Bothalia

A JOURNAL OF BOTANICAL RESEARCH

Vol. 27,1

May 1997

TECHNICAL PUBLICATIONS OF THE NATIONAL BOTANICAL INSTITUTE,

PRETORIA

Obtainable from the National B otanical Institute, Private Bag X 1 0 1 , Pretoria 000 1 , 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 well known botanical artists have

contributed to the series, such as Cythna Letty (over 700 plates), Kathleen Lansdell, Stella Gower,

Betty Connell, Peter Bally, Fay Anderson, Ellaphie Ward-Hilhorst and Gillian Condy. 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 1^9 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 palaeollora on a pattern comparable to that of the Flora of southern Africa. Much of the

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

Now available:

Molteno Fonnation (Triassic) Vol. 1. Introduction. Dicroidium, by J.M. & H.M. Anderson.

Molleno Formation (Triassic) Vol. 2. Gymnosperms (excluding Dicroidium), by J.M. &

H.M. Anderson.

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

Anderson. Obtainable from: A. A. Balkema Marketing, Box 317, Claremont 7735, RSA.

BOTHALIA

A JOURNAL OF BOTANICAL RESEARCH

Volume 27,1

Scientific Editor: O.A. Leistner

Technical Editor: B.A. Momberg

NATIONAL

§8 O T A N I CAL

INSTITUTE

2 Cussonia Avenue, Brummeria, Pretoria

Private Bag X 101, Pretoria 0001

ISSN 0006 8241

May 1997

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

CONTENTS

Volume 27,1

1 . Notes on Plectrantlms (Lamiaceae) from southern Afriea. E. J. VAN JAARS VELD and T.J. EDWARDS 1

2. Live new species of Lachenalia (Hyacinthaceae) from arid areas of South Africa. G.D. DUNCAN . 7

3. Studies in the liverwort genus Fossombronia (Metzgeriales) from southern Africa. 1 . Three new species

from Northern Province, Gauteng and Mpumalanga. S.M. PEROLD 17

4. Studies in the liverwort genus Fossombronia (Metzgeriales) from southern Africa. 2. An amendment

to three species from Western Cape, described by S.W. Arnell. S.M. PEROLD 29

5. Studies in the liverwort genus Fossombronia (Metzgeriales) from southern Africa. 3. An amendment

to F. spinifolia. S.M. PEROLD 39

6. Notes of African plants:

Apiaceae (Umbelliferae). A new name for a South African Peucedanum. B.L. BURTT 51

Asteraceae. New combination in Dicoma. S. ORTIZ, J. RODRIGUEZ-OUBINA and I. PULGAR 48

Boraginaceae. The taxonomic status of Lobostemon horridus. M.H. BUYS and J.J.A. VAN DER

WALT 55

Eabaceae. A survey of antipodals in the gametophyte of the tribes Podalyrieae and Liparieae. A.L.

SCHUTTE 43

Proteaceae. A new species of Leucadendron from the western Little Karoo. J.P. ROURKE .... 52

Rubiaceae. A new species of Vangueria from the Soutpansberg. N. HAHN 45

Thymelaeaceae. New combinations in Lachnaea. J.B.P. BEYERS 45

Vitaceae. A new species of Rhoicissus from the Eastern Cape. E. RETIEE and E. J. VAN

JAARSVELD 49

7. Composition and biogeography of forest patches on the inland mountains of the southern Cape. C.J.

GELDENHUYS 57

8. Cytogenetic studies in some representatives of the subfamily Pooideae (Poaceae) in South Africa. 3.

The tribe Poeae. J.J. SPIES, S.M.C. VAN WYK, I.C. NIEMAN and E.J.L. LIEBENBERG. ... 75

9. Comparative field performance of three different gas exchange systems. G.E. MIDGLEY, M. VESTE,

D.J. VON WILLERT, G.W. DAVIS, M. STEINBERG and L.W. POWRIE 83

10. Obituary: Leslie Charles Leach ( 1909-1996). R.H. ARCHER 91

11. Book reviews 97

Digitized by the Internet Archive

in 2016

https://archive.org/details/bothaliavolume2727unse

Bothalia27,l: 1-6 (1997)

Notes on Plectranthus (Lamiaceae) from southern Afiica

E.J. VAN JAARSVELD* and T.J. EDWARDS**

Keywords: comb, nov., Lamiaceae, Plectranthus spp., southern Africa, spp. nov., stat. nov., taxonomy

ABSTRACT

Four new Plectranthus taxa from South Africa are described: P malvinus Van Jaarsv. & T.J. Edwards, P. saccatus subsp.

pondoensis Van Jaarsv. & S.Milstein, P. purpuratus subsp. tongaensis Van Jaarsv. & T.J.Edwards and P. purpuratus subsp.

montanus Van Jaarsv. & T.J.Edwards. P. aliciae (Codd) Van Jaarsv. & T.J.Edwards and P. lucidus (Benth.) Van Jaarsv. &

T.J.Edwards are given new status, and P. pentheri (Giirke) Van Jaarsv. & T.J.Edwards is transferred to this genus from Coleus

and recognized as a species.

INTRODUCTION

Plectranthus includes 45 southern African species

which are found in the subtropical forests and savannas

of the summer rainfall region. The genus was revised by

Codd (1975, 1985) who recognized 44 species in southern

Africa. Subsequent collections and additional information

have resulted in the recognition of another species and

three new infraspecific taxa which are described here. Two

varieties are raised to species level and one species, origi-

nally described under Coleus, is transferred to Plectran-

thus. The newly described taxa belong to the subgenus

Plectranthus.

1. Plectranthus malvinus Van Jaarsv. & T.J.Ed-

wards sp. nov. a P. ciliato E.Mey. ex Benth. folds firmis,

coriaceis, succulentis, marginibus serratis pagina foliorum

strigosa floribusque malvinis differt.

TYPE. — Eastern Cape, 3129 (Port St Johns); Mount Sul-

livan, (-DA), E. van Jaarsveld & Bingham 10522 (NBG,

holo.).

Decumbent, strigose, mat-forming herb; roots shallow,

fibrous. Stetns 4-angled, purple-green, ± 3 mm in diame-

ter, strigose (white multicellular hairs), punctate; inter-

nodes 10-40 mm apart. Leaves fleshy, firm, ovate to

obovate, 40-90 x 30-50 mm, strigose, serrate with 8-10

pairs of teeth; abaxial surface strigose, veins purple,

densely strigose, hairs white, gland dots colourless,

sunken; base cuneate, apex acute; petiole 5-10 mm long,

purple, densely villose on adaxial surface, decurrent. In-

florescence a raceme or lax panicle, 180-210 mm long;

cymes 3-flowered, 10-15 mm apart; bracts ovate-lanceo-

late, 7x2 mm long; pedicels ± 10 mm long. Calyx 4

mm long (enlarging to 10 mm), upper lobe ovate, 2 mm

long (5 mm in fruit), lower lobes 4, linear, 1.5 mm long

(4 mm after flowering). Corolla ± 12 mm long, pink (vio-

let group 84c), tube laterally compressed, 6 mm long, sac-

cate at base and 3 mm deep, narrowing to 2 mm at throat.

upper lip 7 mm long, 2-lobed, lateral lobes 3 mm long, lower

lip boat-shaped, 5 mm long. Nutlets brown to black, ovoid,

1.5 X 1.0 mm. Flowering time: March to May. Eigure 1.

This species was collected along forest fringes on

quartzitic sandstones of Mount Sullivan (W end). It has

been cultivated as a ground cover at Kirstenbosch Na-

tional Botanical Garden for a number of years under the

name P. ciliatus ‘Bingham’. Codd (1975) related this

* National Botanical Institute. Kirstenbosch, Private Bag X7, Claremont

7735, Cape Town.

** Botany Department, University of Natal, P.O.Box 375, Pieter- EICURE 1. — P. malvinus, E. van Jaarsveld & Bingham 10522, Port St

maritzburg 3200. Johns, Eastern Cape. Habit, x 0.5. Scale bar: 12 mm. Artist: Vicky

MS. received: 1996-04-20. Thomas.

2

Bothalia27,l (1997)

taxon both to P. strigosus and P. ciliatus. It is distinguished

by its firm, succulent, ovate to obovate leaves which are

serrate, have densely pilose purple veins and are densely

punctate underneath. The latter feature immediately sepa-

rates it from P strigosus or P lucidus which have red

gland dots. The flowers are reminiscent of P ciliatus but

are generally smaller and are an attractive mauve. P cili-

atus has soft leaves and usually bears white flowers.

P malvinus is one of the endemics of the region of

quartzitic sandstone in KwaZulu-Natal and northern East-

ern Cape. It is found on Mount Thesiger, Mount Sullivan

and in adjacent territory on forest margins. Associated spe-

cies in the habitat include Mitriostigma axillare, Drimiop-

sis maculata and P ciliatus.

2. Plectranthus purpuratus Harv., Thesaurus cap-

ensis 1: 53, t. 83 (1859). Type: ex Hort., Kew, from seed

sent from Port Natal [KwaZulu-Natal: Durban], R. Vause

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

Procumbent to decumbent, perennial, succulent herb,

up to 200 mm high; roots fibrous. Stems 4-angled, suc-

culent. Leaves broadly trullate, broadly ovate, obovate to

subrotund, 10-15 x 10-15 mm, entire to variably serrate

to crenate with 2 or 3 pairs of teeth, strigose to sub-

glabrous; lower surface occasionally purplish, rubropunc-

tate; tip acute, base truncate to broadly cuneate; petioles

3-15 mm long. Raceme 30-290 mm long, racemose, oc-

casionally with a pair of side branches; cymes 3-flowered,

5-10 mm apart; bracts linear-lanceolate, 3-4 mm long,

persistent beyond flowering stage; pedicels 3-5 mm long.

Calyx 3 mm long; fruiting calyx 5 mm long. Corolla

12-13 mm long, white or pale mauve, tube 5-8 mm long,

basally ventricose, constricted about 3 nun from base and

flared at throat, upper lobes emarginate, 3-5 mm long,

lateral lobes ± 2 mm long, lower lip boat-shaped, 5-7 mm

long. Nutlets brown or black, 1.5 mm long.

Plectranthus purpuratus is widely distributed in the

eastern parts of southern Africa from Durban through

KwaZulu-Natal and Swaziland to Mpumalanga (Eastern

Transvaal), occurring in rocky gorges in savanna and

grassland. The species has a diagnostic, medial constric-

tion of the corolla tube which distinguishes it from P

strigosus, P lucidus and P. oertendahlii.

Key to subspecies of P purpuratus

la Plants erect to decumbent; leaves entire to obscurely crenate,

succulent, subrotund, subglabrous, subimbricate

subsp. purpuratus

lb Plants procumbent; leaves not distinctly succulent, strigose,

lax:

2a Leaves large, trullate (petioles 12-15 mm long), distinctly

serrate, with 3 or 4 pairs of teeth subsp. tongaensis

2b Leaves broadly ovate to obovate (petiole 6-12 mm long),

shallowly serrate with 2 or 3 pairs of teeth . . subsp. rrumtanus

2a. Plectranthus purpuratus Harv. subsp. purpuratus

Stems erect to decumbent, succulent. Leaves subimbri-

cate, subrotund to broadly ovate, succulent, 15^5 x

15-38 mm, grey-green, entire, occasionally shallowly cre-

nate, with 3 pairs of teeth, subglabrous, rubropunctate be-

neath, apex rounded, base truncate to cuneate. Raceme

30-120 mm long, often with a pair of side branches;

cymes 3-flowered, 5-10 mm apart; bracts ovate-lanceo-

late, 2 mm long, persistent; pedicel 2-3 mm long; fruiting

calyx 5 mm long. Corolla 10-11 mm long, white, tube ±

4 mm long, constricted in middle. Nutlets brown, 1 mm

long. Figure 2.

P. purpuratus subsp. purpuratus is confined to the Dur-

ban-Pietermaritzburg region of central KwaZulu-Natal

(Figure 3) and occurs on rocky outcrops or south-facing

cliffs in bushveld. It is commonly found in association

with species such as Aloe arborescens, Gasteria croucheri

and Plectranthus hadiensis var. tomentosus.

Subsp. purpuratus is distinguished from subsp. mon-

tanus and subsp. tongaensis by its decumbent habit and

succulent, glabrescent, subrotund leaves which are

crowded and often subimbricate. Its leaves are entire or,

rarely, obscurely crenate. The typical subspecies was

named for its purple abaxial leaf surfaces. The plants have

an erect to decumbent habit.

2b. Plectranthus purpuratus subsp. tongaensis

Van Jaarsv. & T.J. Edwards subsp. nov. a subspecie typica

FIGURE 2.—P. purpuratus subsp. purpuratus, E. van Jaarsveld 9843,

Mamba Valley, KwaZulu-Natal. Plant, x 0.5. Scale bar: 10 mm.

Artist; Vicky Thomas.

Bothalia27,l (1997)

3

FIGURE 3. — Distribution of P. lucidus, •; P. purpuratus subsp. purpu-

ratus, ■; P. purpuratus subsp, tongaensis, ★; P. purpuratus

subsp. montanus. A; P. pentheri, O.

habitu procumbenti foliisque trullatis non congestis, mar-

ginibus serrratis paribus dentium tribus vel quatuor differt.

TYPE. — KwaZulu-Natal, 2732 (Ubombo): Kosi Bay,

(-BB), Van Jaarsveld 12206 (NBG, holo.).

Procumbent, pubescent to glabrescent, succulent herb,

rooting at nodes. Stems 4-angled, purplish to green, 2-3

mm in diameter, strigose (purplish to white hairs),

rubropunctate; intemodes 20-30(-60) mm long. Leaves

ovate to broadly ovate, 30-34 x 15-18 mm, green or pur-

ple tinged, coarsely serrate, with 3 or 4 pairs of shallow

teeth, abaxial surface strigose to glabrescent, rubropunc-

tate, veins densely strigose, apex acute, base broadly cun-

eate; petiole 12-15 mm long, strigose, rubropunctate.

Raceme 140-290 mm long, occasionally with a pair of

basal side branches; cymes 3-flowered, 8-20 mm apart;

bracts ovate-lanceolate, ± 3 mm long; pedicel ± 4 mm

long. Calyx ± 3 mm long, enlarging to 6 mm, upper lip

ovate, 2 mm long (± 4 mm after flowering), lower lobes

4, linear, ± 1.5 mm long (± 3.5 mm long after flowering),

densely strigose (flushed with blue). Corolla 12-13 mm

long, white, bilabiate, tube basally saccate, 7 mm long,

upper lip 5-6 mm long, 2-lobed, lateral lobes 2-3 mm

long, lower lip boat-shaped, 5-6 mm long. Nutlets black,

ovoid, 1.5 X 1.0 mm. Flowering time: March to May.

Figure 4A.

FIGURE 4. — P. purpuratus-. A, subsp.

tongaensis, E. van Jaarsveld,

Van der Walt & Crous 50,

Kosi Bay, KwaZulu-Natal; B,

subsp. montanus, E. van

Jaarsveld 3386, Barberton,

Mpumalanga. A, B, habit,

X 0.7. Scale bars: 14 mm. Art-

ist: Vicky Thomas.

4

Bothalia27,l (1997)

FIGURE 5, — P. saccatus: A-C, subsp. saccatus; D, subsp. pondoensis,

E. van Jaarsveld 2201, Oribi Gorge, KwaZulu-Natal. Habit, x

0.5. Artist: Vicky Tbomas.

This subspecies is locally common or sporadic in

coastal sand forest from St Lucia to Kosi Bay in northern

KwaZulu-Natal (Figure 3). Plants are often found in as-

sociation with Crassula expansa subsp. fragilis, Plectran-

thus petiolaris and Cussonia arenicola.

2c. Plectranthus purpuratus subsp. montanus Van

Jaarsv. & TJ. Edwards subsp. nov. a subspecie typica foliis

non congestis laminis obovatis marginibus vix serratis

paribus dentium duobus differt.

TYPE. — Swaziland, 2631 (Mbabane); Mbabane, (-AC),

Compton 32207 (NBG, holo.).

Procumbent, pubescent, succulent herb; roots shallow,

fibrous. Stems 4-angled, purplish to green, 2.5 mm in di-

ameter, strigose, rubropunctate; internodes 5-12 mm long.

Leaves ovate to broadly ovate, 12-15 x 10-15 mm, green

to purplish, teeth (2 pairs) shallow, serrate to crenate-den-

tate, abaxial surface sparsely strigose, rubropunctate, veins

densely strigose, base cuneate, apex obtuse to acute; peti-

ole 6-12 mm long, strigose, rubropunctate. Raceme

30-100 mm long, occasionally with a pair of side

branches at base; cymes 3-flowered, 5-12 mm apart;

bracts ovate-lanceolate, ± 2 mm long; pedicels 2-4 mm

long. Calyx 3 mm long, enlarging to 6 mm; upper lobe 2

mm long (4 mm after flowering), ovate and lower lobes

4, linear, ± 1.5 mm long (3.5 mm after flowering), densely

strigose. Corolla 14-15 mm long, white, tube 7 mm long,

upper lip 5-6 mm long, 2-lobed, lower lip boat-shaped,

3-4 mm long. Nutlets black, ovoid, 1.5 x 1.0 mm. Flow-

ering time: March to May. Figure 4B.

The obovate leaves of this subspecies are usually cov-

ered in a grey indumentum. The lamina margin is ob-

scurely serrate with two pairs of teeth. The subspecies

occurs along the Mpumalanga (Eastern Transvaal) Drak-

ensberg (Figure 3), usually among rocks in grassland and

forest margins. Associated species include Aloe suprafo-

liata and R verticillatus.

3. Plectranthus saccatus Benth. subsp. pondoensis

Van Jaarsv. & S.Milstein subsp. nov. a subspecie typica

foliis manifesto succulentis, inflorescentia brevi, tuboque

corollae brevi 6-7 mm longo differt.

TYPE. — KwaZulu-Natal, 3030 (Port Shepstone): Oribi

Gorge, (-CA), E. van Jaarsveld 2201 (PRE, holo.).

Trailing, glutinous succulent herb. Stems obscurely

four-angled; young stems 2-5 mm in diameter, purplish,

minutely glandular pubescent, becoming glabrous; older

stems striate; intemodes 6-40 mm long. Leaves ovate to

broadly trullate, 12-27 x 10-28, both surfaces minutely

glandular pubescent, abaxial surface slightly costate; peti-

ole 5-25 mm long. Racemes secund, 30-50 mm long;

bracts linear, ± 1 mm long, caducous; pedicels 6-7 mm

long. Calyx 3-6 mm long (enlarging to 8 mm), upper lip

ascending, ovate, 1 mm long, lower lobes 4, narrowly

linear-lanceolate, 1 mm long. Corolla saccate, tube 6-20

X 3-7 mm, upper lip 2-lobed, 9-13 x 7-12 mm, erect,

lobes folded back, lower lip horizontal or slightly droop-

ing, blue to pale mauve-pink, inner surface speckled with

purple (often in 4 rows), lower lip 8 mm long, speckled

with purple. Stamens ± 14 mm long, declinate in lower

lip, free for 8 mm; anthers purple, bent upwards. Style

11-12 mm long. Figure 5D.

P. saccatus subsp. pondoensis is distinguished from

the typical subspecies by its distinctly succulent leaves

and decumbent to procumbent habit with flexible stems

up to 4 m long. The secondary growth of the species is

anomalous with many broad collenchymatous rays which

impart flexibility.

The two subspecies are ecologically separated, with

subsp. pondoensis being common in scrub along gorge

lips, and subsp. saccatus being common in forest. Suc-

culence occurs in a number of cliff-dwelling species {P.

ernstii, P purpuratus and P strigosus). The two sub-

species maintain their vegetative characteristics under

uniform cultivation. No hybrids between them have

been observed.

P saccatus subsp. pondoensis is restricted to the quartzitic

sandstone cliff faces of the Msikaba River of northern East-

ern Cape (northern Transkei) and southern KwaZulu-Natal

(Figure 6). Associated species include Cryptocarya wyliei

and Crassula sarmentosa var. sarmentosa.

P saccatus subsp. saccatus is very variable, with sev-

eral local forms (Figure 5A-C). The recognition of var.

longitu- bus Codd within this species is contentious. The

corolla tube length is highly variable and appears to form

a continuum.

Bothalia27,l (1997)

5

FIGURE 6. — Distribution of P. aliciae, •; P. saccatus subsp. pondoen-

sis, □.

4. Plectranthus aliciae (Codd) Van Jaarsv. &

TJ. Edwards stat. nov. Type: Transkei, Butterworth, Ken-

tani, Pegler 909 (PRE, holo.!).

P madagascariensis var. aliciae Codd in Bothalia 11: 404 (1975).

Erect to decumbent, soft, semi-succulent herb to 400

mm high. Leaves 25^0 x 22^0 mm, broadly ovate,

sparsely strigose, adaxial surface rubropunctate, serrate,

teeth in 3 or 4 pairs, base truncate, apex acute. Raceme

up to 130 mm long; cymes 3-6- flowered. Calyx 3 mm

long (enlarging to 5 mm). Corolla white, 5-14 mm long,

tube widening to throat. Nutlets 1 mm long, light to dark

brown. Figure 7.

R aliciae is distinguished by its ascending habit, mem-

branous, sparsely pubescent lamina wih 2 or 3 pairs of

teeth and short corolla tube. It is distributed from East

London to southern KwaZulu-Natal (Figure 6) occurring

in subtropical lowland forest. Associated species at Oribi

Gorge include Cryptocarya woodii, P. oertendahlii, P.

oribiensis, P zuluensis and P petiolaris.

P aliciae is sympatric with P. madagascariensis and is

therefore accorded specific status. P madagascariensis is a

species of open subtropical thickets and forms dense stands,

whereas P aliciae is a forest dweller and retains its habit in

cultivation. P aliciae commemorates Alice Pegler

(1861-1929), teacher and amateur botanist in the E Cape.

5. Plectranthus lucidus Van Jaarsv. & TJ. Edwards

stat. nov. Type: Eastern Cape, Bathurst, Burchell 3924,

(K!, holo.).

P. slrigosus Benth. var. lucidus Benth. 12; 68 (1848).

Procumbent, succulent herb, rooting at nodes, glabrous

or sparsely pubescent; roots fibrous. Stems 4-angled, pur-

plish to green, 2.5 mm in diameter, strigose (white mul-

ticellular hairs), rubropunctate; internodes 5-20 mm long.

Leaves ovate to broadly ovate, 12-15 x 10-15 mm, serrate

to crenate-dentate, with 3 or 4 pairs of teeth, green to

purplish, abaxial surface sparsely strigose, veins densely

strigose, rubropunctate, base cuneate, apex obtuse to

acute; petioles 5-15 mm long, strigose and rubropunctate.

Raceme simple, 30-100 mm long, rarely with a pair of

side branches; cymes 1 -flowered, verticillasters 5-10 mm

apart; bracts ovate-lanceolate, 2-3 mm long; pedicel 4-5

mm long. Calyx at flowering 3-4 mm long, enlarging to

7 mm, consisting of a large, ovate upper lip 2 mm long

(4 mm after flowering) and 4 subulate lower lobes, 3.5

mm long (5 mm long after flowering), densely strigose,

with longer purplish hairs. Corolla 9-11 mm long, white

or mauve; tube straight, 5 mm long, basally saccate, con-

stricted distally; upper lip 5 mm long, 2-lobed, upper lobes

3 mm long, lateral lobes 2 mm long, with long multicel-

lular white hairs, lower lip boat-shaped, margins strongly

involute. Nutlets brown to black, ovoid, 1.5 x 1.0 mm.

Flowering time: March to May. Figure 8.

