis Ker Gawl., V 541

Berkheya armata (Vahl) Druce, V 531

Chrysanthemoides monilifera (L.) Norl. subsp. pisifera (L.) Norl., V 262

Cineraria britteniae Hutch. & RA.Dyer, V&H 39, 71

Cullumia bisulca (Thunb.) Less.

Disparago kraussii Sch. Bip., V 502 ; V&H 58

Elytropappus rhinocerotis (L.f.) Less., V 548

Eriocephalus africanus L., V 244; V&H 33

Felicia

amoena (Sch. Bip.) Levyns, subsp. latifolia Grau, V 536

filifolia (Vent.) Burtt Davy, subsp. bodkinii (Compton) Grau, V 300

Gazania rigens (L.) Gaertn. var. uniflora (L.f.) Rossler, V 294; V&H 75

Gerbera serrata (Thunb.) Druce, V 239; V&H 66

Helichrysum

anomalum Less., V 240, 514

asperum (Thunb.) Hilliard & B.L.Burtt var. glabrum Hilliard, V 532

cymosum (L.) D.Don subsp. cymosum, VFC 227

felinum Less., V 322

odoratissimum (L.) Sweet, V 564

teretifolium (L.) D.Don, V 235

Metalasia

acuta Karis, V 217

pungens D.Don, V 247; V&H 52, 54b

Oedera

capensis (L.) Druce, V 236

imbricata Lam., V 267

Othonna camosa Less., var. camosa, V 370

Plecostachys serpyllifolia (Berg.) Hilliard & B.L.Burtt, V 301

Pnntzia polifolia (L.) Hutch., V 323

Relhania

calycina (L.f.) L’Her. subsp. calycina, V 364

pungens L’Her. subsp. pungens, V 501

Senecio

angulatus L.f, V 348

deltoideus Less., VFC 183

ilicifolius L., V 508

Stoebe

microphylla DC., V 357; V&H 72, 82

plumosa (L.) Thunb., V 304

Syncarpha

canescens (L.) B.Nord.

paniculata (L.) B.Nord.

Tarchonanthus camphoratus L., V 34; V&H 44

Ursinia

heterodonta (DC.) N.E.Br., V 252, 528

saxatilis N.E.Br., V&H 34

BRASSICACEAE

Heliophila subulata DC., V 334, 527

CAMPANULACEAE

Lightfootia

divaricata H.Buek var. debilis (Sond.) Adamson, VFC 95

fasciculata (L.f.) A. DC., V 500

CARYOPHYLLACEAE

Silene vlokii Mass., V 864

*Spergula arvensis L., V 214

CELASTRACEAE

Cassine papillosa (Hochst.) Kuntze

Pterocelastrus tricuspidatus (Lam.) Sond., V 299

Putterlickia pyracantha (L.) Szyszyl., V 549

CHENOPODIACEAE

Sarcocomia natalensis ( Ung.-Sternb.) A.J. Scott var. natalensis

CONVOLVULACEAE

Falkia repens L.f., V 281

CRASSULACEAE

Adromischus caryophyllaceus (Burm.f.) Lem., V 552; V&H 37

Crassula

atropurpurea (Haw.) Dietr. var. atropurpurea, V&H 38

biplanata Haw., VFC 169

decumbens Thunb. var. brachyphylla (Adamson) Tolken, V 215

lactea Soland., V 344

nudicaulis L. var. nudicaulis, V 553

orbicularis L., V 297

rubricaulis Eckl. & Zeyh., V 343

rupestris Thunb. subsp. rupestris, V 551

southii Schbnland subsp. sphaerocephala Tolken, V 223

subulata L. var. fastigiata (Schonland) Tolken, V 382

DROSERACEAE

Drosera sp., V 318

EBENACEAE

Diospyros

dichrophylla (Gand.) De Winter, V 209

lycioides Desf. subsp. lycioides, V&H 49

Euclea

crispa (Thunb.) Guerke subsp. crispa, V 368

natalensis A. DC. subsp obovata F. White, V&H 41

ERICACEAE

Blaeria ericoides L., V 543

Erica

canaliculata Andr., V 212

discolor Andr. var. discolor, V 233, 293

formosa Thunb., V 243, 323, 355; V&H 25, 46

glandulosa Thunb. var. glandulosa, V 555; V&H 48

hispidula L. var. hispidula, V 258, 356

imbricata L., V 354

peltata Andr., V&H 26

speciosa Andr., V 241

triceps Link, V 559

versicolor Wendl.

sp., V&H 24

Salaxis axillaris (Thunb.) G.Don, V 258, 530

EUPHORBIACEAE

Clutia

alatemoides L. var. brevifolia Sond.

laxa Sond., VFC 218

Euphorbia cf. caterviflora N.E.Br., V 272

FABACEAE

Acacia

*cyclops G.Don, V 360

*meamsii De Wild., V 257; V&H 54a

Amphithalea fourcadei Compton, V 358

Aspalathus

alopecurus Benth., V 353; V&H 56

asparagoides L.f. subsp. asparagoides, V&H 21

asparagoides L.f. subsp. rubro-fusca (Eckl. & Zeyh.) R.Dahlgren, V 331

ciliaris L., V 526, V 544

florifera R.Dahlgren, V 248

kougaensis ( R.Dahlgren ) R.Dahlgren, V 266; V&H 22

laricifolia PJ.Bergius subsp. laricifolia, V&H 55

nigra L.