Plectranthus lucidus forms mats on stabilized coastal

sand dunes below trees (with mainly Mimusops obovata

and Allophyllus natalensis), from Bathurst in the south-

eastern part to Port St Johns in the northeastern part of

Eastern Cape. It shares its habitat with other shade-loving

coastal plant species such as P. madagascariensis and

Scadoxus membranaceus.

P. lucidus is closely related to P. verticillatus (L.f.)

Druce and P. strigosus Benth. (section Plectranthus series

Plectranthus)-, it is distinguished by its 2-flowered ver-

ticillasters and short corolla (9-11 mm) of which the lower

lip margins are conspicuously inrolled. In both P. strigosus

and P. lucidus the corolla is constricted at the throat, unlike

P. verticillatus where the corolla is linear.

6. Plectranthus pentheri (Giirke) Van Jaarsv. &

T.J. Edwards comb. nov. Type: Eastern Cape, Albany Dis-

FIGURE 7. — P. aliciae, Umtamvuna, KwaZulu-Natal. E. van Jaarsveld

& Campher 117. Plant, x 0.5; habit much reduced. Scale bar: 20

mm. Artist: Vicky Thomas.

6

FIGURE 8. — P. lucidus, E. van Jaarsveld 3827, Port St Johns, Eastern

Cape. Plant, x 0.6; flowers and calyx, x 1.3. Artist: Vicky

Thomas.

trict, Breakfast Vlei, Kwok in Herb. Penther 1716 (W,

holo.; PRE!).

Coleus pentheri Giirke in Annalen des Naturhistorischen Museums in

Wien 20: 48 (1905).

Decumbent, succulent herb up to 100 mm tall; basal

tuber 30 X 15 mm, white. Stems 4-angled, 2-4- mm in

diameter, sparsely pilose, punctate, glands orange; inter-

nodes 3-6 mm long. Leaves conduplicate, succulent, firm,

ovate to subrotund, 9-14 x 9-14 mm, entire to dentate-

serrate in upper half, teeth shallow, in 4 or 5 pairs; sparsely

pubescent, punctate, glands orange, base broadly cuneate,

apex obtuse; petiole 2 mm long. Raceme 50-70 mm long,

simple; cymes 3-llowered, 5-10 mm apart; bracts broadly

ovate to subrotund, mucronate, 4x4 mm, initially imbri-

cate, caducous; pedicels 5 mm long. Calyx 3 mm long

(enlarging to 5 mm), upper lip broadly ovate, 2.0 x 2.5

mm, lower lobes 4, linear, 2 mm long. Corolla 15 mm

long, white; tube laterally compres.sed, slightly geniculate,

5.0 X 1.5 mm, expanding to 3 mm at throat, upper lip 3

mm long, 2-lobed, lateral lobes 2.5 mm long, lower lip

Bothalia27,l (1997)

FIGURE 9. — P. pentheri, E. van Jaarsveld, Sardien & Peterson 13774,

Kei River, Eastern Cape. Plant and rootstock, x 0.5. Scale bar: 20

rrun. Artist: Vicky Thomas.

boat-shaped, 9 mm long. Nutlets brown to black, ovoid,

1.5 X 1.0 mm. Flowering time: March to May. Figure 9.

Codd (1985) regarded P. pentheri (as Coleus pentheri) as

a synonym of P neochilus. The two entities, however, differ

markedly. Both belong to subgenus Calceolanthus but P.

pentheri is at once distinguished by its oblong tuberous roots

and white flowers. It also lacks the strong aroma of the

closely related P neochilus, which is a widespread species

with fibrous roots and mauve to purple flowers.

P. pentheri is known from a few gatherings. It grows

with Crassula ericoides on lithosols of granite boulders

in grassland.

ACKNOWLEDGEMENTS

Dr O. A. Leistner is thanked for preparing the Latin

diagnoses and editing the text.

REFERENCES

BENTHAM, G. 1848. Labiatae. In A.P. de Candolle, Prodromus 12: 55.

CODD L.E., 1975. Plectranthus (Labiatae) and allied genera in southern

Africa. Bothalia 11: 371-442.

CODD, L.E., 1985. Lamiaceae. In O.A. Leistner, Elora of southern Africa

28,4: 137-172.

GURKE, R E A M. 1905. Coleus pentheri. Annalen des Naturhistoris-

chen Museums in Wien 20: 48.

HARVEY, W.H. 1895. Thesaurus capensis 1 : 53, t. 83.

Bothalia27,l; 7-15 (1997)

Five new species of Lachenalia (Hyacinthaceae) from arid areas of South

Africa

G.D. DUNCAN*

Keywords: Hyacinthaceae, Lachenalia, new species. South Africa

ABSTRACT

Five new species of Lachenalia are described: L. aurioliae G.D. Duncan from the Little Karoo and Great Karoo, L. obscura

Schltr. ex G.D. Duncan from Namaqualand, the Kamiesberg, the western Great Karoo and the Little Karoo, L. inconspicua

G.D. Duncan from the Kamiesberg, western Bushmanland and southern Namaqualand, L. marlothii W.F.Barker ex G.D. Duncan

from the Calvinia-Sutherland region of the western Great Karoo, and L. xerophila Schltr. ex G.D. Duncan from northwestern

and central Namaqualand, and western Bushmanland.

INTRODUCTION

This is the second in a series of papers on new species

of Lachenalia, and serves as a continuation of the recent

work of W.F.Barker (Barker 1978, 1979, 1983a &b, 1984,

1987, 1989) and the current author (Duncan 1987, 1988a

& b, 1989, 1992, 1993, 1994, 1996), towards a revision

of the genus. Material of a number of unpublished

Lachenalia species has languished in local and foreign

herbaria for many years, and in some instances manuscript

names accompany this material; the abovementioned pa-

pers serve as a means of validating these names, where

it is considered expedient. Furthermore, new species are

published which have come to light in recent years. In

most instances, the new species are described from both

dried and living material, the dried material extracted

mainly from the extensive Lachenalia holdings at the

Compton Herbarium at Kirstenbosch (NBG), and the liv-

ing material both from the wild and the large collection

maintained in the nursery at Kirstenbosch National Bo-

tanical Garden.

Lachenalia aurioliae G.D. Duncan, sp. nov. L.

schelpei W.F Barker affmis ob flores similares subspicatos

oblongo-urceolatos, stamina parum exserta foliaque

lanceolata; sed segmentis interioribus perianthii multo

longioribus, staminibus declinatis folioque conduplicato

immaculate plerumque costa distincta differt.

TYPE. — Western Cape, 3222 (Beaufort West): hillside

facing Hesperus Old Age Home, Beaufort West, (-BC),

26-6-1984, A. Batten 468 (NBG, holo.!; PRE).

Deciduous, winter-growing geophyte 45-120 mm

high. Bulb subglobose, 20-25 mm in diam., surrounded

by thin, pale to dark brown spongy outer tunics, produced

into a short neck. Leaves 1 or 2, partially to fully condu-

plicate, lanceolate to ovate-lanceolate with a distinct mid-

rib and faint depressed longitudinal veins on upper

surface, yellowish green, plain or faintly spotted with dull

*National Botanical Institute, Kirstenbosch, Private Bag X7, Claremont

7735, Cape Town.

MS. received: 1996-11-08.

green or purple on upper surface, with darker green

blotches and transverse bands on lower surface, merging

into brownish magenta transverse bands on the loosely

clasping leaf bases. Inflorescence subspicate, moderately

dense, few to many-flowered, 35-90 mm long, with a

short sterile tip; peduncle suberect to erect, fairly sturdy,

45-140 mm long, pale green with distinct, irregularly

scattered brownish magenta blotches in lower half, and

minute spots in upper half; rachis pale green in lower half,

shading to pale greenish brown in upper half, mottled with

tiny brownish magenta specks; pedicels absent or up to

1mm long; bracts ovate to lanceolate, greenish white, with

or without pale brownish magenta tips, 1^ x 1-3 mm.

Flowers patent to cernuous, oblong-urceolate, pale bluish

white to yellowish white, fading to dull reddish brown;

outer perianth segments swollen at base, oblong,

6- 7 X 4-5 mm, minutely spotted with dark blue on keel

and near apex, pale bluish white to yellowish white with

distinct dull reddish brown to purplish brown gibbosities;

inner perianth segments obovate, 9-10 x 5 mm, translu-

cent white with a dark blue or brownish blue keel, with

or without a dull reddish purple zone near apex. Stamens

very slightly exserted, declinate; filaments white, 8-9 mm

long. Ovary ovoid, pale green, 3x2 mm; style white,

7- 8 mm long, protruding beyond stamens as ovary en-

larges. Capsule ovoid, membranous, 6-8 x 5 mm. Seed

ovoid, shiny black, 2 mm long, with a ridged terminal

arillode 1 mm long. Figures lA; 2 & 3.

Etymology: named after Mrs Auriol Batten, whose col-

lection forms the type material of this species, in recog-

nition of her contribution to the knowledge of South

Africa’s flora through her superb watercolour paintings

which have illustrated several books authored or co-

authored by her.

Diagnostic characters

L. aurioliae is characterised by a subspicate inflorescence

of cernuous or patent, oblong-urceolate flowers with the

translucent white inner perianth segments distinctly longer

than the outer ones. The declinate stamens are very slightly

exserted beyond the tip of the perianth and the plant usually

has two lanceolate or ovate-lanceolate leaves which are par-

tially to fully conduplicate, usually with a distinct midrib.

8

Bothalia27,l (1997)

FIGURE 1 . — A, Lachenalia aurioliae. Batten 468\ B, L. obscura, Duncan 108; C, L. inconspicua. Duncan 259. Scale bars: 10 mm.

The upper leaf surface may be plain or spotted with dull

green or purple, and the lower surface with or without darker

green blotches and transverse bands.

L aurioliae is related to L schelpei W.F.Barker, (at pre-

sent known only from the Hantam Mountains at Calvinia in

the Northern Cape), in that both species have patent or cer-

nuous oblong-urceolate flowers borne on very short pedicels,

with very slightly exserted stamens and similar lanceolate

leaves. In L. schelpei, the flowers are subtended by conspicu-

ous, long, narrowly lanceolate bracts, and the inner perianth

segments are only slightly longer than the outer segments,

as compared to the very short, ovate to lanceolate bracts and

distinctly longer inner perianth segments of L. aurioliae. Fur-

thermore, the stamens of L. schelpei are arranged symmet-

rically around the rim of the mouth of the perianth, whereas

in L. aurioliae the stamens are distinctly declinate. The

flower colour of L. aurioliae varies from pale bluish white

to yellowish white whereas L schelpei has greenish white

flowers. The two species are geographically clearly sepa-

rated. Flowering time: June to August.

Distribution and habitat

Material of L. aurioliae was collected for the first time

by C. Thorne in October 1935 at Leeuwkloof in the Nu-

weveld Mountains north of Beaufort West. It is a variable,

early flowering species with a fairly wide distribution in

the southern Great Karoo and the Little Karoo, where it

is found in a variety of arid habitats ranging from sandy

river courses in full sun to south-facing hill slopes in

heavy soil. At a locality near Whitehill Station in the Little

Karoo it grows together with the very distinctive L. white-

hillensis W.F.Barker, another Karoo endemic species

which flowers much later in the year.

Material examined

NORTHERN CAPE. — 3221 (Merweville): between Boschluiskloof

and Prince Albert, (-AB), July 19.‘54, Stokoe s.n. (SAM); west of Steen-

bokkraal, (-BA), June 1986, Bayer 5189 (NBG).

WESTERN CAPE. — 3222 (Beaufort West): Leeuwkloof, Nieuweveld,

(-BA), Oct. 1935, Thome s.n. (SAM); 16 km S of Beaufort West, (-BC),

July 1986, Van Zijl s.n. (NBG); hillside facing Hesperus Old Age Home,

Beaufort West, (-BC), June 1984, Batten 468 (NBG). 3320 (Montagu):

in river course at Whitehill Station, (-BA), Aug. 1986, Duncan 243

(NBG); Ratelfontein, between Montagu and Kareevlakte, (-CB), July

FIGURE 2. — Holotype of Lachenalia aurioliae, Batten 468.

Bothalia27,l (1997)

9

1954, Lewis 4397 (SAM); Wildehondekloof Pass, 36 km E of Montagu,

(-CC), Aug. 1974, Nordenstam & Lundgren 1191 (NBG). 3321 (Ladi-

smith): Mannshoop Farm, 0.8 km from homestead, (-CA), July 1982,

Laidler 186 (NBG); N of Rooiberg, Ladismith, (-CB), Aug. 1954, Wurts

1227 (NBG). 3322 (Oudtshoom); hills near Kammanassie, (-DB), July

1954, Lewis 4396 (SAM).

Lachenalia obscura Schllr. ex G.D. Duncan, sp.

nov. L maximiliani Schltr. ex W.F. Barker affmis ob seg-

mentos interiores perianthii similares ad apicem magen-

teos, folium lanceolatum, bulbum globosum squamis

exterioribus duris brunneis circumcinctum, seminaque

globosa; sed planta omnino grandiora floribus oblongo-

campanulatis pallide flavo-virentibus ad brunneolo-

caeruleis vel cremeis, staminibus parum exsertis foliisque

plerumque pagina inferiora fasciis distinctis viridibus,

brunneolo-purpureis et magenteis differt.

TYPE. — Northern Cape, 3119 (Calvinia): Vogelstruis

Vlakte, Calvinia Division, (-DC), 26-7-1941, R.H. Comp-

ton 11174 (NBG, holo.!).

Deciduous, winter-growing geophyte 55-380 mm

high. Bulb globose, 10-25 mm in diam., white, sur-

rounded by hard, cartilaginous pale to dark brown outer

scales, produced into a short, strawlike neck. Leaves usu-

ally 2, erect to suberect, yellowish green to dark green,

often conduplicate, 25-280 x 5-45 mm, upper leaf sur-

face with faint, depressed longitudinal veins, usually un-

marked, lower leaf surface usually heavily banded with

bright green merging into dull brownish purple and ma-

genta on clasping leaf base. Inflorescence spicate to sub-

spicate, 15-220 mm long, few to many-flowered, with a

short sterile tip; flowers often arranged in distinct whorls

of three at base of inflorescence, becoming less distinctly

whorled towards top of inflorescence; peduncle erect to

suberect, slender or sturdy, 25-100 mm long, pale to dark

green with pale to dark purplish blotches; rachis mottled

with very pale bluish purple; pedicels white, often absent

in lower half of inflorescence, but up to 2 mm long in

upper part; bracts ovate to lanceolate, greenish white,

2-4 X 1-3 mm. Flowers patent to slightly cemuous, ob-

long-campanulate, pale yellowish green to brownish blue

or cream, with or without distinct magenta tips, fading to

dull purple; outer perianth segments oblong, 6-9 x 4 mm,

pale yellowish green to brownish blue or cream with pale

blue speckles or solid blue at base, with dull brown,

brownish purple or green gibbosities and slightly recurved

tips; inner perianth segments protruding well beyond outer

segments, obovate, translucent white with distinct bright

FIGURE 4. — Holotype of Lachenalia obscura. Compton 11174.

10

Bothalia27,l (1997)

FIGURE 5. — Painting of Lachenalia

obscura, drawn from Compton

1 1174, reproduced from the

original watercolour by Miss

WE. Barker. Plant, x 0.7.

green keels, recurved, with or without pale to dark ma-

genta tips, 8-10 X 5-7 mm. Stamens declinate, as long as

or shortly exserted up to 2 mm beyond inner perianth seg-

ments; filaments white, 8-11 mm long. Ovary ovoid, pale

green, 3-4 x 2 mm; style white, 8 mm long, protruding

beyond stamens as ovary enlarges. Capsule ovoid,

7-8 X 4-6 mm. Seed globose, 1 mm long, shiny black,

with a short, ridged terminal arillode 0.3 mm long. Fig-

ures IB; 3-5.

Etymology: named obscura by Schlechter to convey

the obscure and very variable appearance of this species.

Diagnostic characters

L. obscura is characterized by a spicate or subspicate

inflorescence of patent to slightly cernuous, oblong-cam-

panulate, pale yellowish green to brownish blue or cream

flowers, with or without distinct magenta tips. The trans-

lucent white, inner perianth segments are distinctly longer

than the outer segments and have slightly recurved tips.

The flowers are usually arranged in distinct, three-flow-

ered whorls in the lower part of the inflorescence, becom-

ing less distinctly whorled towards the top. The declinate

stamens are included within, or very slightly exserted be-

yond the perianth. The one or two erect to suberect,

lanceolate leaves usually have distinct green to brownish

purple and magenta bands on the lower surface, and the

globose bulb is surrounded by hard, dark brown, cartilagi-

nous outer scales.

L. obscura appears to be most closely related to L.

maximiliani Schltr. ex W.F.Barker, a dwarf species occur-

ring in large colonies and restricted to the Wuppertal-Ced-

erberg area of the Western Cape. Both species have spicate

or subspicate inflorescences with magenta-tipped inner pe-

rianth segments and a globose bulb surrounded by hard,

dark brown outer scales and similar globose seeds with a

terminal, ridged arillode. L. maximiliani differs from L.

obscura mainly in having narrow-urceolate, very pale blu-

Bothalia27,l (1997)

II

ish grey flowers and a single unbanded, canaliculate leaf.

Flowering time: June to October.

Distribution and habitat

Rudolf Schlechter collected this species for the first

time at Papkuilsfontein, southeast of Vanrhynsdorp, in Au-

gust 1897. He distributed material to nine local and for-

eign herbaria under his manuscript name L. obscura,

where it has remained unpublished for one hundred years.

L. obscura is a widely distributed and very variable spe-

cies; it is currently recorded from Steinkopf at the north-

ernmost end of its range, southwards throughout Nama-

qualand to the Kamiesberg and Knersvlakte, eastwards to

the Nieuwoudtville-Calvinia area where it is common, and

further south to Sutherland, and south and southeastwards

to the Montagu and Oudtshoorn areas.

Due to its wide distribution, L. obscura is encountered

in a wide variety of habitats, but is usually found on kar-

roid flats in dry stony clay soil, and less frequently in

sandy soil on moist lower mountain slopes. Plants usually

grow singly or in small clumps in full sun among low-

growing bushes. The typical forms of this species, such

as those occurring in the Nieuwoudtville-Calvinia area,

have pale yellowish green, oblong-campanulate flowers

with distinct magenta tips and leaves with bright green to

brownish purple and magenta bands on the lower leaf sur-

face, whereas certain forms found further north in

Namaqualand have longer, less campanulate flowers

which are pale blue, with or without very pale magenta

tips, and leaves without distinct bands on the lower sur-

face.

Material examined

NORTHERN CAPE. — 2917 (Springbok): Steinkopf Reserve, (-BA),

Aug. 1980, Van Berkel 164 (NBG); Farm Ratelpoort, (-BD), Aug. 1971,

Hall 4087 (NBG); Farm Eksteenfontein, (-DB), Aug. 1986, Duncan 252

(NBG). 3017 (Hondeklipbaai): Kamieskroon, (-BB), Aug. 1980, Van

Berkel 166 (NBG). 3018 (Kamiesberg): Kamiesberg, (-AC), Aug. 1984,

Van Zijl s.n., (NBG); Studer's Pass, E of Garies, (-AC), June 1970,

Stayner s.n. (NBG). 3019 (Loeriesfontein); N of Loeriesfontein, (-CD),

July 1972, Hiemstra s.n. .sub NBG 97199 (NBG). 3118 (Vanrhynsdorp);

Farm Papkuilsfontein, 41 km SE of Vanrhynsdorp, (-DD), Aug. 1897,

Schlechter 10907 (B, BM, G, GRA, K, L, PRE, S, Z). 3119 (Calvinia):

Grasberg, N of Nieuwoudtville, (-AC), Aug. 1961, Barker 9351 (NBG);

between Grasberg & Nieuwoudtville, (-AC), Aug. 1961, Barker 9362

(NBG); Klipkoppies, Nieuwoudtville, (-AC), Aug. 1961, Barker 9379

(NBG); Sept. 1971, Hardich s.n. (NBG); Nieuwoudtville Reserve, (-AC),

July 1983, Perry & Snijnum 2173 (NBG); E of Nieuwoudtville, (-AC),

July 1970, Nordenstam 770 (NBG); Farm Uitkomst, NW of Nieu-

woudtville, (-AC), Sept. 1970, Barker 10750 (NBG); Farm Soefwater,

between Nieuwoudtville & Calvinia, (-AD), Aug. 1974, Botha s.n.

(NBG); Akkerendam Nature Reserve, Calvinia, (-BD), July 1961, Barker

9321 (NBG); Calvinia commonage, (-BD), Aug. 1968, Stayner s.n. sub.

NBG 93582 (NBG); 30 km before Calvinia, on road from Karoopoort,

(-BD), Sept. 1983, Duncan 108 (NBG); E of Calvinia, on road to Wil-

liston, (-BD), July 1973, Thomas s.n. sub. NBG 98484 (NBG); Farm

Vanrhynshoek, Calvinia, (-BD), Thompson 2366 (NBG); Lokenburg,

(-CB), Aug. 1959, Acocks 20602 (NBG); W of Tafelberg, SE of Calvinia,

(-DB), May 1975, Thompson 2445 (NBG); Vogelstmis Vlakte, (-DC),

July 1941, Compton 11174 (NBG). 3120 (Williston); 34 km on road from

Middelpos to Calvinia, (-CA), Oct. 1974, Thomas s.n. sub. NBG 105715

(NBG); Farm Blomfontein, E of Middelpos, (-CC), Aug. 1972, Barker

10784 (NBG). 3220 (Sutherland): Farm Voelfontein, Sutherland, (-AD),

Sept. 1968, Hall 3252 (NBG); S of Sutherland, (-BC), Oct. 1968, Hall

3287 (NBG); near Sutherland, (-BC), Sept. 1969, Stayner .s.n. .sub NBG

93909 (NBG).

WESTERN CAPE. — 3319 (Worcester); E of Worcester on Robertson

road, (-DA), July 1954, Barker 8255 (NBG). 3320 (Montagu): Farm

Mooi-Erfenis, NW of Montagu, (-CA), Oct. 1979, Kriel s.n. sub NBG

120475 (NBG); Flats E of Warmwaterberg, (-DC), Aug. 1971, Boucher

1576 (NBG). 3322 (Oudtshoorn): S slopes of Mannetjiesberg, (-DB),

Oct. 1971, Oliver 3605 (NBG).

Lachenalia inconspicua G.D. Duncan, sp. nov. L.

concordianae Schltr. ex W.F. Barker affinis ob habitum

nanum similarem, inflorescentiam spicatum floribus

plerumque verticillis 3-floris, foliumque singularem

lanceolatum pagina inferiora fasciis brunneolo-viridibus;

sed floribus oblongo-campanulatis, lacteis vel viridi-albis,

folioque coriaceo late patenti pagina superiora maculis

purpureo-brunneis differt.

TYPE. — Northern Cape, 2918 (Gamoep); 500 m be-

yond Gamoep, on road Springbok to Gamoep, in deep

red gravelly sand at side of road, (-CD), 19-8-1986,

G.D. Duncan 259 (NBG, holo.).

Deciduous, winter-growing geophyte 120-160 mm

high. Bulb globose, 15-20 mm in diam., white with thick,

spongy, dark brown outer tunics produced into a short

neck. Leaf usually solitary, lanceolate or occasionally

ovate, widely spreading, deeply channelled, leathery,

85-150 X 15-20 mm, glaucous with depressed longitudi-

nal veins and irregularly scattered purplish brown spots

on upper surface, and broad, brownish green bands on

lower surface, shading to narrower brownish magenta

bands on clasping leaf base. Inflorescence spicate or sub-

spicate, fairly dense, few to many-flowered, 45-80 mm

long, with a very short sterile tip, flowers usually arranged

in distinct three-flowered verticils; peduncle erect, sturdy,

pale green with large purplish brown blotches, 55-80 mm

long; pedicels absent, bracts very much reduced, ovate,

1 mm long. Flowers erecto-patent, oblong-campanulate,

pale bluish or greenish white, fading to dull brownish pur-

ple; outer perianth segments oblong, slightly recurved,

pale greenish blue with darker blue markings at base, and

dull purplish brown or brownish green gibbosities, 7-8

X 4-5 mm; inner perianth segments obovate, translucent

white, with brownish green keels, protruding well beyond

outer segments, slightly recurved, upper two segments

overlapping, lower segment slightly longer, 8-11 x

4-5 mm. Stamens declinate; filaments white, 7-9 mm

long, included within, or slightly exserted up to 1 mm

beyond perianth. Ovary ovoid, pale green, 3 mm long;

style white, 8-9 mm long. Capsule ovoid, membranous,

9-10 X 5 mm. Seed globose, 1 mm long, dull black, with

a ridged terminal arillode 1 mm long. Figures 1C; 6 & 7.

Etymology, named for the inconspicuous, well-camou-

flaged flowers.

Diagnostic characters

L inconspicua is a dwarf species characterized by a

spicate or subspicate inflorescence of pale bluish white or

greenish white oblong-campanulate flowers with purplish

brown or brownish green gibbosities, usually arranged in

distinct three-flowered verticils, and a usually solitary,

lanceolate, widely spreading, deeply channelled leaf with

conspicuous transverse bands on the tightly clasping leaf

base. It is related to L. concordiana Schltr. ex W.F.Barker,

another dwarf species with greenish cream flowers also

arranged in three-flowered verticils; but the latter differs

12

Bothalia27,l (1997)

FIGURE 6. — Holotype of Lachenalia inconspicua, Duncan 259.

in having a linear-lanceolate leaf and widely campanulate

flowers with dark green gibbosities and the tips of the

inner and outer perianth segments all distinctly recurved.

Flowering time: July to August.

Distribution and habitat

L. inconspicua is at present known only from a few

collections made in the Kamiesberg, western Bushman-

land and southern Namaqualand in the Northern Cape.

According to current records, the first collection was made

by F. Archer as recently as August 1982 NE of Gamoep,

and it has since been found at several locations near

Gamoep, and further south near Leliefontein, Kliprand

and Bitterfontein. It is locally plentiful and occurs on open

flats in deep red gravelly sand. An interesting growth fea-

ture of this species in cultivation is the manner in which

healthy bulbs periodically remain completely dormant

during the growing season, a characteristic shared by sev-

eral Lachenalia species from the very arid parts of South

Africa.

Material examined

NORTHERN CAPE. — 2918 (Gamoep): 0.5 km beyond Gamoep, on

road from Springbok to Gamoep, (-CD), Aug. 1986, Duncan 259 (NBG);

Kouberg, NE of Gamoep, (-CD), Aug. 1982, Archer 192 (NBG); Vaal-

koei Farm, SE of Gamoep, (-CD), Aug, 1996, Duncan 381 (NBG). 3018

(Kamiesberg); 2 km S of Paulshoek village, (-AD), Sept, 1996, Petersen

41 (NBG); 5 km on R358 to Kliprand, (-CD), Aug. 1996, Duncan 378

(NBG); 6 km SW of Kliprand, (-DA), Aug. 1995, Symmonds 2 (NBG).

Lachenalia marlothii W.F.Barker ex G.D.Duncan,

sp. nov. L. marginatae W.F. Barker affinis ob folium sin-

gularem simularem ovatum ad ovato-lanceolatum, co-

riaceum, basi amplectenti fasciis distinctis purpureo-

brunneis, flores urceolato-oblongos, segmentosque inte-

riores perianthii longos protrudentes; sed folio sine mar-

gine distincto coriaceo, basi arete circumcincto, segmentis

perianthii distincte recurvatis floribusque valde odoratis

differt.

TYPE. — Northern Cape, 3119 (Calvinia); between

Vlakkraal and Kalkgat Suid, S of Calvinia, (-DC), 23-7-

1961, W.F.Barker 9330 (NBG, holo.!).

Deciduous, winter-growing geophyte 90-160 mm high.

Bulb subglobose, 15-33 mm in diam., white with brown

spongy outer tunics. Leaf solitary, 30-60 x 10-25 mm,

suberect, ovate to ovate-lanceolate, with an undulate,

sometimes crisped margin, blade very leathery, dark green

and unmarked on upper surface, with dark purplish brown

and green transverse bands on lower surface; tightly clasp-

ing leaf base 30-70 mm long, white with very conspicu-

ous purplish brown bands in the upper half, shading to

magenta in the lower half. Inflorescence spicate or sub-

spicate, fairly dense, few- to many-flowered, 40-95 mm

HEIGHT ABOVE SEA LEVEL

Over 1500 m

|~3 900 - 1500 m

r~1 300 - 900 m

I I Under 300 m

FIGURE 7. — Distribution of Lachenalia inconspicua, □; L. marlothii,

•; and L xerophila, A.

Bothalia 27,1 (1997)

13

FIGURE 8. — A, Lachenalia mar-

lothii, Botha s.n.; B, L xero-

phila, Botha s.n. Scale bars:

10 mm.

long, with a very short sterile tip; peduncle slender, erect,

40-100 mm long, pale green with brownish purple

blotches; pedicels absent or up to 2 mm long; bracts

white, ovate, 1-2 mm long. Flowers strongly scented, pa-

tent or suberect, urceolate-oblong; outer perianth segments

ovate, pale blue and green with darker blue bases, 8-9 x 5

mm, with greenish purple or purplish brown gibbosities

and recurved tips; inner perianth segments obovate,

8-11 X 3-6 mm, white or brownish yellow with purplish

green keels, protruding well beyond outer segments, re-

curved, upper two segments overlapping, lower segment

longer and narrower. Stamens declinate; filaments white,

8-10 mm long, included within perianth. Ovary ovoid,

pale green, 4-5 x 3-i mm; style white, 6-7 mm long.

Capsule ovoid, 5-6 x 4 mm. Seed ovoid, 1.5 mm long,

shiny black with a ridged terminal arillode 0.5 mm long.

Figures 7, 8A & 9.

Etymology: L marlothii is named after the famous Ger-

man chemist and botanist Rudolf Marloth, who made the

first recorded collection of this species in October 1920.

Diagnostic characters

L. marlothii is characterized by a spicate or subspicate

inflorescence of patent or suberect, pale blue and yel-

lowish green urceolate-oblong flowers with recurved tips

and included stamens, and a very distinctive, single, ovate

to ovate-lanceolate, coriaceous, suberect leaf with a heav-

ily banded, tightly clasping leaf base. The peduncle is

heavily marked with brownish purple blotches and the

flowers have a strong sweet scent.

L. marlothii appears to be most closely related to L.

marginata W.F. Barker which also has an ovate or ovate-

lanceolate, leathery leaf with distinct purplish brown

bands on the clasping base, urceolate-oblong flowers with

brownish gibbosities on the outer perianth segments and

long protruding inner perianth segments. L. marginata dif-

fers in having a distinctly coriaceous leaf margin, the

clasping leaf base is not tight, the inner and outer perianth

segments are not recurved, and the flowers are not heavily

scented. Flowering time: July to September.

FIGURE 9. — Holotype of Lachenalia marlothii. Barker 9330.

14

Bothalia27,l (1997)

FIGURE 10, — Holotype of Lachenalia xerophila, Botha s.n.

Distribution and habitat

L. marlothii was collected for the first time by Rudolf

Marloth in October 1920 at Waterkloof in the Sutherland

Roggeveld, which forms the southern boundary of its dis-

tribution. It has since been collected mainly in the Calv-

inia District, and is currently known from as far north as

Brandkop, north of Nieuwoudtville. Plants occur singly

or in groups and are usually associated with south-facing

hillslopes in clay soil.

Material examined

NORTHERN CAPE. — 3119 (Calvinia): Brandkop, N of Nieuwoudt-

ville, (-AA), Aug. 1950, Barker 6486 (NBG); Calvinia road towards

Soetwater turn-off, (-AD), July 1971, Botha s.n., (NBG); Van Rhynshoek

Farm, Calvinia, (-BD), Oct. 1986, Thomas 275, (NBG); between

Vlakkraal and Kalkgat Suid, S of Calvinia, (-DC), July 1961, Barker

9330 (NBG). 3220 (Sutherland): Gannaga Pass, SW of Middelpos on

Calvinia-Ceres road, (-AA), Sept. 1971, Hardich s.n. sub. NBG 93862

(NBG); Waterkloof, Sutherland Roggeveld, (-BB), Oct. 1920, Marloth

9661 (PRE).

Lachenalia xerophila Schltr. ex G.D. Duncan, sp.

nov. L. klinghardtianae Dinter et L physocaulotis W.F.

Barker affinis ob stamina similares bene exserta declinata,

pedunculum rachidemque conspicue tumidum, inflores-

centiamque subspicatam vel racemosam floribus oblongo-

campanulatis; sed inflorescentia multo densiora floribus

minoribus cemuis gibbis distinctissimis magnis brunneis,

folioque lanceolato-acuto canaliculato differt.

TYPE. — Northern Cape, 2918 (Gamoep): Kouberg

Farm, off R355 Springbok to Gamoep, western Bushman-

land, (-CD), 29-8-1972, M.C.Botha s.n. sub. NBG 95451

(NBG, holo.!).

Deciduous, winter-growing geophyte 100-250 mm

high. Bulb usually deep-seated, globose, 15-25 mm in

diam., white with membranous dark brown outer tunics.

Leaves one or two, 25-180 x 10-20 mm, lanceolate,

acute, canaliculate, glaucous, with an undulate and crisped

margin, clasping leaf base white, up to 200 mm long. In-

florescence subspicate or racemose, dense, many-flow-

ered, 55-130 mm long with a short sterile tip; peduncle

erect to suberect, very sturdy, 30-110 mm long, conspicu-

ously inflated below and at base of inflorescence, gradu-

ally becoming less inflated towards top of rachis, pale

green; pedicels 2-5 mm long; bracts small and membra-

nous, ovate, 1-2 mm long. Flowers cemuous to spread-

ing, oblong-campanulate; outer perianth segments ovate-

oblong, 6-7 X 3 mm, very pale blue at base, shading

to white above, with very large dull red to dark brown

gibbosities; inner perianth segments obovate, 8-9 x

3-4 mm, protruding beyond outer segments, white with

a greenish brown marking at apex. Stamens well exserted,

declinate; filaments white, up to 12 mm long. Ovary

obovate, pale green, 3^ x 2 mm; style white, up to

9 mm long, protruding beyond stamens as ovary enlarges.

Capsule obovate, membranous, 8-9 x 5 mm. Seed ob-

long, 2-3 mm long, with a narrow decurrent, inflated ter-

minal arillode 0.2 mm long. Figures 7, 8B & 10.

Etymology: named L. xerophila by Schlechter to de-

scribe the preference this species has for growing in dry

places.

Diagnostic characters

L. xerophila is characterised by a dense subspicate or

racemose inflorescence of small oblong-campanulate,

white flowers with very large dull red to dark brown gib-

bosities, and well-exserted, declinate stamens. The sturdy

peduncle and rachis are distinctly swollen, and the erect,

lanceolate, acute, canaliculate leaf has a distinctly undu-

late, and sometimes crisped margin. The seed is unique

within the genus in having a narrow decurrent, inflated

terminal arillode.

L. xerophila is closely related to L. klinghardtiana Din-

ter and L. physocaulos W.F.Barker, which fall into the

group of species having well exserted stamens and a con-

spicuously swollen peduncle and rachis. L. klinghardtiana

also occurs in northwestern Namaqualand, as well as in

the Richtersveld and the southwestern corner of Namibia,

but differs in having larger oblong-campanulate flowers

with much smaller greenish brown to reddish brown gib-

bosities, borne on a less dense, usually shorter inflores-

cence. Further, it differs in having a lanceolate-falcate leaf,

the peduncle usually has greenish brown blotches or spots,

and its seed has a different shape, being globose with an

inflated terminal arillode. L. physocaulos differs from L.

Bothalia27,l (1997)

15

xerophila in having a linear-conduplicate leaf which wid-

ens suddenly into a subterranean clasping base, its flowers

are pale magenta and its very small seeds are globose

with an inflated terminal arillode. The two species are

well separated geographically, as L. physocaidos occurs

only in the Robertson and Swellendam areas of the south-

ern Western Cape. Flowering time: July to September.

Distribution and habitat

Material of this species was first collected by Rudolf

Schlechter on 21st September, 1897 at Leeuwpoort just

north of Concordia, in Namaqualand. He appended the

manuscript name xerophila to this material, and distrib-

uted it to seven overseas and local herbaria (Barker 1983),

where it has languished unpublished for one hundred

years. L xerophila is restricted to the dry northwestern

and central parts of Namaqualand, and western Bushman-

land, where it occurs singly or in colonies in deep red

sand in full sun. The fleshy bulb is deep-seated in order

to survive the harsh dry summer conditions, and may re-

main dormant during the winter growth period if rainfall

is insufficient. In western Bushmanland, an area of pre-

dominantly summer rainfall, L. xerophila nevertheless fol-

lows the typical pattern of winter rainfall growth and

summer dormancy characteristic of the vast majority of

species belonging to this genus.

Material examined

NORTHERN CAPE. — 2816 (Oranjemund): Holgat, Namaqualand,

(-DD), Aug. 1952, Hall 558 (NBG). 2917 (Springbok); Leeuwpoort, 14

km N of Concordia, (-DB), Sept. 1897, Schlechter 11366 (BM, BOL,

G, GRA, K, LD, Z). 2918 (Gamoep): Kennedy’s Farm, 40 km E of

Springbok, (-CA), Sept. 1967, Eliovson 13 (NBG); near Ratelkraal,

Namaqualand, (-CA), Sept. 1950, Barker 6759 (NBG); Kouberg Farm,

off R355 Springbok to Gamoep, Bushmanland, (-CD), Aug. 1972, Botha

s.n. sub. NBG 95451 (NBG). 3018 (Kamiesberg): Vaalputs Farm, 7 km

E of Stofkloof, (-AB), Sept. 1983, Schelpe s.n. sub. NBG 127341 (NBG).

ACKNOWLEDGEMENTS

I thank the staff of the Compton Herbarium, Kirsten-

bosch, especially Dr D.A. Paterson-Jones, Dr J.C. Man-

ning, Mrs J. Beyers and Mrs S.E. Foster for their friendly

assistance at various stages of this study. I am also very

grateful to Dr O. A. Leistner for kindly preparing the Latin

diagnoses, Mrs J. Loedolff for taking the black and white

photographs of the herbarium sheets, and Prof. M.C.

Botha and Mr R. Symmonds for assistance in the field.

REFERENCES

BARKER, W.F. 1978. Ten new species of Lachenalia (Liliaceae). Jour-

nal of South African Botany 44: 391^18.

BARKER, W.F. 1979. Ten more new species of Lachenalia (Liliaceae).

Journal of South African Botany 45: 193-219.

BARKER, W.F. 1983a. A list of the Lachenalia species included in

Rudolf Schlechter’s collections made on his collecting trips in

southern Africa, with identifications added. Journal of South Afri-

can Botany 49: 45-55.

BARKER, W.F. 1983b. Six more new species of Lachenalia (Liliaceae).

Journal of South African Botany 49: 423-444.

BARKER, W.F. 1984. Three more new species of Lachenalia and one

new variety of an early species (Liliaceae). Journal of South

African Botany 50: 535-547.

BARKER, W.F. 1987. Five more new species of Lachenalia (Liliaceae-

Hyacinthoideae), four from the Cape Province and one from

southern South West Africa/Namibia. South African Journal of

Botany 53: 166-172.

BARKER, W.F. 1989. New taxa and nomenclatural changes in

Lachenalia (Liliaceae) from the Cape Province. South African

Journal of Botany 55: 630-646.

DUNCAN, G.D. 1987. Lachenalia macgregoriorum. The Flowering

Plants of Africa 49: 1. 1 95 1 .

DUNCAN, G.D. 1988a. The Lachenalia handbook. Annals of Kirsten-

bosch Botanic Gardens 17. National Botanical Institute, Cape

Town.

DUNCAN, G.D. 1988b. Lachenalia arbuthnotiae. The Flowering Plants

of Africa 50: t.mi.

DUNCAN, G.D. 1989. Lachenalia. In B. Jeppe, Spring and winter flow-

ering bulbs of the Cape. Oxford University Press, Cape Town.

DUNCAN, G.D. 1992. Lachenalia: its distribution, conservation status

and taxonomy. Acta Horticulturae 325: 843-845.

DUNCAN, G.D. 1993. Lachenalia thomasiae. The Flowering Plants of

Africa 52: t. 2061.

DUNCAN, G.D. 1994. The genus Lachenalia. and the discovery of a

beautiful new species from the Western Cape Province of South

Africa. Shin-Kaki 163: 32-35.

DUNCAN, G.D. 1996. Four new species and one new subspecies of

Lachenalia (Hyacinthaceae) from arid areas of South Africa.

Bothalia 26; 1-9.

£1

fa^? ;? .'(iwiis nAi la- ^n1^ fiiortsv je 3r>mif«ie0e l^iu

uSMJ-art? gnheqi«a"(i^'ii «*1 .0 iq

ig/iw iS^’i ^fiiii«;fO^ I

i»bi| ortJ, ju 9t<*iB'»iaaB ■>vS«^'’fil'

^4. e .%;- '%

H3£ V, ^ i • ,' ■'- ~h* ^4. *..: ’- i«ir^' ■^^yn. : ' . ■ ^

ms

' .33X11A8 I

i'ta»

‘^'' tWi^.fei- s*H «s vAip/ ©qf^f-sa irtfeWsStti^ W.fcwl ®'i''-

*" * ' * -?£-’~^ <fJ>^{W>^ «v<

asuMis^Wa flto'uKSKvitl

>3 .«{

JKKI , tVffi «’

.«5„

§^::

MW

•yvjo(a{i.J)^m»iA'j&:l^ asSjxjir WtMs VkiIb w.?i*f Aw

Hffrf «ip Ima 80»vy^ 4iqi!£>-'wj» " 4tqn W a'>ir rk hndfftfijliiJMiS Rii5ti«sW irf

ai^iiwtA^ Mtf isfiu^«(£ak.V iJ^b utt iHADHy^

~ '<<^<{6'i|^)ibwiA jkKidtiiiirt jsStoMAn^i «<T ;^fSAl M,0 r«A'

&■.

mcA

'i-V..".*

a ^

w :rMVi->C;

4' ^-4 \a"'V

01 .>-v u,^.’. ;

flAWotiig

JW/<^9^‘?,3a5fl t#*Wi«U;<)t^ r ,fcsTif5Mi>V t«s« .^08^ *,^(^1 , .4^

" %^..:<K.?3 -

i, , , s»'p;.' .-■4,1 4i***-v -I*

?.!i

V '»

»<DS» ,! '->

V-Sf--

1%,- , , ,;£"'^

_ fil

. i93aji4\PA A

■:’» ' -A-'

. ^ :

#•': ^

—.n l' >•”!•■ *• ». r..< . «#T*Ti»' '^..♦'t’

■T-iw* wax-U-f «u| ■ DiMte-* 4mj>if^:

;*'■

n:

:-^: -• ■■• :

>f, !»*’■•«.-.

Jt » _•. . :j^

■ ^-.. .- '■>■,'*’.» i’f|> •■ ';

■*iy 3,1

1 ■- ft J

- K 'S-V-S t» -

«Xt

Bothalia27,I: 17-27 (1997)

Studies in the liverwort genus Fossombronia (Metzgeriales) from southern

Africa. 1. Three new species from Northern Province, Gauteng and Mpu-

malanga

S.M. PEROLD*

Keywords: Fossombronia, Fossombroniaceae, F. gemmifera sp. nov., F. glenii sp. nov., F. straussiana sp. nov., liverworts, Metzgeriales

ABSTRACT

Three new species of Fossombronia from Northern Province, Gauteng and Mpumalanga (formerly Transvaal) are described:

F. gemmifera, F. glenii and F. straussiana. F. gemmifera is distinguished by a highly convoluted pseudoperianth, the frequent

presence of gemmae produced dorsally on the stem, lamellate spores and mostly rather short elaters; F glenii can be recognized

by a smallish, deeply lobed pseudoperianth, perigonial bracts with finger-like projections shielding the antheridia and by spinous

spores; F. straussiana is distinct by its hyaline or brownish rhizoids, by its dense, frilly leaves, its pseudoperianth with lamellate

lateral outgrowths, by spore ornamentation that usually has inclusions in the incomplete areolae and by the elaters which are

finely papillose.

INTRODUCTION

Within the order Metzgeriales the cosmopolitan sim-

ple-thalloid liverwort genus Fossombronia Raddi is clas-

sified in the subfamily Fossombronioideae Engl, emend.

R.M.Schust. (family Fossombroniaceae Hazsl. ( = Co-

doniaceae H.Klinggr.). The Fossombroniaceae is the only

family under the suborder Fossombroniineae R.M.Schust.

and is regarded as phylogenetically pivotal (Solomon

1995). It comprises four genera: Petalophyllwn Nees &

Gottsche ex Fehm., Sewardiella Kashyap, Austrofossom-

broriia R.M.Schust. and Fossombronia, of which only the

latter is known to occur in southern Africa.

Plants of the genus Fossombronia are small, usually

gregarious with mostly prostrate, fleshy stems, laterally

bearing succubously inserted, undulating or crisped thallus

wings, dissected into leaf-like segments and free to the

base, where they are bistratose, but unistratose elsewhere.

The stems are anchored to the substrate by rhizoids which

are usually purple, but hyaline in a few rare species.

The gametangia develop acropetally and are situated

dorsally along the stem between the leaf insertions. They

are either intermingled or borne on separate shoots in

monoicous species, or they may occur on separate plants

in dioicous species. The antheridia are short-stalked, ±

globose and either naked or shielded by perigonial bracts.

The archegonia are exposed, but after fertilization the de-

veloping sporophyte becomes surrounded by a campanu-

late pseudoperianth, constricted at the sometimes stipitate

base, and flaring at the plane, lobulate or crispate mouth.

The globose capsule is raised on a moderately short or

long seta, 6 or 7 to 10 cells in diameter; the capsule wall

is bistratose, the outer wall with delicate, hyaline cells and

the inner wall with irregularly quadrangular cells, contain-

ing nodular and semi-annular thickenings. The walls rup-

ture irregularly into small plates, releasing the spores and

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

MS. received: 1996-06-25.

elaters. The spores are relatively large and highly ornamented

with spines, lamellae or areolae. The ornamentation on the

outer face is generally considered species-specific and re-

garded as essential for identification, although ‘so variable

that patterns of ornamentation . . . must be used to define

taxa’ (Schuster 1992). The elaters may be well or occasion-

ally poorly developed and are usually 2- or 3-spiral. The

determination of sterile material is extremely difficult, if not

impossible because of the plasticity of the gametophytes, the

morphology of which varies considerably according to en-

vironmental conditions.

HISTORICAL NOTES

The genus Fossombronia has been relatively poorly

studied in southern Africa. Lehmann (1829) described

Fossombronia (sub Jimgennannia) leucoxantha, collected

by Ecklon on Table Mountain, Devil’s Peak and Lion’s

Head and also reported the presence of F (sub Jiinger-

mannia) pusilla L. on Table Mountain. Gottsche et al.