Indigofera heterophylla Thunb., V271; V&H 59, 77

*Lotus subbiflorus Lag., V 505

Rhynchosia

capensis (Burnt.) Schinz, V 315

ciliata (Thunb.) Schinz

*Vicia sativa L., V 381, 560

FUMARIACEAE

*Fumaria muralis Koch subsp. muralis, V 211

GENTIANACEAE

Chironia baccifera L., V 336

GERANIACEAE

Pelargonium

candicans Spreng., V 280

capitatum (L.) L’Her., V&H 74

caucalifolium Jacq. subsp. convolvulifolium ( Kunth ) J.JA.van der

Walt, V 538

fruticosum (Cav.) Willd., V 311

LAM I ACE AE

Stachys graciliflora Presl, VFC 164

LAURACEAE

Cassytha ciliolata Nees, V 227; V&H 76

LOBELIACEAE

Lobelia

bicolor Sims, V&H 61

coronopifolia L., V 255, 537

erinus L., V 520

tomentosa L.f.

Monopsis unidenlata (Dryand.) E.Wimm. subsp. unidentata, V 515

MALVACEAE

Anisodontea scabrosa (L.) Bates, V 279

Hibiscus aethiopicus L. var. ovatus Ha rv., V&H 57

Bothalia 30,1 (2000)

101

MESEMB RYANTHEM ACEAE

Carpobrotus

deliciosus (L. Bolus) L. Bolus, V 305

edulis (L.) L. Bolus, V 210

Delosperma

edwardsiae L.Bolus, V 218 , 221, 269, 270; V&H 31

litorale (Kensit) L.Bolus, V 219, 550; V&H 79

Disphyma crassifolium (L.) L.Bolus, V 268

Drosanthemum

brevifolium (Alton) Schwantes, V 222

sp„ V&H 63

Lampranthus

conspicuus (Haw.) N.E.Br., V 260; V&H 64

dependens (L.Bolus) L.Bolus, VFC 180

sociorum (L.Bolus) N.E.Br., V&H 32

stipulaceus (L.) N.E.Br., V 230

sp„ V220

Conophytum bilobum (Marlolh) N.E.Br., VFC 176

Ruschia tenella (Haw.) Schwantes V 565; V&H 65

MYRICACEAE

Myrica quercifolia L.

MYRTACEAE

*Leptospermum laevigatum ( Caertn.) F.Muell., V 264

OLEACEAE

Olea exasperata Jacq.

OXALIDACEAE

Oxalis spp., V 310, 224

PLANTAGINACEAE

*Plantago lanceolata L., V 525

PLUMBAGINACEAE

Limonium scabrum (Thunb.) Kuntze var. scabrum, V&H 60

POLYGALACEAE

Muraltia ericoides (Burm.f.) Steud., V 499

Polygala

fruticosa PJ.Bergius, V 295

microlopha DC. var. gracilis Levyns, V 314

myrtifolia L., V 346

PRIMULACEAE

*Anagallis arvensis L., V 291, 519

PROTEACEAE

*Hakea sericea Schrad., V 361

Protea neriifolia R.Br., V 237

Leucadendron salignum P J.Bergius, V 263

RANUNCULACEAE

Knowltonia vesicatoria (Lf) Sims subsp. grossa H.Rasm., V 340

RHAMNACEAE

Phylica

axillaris Lam.

var. axillaris, VFC 1 74

var. maritima Pillans, VFC 228

confusa Pillans, V&H 50

purpurea Sond. var. pearsonii Pillans, V 245

strigulosa Sond., V&H 29

ROSACE AE

Cliffortia

falcata L.f., V 265

serpyllifolia Cham. & Schlechtd., V 228, 542

sp„ V313

RUB I ACEAE

Anthospermum

aethiopicum L., V 249, 512

prostratum Sond., V 251

RUTACEAE

Agathosma

apiculata G.F.W.Me y., V 275

capensis (L.) Diimmer, V 312

ovata (Thunb.) Pillans, V 232, 337

Euchaetis burchellii Diimmer, Wi 2310

S ANTAL ACEAE

Colpoon compressum PJ.Bergius, V 229, 342

Thesidium

fragile (Thunb.) Sond.