(1844-1847) confirmed the presence of F pusilla Nees

and described F. crispa Nees from Promontorio Bonae

Spei. Krauss (1846) reported F. angulosa ‘in rupibus ri-

vulor montium natalensium’. Mitten (1878) described F

tumida collected by the Rev. A.E. Eaton at the foot of

Lion’s Head and endorsed the records of F. crispa and F

leucoxantha from localities near Cape Town. Stephani

(1900) described F zeyheri and F. spinifolia. Sim (1926)

added no new species, but Amell (1952) described three

from the Cape: F capensis, F. densilamellata and F mon-

taguensis (Perold 1997: 29). Amell (1963) stated that ‘the

South African species of the genus are in great need of

revision’. Scott & Pike (1984, 1987a-c, 1988a, b), who

studied the genus in Australia (i.e. the western parts of

the State of Victoria and adjacent regions of South Aus-

tralia and New South Wales) for six years and described

many new species from there, expected a comparable spe-

cies richness to emerge from a long-term study of the

genus in South Africa (and South America), since they

regarded its origins to have been mainly in Gond-

wanaland.

18

Bothalia27,l (1997)

FIGURE I . — Fossomhronia gemmifera. A-D, leaves; E, opened pseudoperianth; F, pseudoperianth from side; G-J, perigonial bracts; K, cross section

of stem; L, detail of marginal area of leaf, with slime papilla (see arrow); M, median leaf cells with oil bodies and chloroplasts; N, cells in

capsule wall. A, D. Strauss & Relief CHI 3655, B, C, Perold & Koekemoer 3116b-, E, F, Perold & Van Rooy 3559a', G-N, Strauss 132. Drawn

by G. Condy. Scale bars: A-F, 5(X) pm; G-J, 250 pm; K, 100 pm; L-N, 50 pm.

Bothalia27,l (1997)

19

This paper is the first of a series dealing with southern

African Fossombronia species.

METHODS

Samples of field-collected specimens were transferred

to a conically shaped fme-mesh sieve and washed in a jet

of running water to clear away soil particles. Remaining

particles were manually removed by using fine-tipped for-

ceps. Cross sections were cut of some cleaned stems; several

leaves and a pseudoperianth were detached and all trans-

ferred to a drop of water on a clean slide. Finally a coverslip

was applied to the preparation, which was stored in a covered

plastic dish lined with damp filter paper. During examination

of the slide the evaporated water was periodically replen-

ished. The structures were measured and also photographed

under a compound light microscope.

The remaining portion of the cleaned specimen was

fixed in FAA (formaldehyde/alcohol/glacial acetic acid

and distilled water in proportion of 2:1:1:20). For later

reference some thalli were permanently preserved in FAA,

the remainder only for several hours and then dehydrated

in an ascending series of acetone to 100% and critical

point dried in a Balzers Union dryer, using liquid CO2 as

the transitional fluid. The thalli were mounted on alu-

minium stubs with double-sided Sellotape, gold-coated,

then viewed and photographed, using an ISI SX 25 scan-

ning electron microscope (SEM).

There are some advantages to studying and photo-

graphing Fossombronia material, treated as described

above, with the aid of SEM. Scott & Pike (1984) stated

that the form, colour and orientation of the leaves (of dif-

ferent species) are distinct but ‘beyond our powers to il-

lustrate and exceedingly difficult to describe’. Although

the colour cannot be recorded by SEM, the form and cer-

tainly the orientation of the leaves can. Scott & Pike

(1984) also observed that ‘male plants have often shriv-

elled and disappeared by the time the corresponding

spores are ripe and the gametophyte is then so desiccated

that its pristine vegetative appearance is irrecoverable’.

Samples taken at different stages and treated as above,

would have made comparisons easier for them. Scott &

Pike did not illustrate the plants of any of their new spe-

cies, except F mdis (Scott & Pike 1988b) and relied heav-

ily on descriptions and the spore ornamentation to

distinguish between species.

The spores and elaters were mounted on slides in

Hoyer’s fluid for examination and measurement by LM.

Eor SEM microscopy some spores and elaters from the

same capsule were allowed to air-dry, mounted on stubs

with double-sided Sellotape, gold coated, viewed and

photographed. SEM micrographs of Fossombronia

spores have frequently been published and have been

used as an aid to identification. Features of spore mor-

phology should be used with caution, however, because

they can be rather variable, and the spores need to be

fully mature. The Degree Reference System was used

for recording distribution data (Edwards & Leistner

1971).

I. Fossombronia gemmifera Perold, sp. nov.

Plantae repentes, gregariae vel dispersae; saepe gem-

mas dorsaliter secus caulem ferens. Folia imbricata, fim-

briata, apicem versus aliquando longiora quam latiora,

basin versus plerumque breviora quam latiora, nonnulla

leviter bilobata, alia apicem versus undulata. Rhizoidea

purpurea. Monoicae, interdum ut videtur dioicae.

Antheridia bracteis tecta. Pseudoperianthium breviter

setatum, orificio intricate convoluto, in lobis multis di-

viso. Sporae 55.0-62.5 pm diametro, superficie distali

lamellis minimum 14 irregularibus duplicato-parietatis,

aliquando areolas imperfectas facientibus; superficie

proximali sine nota triradiata distincta, areolis imperfec-

tis, parietibus aids circumcinctis. Elateres 100-160 pm

longi, medio 7.5 pm lati, extremitates versus decre-

scentes, bis vel ter spirales, sed interdum arete spirales,

50.0- 62.5 pm longi, 12.5 pm lati.

TYPE. — Northern Province, 2427 (Thabazimbi):

Kransberg, Farm Geelhoutbosch, on streambank, directly

south of rondavel, (-BC), S. Strauss 132 (PRE, holo.).

2529 (Witbank): ± 80 km E of Pretoria, on PretoriaAVit-

bank road, facing Balmoral turnoff, left side of road at

seepage area, (-CC), Perold & Van Rooy 3559a (PRE,

para.).

Plants smallish to medium-sized, creeping, gregarious

or scattered, green, proximal leaves sometimes clasped

around stem, occasionally with pink margins; shoots sim-

ple, 5.0-9. 0 mm long, 0.85 mm high, 2.6 mm wide, or

once/twice furcate, apical segments moderately divergent

(Figure 2A) 2.0-5.0 mm long, basally ± 3 mm long. Stems

prostrate, in cross section apically 300-325 pm (12 or 13

cell rows) high, 400-500 pm wide, basally 250 x 300 pm,

plano-convex (Figure IK). Rhizoids purple, 12.5-20.0 pm

wide, some with internal mycorrhizal hyphae. Leaves

overlapping, frilly, obliquely inserted succubously, vari-

ously shaped, toward apex sometimes longer than wide,

but more proximally usually shorter than wide, 925-1525

X 1325-1625 pm, some slightly bilobed (Figure lA, D),

others ruched above (Figure IB, C); lateral margins with

3 or 4 uni- or bi-celled slime papillae (Figure IL). Leaf

cells thin-walled, at upper margins subquadrate to rectan-

gular across, 32.5-50.0 x 22.5-30.0 pm, at lateral margins

long-rectangular, up to 75 x 30 pm; upper laminal cells

(4-)5- or 6-sided or polygonal, 32.5-50.0 x 22.5-30.0 pm,

middle laminal cells 62.5-72.5 x 37.5-50.0 pm, basal

cells 112.5-137.5 x 37.5-62.5 pm. Oil bodies glistening,

13-36 per cell, rounded, up to ± 3.5 pm in diameter; chlo-

roplasts numerous, round or oblong, 5.0-7.5 pm in diame-

ter (Figure IM).

?Monoicous, some specimens with shoots bearing both

antheridia and archegonia, but occasionally only

antheridia or often only archegonia present. Antheridia

dorsal on stem, interspersed between archegonia, short-

stalked, globose, ± 185 pm in diameter, shielded by an

irregularly shaped perigonial bract (Figure 2C), 500-700

X 380^50 pm, margins toothed, with projecting cells and

slime papillae (Figure IG-J), cells in interior 4-6-sided,

80.0- 90.0 X 30.0-37.5 pm; sometimes antheridia in

groups between leaves and then lacking bracts or with

reduced bracts (Figure 2B). Archegonia numerous along

stem, naked, up to 3 per shoot becoming fertilized, soon

20

Bothalia27,l (1997)

FIGURE 2. — Fossombronia gemmifera. A, stem branching near apex, bracts and pseudoperianth with capsule shown; B, antheridia between leaves;

C, detail of perigonial bracts with mostly obscured antheridia and young fertilized archegonia; D, convoluted pseudoperianth from above; E,

archegonia and gemmae between leaves; F, close-up of gemmae. A, C, D, Strauss 132\ B, Perold & Koekemoer 3116b\ E, F, Perold & Van

Rooy 3557. A, x 13; B, x 26; C, x 21; D, E, x 20; F, x 54.

forming young pseudoperianths. Pseudoperianths often

crowded together, in acropetal sequence, raised on a short

stalk, ± 210 X 470 pm, widely flaring above (Figure IF),

± 900 pm long, up to 2000 pm wide across intricately

convoluted mouth (Figure 2D), divided into numerous

lobes (Figure IE), 200-500 x 150-440 pm, some apically

rounded, others with an acute apex ending in a papilla;

cells comparable in shape and size to those of leaves.

Capsules globose (Figure 2D), up to ± 600 pm in diame-

ter, enveloped in a calyptra which is later shed, capsule

wall bistratose, cells in inner layer irregularly shaped,

27.5-50.0 X 25.0-37.5 pm, each cell wall with (1)2^

nodular and occasionally semi-annular thickenings (Figure

IN). Seta delicate, 1.2-3. 2 mm long, 270-350 pm in di-

ameter, 6 cells across. Spores brown, hemispherical (Fig-

ure 3C), 55.0-62.5 pm in diameter, including lamellae

projecting around circumference, distal face convex

(Figure 3A, B), with at least 14 irregular, mostly double-

walled lamellae running across, ± 7.5 pm apart, occasion-

ally anastomosing and forming incomplete areolae;

proximal face (Figure 3D, E) lacking distinct triradiate

mark, ornamentation seemingly ‘raised’ from an encir-

cling, marginal furrow (Figure 3E), tall walls forming

small, irregular, incomplete areolae, up to 5 pm wide,

around periphery 19-22 projecting ‘ends’ of lamellae, ±

5 pm long. Elaters (Figure 3F) yellow-brown, smooth,

mostly rather small and sometimes poorly formed,

(50.0-)62.5-85.0 pm long, medianly 12.5 pm wide, 3(^)-

spiral and loosely coiled, tapering toward tips, 5 pm wide

and ending in bispiral loops, rarely 100-160 pm long, 7.5

pm wide medianly, 3-spiral. Vegetative reproduction by

gemmae (Figure 2E, F), abundantly borne dorsally along

stem, short-stalked, spindle-shaped, up to 300 x 170 pm,

green when fresh, turning brown with age.

Fossombronia gemmifera grows on sandy loam soil on

stream banks or at seepages, often mixed with mosses or

other Fossombronia species, particularly F. straussiana,

which has hyaline rhizoids and is thus easily distinguished

from it. This new species is quite distinctive with its

highly convoluted pseudoperianth and by the frequent

presence of dorsal gemmae for which it has been named.

Its spore ornamentation with double-walled lamellae on

the distal face and ‘raised’ on the proximal face above a

clear marginal furrow, is also unique. There are two other

species with purple rhizoids from the same distribution

area: F glenii which has spinous spores and F zeyheri

which has reticulate spores. Fossombronia gemmifera has

been collected at several localities in Northern Province

and Gauteng (Figure 4).

2. Fossombronia glenii Perold, sp. nov.

Plantae repentes, gregariae vel crebrae in coloniis. Fo-

lia imbricata, undulata, valde obliquiter inserta, plerumque

oblonga, longiora quam latiora, apicem versus latiora

quam basi, apice subtruncata, lobis vadosis, angulatis. Rhi-

zoidea purpurea. Dioicae. Antheridia bracteis cum laciniis

digitiformibus tecta. Pseudoperianthium infundibulifor-

me, orificio in lobis pluribus diviso, aliquot ad basin di-

visis. Sporae 40.0-52.5 pm diametro, superficie distali

cum spinis multis altis conicis, raro fractis vel cristis bre-

vibus junctis; superficie proximali sine nota triradiata

distincta, cum tuberculis multis tenuibus vel grossis in-

spersis. Elateres 70.0-137.5 x 7.5-10.0 pm, bis vel ter

spirales.

TYPE. — Northern Province, 2427 (Thabazimbi): Water-

berg, Welgevonden Estate, cliffs at drift over Sterkstroom

above farmhouse, in partial shade, with Fissidens erosulus

(Mull.Hal.) Paris, (-BD), H.F Glen 2146 (PRE, holo.);

H.F Glen 2134, same locality (PRE, para.).

Bothalia27,l (1997)

21

FIGURE 3. — Fossombronia gemmifera. Spores and elater. A, B, distal face; C, side view of distal face; D, E, proximal face; F, elater. A, B, D, S.M.

Perold2017\ C, E, F, Strauss 132. A, x 734; B, D, x 688; C, E, F, x 535.

Plants medium-sized, creeping, gregarious or in

crowded colonies, green; shoots mostly simple, 5.0-7.0

mm long, 1.8 mm high, 3.0 mm wide, occasionally once

furcate close to apex (Figure 6A) or to base (Figure 6B),

moderately to widely divergent, apical segments 2-5 mm

long. Stems prostrate, fleshy, in cross section at apex

400-500 pm (13 cell rows) high, 530-650 pm wide, ta-

pering proximally, at base (300-) 400^50 x 400-480 pm,

plano-convex (Figure 51). Rhizoids purple, 12-20 pm

wide. Leaves overlapping, undulating, very obliquely in-

serted succubously, generally oblong, longer than wide

and wider above than basally, 2000-2500 pm long, width

1325-2500 pm above, 1050-1400 pm below, apex sub-

truncate, with shallow angular lobes (Figure 5A-F); mar-

gins with or without 1 or 2 slime papillae, the lower one

on proximal edge of leaf below midline often 2-celled

(Figure 5N). Leaf cells thin-walled, at upper margins

subquadrate, 25-30 x 32^5 pm, at lateral margins long-

rectangular, up to 87.5 X 22.5 pm; upper laminal cells 5-

or 6-sided, 57.5-75.0 x 27.5-37.5 pm; middle laminal

cells 62.5-102.5 x 42.5 x 47.5 pm; basal cells 75.0-87.5

X 37.5^0.0 pm. Oil bodies glistening, numerous, more

than 50 per cell, up to 2.5 pm in diameter; chloroplasts

densely scattered in cells when fresh, but later tending to

clump together, ± 5 pm in diameter (Figure 50).

Dioicous. Male plants hardly smaller than females.

Antheridia dorsal between leaves, in a row, white when

immature, later turning yellow, globose (Figure 6C), short-

stalked, ± 210 pm in diameter, posteriorly shielded by

perigonial bracts (Figure 5J-M), ± 550 x 370 pm, cells

in interior subquadrate to hexagonal, 45.0-62.5 x

40.0-42.5 pm, apices divided into (2)3 or 4 finger-like

projections (Figure 6C), 150-300 pm long and 1 or 2 cell

rows wide. Archegonia naked, in an interrupted, irregular

row or in groups dorsally along the stem (Figure 6D).

Pseudoperianths produced in acropetal sequence, 1 or 2

per shoot, the younger one near the apex and the other

(if present) usually more proximally, but occasionally in

close proximity; sessile, ± funnel-shaped (Figure 5H), at con-

stricted base 425-525 pm wide, up to 1750 pm long, width

across flaring mouth 2250-2375 pm, consisting of (4)5 or 6

lobes (Figure 6E), with 1-4(5) deep clefts to near the base

(Figure 5G), apices truncate, entire, or with sharply pointed

projections; cells comparable in shape and size to those of

leaves. Capsules globose, ± 625 pm in diameter, initially

entirely enveloped in calyptra, which is later shed, revealing

capsule wall (Figure 6F), the latter bistratose, the inner cell

layer with irregularly quadrangular cells, 32.5-37.5 x

20.0-37.5 pm, at each cell wall 2 or 3 nodular and sometimes

semi-annular thickenings, (Figure 5P). Seta delicate, up to

5.6 mm long, 110 pm in diameter. Spores brown, ± hemi-

spherical (Figure 7D), 40.0-52.5 pm in diameter; distal face

convex, usually covered with numerous tall, conical spines

up to 5 pm long (Figure 7A), rarely broken (Figure 7B), and

FIGURE 4. — Map showing distribution of F. gemmifera, □; F. glenii, •;

and F. straussiana, O, in Northern Province, Gauteng and Mpu-

malanga.

22

Bothalia27,l (1997)

FIGURE 5. — Fossomhwnia f’lenii. A-F, leaves; G, opened pseudoperianth; H, pseudoperianth from side; I, stem in cross section; J-M, perigonial

bracts; N, detail of marginal area of leaf, with slime papilla (see arrow); O, median leaf cells with oil bodies and chloroplasts; P, cells in

capsule wall. A-C, E-K, M-P, S.M. Perold 3052', D, L, Glen 2134. Drawings by G. Condy, Scale bars: A-H, 500 pm; J-M, 250 pm; 1, 100

pm; N-P, 50 pm.

Bothalia27,l (1997)

23

FIGURE 6. — Fossombronia glenii. A, stem branching near apex; B, stem branching near base; C, antheridia shielded by perigonial bracts; D,

archegonia naked along stem, young pseudoperianth near apex; E, pseudoperianth with capsule from above; F, close-up of capsule in

pseudoperianth. A, Perold & Van Rooy 3569\ B-E, Perold & Van Rooy 3568', F, Glen 2134. A, B, D, E, x 7; C, x 15; F, x 23.

sometimes connected by short ridges, even occasionally

forming some high-walled areolae (Figure 7C); proximal

face ± flat, lacking a distinct triradiate mark, sprinkled

with numerous fine to coarse tubercles (Figure 7E), which

are sometimes rather flattened; circumference with numer-

ous, up to 45, projecting spines. Elaters yellow, smooth,

70.0-137.5 pm long, 7.5-10.0 pm wide in middle and

tapering slightly toward tips, 2- (Figure 7F) and some-

times 3-spiral in the same elater.

Fossombronia glenii grows on rather sandy soil in dry

stream beds or on stream banks, or else on soil pockets

in exposed roeky cliffs above streams. So far, it has only

been collected at a few localities in Northern Province,

Gauteng and Mpumalanga (Figure 4), but it is surely more

widespread. It is easily recognized by the rather small

pseudoperianth split into several lobes, by perigonial bracts

with finger-like projections and by spinous spores. Beeause

of its spinous spores, a specimen collected by Mogg (CH

FIGURE 7. — Fossombronia glenii. Spores and elater. A-C, distal face; D, side view of distal face; E, proximal face; F, elater. A, F, Glen 2146', B,

D, Mogg CHI 57. C, S.M. Perold 3052', E, Glen 2134. A. x 562, B, x 700; C, x 737; D, x 694; E, x 582; F, x 575.

24

Bothalia 27,1 (1997)

157) at Wonderboom Poort, Pretoria, was previously misi-

dentified as the so-called F. crispa, but the name had been

misapplied (Perold b in press). Fossombronia glenii is dis-

tinguished from the winter rainfall species, F leucoxantha,

by the dentate leaves and pseudoperianths (Perold a in press).

Both species have spinous spores.

This species is named in honour of an esteemed col-

league at the National Botanical Institute, Dr H.F. Glen,

who has often collected liverworts, together with his wife,

Mrs R. Glen (another colleague) and their young daughter,

Melissa.

3. Fossombronia straussiana Perold, sp. nov.

Plantae repentes, crebrae in coloniis. Folia plerumque

valde convoluta, dense imbricata, irregulariter formata,

longiora vel breviora quam latiora, interdum cum appen-

diculo oblongo in basi proximali. Rhizoidea hyalina vel

brunnescentia. Monoicae. Antheridia archegoniaque con-

ferta, appendiculo foliari basali partialiter circumdata.

Pseudoperianthium campanulatum, orificio patent! lobati,

processibus plures lamellatis lateraliter procurrentibus.

Sporae 35.0-42.5 pm diametro distaliter usque ad 10

lamellis discontinuis, cristis tenuibus interjunctis,

aliquando areolas imperfectas facientibus, saepe inclu-

sionibus papilliformibus vel cruciformibus; superficie

proximali cum nota triradiata imperfecta, non valde

distincta, superficiebus cum papillis humilibus et cristis

irregularibus brevibus tectis. Elateres 107.5-175.0 pm

longi, ter vel bis spirales, papillis tenuibus humilibus tecti.

TYPE. — Northern Province, 2427 (Thabazimbi); Krans-

berg. Farm Geelhoutbosch, on streambank, directly south

of rondavel, (-BC), S. Strauss 133 (PRE, holo.); S. Strauss

134, same locality (PRE, para.).

Plants medium-sized, creeping, in crowded colonies,

green; shoots simple, up to 9.5 mm long, 1.4 mm high,

2.3-3. 0 mm wide, or once-furcate, apical segments nar-

rowly divergent (Eigure 9A), 4.5-7 .5 mm long, basal part

3.0- 4. 5 mm long. Stems prostrate, fleshy, in cross section

at apex (300-)400-500 pm (13 cell rows) high, 550-680

pm wide, tapering basally to 300 x 320 pm, plano-convex

(Figure 8K). Rhizoids of all plants entirely hyaline or

brownish, 10.0-12.5 pm wide. Leaves mostly highly con-

voluted, densely imbricate (Figure 9B), obliquely inserted

succubously, irregularly shaped, longer than wide (Figure

8B, C) to shorter than wide (Figure 8A, E, E), (575-)

1300-1500 X (850-) 1025- 1500 pm, sometimes with an

oblong appendage at proximal base (Figure 8C, D), ± 675

X 500 pm; margins with up to 8 slime papillae at angu-

lations or in between, 22.5-27.5 x 22.5-25.0 pm. Leaf

cells thin-walled, at upper margins subquadrate or rectan-

gular across (Figure 8L), 25.0-27.5 x 30.0-40.0 pm; at

lower lateral margins long-rectangular, up to 47.5 x 22.5

pm; upper laminal cells subquadrate, 25.0-32.5 x

30.0- 32.5 pm; middle laminal cells 5- or 6-sided,

47.5-62.5 X 27.5-40.0 pm; basal cells 70.0-75.0 x 50.0

pm. Oil bodies glistening, faintly granular, 10-25 per cell,

up to 2.5 pm in diameter; chloroplasts densely scattered

in cells, 3-5 pm in diameter (Figure 8M).

Monoicous. Antheridia short-stalked, globose, 1 10-135

pm in diameter, dorsal on stem between leaves, basal leaf

appendage partly curved around 1 or 2 (Figure 9D, E),

sometimes leaf appendage detached, forming an oblong

perigonial bract (Figure 8J), 430-620 x 300-420 pm, mar-

gin with 1 or 2 papillae. Archegonia in close proximity

to antheridia. Pseudoperianths soon forming (Figure 9C)

after fertilization in acropetal sequence, often 2 per shoot,

close together near apex, older more proximal one some-

times with capsule already dehisced; sessile, campanulate

(Figure 9F), at constricted base ± 600 pm wide, 1250 pm,

rarely to 1825 pm long, width across flaring mouth 2250

pm, margin ± scalloped, consisting of 4-7 shallow,

rounded lobes (Figure 8H), several lamellate outgrowths

projecting laterally from sides (Figure 81); cells compara-

ble in shape and size to those of leaves. Capsules globose,

500-610 pm in diameter, capsule wall bistratose, cells in

inner layer irregularly shaped, 30.0-37.5 x 15.0-20.0 pm,

crowded with nodular and some semi-annular thickenings

(Figure 8N). Seta 2.25-4.0 mm long, 140-150 pm in di-

ameter, 6 ox 1 cells across (Figure 8G). Spores light

brown, hemispherical (Figure lOB), 35.0^2.5 pm in di-

ameter, including marginally projecting lamellae; distal

face convex (Figure lOA, C, D), ornamented with up to

10 discontinuous lamellae, ± 2.5 pm long and 5. 0-7.5 pm

apart, but interconnected with faint cross ridges, some-

times forming incomplete areolae and frequently with pa-

pilla-like or cross-like inclusions; proximal face (Figure

lOE) with incomplete and not very distinct triradiate mark,

facets covered with low papillae and short irregular ridges,

up to 25 lamellae projecting around periphery. Elaters

light brown, 107.5-175.0 pm long, 7.5 pm wide in middle,

tapering to ends, 3-spiral or partly 2-spiral, covered with

fine, low papillae (Figure lOF).

Fossombronia straussiana is often mixed with other

Fossombronia species and grows on streambanks and in

seepage areas at several localities in the Northern Province

and Gauteng (Figure 4). More specimens of it, J. Braggins

91/191 and S.M. Perold 2654, were collected in 1991 in

Malawi on the Zomba Plateau, as well as at Nyika Nat.

Park, S.M. Perold 2663] thus it appears to be widespread.

Vanden Berghen (1965) reported F. husnotii (with hyaline

rhizoids) from the Congo Republic, Symoens 4329, and

from Tanzania ( = Tanganyika), Bryan 1036. The spores

of these specimens have much taller lamellae than those

F. straussiana and also appear to differ from those of F.

husnotii. In 1978 Vanden Berghen reported F. husnotii

from Shaba, Zaire (Malaisse 9039), without mentioning

the rhizoids, but presumably referred here because they

are hyaline. Scott & Pike (1988a) comment that the spores

of F. husnotii are extraordinarily variable and that more

research is required. Beside F. husnotii, which has mostly

hyaline rhizoids, Scott & Pike (1984) also described three

new Australian species, F. punctata, F. scrobiculata and

F. vermiculata, as having hyaline rhizoids on all or on

most plants.

Fossombronia straussiana has been named in honour

of Mrs Susan Strauss, owner of the Farm Geelhoutbosch,

where she has collected it a number of times, together

with other Fossombronia species.

This species is easily distinguished by its hyaline or

brownish rhizoids and dense, frilly leaves, its pseudope-

rianth with lamellate, lateral outgrowths, by the spore or-

namentation that usually has incomplete areolae with

Bothalia27,l (1997)

25

FIGURE 8. — Fossombronia straussiana. A-F, leaves; G, cross section of seta; H, opened pseudoperianth; 1, pseudoperianth from side; J, perigonial

bract and 2 antheridia; K, cross section of stem; L, detail of leaf margin; M, median leaf cells with oil bodies and chloroplasts; N, cells in

capsule wall. A-C, E, F, H, I, K, N, Strauss & Relief CHI 3655; D, J, Perold & Koekemoer 3116a; G, Strauss CHI 3653; L, M, Strauss

CH13651. Drawings by G. Condy. Scale bars; A-F, H, I, 500 pm; G, K, 100 pm; J, 250 pm; L-N, 50 pm.

26

Bothalia27,l (1997)

FIGURE 9. — Fossomhronia strcmssiana. A, stem branching near apex; B, leaves crowded on stem; C, antheridium and archegonium in close proximity

between leaves; D, proximal leaf appendage curved around gametangia; E, close-up of same; F, pseudoperianth with seta emerging from it.

A, F, Strauss 133', B, Perold & Koekemoer 3124a', C-E, Perold & Koekemaer 3116a. A, x 7; B, x 10; C, x 66; D, x 15; E, x 106; F, x 23.

inclusions and by the usually finely papillose elaters. The

specimen, Scott 13, has a larger and more elaborate

pseudoperianth than those usually encountered in this spe-

cies. Its spore ornamentation is very similar, however, and

the rhizoids are hyaline.

ACKNOWLEDGEMENTS

I wish to thank the referees of this article. Dr E.O.

Campbell and Prof R. Stottler for their constructive com-

ments as well as the kind people who have collected

specimens for me, particularly Dr H.F. Glen and Mrs

Susan Strauss, also my colleagues at the National Bo-

tanical Institute, Ms Marinda Koekemoer and Mr Jac-

ques van Rooy who helped me with fieldwork. Many

thanks to Dr Glen for translating the diagnoses into

Latin, Mrs J. Mulvenna and Ms D. Maree for typing

the manuscript, Ms G. Condy for the drawings and Mrs

A. Romanowski for developing and printing many pho-

tographs.

FIGURE 1 0, — Fossomhronia straussiana. Spores and elater. A, C, D, distal face; B, side view of distal face; E, proximal face; F, elater. A, B, E, Scott 13',

C, Strauss & Retief CHI 3655, D, S.M. Perold 32H0', F, Perold & Koekemoer 3124a. A, C, x 121', B, x 1084; D, x 588; E, x 674; F, x 839.

Bothalia27,l (1997)

27

REFERENCES

ARNELL, S.W, 1952. Hepaticae collected in South and West Africa

(1951). New and little known species. Botanislai Notiser 105;

307-329.

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

ence Council, Stockholm.

EDWARDS, D. & LEISTNER, O.A. 1971 . A degree reference system for

citing biological records in southern Africa. Mitteilungen der

Botanischen Staatssammlung, Munchen 10: 501-509.

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

C.G. 1844-1847. Synopsis hepaticarum. Hamburg, Meissner. Re-

printed 1967, Cramer, Lehre.

KRAUSS, F. 1846. Pflanzen des Cap- und Natal-Landes, gesammelt und

zusammengestellt von Dr Ferdinand Krauss. Flora 29: 135, 136.

LEHMANN, J.G.C. 1829. Hepaticarum capensium a C.F Ecklon. Lin-

naeaA-. 357-371.

MITTEN, W. 1878. List of Hepaticae collected by Rev. A.E. Eaton at the

Cape of Good Hope (Aug. & Sept. 1874). Journal of the Linnean

Society (Bot.) 16: 187-197.

PEROLD, S.M. 1997. Studies in the liverwort genus Fossombronia

(Metzgeriales) from southern Africa. 2. An amendment to three

species from Western Cape, described by S.W. Amell. Bothedia

27: 29-38.

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

(Metzgeriales) from southern Africa. 4. A re-examination of F.

crispa, F. leucoxantha and F. tumida. Bothalia.

PEROLD, S.M. b in press. The lectotypification of Fossombronia crispa.

Bothalia.

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

America. Vol.5. Field Museum of Natural History, Chicago.

SCOTT, G.A.M. & PIKE, D.C. 1984. New species of Fossombronia from

Australia. Journal of the Hattori Botanical Laboratory 56:

339-349.

SCOTT, G.A.M. & PIKE, D.C. 1987a. The Fossombronia foveolata

complex. Lindbergia 13: 79-84.

SCOTT, G.A.M. & PIKE, D C. 1987b. Studies on Fossombronia in

Australia. 11. Fourteen more new species. Journal of the Hattori

Botanical Laboratory 62: 367-386.

SCOTT, G.A.M. & PIKE, D.C. 1987c. Studies on Fossombronia in

Australia. III. Taxonomic and nomenclatural problems. Journal of

the Hattori Botanical Laboratory 63: 99-105.

SCOTT, G.A.M. & PIKE, D.C. 1988a. Revisionary notes on Fossom-

bronia. The Bryologist 9\ : 193-201.

SCOTT, G.A.M. & PIKE, D.C, 1988b. A new species of Fossombronia

from Australia. Beiheft zur Nova Hedwigia 90: 109-112.

SIM, TR. 1926. The Bryophyta of South Africa. Transactions of the

Royal Society of South Africa 15: 1^75. Cape Town.

SOLOMON, J.C. 1995. (ed.) Herbarium News 15 (11/12): 66. Missouri

Botanical Garden.

STEPHANI, F \900. Species hepaticarum No\. 1. Geneva.

VANDEN BERGHEN, C. 1965. Hepatiques recoltees par le Dr J.-J.

Symoens dans la region peri-tanganyikaise. Bulletin de la Societe

Royale de Botanique de Belgique 98: 129-174.

VANDEN BERGFIEN, C. 1978. Hepatiques du Shaba. Corrections et

additions. Bulletin du Jardin botanique rmtioruil de Belgique 48:

367-372.

SPECIMENS EXAMINED

Held at PRE, unless otherwise indicated. Bracketed

numbers after citation of collector’s name and collecting

number refer to the species described in the text in alpha-

betical order, namely: F. gemmifera (1); F. glenii (2); and

F straussiana (3).

Braggins 91/191 (3) (Malawi).

Glen 2134 (2) (paratype), 2146 (2) (holotype).

Koekemoer 976 (2).

Mogg CH157 (2) BOL, PRE.

Perold S.M. 2017 (1); 3052 (2); 3280, 3281, 2654 (Malawi), 2663

(Malawi) (3). Perold & Koekemoer 3116a, 3124a (3); 3116b, 3124b,

3129 (1). Perold & Van Rooy 3555, 3559a (paratype). 3564, 3565 (1);

3568, 3569 (2).

Scott 13 (3) CH3697. Strauss 132 (1) (holotype), 133 (3) (holotype), 134

(3) (paratype), CH13651, CH13653, CH13654 (3). Strauss & Retief

CH13655 (3).

Wager 14 (3).

,■ ^(Mjrin, so».tav«w««wiS ■WCl'A'^ »»l'w''^'v»«rt to .js.wq iu CM/2

-- - - ^ - - — - ■ ■^' -■ — - - - - — — -”'“"^::aSrkt^to3'

• '•• ' kv ' K' «. kM ft • ■

•1. -i—riS^.-.^-a ^t»* »A'&f :5.^rs »Vta A hA-A.ti TTVY

■Sis

Bothalia 27,1:29-38 (1997)

Studies in the liverwort genus Fossombronia (Metzgeriales) from southern

Afilca. 2. An amendment to three species from Western Cape, described by

S.W.AmeU

S.M. PEROLD*

Keywords: Fossombronia, F. capensis, F. densilamellata, F. montaguensis, Hepaticae, Metzgeriales, southern Africa, Western Cape

ABSTRACT

Subsequent to his visit to South Africa in 1951, S.W. Arnell ( 1 952), described three new Fossombronia species from Western

Cape, namely F. capensis, F. densilamellata and F. montaguensis. Unfortunately, however, they were not described in detail, nor

were they fully illustrated. An attempt is hereby made to augment Arnell’s descriptions and to illustrate his species more

completely, with the aid of drawings and SEM micrographs. A distribution map is also provided. Scott & Pike (1988), after

examining many Fossombronia specimens of world-wide origin, concluded that the above three species were good species, a

conclusion 1 support.

1. Fossombronia capensis S.W. Arnell in Botaniska

Notiser 3: 314 ( 1952); S.W. Arnell: 81 (1963). Type: Western

Cape, 3423 (Knysna): Bracken Hill Forest, (-AA), roadside,

5.W Arnell 1376 (S, holo.!; PRE, iso.).

Plants in crowded colonies, green; shoots medium-

sized in male plants, 10-15 mm long, 1. 3-2.0 mm high,

2. 8-3.0 mm wide; female plants more common and

rather larger, simple, up to 18 mm long, 1.5-2. 5 mm

high, 2. 5^.0 mm wide, or once/twice to repeatedly fur-

cate, segments moderately to widely divergent, 4.0-6.0

mm long. Stems prostrate, tapering proximally, chloro-

phyllose, occasionally ventral row of cells purple, some-

times with a lateral bud or side branch, plano-convex in

cross section, in male plants (Figure II) 250-350 pm (11

cell rows) high, 420-610 pm wide, in female plants (Fig-

ure IJ) 270-350 pm (10-12 cell rows) high, 400-610

pm wide. Rhizoids purple, ± 15 pm wide. Leaves over-

lapping, widely spreading, succubously inserted (Figure 2A),

apically small, free margin rounded, soon becoming larger,

obovate, short- or long-rectangular, or irregularly shaped, oc-

casionally slightly notched and shortly bilobed; in male

plants rather smaller (Figure lA-D), 1125-1375 x

1225-1350 pm; in female plants (Figure lE-G) mostly

larger, 1000-2750 x 1150-2575 pm above, sometimes,

when sides not parallel, narrower below, 1075-1750 pm

wide; margins almost entire or with ± 6 well-spaced slime

papillae, ± 25.0 x 17.5 pm. Leaf cells thin-walled, in male

plants not appreciably different from those of females, at

upper margins (Figure IH) rectangular across, 22.5-32.5

X 37.5-45.0 pm, at lateral margins long-rectangular,

40.0-62.5 X 20.0-25.0 pm, upper laminal cells 5- or 6-

sided, 37.5-57.5 x 35.0-50.0 pm, middle laminal cells

(Figure IK) 65.0-87.5 x 50.0-57.5 pm, basal cells

67.5-87.5 X 50.0-62.5 pm. Oil bodies quite variable in

number, 17-37 per cell, larger ones ± 5 pm in diameter

and granular, others much smaller and smooth; chloro-

* National Botanical Institute, Private Bag X 10 1, Pretoria 0001 .

MS. received: 1996-10-08.

plasts numerous, mostly rounded, ± 5 pm in diameter,

sometimes elongate, 7.5 pm long (Figure IK).

Dioicous. Antheridia dorsal on stem, generally in 2

crowded rows (Figure 2B), short-stalked, globose or ovoid,

160-250 pm in diameter, each shielded by a bract (Figure

IM-P), 480-770 X 330-640 pm, sometimes 2 adjacent

ones joined together, margins with 3 or 4 projecting pa-

pillae or processes, cells in interior 4- or 5-sided, 42.5-75.0

X 37.5-67.5 pm. Archegonia in 1 or 2 rows (Figure 2C,

D) along stem, naked; sometimes several per branch, at

intervals (Figure 2E) or 2 adjacent, becoming fertilized.

Pseudoperianth (Figure IQ, R) campanulate, proximal to

apex, as tall as leaves or projecting somewhat above them,

raised on a short stalk, then widely Haring above,

1875-2125 pm long, 1625-2125 pm wide across mouth,

margin with 10-15 angular projections, each with a pa-

pilla, ± 20.0 X 17.5 pm, often with winged outgrowths on

outside (Figure 2F); cells comparable in shape and size

to those of leaves. Capsules globose, ± 850 pm in diame-

ter, wall bistratose, cells in inner layer irregularly shaped

(Figure IT), 32.5-50.0 x 27.5-35.0 pm, each cell wall

with 1-3 dark brown, nodular and sometimes semi-annu-

lar thickenings. Seta 2. 8-4.0 mm long, 250-300 pm in

diameter, 6-8 cells across (Figure IF). Spores golden

brown to brown, hemispherical, 42.5-55.0 pm in diameter,

including lamellae projecting at margin; distal face (Figure

3A, B) convex, with up to 8 lamellae, ± 5 pm high and

5-10 pm apart running across face, sometimes in different

directions or parallel to each other (Figure 3C), occasion-

ally anastomosing and forming a few to several areolae,

surface between lamellae with fine cross striations (Figure

3D); proximal face (Figure 3E) lacking triradiate mark,

flat, covered with irregularly shaped papillae and short

ridges, sometimes with scattered granules, 12-16 ‘spines’

(i.e. ‘end-on’ view of terminations of lamellae from the

distal face) projecting around spore periphery and

joined by a 5 pm wide, incomplete membranous wing

or perispore. Elaters (Figure IS) mostly delicate, outer

wall collapsing on drying, rather short and stout, blunt at

tips, 62.5-125.0 x 10.0-12.5 pm, loose spirals faintly

greenish yellow, strands often difficult to distinguish, but

30

Bothalia27,l (1997)

FIGURE 1 -Fossombronia capensis. A-D, male leaves; E-G, female leaves; H. detail of upper margin of leaf; I, cross section of male stem J cross

section of female stem; K, median leaf cells with oil bodies (solid lines) and chloroplasts (dotted lines); L, cross section of seta M P bracte,

Q, opened pseudoperianth; R, pseudoperianth from side; S,. S2, elaters; T, cells in capsule wall. A, ^,Koekemoer 998, C-K M-R, S.M.

Perold 3494\ L. S. T, S.M. Perold 3492. Scale bars; A-G, Q, R, 500 pm; I, J, M-P, 250 pm; L, 100 pm; H, K, T. 50 pm, Si, S2, 25 pm.

Bothalia27,l (1997)

31

FIGURE 2. — Fossombronia capensis. A, thallus branches; B, male plants with rows of bracts; C, female plant with 1 or 2 rows of archegonia; D,

detail of female plant with archegonia; E, female plant with 2 pseudoperianths; F, young pseudoperianth from above with outgrowth (see

arrow). A, S.M. Perold 3497'. B-F, Koekemoer 998. A, B, C, E, x 7; D, F, x 30.

sometimes better developed, with brown rings or even

3 spirals and narrowed at tips, 5 pm wide (Figure

3F).

The correct collector’s number of the holotype speci-

men is Arnell 1376 and not 1876 (Amell 1952). Fossom-

bwnia capensis is confined to the southeastern Western

Cape, in the winter rainfall region (Figure 4). It grows on

soil, rarely extending onto slate, in forests, at roadsides,

or in clearings, on streambanks or on vertical, shaded rock

walls in soil pockets. It is distinguished by the relatively

large size of the female plants with overlapping leaves

and exposed stems, by fairly common, smaller male plants

with large bracts subtending 2 rows of adjacent antheridia,

by the spore ornamentation with widely separated lamel-

lae and marginally by 12-16 projecting ‘spines’ joined by

FIGURE 3. — Fossombronia capensis. Spores. A, B, distal face; C, side view of distal face; D, detail of lamellae and spore surface on distal face; E,

proximal face; F, elaters. A, Arnell 1783', B, Arnell 1555', C, S.M. Perold 3494', D, Arnell 1470', E, Koekemoer 998', F, Arnell 1477. A, x 612;

B, X 672; C, x 625; D, x 2755; E, x 739; F, x 406.

32

Bothalia 27,1 (1997)

FIGURE 4. — The distribution of Fossombronia capemis, •; F. densi-

lamellata, ▲; and F. montaguensis, ■, in southern Africa.

an incomplete membranous wing or perispore and also by

generally reduced, delicate elaters. Arnell (1952, 1963) re-

ferred to the latter as ‘leaf-like’. Poorly formed elaters are

also present in F. cristula (Piippo 1991; Scott & Pike

1987) and in F. foveolata var. cristula (Schuster 1992).

Schuster treated cristula as a variety of F. foveolata, al-

though it is regarded by Scott & Pike (1987) as a complex

[in which they include the southern African F. zeyheri

(Perold in press)]. According to Scott & Pike (1987), cu-

rious and variously malformed elaters are quite common

and they found cristula-type. elaters occurring in individu-

als apparently belonging to a number of other taxa, with

spores unlike those of the F foveolata complex. They ad-

mit, however, that F cristula and F foveolata are perhaps

not distinct, but on the evidence of dioicism versus

monoicism, they presently maintain them as separate spe-

cies. All F capensis specimens are dioicous and those

from the George/Knysna/Brackenfell/Gouna and Deep

Walls Forests and Diep River areas have poorly formed elat-

ers. Collections from the nearby Bloukranz Pass {LUbenau-

Nestle SA 139/2 and S.M. Perold 3534, 3539-3541) have

elaters with well-formed spirals, but have been referred here

because the spores and plants are closely similar.

The strong aromatic smell referred to by Arnell was

not observed. Arnell (1952) stated that the spores of F.

capensis and F. pusilla had the same appearance, but he

distinguished F capensis from the latter because it was

dioicous and had large bracts subtending the antheridia in

the male plants. Initially Arnell (1953) thought that F

pusilla did not grow in South Africa, hut in his Hepaticae

of South Africa (Arnell 1963) he included it. Its presence

here still needs to be confirmed. Earlier reports of it

(Lehmann 1829; Gottsche et al. 1846; Sim 1926) are most

probably based on misidentifications. Best (1990) lists F

pusilla as present in Zimbabwe. Vina et al. (1979) re-

ported the presence of F capensis and E pusilla in

Rwanda and Burundi, but they doubted the determination

of F. capensis for a specimen from Rwanda. Examination

of the spores of De Sloover 18574 (BR) from Rwanda

shows it to belong to a different species. De Sloover

13.345 and 19.118 are sterile. F. capensis appears to be

confined to a relatively small area in the southeastern part

of Western Cape which has winter rainfall. Its spores ripen

in spring and summer.

2. Fossombronia densilamellata S.W. Arnell in Bo-

taniska Notiser 1952: 317 (1952); S.WArnell: 80 (1963).

Type: Western Cape, 3318 (Cape Town): Lion’s Head near

IGoofnek, (-CD), S.W. Arnell 295 (S, lecto.!, here desig-

nated; PRE, isolecto.).

Plants in dense colonies or overlying mats, pale green

to yellow-green, older leaves dying, turning yellow-brown

and translucent, lower part of stem occasionally almost

denuded of leaves; shoots smallish to medium-sized in

male plants, up to 12 mm long, 1.1 mm high, 2.5 mm

wide; female plants far more common and rather larger,

shoots sometimes simple, 9-14 mm long, 1.4-1 .7 mm

high, 1.6-3. 8 mm wide, at pseudoperianth up to 4.4 mm

wide, mostly bifurcate with terminal segments (Figure 6A)

closely to moderately divergent and of unequal length,

2-4- mm long. Stems prostrate, green, occasionally central

core purple, sometimes apically very shortly branched,

with dorsal bud-like layers of small leaves at tips, lateral

branches often developing from latero-ventral buds,

plano-convex in cross section, in male plants (Figure 51)

apically 210 pm (9 cell rows) high, 400 pm wide, basally

350 pm high, 350 pm wide, in female plants apically swol-

len (Figure 5J), 460 pm (± 16 cell rows) high, 700 pm

wide, gradually tapering toward base (Figure 5K), 280 pm

high, 430 pm wide. Rhizoids purple, 12.5-20.0 pm wide,

sometimes with flat tips. Leaves erect, imbricate, undulate

along upper margin, succubously inserted, markedly

decurrent on stem (Figure 6B), subquadrate to rectangular,

sometimes wider above than below, apex truncate or with

several low triangular or toothed projections, in male

plants (Figure 5A-C) 1125-1750 x 1125-1625 pm, in fe-

male plants mostly larger (Figure 5D-G), 1625-2750 pm

long, width above 1250-2125 pm, below 825-1250 pm;

margins with up to 13 slime papillae, 25.0 x 22.5 pm,

mostly at angulations and often more numerous on distal

(leading) edge (Figure 5C) than on proximal (trailing)

edge. Leaf cells (Figure 5H) above somewhat thicker-

walled than below, in male plants not appreciably different

from those of females, at upper margins rectangular

across, 27.5-35.0 x 32.5-57.5 pm, at lateral margins long-

rectangular, 52.5-75.0 X 20.0-22.5 pm, mostly longer at

proximal edge, up to 140.0 x 17.5 pm, upper laminal cells

4- or 5-sided, 45-50 x 45-50 pm, middle laminal cells

6- or 7- sided, walls bulging, 82.5-95.0 x 50.0-55.0 pm,

basal cells 100-125 x 25-60 pm. Oil bodies (Figure 5L)

very variable in number, in young leaves some cells with

8-10, in others much more numerous, round or bean-

shaped, up to 2 pm in diameter; chloroplasts numerous,

rounded, ± 5 pm in diameter.