microcarpum (A.DC.) A.DC., V 231, 298

podocarpum (A.DC.) A.DC., V 278

Thesium

lisae-mariae Stauffer, V 330

nigromontanum Sond., V&H 51

sertulariastrum A.W.Hill, V 226

virgatum Lam., V&H 28

SAPOTACEAE

Sideroxylon inerme L. subsp. inerme, V&H 67

SCROPHULARIACEAE

Graderia scabra (L.f.) Benth., Bo 8157

Phyllopodium rustii (Rolfe) Hilliard, V 284

Sutera

hispida (Thunb.) Druce, VFC 163a

sp„ V277, 339

SELAGINACEAE

Selago

corymbosa L., V 554, 563

dregei Rolfe, VFC 209

sp. nov., V 234, 496

SOLANACEAE

Solanum rigescens Jacq., VFC 178

STERCULIACEAE

Hermannia

althaeifolia L., V 250

angularis Jacq., V 256, 535

holosericea Jacq., V 539

lavandulifolia L., V 283

salviifolia L.f. var. salviifolia, V 507

THYMELAEACEAE

Passerina

falcifolia C.H. Wright, V 556

vulgaris Thoday, V 498

Struthiola

ciliata (L.) Lam. subsp. angustifolia (Lam.) Peterson, V 216

parviflora Meisn., V 497

striata Lam., V 504; V&H 73

TILIACEAE

Grewia occidentals L., V 521

VERBENACEAE

*Verbena bonariensis L., V 506

VISCACEAE

Viscum capense L.f. subsp. hoolei Wiens, V 225

ZYGOPHYLLACEAE

Zygophyllum morgsana L., V 274, 369

■

Bothalia 30,1: 103-110(2000)

Miscellaneous notes

POACEAE

APOMICTIC EMBRYO SAC DEVELOPMENT IN CENCHRUS CILIARIS (PANICOIDEAE)

Apomixis is distributed throughout the plant kingdom

and is known in over 300 species of at least 35 different

families (Hanna & Bashaw 1987). Research on apomix-

is in grasses began in the early 1 930’s with Poa praten-

sis L. (Miintzing 1933). Apomixis is especially prevalent

among perennial forage grasses and has been reported in

more than 125 species representing most of the tribes.

Well-known apomictic grasses include Paspalum dilata-

tum Poir. (Snyder 1957), P. notatum Fluegge (Snyder

1957), Setaria P.Beauv. (Emery 1957) and Cenchrus cil-

iaris L. (Fisher et al. 1954; Snyder et al. 1955).

Apomixis in the angiosperms means asexual reproduc-

tion by seeds (Nogler 1984). In this study reduced

embryo sacs (embryo sacs in which meiosis occurs)

imply sexual reproduction, whereas unreduced embryo

sacs imply asexual or apomictic reproduction.

The aim of this study is to determine whether reduced

embryo sacs are formed in C. ciliaris specimens includ-

ed in this study. The type of embryo sac will give an indi-

cation of the sexual or apomictic nature of the species

and will suggest to what extent cross-fertilisation con-

tributed to the genetic variation in this species.

MATERIAL AND METHODS

The plant material used in this study was collected in

the field (Table 1). A selection from the material, col-

lected for meiotic analysis, was used to study embryo

sac development. Florets in various developmental

stages were used. Inflorescences were dehydrated with

ethanol and tertiary buthanol before being embedded in

pastulated Paraffin wax. Sections (6 |im) were cut with

a rotary microtome, mounted and stained with a modi-

fied (Spies & du Plessis 1986) safranin (Johansen 1940)

and fast green (Sass 1951) procedure. A minimum of

twenty embryo sacs per developmental stage were stud-

ied in each specimen. A Nikon photomicroscope and

Ilford Pan-F film (ASA 50) were used for the photomi-

crographs.

RESULTS

Embryo sac development was studied in 27 speci-

mens, representing four different ploidy levels (Table 1)

The nucellus was multiseriate and protected by various

TABLE 1. — List of Cenchrus ciliaris specimens with Panicum type of embryo sacs, their localities and their

chromosome number (B-chromosomes are abbreviated as B)

104

Bothalia 30,1 (2000)

• A*; . ' *1#'V * ■' n

tr jr, * .4 -

'&<•$ 4m

* w. E :»••**%

'* *

.

‘as* >.

CM*

.in

Hk

' vi

FIGURE 1 . — Photomicrographs of unreduced embryo sac development of the Panicum type in ovules of Cenchrus ciliaris. A, B, Spies

5240 A, three embryo sacs in ovule with only two visible in this section; two nuclei in micropylar region of one embryo sac

resembling either an egg nucleus and a synergid or two synergids, with one chalazal polar nucleus; B, next section of ovule in

A, three embryo sacs visible; third nucleus of embryo sac in A in micropylar region, representing either an egg or a synergid.

C-E, Spies 5232: C, two embryo sacs per ovule; two nuclei in micropylar region, resembling an egg nucleus and a single syn-

ergid, with one chalazal polar nucleus; D, second section of ovule in C with a second polar nucleus in chalazal region of

embryo sac; E, at least four unreduced embryo sacs in mature ovule. F, Spies 5488 , at least five unreduced embryo sacs in

mature ovule. Scale bar; lOpm.

multicellular integuments. Some of the integuments

completed their development at a relatively late stage.

The archesporial cell functioned directly as the primary

megasporocyte. One or more somatic cells with promi-

nent nuclei, lying in the centre of the nucellus, also

enlarged. These cells soon obscured all traces of the

degenerated gametophyte. These cells had dense cyto-

plasms and were usually adjacent to the sporogenous

material The nuclei of the aposporous cells underwent

one to several divisions and formed a well-defined dyad.

The aposporic embryo sacs are extremely vacuolated

and mature embryo sacs usually contained four nuclei.

Some of the nuclei resembled the egg and polar cells of

a reduced embryo sac. These embryo sacs usually

included one polar nucleus, an egg nucleus and two syn-

ergid cells (Figure 1A, B). Some embryo sacs included

an egg nucleus, one synergid cell and two polar nuclei

(Figure 1C, D). These unreduced embryo sacs were clas-

sified as four nucleated Panicum type aposporic embryo

sacs (Figure 1 A, D).

Bothalia 30, 1 (2000)

105

The total number of aposporic embryo sacs per ovule

varied among the specimens studied (Figure IE, F). A

maximum of eight aposporic embryo sacs per ovule was

observed ( Spies 5239). These multiple embryo sacs were

usually concentrated in the central part of the ovule and,

at maturity, occupied most of the former region of the

nucellus (Figure IE, F).

DISCUSSION

The mechanism of apospory in C. ciliaris involves the

development of the embryo from an unreduced nucleus,

in an embryo sac derived from a somatic cell in the

ovary. These aposporous embryo sacs may develop in the

nucellus of the ovule or in some species in the integu-

ments and ovary wall (Bashaw & Hanna 1990).

Apospory is the apomictic mechanism most common in

the grasses, particulary in the tribe Paniceae, which

accounts for more than 95% of known apomictic species

(Bashaw & Hanna 1990). The origin of aposporous cells

in the grass ovule is quite different from the normal pat-

tern of sexual megasporogenesis (Bashaw 1980). Early

development of the megaspore mother cell is usually

identical in aposporous and sexual ovules. In both cases,

the megaspore mother cell differentiates in the hypoder-

mal layer of the nucellus in the micropylar region, during

the enlargement stage of the young ovule. Meiosis in

both aposporous and sexual ovules generally results in a

linear tetrad. The similarity between apospory and sexu-

al embryo sac development ends at this point (Bashaw &

Hanna 1990).

Aposporic development in C. ciliaris is initiated by

the unusual enlargement of one or more somatic (nucel-

lus) cells. These cells, in comparison to the normal

nucellar cells, usually have prominent nuclei and dense

cytoplasms. The nucleus of the aposporous cell initially

undergoes one to several mitotic divisions. The degree of

differentiation of aposporous embryo sacs in C. ciliaris

varies. In some grass species, one or more of the

aposporous sacs may develop to the extent that they

closely resemble the typical sexual embryo sac (Bashaw

& Hanna 1990). This is not the case in C. ciliaris , for a

four-nucleated embryo sac was most often observed

(Figure 1A-D). This included a polar nucleus, an egg

nucleus and two synergid cells (Figure 1A, B) or one

egg, one synergid and two polar nuclei (Figure 1C, D).

Antipodal cells were completely lacking. The embryo

sacs observed were, therefore, classified as unreduced

aposporic embryo sacs, of the Panicum type. This find-

ing corresponds with previous reports by various authors

(Bashaw & Holt 1958; Brown & Emery 1958; Bashaw

1962) and is typical of the Panicoideae (Bashaw &

Hanna 1990; Mogie 1992). The number of embryo sacs

of nucellar origin varied from two to eight in Spies 5239.

Occasionally observed, was a cluster of unreduced

embryo sacs which could not be accurately counted.

Various authors have reported a number of ovules in

C. ciliaris containing a single aposporous embryo sac in

the nucellus (Fisher et al. 1954; Bashaw & Hanna 1990).

According to their reports, these sacs were centrally

located, in the micropylar region of the ovule, and occu-

pied more or less the same location and total area as the

sexual gametophyte in sexual plants of this species. This

type of embryo sac development was not observed in the