Dioicous. Antheridia dorsal on stem, in 1 or 2 rows,

globose, ± 180 pm in diameter, each shielded by a bract

(Figures 5M-0; 6C), 300-450 x 320-400 pm, with sev-

eral projections, mostly topped by a mucilage papilla, cells

in body 4-7-sided, ± 62.5 x 25.0 pm. Archegonia (Figure

6D) in a row dorsally along stem, naked, sometimes sev-

eral per branch becoming fertilized, occasionally next to

each other or even surrounded by the same pseudoperi-

anth. Pseudoperianth (Figures 5P, Q; 6F) campanulate,

proximal to apex (Figure 6E), projecting ± 750 pm above

top of leaves, raised on a short stalk, ± 425 x 500-700

pm, then widely flaring above, 2250-2750 pm long, 2375

pm wide across mouth, its margin with toothed or angular

projections, 200-250 pm long, crowned with papillae.

Bothalia27,l (1997)

33

FIGURE 5. — Fossombronia densilamellata. A-C, male leaves; D-G, female leaves; H, detail of distal margin of male leaf; I, cross section of male stem;

J, cross section of apical part of female stem; K, cross section of basal part of female stem; L, median leaf cells with oil bodies and chloroplasts;

M-O, bracts; P, opened pseudoperianth; Q, pseudoperianth from side; R, cross section of seta; S, cells in capsule wall. A, B, E-G, I-K, M, O, P-S,

S.M. Perold 3349', C, H, N, Garside 6510', D, L, S.M. Perold 629. Scale bars; A-G, P, Q, 5(X) pm; H, R, 100 pm; 1-L, S, 50 pm; M-O, 250 pm.

34

Bothalia27,l (1997)

FIGURE 6. — Fossombronia densilamellata. A, thallus branches; B, apical leaves; C, male plant with rows of bracts; D, female plant with row of

archegonia; E, female plant with 2 pseudoperianths (see arrows); F, pseudoperianth. A-C, S.M. Perold 3346\ D-F, S.M. Perold 3349. A, D,

X 7; B, X 17; C, x 13; E, x 9; F, x 20.

± 22.5 X 17.5 |jm, sometimes with a winged outgrowth

at the side, ± 1925 |im long, 775 |im wide at apex; cells

comparable in shape and size to those of leaves. Capsules

globose, 700-875 pm in diameter, cells in inner layer of

bistratose wall (Figure 5S) irregularly shaped, 35.0-50.0

X 27.5-37.5 pm, each cell wall with 2 or 3 nodular and

sometimes semi-annular thickenings. Seta (Figure 5R)

5.0-9.5 mm long, 140-150 pm in diameter, up to 8 cells

across. Spores light brown or yellow-brown, hemispheri-

cal, 40-45 pm in diameter, including ‘spines’ projecting

± 2.5 pm at margin, these not connected by a wing; distal

face convex, with 12-16 thin, parallel, curving lamellae,

2.5 pm high (Figure 1C) running across (Figure 7A), cen-

tral ones usually longer, sometimes branched, lateral ones

(Figure 7D) shorter and ± radiating, separated by ± 2.5

pm, sometimes interconnected by slender threads (Figure

7B); proximal face (Figure 7E) mostly lacking a distinct

triradiate mark, rarely more pronounced, generally oma-

FIGURE 7. — Fossombronia densilamellata. Spores. A, B, distal face; C, detail of lamellae on distal face; D, side view of distal face; E, proximal

face; F, elater. A. Garside 65 W' B, Duthie CH 165P, C, D, Arnell & Garside 260; E, S.M. Perold 629; F, S.M. Perold 2355. A, x 7 1 9; B, E,

X 772; C, X 2795; D, x 865; F, x 805.

Bothalia27,l (1997)

A

35

FIGURE 8. — Fossombronia montaguensis. A-D, young apical leaves; E, F, K, older leaves; G-J, proximal leaves; L, detail of upper margin of leaf;

M, median leaf cells with oil bodies and chloroplasts mostly clumped together; N, cross section of stem; O. opened pseudoperianth; P,

pseudoperianth from side; Q, cells in capsule wall. A, F-J, L-Q, S.W. Amell 73 1\ B, E, K, 5. IV. Arnell 724\ C, D, SM. Perold 3454. Scale

bars: A-K, O, P, 500 pm; L, 100 pm; M, Q, 50 pm; N, 250 pm.

36

Bothalia27,l (1997)

FIGURE 9. — Fossombronia montaguensis. A, thallus branches; B, apical leaves; C, female plant with archegonia; D, female plant with

pseudoperianth near apex (see arrow); E, pseudoperianth from the side; F, pseudoperianth from above. A, B, 5. M Perold 3453\ C-F, S. W.

ArneU73l. A, x 12;B, x24; C, x 8; D, x 7; E x 9; F, x 11.

merited with coarse or slender pointed processes or short,

uneven ridges, around circumference up to ± 30 projecting

lamellar ‘spines’. Elaters (Figure 7F) light brown,

120-180 X 7.5-10.0 pm, tapering to tips, smooth, bispiral,

rarely trispiral, occasionally branched.

Fossombronia densilamellata is known only from

Western Cape and grows on partially shaded earth banks

at roadsides or on river banks. It has been collected at

Kloofnek, Round House, Lion’s Head, Newlands in Cape

Town and at Camps Bay, as well as at Franschhoek, Stel-

lenbosch and Algeria Forest (Figure 4). At PRE, the speci-

men Arnell 762 from Cogman’s Kloof, was labelled F.

densilamellata, but the packet contains no plant material,

only a slide preparation of the capsule wall without any

spores; the determination could thus not be verified and

Cogman’s Kloof cannot with certainty be included in its

distribution range.

FIGURE 10. — Fossombronia montaguensis. Spores. A, B, distal face; C, detail of lamellae and spore surface on distal face; D, side view of distal

faee; E, proximal face; F, elater. A-F, Arnell 731. A, x 745; B, x 699; C, x 1637; D, x 779; E, x 759; F, x 852.

Bothalia27,l (1997)

37

Fossombronia densilamellata is distinguised by its un-

dulating, decurrent leaves, rare and somewhat smaller

male plants with bracts shielding the antheridia and by

the spore ornamentation with 12-16 narrowly spaced,

thin, parallel lamellae on the distal face. The species fruits

in late winter and early spring and soon dies off, and only

the tuberous stem apices survive the dry summers. In

specimen SM. Perold 2356, from Algeria Forest, some of

the capsules bore spores with thick, granular ridges, not

thin lamellae; repeated samplings eventually turned up

spores with typical lamellae. Arnell had previously named

this species F. confertilamellata and his specimen 265,

held at S, still bears this epithet in his handwriting. Arnell

(1952) did not designate a holotype from the syntypes he

cited, Arnell 257, 265, 295, 762. The specimen Arnell 295

(S), is selected as lectotype because it closely matches the

protologue and a duplicate is held at PRE.

Arnell (1952, 1963) seems to have overestimated the

number of spines at the periphery of the spore. On SEM

micrographs they appear to be nearer to 30 than to 50.

Arnell (1963) refers to some similarity between the spores

of F. densilamellata and F. wondraczekii. In the latter they

frequently have papillae between the lamellae or some-

times the lamellae anastomose to form a few areolae in

the centre. Curiously, Arnell (1952) placed ‘E tiimida Sim’

in synonymy under F densilamellata. He must have meant

‘sensu Sim’ and he seems to have misinterpreted Sim’s

(1926) drawing and description of the spores of F. tiimida

Mitt. The drawing correctly illustrates the significant fea-

tures (although not well) and the description reads ‘lamel-

lae radiating from a few central areolae, and showing as

twenty-four to thirty spines on the margin’. Arnell (1963)

did not repeat these observations.

3. Fossombronia montaguensis S.W.Aniell in Bo-

taniska Notiser 1952: 316 (1952); S.W. Arnell: 83 (1963).

Type: Western Cape, 3320 (Montagu): Bath Kloof, (-CC),

S.W. Arnell 731 (S, lecto.!, here designated).

Plants in crowded overlying mats or more loosely ag-

gregated; leaves light green, becoming translucent or not,

and then mostly darker green, later on turning yellow at

margins or entirely so, sometimes juvenile leaves at apex

deep red; shoots medium-sized to large, up to 10 mm

long, 1. 1-2.0 mm high, 1. 8-3.5 mm wide, mostly repeat-

edly furcate, terminal segments (Figure 9A) 1-5 mm long,

moderately divergent. Stems prostrate, green or outer layer

distally purple, lateral branches occasionally developing

from latero-ventral buds, plano-convex in cross section

(Figure 8N), 280-380 pm high (12-14 cell rows),

490-750 pm wide, tapering toward base. Rhizoids purple,

10-25 pm wide. Leaves erect to spreading, frilly and

densely crowded apically (Figure 9B), becoming spaced

and lax proximally, succubous, subquadrate to long-rec-

tangular, sometimes irregularly shaped and wider above

than below, 1375-2500 x (600-) 1175-1950 pm, apical

leaves (Figure 8A-D) generally smaller than more proxi-

mal ones (Figure 8E-K), frequently shorter than wide,

900-1675 X 1250-1900 pm; margins with triangular or

irregular projections, with 6-16 slime papillae, ± 20 x 15

pm, often raised on a basal cell, 30-50 x 20-35 pm. Leaf

cells at upper margins (Figure 8L) quadrate, rectangular

across or irregular, 22.5-37.5 x 35.0-52.5 pm, at lateral

margins long-rectangular, 37.5-90.0 x 15.0-30.0 pm, up-

per laminal cells 5- or 6-sided, 32.5-45.0 x 32.5-55.0 pm,

middle laminal cells 6-sided (Figure 8M), 45.0-125.0 x

25.0-62.5 pm, basal cells 50.0-155.0 x 37.5-75.0 pm. Oil

bodies and chloroplasts were clumped together and could

not be studied adequately in the available material.

?Dioicous. No male plants seen. Archegonia in one or

two rows dorsally along stem (Figure 9C), naked, some-

times up to three per shoot becoming fertilized. Pseudo-

perianth (Figures 80, P; 9D-F) campanulate, ± same

height as leaves, sometimes slightly stalked, basally ± 875

pm wide, flaring widely above, up to 2575 pm long,

2375-3000 pm wide across cup-like mouth, its margin

with ± 30 triangular protrusions on a broad base, ± 160

pm long, and topped with a papilla, 17.5 x 17.5 pm, some-

times open at side, or where two component leaves are

joined, with a winged outgrowth; cells comparable in

shape and size to those of leaves. Capsules globose or

slightly flattened at the poles, 775-1050 pm in diameter,

cells in inner layer of bistratose wall (Figure 8Q) irregu-

larly shaped, 40.0-60.0 x 32.5-37.5 pm, each cell wall

with 1-3 nodular and sometimes semi-annular thicken-

ings. Seta 6.5-7. 5 mm long, 200-275 pm in diameter.

Spores brown, hemispherical, 40.0-47.5 pm in diameter,

including lamellae projecting ± 2.5 pm at margin, not

joined by a wing; distal face (Figure lOA-D) convex, with

up to 10 irregularly branched, long or short, sinuous la-

mellae, some breaking up into spines, others intercon-

nected by fine ridges running across, sometimes forming

incomplete areolae, ±5x5 pm; proximal face (Figure

lOE) lacking a distinct triradiate mark, ornamented with

low, irregular, rather short, branched ridges and with ± 27

irregularly shaped, variously sized, blunt, spine-like papil-

lae projecting at circumference. Elaters (Figure lOF) yel-

low-brown, 137.5-175 pm long, ± 7.5 pm wide in centre,

tips looped, ± 5 pm wide, smooth, bispiral throughout or

trispiral in centre.

Fossombronia montaguensis is most frequently found

on rather dryish soil banks next to footpaths in Western

Cape at Bath Kloof and Cogman’s Kloof, as well as at

Genadendal (Figure 4), where its growth is stunted and

far less luxurious than that of the lectotype specimen, S. W.

Arnell 731 (S), which must have grown in shady, damper

conditions, close to water. In the specimen S.W. Arnell

724 (S), both forms are represented: the proximal leaves

are large, lax and translucent, whereas the distal leaves

are smaller, firm and green, often partly stained with red.

Arnell’s (1952, 1963) measurements of the spores, 30-32

pm and 30-34 pm respectively, are rather less than mine.

Unfortunately my own collections of F. montaguensis

from Bath Kloof, S.M. Perold 3453 and 3454 p.p., are

sterile.

ACKNOWLEDGEMENTS

I wish to sincerely thank my colleagues at NBI, par-

ticularly Ms M. Koekemoer and Mrs C. Bredenkamp for

their kind assistance with fieldwork; also the curators of

BOL, BR, G and S as well as Dr Liibenau-Nestle for the

loan of specimens. My thanks to Ms G. Condy for the

drawings, Mrs A. Romanowsky for developing and print-

ing many photographs and Mrs J. Veldman for her part

in typing the manuscript.

38

Bothalia27,l (1997)

REFERENCES

SPECIMENS EXAMINED

ARNELL, S.W. 1952. Hepaticae collected in South and West Africa

(1951). New and little known species. Botaniska Notiser 1952:

307-329.

ARNELL, S.W. 1953. Notes on South African Hepaticae. Extrait de la

Revue Bryologique et Lidienologique 22, fasc. 1, 2.

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

ence Council, Stockholm.

BEST, E.B. 1990. The Bryophyta of Zimbabwe — an annotated checklist.

Kirkia 13:293-318.

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

C.G. 1844—1847. Synopsis hepaticarum. Hamburg, Meissner.

LEHMANN, J.G.C. 1829. Hepaticarum capensium a C.F. Ecklon. Lin-

naea4: 351-31

PEROLD, S.M. in press. Studies in the liverwort genus Fossombronia

(Metzgeriales) from southern Africa. 4. A re-examination of F.

crispa, F. leucoxantha and F. tumida. Bothalia.

PIIPPO, S. 1991. Bryophyte flora of the Huon Peninsula, Papua New

Guinea. XXXIX. Acta Botanica Fennica 143: 1-22.

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

America. Vol. 5. Field Museum of Natural History, Chicago.

SCOTT, G.A.M. & PIKE, D.C. 1987. The Fossombronia foveolata com-

plex. Lituibergia 13: 79-84.

SCOTT, G.A.M. & PIKE, D.C. 1988. Revisionary notes on Fossom-

bronia. The Bryologist 91: 193-201.

SfM, T.R. 1926. The Bryophyta of South Africa. Transactions of the

Royal Society of South Africa 15: 1^75. Cape Town.

VANA, J., P(DCS, T. & DE SLOOVER, J.L. 1979. Hepatiques d’ Afrique

tropicale. Lejeunia 98: 1-23.

Held at PRE, unless otherwise indicated. Bracketed

numbers after collectors’ name and number refer to the

species in the text in alphabetical order, namely: F. capen-

sis (1); F. dens Hamel lata (2) and F. montaguensis (3).

S.W. Amell 265 (2) S; 275, 295 (2) S (lectotype), BOL (isolectotype);

724 (3) S; 731 (3) S (lectotype); 785 (3) PRE, S; 1376 (1) S (holotype),

BOL (isotype); 1470 (1) PRE, S; 1474 (1) BOL; 1477 (1) G; 1528, 1555

(1) BOL; 1678 (1) S; 1694 (1) BOL; 1715, 1716 (1) S; 1756 (1) BOL;

1757 (1) S; 1783 (1); 1851 (1) S. 5.W Amell & Garside 215 (2) BOL,

S; 259 (2) BOL; 260 (2).

Cholnoky 388 (1) S.

Duthie CH 1651 (2).

Ecklon 7691 (2) W.

Garside 6226, 6456, 6489, 6575, 6586 (2) BOL.

Koekemoer 998 (1).

Lubenau-Nestle SA 139/2 (1) pte.herb.

S.M. Perold 629 (2); 919 (1); 2355, 2356 (2); 3343, 3346, 3347, 3349

(2) ; 3453, 3454 pp. (3); 3492, 3494, 3497, 3498 (1); 3534, 3539,

3540-3541 (1).

S. Russell 2530 (1).

Van Zanten et al 7609809 pp. (2).

Bothalia 27,1: 39-42 (1997)

Studies in the liverwort genus Fossombronia (Metzgeriales) from southern

Africa. 3. An amendment to F. spinifoUa

S.M. PEROLD*

Keywords: Fossombronia, F. spinifoUa, Hepaticae, Metzgeriales, southern Africa

ABSTRACT

Fo.ssomhronia <;pinifolia was described by Stephan! (1900) from a specimen collected by Breutel at Genadendal (Gnaden-

thal) during his visit to the Cape, which lasted from November 1853 to April 1854 (Gunn & Codd 1981). This species was not

adequately described and illustrated by Stephani or by subsequent workers. Moreover, Scott & Pike (1987) misapplied the

epithet, F .spinifoUa, to specimens Amell had identified as F. leucoxantha, because they (Scott & Pike 1988) overlooked a

capsule with ripe spores in the type specimen. An attempt is hereby made to describe and illustrate this species in greater detail

in order to prevent more confusion in future.

Fossombronia spinifoUa Steph. in Species hepati-

carum 1; 389 (1900); Sim: 36 (1926); S.W.Amell: 84

(1963). Type: Africa Australis, Gnadenthal, leg. Breutel

s.n. G 22186, ex herb. K. Miiller Halensis (holo.!).

Plants in dense, overlying mats, green to dark green,

leaf margins occasionally yellowed or those of proximal

ones tinged with red to purple, rather small and slender;

shoots mostly simple, 3.0-7.0 mm long, 1.0-1 .2 mm high,

1. 7-2.5 mm wide, generally arising from fleshy, tuberous

apices (Figure 2A) and laterally from sides of older,

mostly leafless stems of various length (Figure 2B), that

taper proximally to thin, narrow bases. Stems bearing

leaves prostrate, these in turn tapering distally, mostly

plano-convex in cross section, in male plants at apex (Fig-

ure IM) ± 170 pm (10 cell rows) high, ± 210 pm wide,

at base (Figure IN) 230-240 pm (10 cell rows) high, ±

250 pm wide; in female plants at apex (Figure 10)

230-240 pm (12 cell rows) high, up to 380 pm wide, at

base (Figure IP) ± 360 pm (14 cell rows) high, ± 500

pm wide. Rhizoids purple, 12.5-22.5 pm wide, absent

from shoot apices. Leaves in male plants (Figure lA-E)

closely spaced, but rarely overlapping, spreading, suc-

cubously inserted on stem, oblong to irregularly shaped,

700-1175 X 500-1500 pm, sometimes narrower below,

upper margin mostly markedly dentate, with up to 10

teeth, these occasionally expanded into 1 or 2 triangular

lobes; in female plants, leaves (Figure IF-J) spreading to

suberect, overlapping (Figure 2E), frequently ‘ruched’ above,

very obliquely succubously inserted, leaving stem exposed,

shape irregular, as long as wide, 875-1000 x 800-1075 pm,

but sometimes shorter than wide, 1050-1200 x 1300-1625

pm, upper margin (Figure IK) dentate but often less pro-

nouncedly so than in male leaves, more often with Piangular

lobes. Leaf cells not appreciably different in male and female

plants, thin-walled, at margins subquadrate to rectangular

across, 20.0-35.0 x 32.5-57.5 pm, in male plants teeth

112.5-270.0 pm long, composed of an apical slime papilla,

± 12.5 X 12.5 pm, followed below by 1-3 single, long-rec-

tangular cells arranged end to end, 27.5-50.0 x 17.5^0.0

pm, gradually broadening to base, usually with 2 or 3 (oc-

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

MS. received: 1997-01-10.

casionally more) cells alongside each other, margins in

female plants with up to 12 well-spaced papillae, at lower

lateral margins cells long-rectangular, 42.5-50.0 x

17.5- 25.0 pm, upper laminal cells 5- or 6-sided, 30.0-37.5

X 32.5-40.0 pm, middle laminal cells 47.5-52.5 x

32.5- 37.5 pm, basal cells 50.0-62.5 x 40.0-50.0 pm. Oil

bodies absent from most cells; chloroplasts clumped to-

gether along cell walls, ± 5 pm in diameter (Figure IF).

Dioicous. Antheridia dorsal on stem (Figure 2C, D),

in a row, short-stalked, globose, ± 280 pm in diameter,

each shielded by a perigonial bract (Figure IQ-T),

300-^50 X 200-400 pm, margins with 1 or 2 (3) teeth

and 2-5 papillae, cells in interior 5- or 6-sided, 52.5-75.0

X 27.5-37.5 pm. Archegonia in a well-spaced row along

stem (Figure 2F), ± 280 pm long. Pseiidoperianth (Figure

lU, V) turbinate, near stem apex, mostly projecting some-

what above leaves, from narrow base flaring widely

above, ± 1625 pm long, 1500 pm wide across mouth,

margin with several very long teeth, 7-12 cells or up to

385 pm long, consisting of an apical papilla followed be-

low by 4 single, long-rectangular cells, arranged end-to-

end and then basally by 3-5 cells in pairs, the rest of the

cells comparable in shape and size to those of leaves.

Capsule globose, diameter could not be measured as it

was no longer intact, wall bistratose, cells in inner layer

(Figure IW) irregularly shaped, 35.0-42.5 x 20.0-30.0

pm, each cell wall with 1 or 2 yellow or brown nodular

and sometimes semi-annular thickenings. Seta 1.4-3. 5

mm long, ± 100 pm in diameter. Spores yellow-brown,

hemispherical, 36.0-45.0 pm in diameter, including spines

projecting around periphery; distal face convex, with 6 or

7 short irregular ridges ± 5 pm apart and sometimes

branched (Figure 3B, C), sides of ridges and surface be-

tween them with fine cross striations and granules (Figure

3D), papillae occasionally interspersed in between (Figure

3A); proximal face (Figure 3E) lacking triradiate mark,

flat to slightly concave, covered with numerous irregular

papillae, around spore periphery up to 25 projecting,

‘spines’, some conical, others truncate. Elaters (Figure 3F)

light brown, 105.0-150.0 x 7.5 pm, tapering to looped

tips, 3-5 pm wide, some completely bispiral, others partly

bi- and trispiral.

40

Bothalia27,l (1997)

FIGURE I . — Fossomhronia spimfolia. A-E, male leaves; F-J, female leaves; K, detail of leaf margin of female leaf; L, median leaf cells with remains

of oil bodies (solid lines) and chloroplasts (dotted lines); M, cross section of male stem apex; N, cross section of male stem base; O, cross

section of female stem apex; P, cross section of female stem base; Q-T, bracts; U, opened pseudoperianth; V, pseudoperianth from side;

cells in capsule wall. A, F-l, U-W, Breutel s.n. G 22I86\ B-E, J-T, Breutel s.n. (W). Scale bars: A-J, U, V, 500 pm; M-T, 250 pm; K.

W, 50 pm.

r ^

Bothalia27,l (1997)

41

FIGURE 2. — Fossombronia spinifolia. A, male branch which arose from apex of older stem; B, two branches arising from side of older, mostly

leafless stem; C, D, close view of antheridia and bracts; E, F, larger female branches with suberect, overlapping leaves and archegonia along

stem. A-F, Brenre/im. (W). A, X 13.5; B,x 15.4; C, D, x 24.3; E, x 18.2; F,x 14.5.

Fossombronia spinifolia is known only from Genaden-

dal, in Western Cape (Figure 4), where it was collected

by Breutel, probably in November or early December

1853, at the start of his visit to the Cape, which included

various halts along the way, ending at Hankey in the East-

ern Cape (Gunn & Codd 1981). The holotype specimen,

Breutel s.n. G 22186, consists of a few fragments, which

were originally housed in Muller’s herbarium and upon

his death in 1899, were acquired by Stephani. The packet

in G contains Stephani ’s well-preserved preparations of

several leaves, leaved branches and pseudoperianths,

which I measured and photographed. The leones No.

003053 (Stephani 1985) illustrate, however, only a single

dentate leaf, a spore with numerous spines around the cir-

cumference and an elater. One of the fragments of the

type specimen contains a single antheridium, 2 fragments

have immature capsules and another consists of a branch

with a pseudoperianth containing a broken capsule with

FIGURE 3. — Fossombronia spinifolia. Spores. A, B, distal face; C, side view of distal face; D, detail of lamellae at margin of distal face; E, proximal

face; F, elaters. A-F, Breutel s.n. G 22186. A, x 870; B, x 987; C, x 940; D, x 2434; E, x 856; F, x 866.

42

Bothalia27,l (1997)

mature spores. The spores were overlooked by Scott &

Pike (1988) and they (Scott & Pike 1987) had previously

misapplied the epithet F. spinifolia to specimens Amell

had identified as F. leucoxantha (Perold 1997 in press).

Sim (1926) reports F. spinifolia from Jonker’s Hoek, Stel-

lenbosch, Genadendal and other southwestern localities.

The Jonker’s Hoek specimen has not been seen, but the

two Stellenbosch specimens, H.A. Wager 34 (CH1650)

and Duthie 3 (CHI 651), are held at PRE. They have, how-

ever, been misidentified. Another Breutel s.n. specimen

from Genadendal is on loan to PRE from W. It had been

determined as F pusilla, but there is no doubt that it is

F spinifolia, although unfortunately it lacks pseudoperi-

anths and spores. The specimen is mixed with a fragment

of another Fossombronia species; it has pseudoperianths

with a few low spines at the margin and spores unlike

those of F spinifolia. The material of F. spinifolia from

W comprises numerous specimens, but the leaf cells are

rather less well preserved than those in Stephani’s prepa-

rations. There are several male plants with mature

antheridia and a few female plants with archegonia only.

Preparations from this specimen were also measured and

photographed by EM as well as SEM.

In an attempt to collect fresh material of F spinifolia,

I paid a brief visit to Genadendal in October 1995, but

failed to find any. The streambanks, which are the most

likely site where the species would be found (the exact

locality was not stated on the label), are much disturbed

and there has also been considerable infestation by alien

plants.

SPECIMENS EXAMINED

Breutel s.n. G22186 (holo.!) ex herb. K. Miiller Halen-

sis; W s.n.

ACKNOWLEDGEMENTS

I wish to sincerely thank Mrs Diana Harvey, honorary

horticulturalist at Genadendal, for her hospitality and as-

sistance with fieldwork; also Mrs C. Bredenkamp, assis-

tant curator, National Botanical Institute, for all her help

and support. The curators of G and W are warmly thanked

for the loan of specimens and I extend my sincere thanks

to Ms G. Condy for the drawings, Mrs A. Romanowski

for developing and printing the photographs and to Ms D.

Maree for typing the manuscript.

REFERENCES

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

ence Council, Stockholm.

GUNN, M. & CODD, L.E. 1981. Botanical exploration of southern

Africa. Balkema, Cape Town.

PEROLD, S.M. 1997 in press. Studies in the liverwort genus Fossom-

bronia (Metzgeriales) from southern Africa. 4. A re-examination

of F. crispa, F. leucoxantha and F. tumida. Bothalia.

SCOTT, G.A.M. & PIKE, D.C. 1987. Studies on Fossombronia in Aus-

tralia 3. Taxonomic and nomenclatural problems. Journal of the

Hattori Botanical Laboratory 63: 99-105.

SCOTT, G.A.M. & PIKE, D.C. 1988. Revisionary notes on Fossom-

bronia. The Bryologist 91 : 193-201 .

SIM, TR. 1926. The Bryophyta of South Africa. Transactions of the

Royal Society of South Africa 15: 1-475. Cape Town.

STEPHANI, F. 1900. Species hepaticarum Vol. 1. Geneva.

STEPHANI, F. 1985. kones hepaticarum, microfiche. Inter Documenta-

tion Company bv, Leiden.

Bothalia 27,1; 43-56 (1997)

Notes on African plants

VARIOUS AUTHORS

FABACEAE

A SURVEY OF ANTIPODALS IN THE GAMETOPHYTE OF THE TRIBES PODALYRIEAE AND LIPARIEAE

Antipodal cells in the female gametophyte of the

Eabaceae are variable (Prakash 1987). In the Mimosoideae

and Caesalpinioideae, antipodals tend to persist at least

until fertilization, whereas they are mostly ephemeral in

the Papilionoideae (George et al. 1979; Prakash 1987).

The variation found in papilionoid antipodals up to 1990,

is summarized by Cameron & Prakash (1990, 1994).

In a detailed study of antipodal behaviour in the Aus-

tralian Bossiaeeae and Mirbelieae, Cameron & Prakash

(1990, 1994) established that antipodals are of consider-

able taxonomic value at the tribal level. They found giant

antipodals in the gametophytes of some genera of the

tribes, and ephemeral or no antipodals in the remaining

genera. These results evidently supported tribal rearrange-

ments proposed earlier by Crisp & Weston (1987).

Relationships amongst the genera of the tribes Poda-

lyrieae and Liparieae have recently been investigated

(Schutte 1995; Van Wyk & Schutte 1995; Schutte & Van

Wyk 1997a). The study showed that the two tribes are

monophyletic, but that the genus Hypocalyptus is incon-

gruous. In view of these results and suggestions by Crisp

& Weston (1987), that the Australian tribes may be closely

related to the Podalyrieae and Liparieae, it was decided

to examine the antipodals of the two tribes as an additional

character. The results are presented and discussed here.

MATERIALS AND METHODS

At least one species of each genus of the Podalyrieae

and Liparieae sensu Polhill (1976, 1981a, b) was included

in the study. Voucher specimens of the material examined

are listed in Table 1.

Buds (just prior to anthesis) were fixed in an ethyl

alcohol-water-glycerol (70:29:1, v/v) mixture. Ovules

were dissected from the buds and serially dehydrated and

embedded in glycol methacrylate (GMA) according to the

method of Eeder & O’Brien (1968). These were sectioned

with an ultramicrotome, stained in Toluidine Blue and

mounted in Eukitt.

RESULTS AND DISCUSSION

From the results summarized in Table 1 , it is clear that

antipodal cells are present in both tribes. The antipodals

are prominent and persistent at least until anthesis in Li-

paria, Xiphotheca, Amphithalea, Coelidium and the gen-

era of the Podalyrieae (Figure 1). Cyclopia in particular,

has large and deeply stained nuclei in the cells (Figure

1C). Hypocalyptus, however, has inconspicuous and

ephemeral antipodal cells (Figure IF), which degenerate be-

fore anthesis (several ovules had to be sectioned before the

antipodals could be traced).

Unlike the Bossiaeeae and Mirbelieae, antipodals are

neither gigantic nor totally absent in the Podalyrieae and

Liparieae. In the latter two tribes they are less than 0.25

times the length of the gametophyte cavity (Figure 1),

compared with the Australian genera, where they are

more than 0.5 times the length of the gametophyte cavity

(see figures in Cameron & Prakash 1990, 1994). These

giant antipodals persist until well after fertilization. A di-

rect link between the Australian and South African tribes

therefore seems unlikely.

The presence of persistent antipodals in the female game-

tophyte of the Podalyrieae and Liparieae (excluding Hypo-

TABLE 1. — ^List of species examined and the characteristics of the antipodal cells. Voucher specimens are housed in JRAU

44

Bothalia 27,1 (1997)

FIGURE 1 . — Female gametophytes of some species of Podalyrieae and Liparieae, showing the antipodals (indicated with arrows). A, Stirtonanthus

taylorianus', B, Virgilia oroboides subsp. oroboides; C, Cyclopia sessiliflora\ D, Amphithalea tomentosa\ E, Coelidium vlokii\ F, Hypocalyptus

sophoroides. Scale bar; 50 pm.

calyptus ) is a synapomorphy for the two tribes. In the tribe

Crotalarieae, which is the sister group of the Podalyrieae and

Liparieae (Schutte & Van Wyk 1997a), antipodals are

ephemeral fNarang 1978; Schutte unpubl.). The proposed

amalgamation of the two tribes is therefore clearly sup-

ported by the antipodal characteristics. Other characters,

such as the strongly reduced bracteoles and the accumu-

lation of esters of anthocyanins in the pink, purple or

orange-flowered species, also support this notion (Schutte

& Van Wyk 1997a).

Hypocalyptus not only deviates from the other genera

in its antipodal characteristics, but also in at least eight

other significant morphological, cytological and chemical

characters (Schutte & Van Wyk 1997b). This undoubtedly

indicates that the genus does not fit in the Podalyrieae

sensii lato. The tribal position of Hypocalyptus within the

Papilionoideae will be re-assessed and discussed elsewhere

(Schutte & Van Wyk 1997b).

ACKNOWLEDGEMENTS

I am very grateful to Dr PM. Tilney (Department of Bot-

any, RAU) for assisting with the preparation of the micro-

scope slides and to Prof. B.-E. van Wyk (Department of

Botany, RAU) for commenting on an earlier draft of the

manuscript. Financial support from the Rand Afrikaans Uni-

versity is acknowledged.

REFERENCES

CAMERON, B.G. & PRAKASH, N. 1990. Occurrence of giant antipo-

dals in the female gametophyte of Australian Bossiaeeae, In-

Bothalia27,l (1997)

45

digofereae and Mirbelieae (Leguminosae). Australian Journal of

Botany 38: 395-401.

CAMERON, B.G. & PRAKASH, N. 1994. Variations of the megagame-

tophyte in the Papilionoideae. In l.K. Ferguson & S. Tucker,

Advances in legume systematics 6: 97-115. Royal Botanic Gar-

dens, Kew.

CRISP, M.D. & WESTON, PH. 1987. Cladistics and legume systematics,

with an analysis of the Bossiaeeae, Brongniartieae and Mirbelieae

(Papilionoideae, Leguminosae). In C.H. Stirton, Advances in leg-

ume systematics 3: 65-130. Royal Botanic Gardens, Kew.

FEDER, N. & O’BRIEN, TP. 1968. Plant microtechnique: some princi-

ples and new methods. American Journal of Botany 55: 123-142.

GEORGE, G.P., GEORGE, R.A. & HERR, J.M. 1979. A comparative study

of ovule and megagametophyte development in field-grown and

greenhouse -grown plants of Glycine max and Phaseolus aureus

(Papilionaceae). American Journal of Botany 66: 1033-1043.

NARANG, A.K. 1978. Contribution to the embryology of Crotalaria

species. Journal of the Indian Botanical Society 57: 322-331.

POLHILL, R.M. 1976. Genisteae (Adans.) Benth. and related tribes

(Leguminosae). Botanical Systematics 1: 143-368.

POLHILL, R.M. 1981a. Tribe 27. Podalyrieae Benth. In R.M. Polhill &

P.H. Raven, Advances in legume systematics 1: 396, 397. Royal

Botanic Gardens, Kew.

POLHILL, R.M, 1981b. Tribe 28. Liparieae (Benth.) Hutch. In R.M.

Polhill 8l P.H. Raven, Advances in legume systematics 1: 398.

Royal Botanic Gardens, Kew.

PRAKASH, N. 1987. Embryology of the Leguminosae. In C.H. Stirton,

Advances in legume systematics 3: 241-278. Royal Botanic Gar-

dens, Kew.

SCHUTTE, A.L. 1995. A taxonomic study of the tribes Podalyrieae and

Liparieae (Fabaceae). Ph.D. thesis, Rand Afrikaans University,

Johannesburg.

SCHUTTE, A.L. & VAN WYK, B.-E. 1997a. Evolutionary relationships

in the Podalyrieae and Liparieae based on morphological, cyto-

logical and chemical evidence. Plant Systematics & Evolution. In

press.

SCHUTTE, A.L. & VAN WYK, B.-E. 1997b. The tribal position of

Hypocalyptus Thunb. (Fabaceae). Novon (submitted).

VAN WYK, B.-E. & SCHUTTE, A.L. 1995. Phylogenetic relationships

in the Podalyrieae, Liparieae and Crotalarieae. In M.D. Crisp &

J.J. Doyle, Advances in legume systematics 7: 283-308. Royal

Botanic Gardens, Kew.

A.L. SCHUTTE*

* Department of Botany, Rand Afrikaans University, P.O. Box 524, 2006

Auckland Park, Johannesburg. Present address: Compton Herbarium,

National Botanical Institute, Kirstenbosch, Private Bag X7, 7735 Clare-

mont, Cape Town.

MS. received: 1996-08-08.

THYMELAEACEAE

NEW COMBINATIONS IN LACHNAEA

The genus Cryptadenia Meisn. comprising five spe-

cies, was established by Meisner in 1840 and based on

Drege’s collections (Meisner 1840). In De Candolle’s

Prodromus the same five species were recognized by

Meisner (1857). Wright (1915) in his treatment of the

genus recognized four of these species, reduced one to

synonymy and described a new species. Beyers & Van

der Walt (1995) concluded that Cryptadenia and Lach-

tiaea L. are congeneric and that Cryptadenia should be

included within Lachnaea. In accepting these findings,

the necessary nomen- clatural changes are made to the

five species which are currently recognized (Van Wyk

1993) following Wright’s treatment of the genus.

Lachnaea filicaulis (Meisn.) Beyers comb. nov.

Cryptadenia filicaulis Meisn.: 407 (1840); Meisn.: 574 (1857); C.H.Wright:

17 (1915).

Lachnaea grandiflora (Lf.) Baill: 109, t. 77 (1880).

Passerina grandiflora L.f.: 226 (1782). Cryptadenia grandiflora (L.f.)

Meisn.: 405 (1840); Meisn.: 573 (1857); C.H.Wright: 16 (1915).

Cryptadenia breviflora Meisn.: 406 (1840); Meisn.: 573 (1857);

C.H.Wright: 17 (1915). Type: Ecklon 360 (?holo, K!; NBG!, iso.).

Lachnaea laxa (C.H.Wright) Beyers comb. nov.

Cryptadenia laxa C.H.Wright: 17 (1915).

Lachnaea uniflora (L.) Beyers comb. nov.

Passerina uniflora L.: 560 (1753). Cryptadenia uniflora (L.) Meisn.:

406 (1840); Meisn.: 573 (1857); C.H.Wright: 16 (1915).

REFERENCES

BAILLON, H. 1 880. The natural history of plants 6. Reeve, London.

BEYERS, J.B.P & VAN DER WALT, J.J.A. 1995. The generic delimi-

tation of Lachnaea and Cryptadenia (Thymelaeaceae). Bothalia

25: 65-85.

LINNAEUS, C. 1753. Species plantarum, edn 1. Salvius, Stockholm.

LINNAEUS, C. (fil). 1782. Supplementum plantarum. Braunschweig.

MEISNER, C.F. 1840, Synopsis Thymelaearum, Polygonearum et Bego-

niarum africae australis, imprimus a cl. J.J. Drege lectarum. Lin-

naea 14: 385-516.

MEISNER, C.F. 1857. Thymelaeaceae. In A.P. de Candolle, Prodromus

systematis naturalis regni vegetabilis 14: 573-580. Victoris Masson,

Paris.

VAN WYK, C.M. 1993. Thymelaeaceae. In T.H. Arnold & B.C. de Wet,

Plants of southern Africa: names and distribution. Memoirs of the

Botanical Survey of South Africa No. 62: 515-519.

WRIGHT, C.H. 1915. Thymelaeaceae. Elora capensis 5,2 part 1: 1-80.

Reeve, London.

J.B.P. BEYERS*

* Compton Herbarium, National Botanical Institute, Private Bag X7,

Claremont 7735, Cape Town.

MS. received: 1996-08-28.

RUBIACEAE

A NEW SPECIES OF VANGUERIA FROM THE SOUTPANSBERG

Vangueria soutpansbergensis N. Hahn sp. nov. Vf TYPE. — Northern Province, Soutpansberg, 2230

parvifoliae Sond. [= Tapiphyllum parvifolium (Sond.) (Messina), Farm Studholme, 22° 56’ 52.4” South and

Robyns ex Good] affinis sed foliiis glabratis. Figure 2. 30° 01’18.8” East (Cape Mapping Datum), (-CC), 1 440

46

Bothalia 27,1 (1997)

m, 28-11-1995, (in flower), N. Hahn 1112 (PRU, holo.; K,

PRE, iso.).

A deciduous shrub or small tree up to 2.5 m high,

growing in mixed woodlands in soils derived from Sout-

pansberg Group quartzites. Bark dark brown to grey-

brown. Branches glabrous. Leaves opposite or fascicled,

if fascicled usually on dwarf lateral branches; lamina el-

liptic to almost circular, (13.6-)16.2-25.2(-26.7) x

(9-)12.9-18.9(-22.8) mm; base obtuse to rounded; apex

obtuse to rounded; glabrous above and below, seldom very

sparsely hairy when young, dark green above, paler be-

low; petiole short, (0.5-)l.l-2.4(-2.7) mm long, glabrous

to rarely sparsely hairy; margins entire; lateral veins 3-5,

opposite to alternate near the leaf base, otherwise alternate.

Inflorescence: dense 2-15-flowered fascicles or peduncu-

late cymes. Peduncle and pedicel glabrous or very

sparsely hairy, pedicel (1.5-)2.1-2.9(-3.7) mm long.

Flowers 5-merous, greenish to lime-green. Calyx lobes

(1.2-)1.7-1.9(-2.6) X (1-)1.3-1.4(-1.9) nun, glabrous to

sparsely hairy. Corolla glabrous to sparsely hairy on out-

side, with a distinct ring of reflexed hairs in throat; tube

(2.3-)2.5-2.7(-3.2) mm long, (2.1-) 2.5-2. 8(-3.4) mm

in diameter at mouth; lobes , -elliptic-oblong (2.9-)3.3-3.5

(-4.1) mm long, occasionally mucronate abaxially at

apex. Stamens inserted in corolla mouth. Anthers ex-

serted, (0.9-)1.3-1.4(-1.6) mm long. Style (3.1-)3.6-3.9

(^.6) mm long, glabrous, conversely curved so as to

touch throat of tube between two anthers. Disc

(1.8-)2.2-2.3(-2.8) mm in diameter, depressed or tumid.

Hypanthium ( 1 . 1-) 1 .6-1 .7(-2. 1 ) mm long. Fruit a

glabrous, subglobose drupe, length (15. 3-)17. 9-24.1

FIGURE 2. — Vangueria soutpansbergensis. A, flowering branch; B, opened corolla, showing hair fringes and position of anthers; C, flower with

tube removed, showing calyx, disc and conversely curved style; D, fruit seen from the base, showing the remains of the calyx ring; E, side

view of fruit; F, pyrene; G, cross section through pyrene, showing shape and position of embryo. Scale bars: A, B, ± 1 mm; C, ± 2 mm; D-G,

± 10 mm.

Bothalia27,l (1997)

47

(-27.5) mm, width (14.7-)17.5-25.9(-29.4) mm, breadth

(13.^)16.3-23.7(-27.2) mm; pedicel (0.5-)0.8-1.9 (-2.3)

mm long, with 1-4 pyrenes, edible, tasting similar to those

of V. infausta; seeds bean-shaped. Flowering period: No-

vember to December. Fruiting period: March to April.

Specimens examined

NORTHERN PROVINCE. — 2229 (Waterpoort): Muswiru, Schlesingers

sawmill, (-DC), G. Gerstner 5912 (K n.v., PRE); Farm Surprise, (-DC),

N. Hahn 454 (Herb. Sout.); Farm Uniondale, (-DC), N. Hahn 329 (PRU,

Herb. Sout.); Farm Clydesdale, (-DD), N. Hahn 613 (Herb. Sout ); Farm

Rushton, (-DD), N. Hahn 650 (Herb. Sout,); Farm Zwarthoek, (-DD), N.

Hahn 109 (PRU, Herb. Sout.). 2230 (Messina): Piesanghoek, (-AA), G.

Gerstner 5736 (PRE); Farm Studholme, (-AA), N. Hahn 1112 (K, PRE,

PRU, Herb. Sout ); N. Hahn 1164 (K, PRU, Heib. Sout.).

Habitat

The geographic distribution of this species correlates

with other endemic taxa of the Soutpansberg flora. A phy-

togeographical survey of the endemic flora of the Sout-

pansberg (Figure 3) being undertaken by the author has

shown that the endemic species of the Soutpansberg can

be divided into two broad groups according to their habi-

tat preferences, namely: 1, species occurring in a rela-

tively restricted area and displaying little variation in their

habitat preference; 2, species distributed over most of the

mountain range showing a relatively large habitat toler-

ance. Vangueria soutpansbergensis falls within the latter

group, occurring in a variety of habitats ranging from

mountain mistbelt to Androstachys woodland.

Generic disposition

Having studied all members of the genera Pachy-

stigma, Vangueria, Lagynias and Tapiphyllum in the con-

text of the Soutpansberg, I conclude that the flowering

structures and fruiting structures of Tapiphyllum parvifo-

lium are identical to those of Vangueria. Bridson (1996)

expressed doubt as to the generic dispensation of Tapi-

phyllum parvifolium as it was atypical of the genus in

many respects: ‘small leaves, few-flowered inflorescence,

larger glabrescent fruit and occurring outside the main dis-

tribution area of the rest of Tapiphyllum’. Tapiphyllum

parvifolium and V. .soutpansbergensis are without doubt

closely related. This supports the argument that they

should be placed under the genus Vangueria.

Key to genera of the tribe Vanguerieae in the

Soutpansberg region

la Calyx lobes short, linear or triangular, shorter than the corolla

tube Vangueria

lb Calyx lobes long and leafy, as long as or longer than the

corolla tube:

2a Calyx lobes spathulate; fruit narrowing towards apex . . Lagynias

2b Calyx lobes linear; fruit not narrowing towards apex . . Pachystigma

Specific disposition

The tribe Vanguerieae is notorious for its taxonomic

complexities. The genera are poorly defined, and at spe-

cific level, characters available for the separation of taxa

are very few and at best can be seen as very artificial.

The morphological differences between Vangueria sout-

pansbergensis and V. parvifolia are very slight and con-

cern mainly the degree of hairiness of various organs

(Table 2). V. soutpansbergensis is nevertheless recognized

at species level for the following reasons: a) other simi-

larly closely related species pairs are widely recognized

in Rubiaceae, for example, Vangueria infausta subsp. in-

fausta and V. cyanescens and Canthium mundianum and

Canthium gilfillanir, b) even though V. soutpansbergensis

and V. parvifolia are sympatric in some plaees no inter-

mediate forms have been found; c) V. soutpansbergensis

is endemic to the Soutpansberg, a region with a high oc-

currence of endemic plants and animals.

Vangueria soutpansbergensis is by no means a rare

plant, usually occurring within mixed woodlands, on

rocky slopes where it may be common. The species has

so far only been found growing on soils derived from

quartzite, an attribute shared with most endemic plant spe-

cies of this region.

ACKNOWLEDGEMENTS

I hereby would like to thank the National Botanical

Institute Pretoria for allowing me access to their library and

herbarium. I am grateful to Prof. Braam van Wyk (Cura-

tor, Schweickerdt Herbarium, University of Pretoria) for his

TABLE 2, — Summary of characters separating the two species of

Vangueria

48

Bothalia 27,1 (1997)

assistance and for allowing me access to that herbarium. I

would also like to thank Ms Diane Bridson for scmtinizing

the original manuscript and for her invaluable comments

given. Lastly I would like to thank Karen Marais for her

beautiful water colour plate accompanying this article.

REFERENCES

BRIDSON, D. 1996. The tropical African genus Ancylanthos (Rubiaceae-

Vanguerieae). Kew Bulletin 51: 343-352 .

COATES PALGRAVE, K. C. 1992. Trees of southern Africa. 2nd revised

edn, 6th impr. Struik Publishers, Cape Town.

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

1 Dicotyledons. Department of Agricultural Technical Services,

Pretoria.

PHILLIPS, E. P. 1951.ThegeneraofSouth African flowering plants, 2nd

edn. Memoir of the Botanical Survey of South Africa No. 25.

N. HAHN*

* P. O. Box 1734, 0920 Louis Trichardt, South Africa.

MS. received: 1996-10-17.

ASTERACEAE

NEW COMBINATION IN DICOMA

In the course of extensive studies of herbarium material

of various genera of the tribes Mutisieae and Inuleae (As-

teraceae) from sub-Saharan Africa, we have noted new

localities and other data for species of Dicoma Cass. (Mu-

tisieae).

Dicoma membranacea S. Moore

Although D. membranacea S. Moore has been consid-

ered by various authors (Moore 1904; Wilson 1923;

Merxmiiller 1967) to be closely related to D. sessiliflora

Harv., nobody to date has questioned the species status of

this taxon; this may be because the currently accepted

distribution of D. membranacea (northwest Namibia and

southern Angola) (Figure 4) does not overlap either with

that of D. sessiliflora subsp. sessiliflora — Malawi, Tanza-

nia, Mozambique and parts of Zaire (Pope 1992) — or with

that of the recently described D. sessiliflora subsp. steno-

phylla Pope in West Africa (Pope 1991). However, we

have examined the material from Mozambique cited be-

low, and currently referred to subsp. sessiliflora, which is

morphologically indistinguishable from D. membranacea

from Angola and Namibia.

We found D. sessiliflora and D. membranacea to differ

only in length of stem (less than 150 mm tall in the latter).

Moore (1904) considered the presence of pedunculate ca-

pitula to be diagnostic for D. membranacea, but we have

examined specimens of this taxon in which the capitula

are sessile [Voucher: Angola, Rui Correia 2589 (LUAI)]

or subsessile [Vouchers: Angola, Borges 123 (LUAI);

Mozambique, Gomes e Sousa 2157 (COI)], and further-

more the capitula of D. sessiliflora are not always sessile.

Moore (1904) considered corolla size to discriminate be-

tween the two taxa, but the size cited by this author for

D. membranacea is the same as that given by Pope (1992)

for D. sessiliflora. Similarly, Moore (1904) stated that the

corolla lobes are the same length as the tube in D. mem-

branacea-, this is not the case in a number of specimens

examined by us in which the lobes are longer than the

tube [including those of Rui Correia 2589, Borges 123

and Giess 8969 (K) from Namibia]. In our opinion the

putatively distinguishing characters included in Pope’s

(1991) key to the section Pterocoma are, with the excep-

tion of length of stem (maximum 150 mm in D. mem-

branacea), likewise of limited value. This author cites

stem hairiness as a distinguishing character and states that

only the stems of D. sessiliflora can be glabrescent. We

have not been able to identify clear differences between

the two taxa in this respect, and have found specimens

of D. membranacea with glabrescent stems [Voucher: An-

gola, Borges 123 (LUAI)]. We did not find significant

differences with regard to length of leaves (more than 120

mm long in D. sessiliflora versus up to 100 mm long in

D. membranacea)-, indeed Moore (1904) described D.

membranacea as having leaves up to 140 mm long.

MOZAMBIQUE. — 1235: Inhambane, Massinga-Vilanculos, Govuru

River, 7-1938, Gomes e Sousa 2157 (COI, K, LISC). 2335: Niassa,

Administrative Post of Mujoco, 30-9-1948, Pedro & Pedrdgdo 5449

(LMA).

FIGURE 4. — Distribution of Dicoma sessiliflora subsp. sessiliflora var.

membranacea. Known distribution based on Moore (1904),

Merxmiiller (1967) and herbarium material: dotted area. New

localities: triangles.

Bothalia27,l (1997)

49

Since the morphological differences between these two

taxa are minimal, and since the range of D. membranacea

is not geographically continuous, we consider that this

taxon should be viewed as a variety of D. sessiliflora.

Dicoma sessiliflora Harv. subsp. sessiliflora var.

membranacea (S.Moore) S. Ortiz & Rodr.Oubina, comb,

et stat. nov.

D. membranacea S.Moore in Bulletin de I’Herbier Boissier, ser. II, 4:

1025 (1904).

ACKNOWLEDGEMENTS

Many thanks to the curators of herbaria from which

material has been loaned for the present study and to Guy

Norman for the English translation.

REFERENCES

MERXMULLER, H. 1967. Asteraceae. In H. Merxmiiller, Prodromus

einer Flora Sudwestafrikas 139: 1-185. Cramer, Lehre.

MOORE, S. 1904. Beitriige zur Kenntnis der Afrikanischen Flora. Com-

positae. Bulletin de I’Herbier Boissier, ser. 77, 4: 1011-1025.

POPE, G.V 1991. Notes on Dicoma Cass. (Compositae). Kew Bulletin

46: 699-709.

POPE, G.V. 1992. Flora zambesiaca, Vol. 6, Part I. Flora Zambesiaca

Managing Committee, London.

WILSON, F.C. 1923. Revision of the genus Dicoma. Bulletin of Miscel-

laneous Information, Kew 1923: 377-388.

S. ORTIZ*, J. RODRIGUEZ-OUBINA* and I. PULGAR*

*Laboratorio de Botanica, Facultade de Farmacia, Universidade de San-

tiago, 15706 Santiago de Compostela, Galicia, Spain.

MS. received: 1996-02-01.

VITACEAE

A NEW SPECIES OF RHOICISSUS FROM THE EASTERN CAPE

Rhoicissus kougabergensis Relief & Van Jaarsv.,

sp. nov., R. microphyllae (Turcz.) Gilg & M.Brandt et R.

laetantis Retief affinis, sed distributione et lamina folii

ambitu anguste obovata, non ovata vel elliptica (ut in R.

microphylla) nec elliptico-obovata (ut in R. laetanti) dif-

fert.

TYPE. — Eastern Cape, 3324 (Steytlerville): Kouga

Dam, NW of chalets, (-DA), Van Jaarsveld 13796 (PRE,

holo.; E, G, K, MO, NBG, iso.).

Spreading shrub with tendency to scramble, covered

with equally two-armed, unicellular hairs. Roots thick and

fleshy. Tendrils absent. Branches with leaves aligned on

one side; old bark greyish white, rough. Leaves simple,

petiolate; blade narrowly obovate, (22-)30-55(-65) x

(6-)10-22(-27) mm, entire, apex obtuse or slightly emar-

ginate, thick in texture, discolorous with lower surface

reddish brown, upper surface glaucous green or both sur-

faces more or less of the same colour, lower surface more

densely hairy; base of blade cuneate or asymmetric; young

leaves covered throughout with reddish brown hairs; older

leaves pale white or transparent on upper surface; petiole

2-4(-7) mm long; stipules present, soon deciduous. Inflo-

rescences leaf-opposed, ± condensed, reddish brown,

bracteate, thyrsoid cymes. Flowers regular, bisexual,

pedicellate, globose in bud; pedicels 0.5-1. 0 mm long.

Calyx 5-lobed, cup-shaped, ± 1 mm high; lobes broadly

ovate. Corolla: petals 5, ovate, 1. 5-2.0 mm long, greenish

yellow. Stamens 5, opposite petals, bending over gynoecium;

filaments 1 mm long; anthers dorsifixed. Disc entire with

ovary immersed in it. Style simple, entire; stigma inconspicu-

ous. Fruit a 1 -seeded berry, globose, 8-10 mm in diameter,

stalk ±1.5 mm long (Eigures 5 & 6).

The globose flower bud and the thick, entire disc of

Rhoicissus kougabergensis indicate that it belongs to the

genus Rhoicissus (Retief 1993). It is the fourth southern

African member of the genus with simple leaves. All other

known species have 3-or 5-foliolate leaves. R. kougaber-

gensis differs from R. microphylla in the outline of its

leaves which are narrowly obovate (Eigure 7A). Table 3

summarizes differences between the simple-leaved spe-

cies known from southern Africa. The other three species

with simple leaves can be distinguished as follows; 1, R.

microphylla has ovate to elliptic leaves (Figure 7B) and

equally two-armed, reddish brown, unicellular hairs (Fig-

ure 8), somewhat more slender than those of R. kou-

EIGURE 5. — Holotype of Rhoicissus kougabergensis. Van Jaarsveld

13796 (PRE).

50

Bothalia 27,1 (1997)

FIGURE 6. — Rhoicissus kougabergensis Relief & Van Jaarsv.: A, flower bud; B, upper surface of leaf; C, undersurface of leaf; D, stamen; E, stamen

bending over gynoecium; F, pollen grain, SEM micrographs from Van Jaarsveld 13796. A, x 169; B, x 53; C, x 53; D, x 40; E, x 1 8; F, x 1 540.

gabergensis, giving a rusty brown appearance to the spe-

cies. R. micwphylla occurs in the Queenstown-Cathcart

area, whereas R. kougabergensis is found only in the vi-

cinity of the Kouga Dam (Figure 9); 2, R. laetans, which

has elliptic, glabrous leaves, is endemic to the Blydepoort

Nature Reserve area in Mpumalanga (Figure 9); 3, R.

tomentosa (Lam.)Wild & R.B.Drumm. has shallowly

lobed, broadly transversely elliptic or reniform leaves and

is a high-climbing liane in contrast to R. kougabergensis,

a spreading shrub.

The pollen grains of R. kougabergensis are typically

isopolar and radially symmetrical. In polar view the grains

are triangular to circular; the mesocolpia are convex and

the equatorial view is elliptic. The colpi are long, narrow

and granular with pointed ends; the tectum is reticulate

with densely spaced lumina (Figure 6F).

Rhoicissus kougabergensis is endemic to the Kouga

Dam area (only locality known) within the Kouga-

Baviaanskloof Wilderness Area in the southern Eastern

Cape where it occurs on steep slopes (Figure 9). The vege-

tation in the region consists of subtropical thicket, domi-

nated by the spekboom, Portulacaria afra Jacq., with fyn-

bos on the upper slopes. The climate is hot in summer

and mild in winter. The rainfall varies between 300 and

400 mm annually and occurs in summer and winter, but

the winters tend to be drier. The terrain is rugged and the

acid, quartzitic, sandstone soils are poor in minerals. Ac-

cording to Archer & Van Wyk (1993), R. digitata (L.f.)

Gilg & M.Brandt, R. revoilii Planch, and R. tridentata

(L.f.) Wild & R.B.Drumm. subsp. tridentata occur in the

Kouga-Baviaanskloof Wilderness Area. R. kougabergensis

with simple leaves is easily distinguished from these spe-

cies which all have 3- or 5-foliolate leaves. R. kougaber-

gensis starts flowering ± in October.

The Kouga-Baviaanskloof Wilderness Area does not

appear to contain an exceptionally rich flora, nor is it very

rich in endemic species. The area is still undercollected

and future floristic surveys will undoubtedly add many

new records and even new taxa (Archer & Van Wyk

1993). Some endemic species occurring in the vicinity of

R. kougabergensis include two tree species, Sterculia al-

exandri Harv. and Atalaya capensis R.A.Dyer. Smaller en-

demic species include Aloe pictifolia D.S. Hardy, Gasteria

TABLE 3. — Southern African species of Rhoicissus with simple leaves

Bothalia27,l (1997)

51

FIGURE 7. — Characteristic leaf blade outlines

of: A, Rhoicissus kougabergensis', B, R.

microphylla, x 1 .

FIGURE 9. — Known distribution of Rhoicissus kougabergensis Retief

& Van Jaarsv., •; R. microphylla (Turcz.) Gilg & Brandt, A; R.

laetans Retief, ♦.

FIGURE 8. — SEM micrographs of upper surface of leaf with two-armed

hairs; A, Rhoicissus kougabergensis, x 234; B, R. microphylla, x 198.

glomerata Van Jaarsv. and G. ellaphieae Van Jaarsv., Cyr-

tanthus flammosus Snijman & Van Jaarsv. and C. labiatus

R. A. Dyer.

EASTERN CAPE. — 3324 (Steytlerville); Kouga Dam, sheer slope

above wall, (-DA), Van Jaarsveld 9902 (NBG, PRE); Kouga Dam, NW

of chalets, (DA), Van Jaarsveld 13796 (E, G, K, MO, NBG).

ACKNOWLEDGEMENTS

The authors wish to thank Mr Rob Welsh for his help

in searching for fruits at the Kouga Dam.

REFERENCES

ARCHER, R. & VAN WYK, B. 1993. Checklist of the vascular plants of

the Kouga-Baviaanskloof Wilderness Area. Plantlife 9: 25-31.

RETIEF, E. 1993. A new species of Rhoicissus from the Transvaal.

Bothalia 23:231-237.

E. RETIEF* and E.J. VAN JAARSVELD**

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

** National Botanical Institute, Private Bag X7, Claremont 7735.

MS. received: 1996-05-10.

APIACEAE (UMBELLIFERAE)

A NEW NAME FOR A SOUTH AFRICAN PEUCEDANUM

Conservation of the name Peucedanum capense (Thunb.)

Sond. (1862) against P. capense (Eckl. & Zeyh.) D.Dietr.

(1840) was proposed some years ago (Burtt 1989; 1991: 271).

Under the Code then current this might have been a contentious

case, and the Committee for Spermatophyta therefore decided

to postpone its decision until after the Tokyo Congress,

when changes to that part of the Code were expected.

These were, indeed, made and the retention of

Peucedanum capense (Thunb.) Sond. has been approved by

the Committee [votes 12:0 — see Taxon 45: 671 (1996)].

52

Bothalia27,l (1997)

There remains the question of the correct name for

the rejected P. capense (Eckl. & Zeyh.) D.Dietr., which

is currently known as P. midtiradiatum Drude. That

name, however, is illegitimate. Under the Tokyo Code

it would be possible to propose P. midtiradiatum for

conservation, but P. multiradiatum is a much less com-

mon plant than P. capense (Thunb.) Sond. and I am not

alone in thinking that conservation in this case is un-

warranted. A change of name will cause only a tiny

ripple in South African plant nomenclature and a pro-

posal for conservation means further work for the Com-

mittee, further delay in a final decision, and no certainty

that conservation would eventually be granted. A simple

name change now will close the issue, and the follow-

ing new name is therefore established:

Peucedanum polyactinum B.L.Burtt, nom. nov.

Type. — Cape, Stellenbosch, Klapmuts, Ecklon & Zeyher

2239 (S).

Oreoselimim capense Eckl. & Zeyh., Enumeratio plantarum africae aus-

tralis extratropicae: 350 (1837). Peucedanum capense (Eckl. & Zeyh.)

D.Dietr.: 967 (1840), non P. capense (Thunb.) Sond. nom. conserv. Bubon

multiradiatum E.Mey. in Drege: 169 (1843) nom. nud. Bubon capense

(Eckl. & Zeyh.) Sond.: 561 (1862). Peucedanum multiradiatum Drude:

237 (1898), nom. illegit.

REFERENCES

BURTT, B.L. 1989. Proposals to conserve 960: Peucedanum capense

(Thunb.) Sond. Taxon 38: 525, 526.

BURTT, B.L. 1991. Umbelliferae of southern Africa: an introduction and

annotated checklist. Edinburgh Journal of Botany 48: 133-282.

DIETRICH, D.N.F. 1840. Synopsis plantarum. 2. Voigt, Weimar.

DREGE, J.F. 1843. Zwei pflanzengeographische Documente nebst einer

Einleitung von E. Meyer. Besondere Beigabe zur Flora 1843

Band 11.

DRUDE, C.G.O. 1898. Umbelliferae. In A.Engler & K.Prantl, Die nattir-

lichen Pflanzenfamilien 3: 237. Wilhelm Engelmann, Leipzig.

ECKLON, C.F. & ZEYHER, K.L.P. 1837. Enumeratio plantarum africae

australis extratropicae. Hamburg.

SONDER, O.W. 1862. Umbelliferae. In W.H. Harvey & O.W.Sonder,

Flora capensis 2: 524-567. Hodges & Smith, Dublin.

B.L. BURTT*

* Royal Botanic Garden, Edinburgh, EH3 5LR, UK.

MS. received: 1996-10-28.

PROTEACEAE

A NEW SPECIES OF LEUCADENDRON FROM THE WESTERN LITTLE KAROO

This large distinctive Leucadendron was unknown in

South African herbaria until specimens collected by Mr

David Osborne of the Cape Dept of Nature Conservation,

Ladismith, were submitted for identification in 1994.

Prior to that it had been observed at several sites on Anys-

berg at the western end of the Little Karoo in the Western

Cape by officers of the Cape Dept of Nature Conserva-

tion although the taxonomic status of these populations

was uncertain.

Subsequent field studies have established that it is fairly

widely dispersed at the western end of the Klein Swartberg

and Little Karoo. It is here described as new and commemo-

rates David Osborne, who made the first recorded herbarium

collections of this species and whose thorough collecting for

the Protea Atlas Project has greatly increased our knowledge

of the Proteaceae in the Little Karoo region.

Leucadendron osbornei Rourke sp. nov.

Frutex erectus robustus ad 4 m altus, Leucadendroni

teretifolio affine, sed statura grandiore, foliis glabris aceroso-

teretibus 15-28 mm longis, inflorescentiis masculinis 20-35

mm longis, et strobilis femineis maturis 30-40 mm longis

differ!.

type.— Western Cape, 3320 (Montagu). Witteport,

extreme western end of Klein Swartberg, (-BD), 9-11-

1995, J.P Rourke 2110 (NBG, male specimen, holo.!,

BOL, K, MO, NSW, PRE, S, iso.!).

Robust rigid shrub 1 .5-4.0 m tall with a stout main tmnk

to 75 mm in diam. Branches stiffly erect, rigid, glabrous,

5-10 mm in diam. Leaves acicular terete, 15-28 x 1. 5-2.0

mm, ascending, hard and cartilaginous, glabrous, slightly

glaucescent, upper surface minutely canaliculate; slightly

shorter in male plants. Male inflorescences densely clus-

tered in groups of 8 to 16 on short (30-60 mm long)

branchlets on flowering shoots. Inflorescence cylindric,

20-35 X 10 mm, pedunculate; peduncle 10 x 2 mm,

sparsely sericeous, covered with tightly adpressed subu-

late bracts 2-3 mm long, glabrous, but margins ciliate.

Floral bracts ovate, 1 x 1 mm, tightly clasping perianth,

glabrous but margins ciliate. Perianth sessile, 5-6 mm

long, straight, glabrous, bright yellow; perianth claws

equally recurved at anthesis; tube cylindrical. Anthers 4;

pollen powdery. Style filiform, 6 mm long, glabrous. Pol-

len presenter clavate-acute, 1 mm long. Hypogynous

scales 2 mm long, projecting to top of tube. Female in-

florescences free-standing, surrounded by a loose pseudo-

whorl of patent involucral leaves, greenish ivory to

yellow at anthesis. Flowering cone ovoid-clavate, obtuse,

30-40 X 12-14 mm, shortly pedunculate; peduncle 10-15

X 10 mm. Involucral bracts dark brown, very narrowly

lanceolate-acuminate to subulate, 8-12 x 1.5 mm, tightly

adpressed to peduncle, glabrous but margins ciliate. Flo-

ral bracts broadly ovate, acute, 3x5 mm, glabrous. Pe-

rianth 3 mm long, laterally compressed; tube region 2

mm long, densely sericeous; claws and limbs recurved,

glabrous. Staminodes generally 3, anterior staminode

usually absent. Style glabrous, 5 mm long, slightly

abaxially deflexed in upper half. Ovary ovoid to spheri-

cal, glabrous, 1 mm long. Pollen presenter minutely

capitellate, inconspicuously bilobed with glandular hairs

on the abaxial face. Hypogynous scales ovate-acute, 1

mm long. Mature female cone ovoid-acute, 35-60 x

25-30 mm, brown, becoming silvery grey with age. Fruit

a flattened black samara, 70 x 50 mm, apically retuse

(Figure 10).

Bothalia27,l (1997)

53

FIGURE 10. — Leucadendron osbor-

nei: A, young female cones

and a mature fruiting cone; B,

branches on an old female

plant showing serotinous

cones retained for up to eight

years; C, David Osborne next

to a mature female plant; D,

densely clustered male inflo-

rescences. B, C, & D taken at

type locality, Witteport, Klein

Swartberg.

Diagnostic characters

Apart from the obvious differences in the size and stat-

ure of the mature shrubs, L. osbornei is distinguished from

L. teretifolium by its longer leaves, 1 5-28 mm long; longer

male inflorescences, 20-35 mm long and by the much

longer (35-60 mm long) mature female cones. In L nobile

the male inflorescences are produced more sparsely in

TABLE 4. — Differences in leaf and inflorescence dimensions between

L. osbornei and related species

smaller numbers. The male inflorescences in L. nobile

(30-70 mm long) are very much larger than in L osbornei,

whereas the male flowers are loosely arranged in a lax spike

and are pubescent basally in the tube region. The male flow-

ers in L teretifolium usually have a bright red spot at the

apex of each bud but in both L osbornei and L. nobile they

are uniformly yellow (Table 4).

Key to species

la Male inflorescences globose, 5-10 mm long; mature female

cones up to 35 mm long teretifolium

lb Male inflorescences cylindric, 20-90 mm long; mature female

cones 30-90 mm long:

2a Male flowers completely glabrous; male inflorescences nu-

merous, 20-35 mm long, clustered together on short

branchlets; Western Cape, western Little Karoo . . . .osbornei

2b Male flowers minutely pubescent in tube region; male in-

florescences solitary, 30-70 mm long, sparsely pro-

duced; Eastern Cape, Kouga, Baviaanskloof &

Willowmore Mountains nobile

54

Bothalia 27,1 (1997)

Affinities

Leucadendron nobile Williams, L. teretifolium (Andr.)

Williams and the species here described as L. osbornei

Rourke form a well-defined group within Leucadendron

subsection Compressa, characterised by their uniformly

acicular glabrous leaves and glabrous cones.

The robust vegetative growth, large stature of the mature

shrub and impressive size of the mature female cones in L

osbornei initially suggested an affinity with L. nobile, a spe-

cies from the Humansdorp, Willowmore and Steytler-ville

Districts of the Eastern Cape. However, in the morphology,

arrangement and number of male inflorescences, L osbornei

is clearly more closely allied to L. teretifolium than to L

nobile. In both L. teretifolium and L. osbornei the male plants

are covered with masses of male inflorescences on short

branchlets borne at the ends of flowering shoots, and in botli

species the male flowers are completely glabrous and borne

in tightly congested cylindric inflorescences. This is unlike

L. nobile where the male inflorescences are relatively

sparsely produced on the ends of long shoots.

At one locality, a site south of trig, beacon 66 on

Matjiesgoedberg, L. osbornei and L. teretifolium occur

sympatrically (2-3 m apart), with no evidence of hybri-

dization. From observations made at this site, it appears

that there is little or no overlap in their flowering pe-

riods. During a site visit on 5-11-1996, L. teretifolium

was observed to be nearly past its flowering period

while L. osbornei was still in bud, about two weeks

from flowering.

Significantly, the male flowers in both species produce

quite different odours which may attract different pollina-

tors. The male flowers in L. teretifolium give off a sweet,

slightly lemon-scented odour with faint sulphurous under-

tones while in L. osbornei they produce a smell reminis-

cent of fresh human semen. Prominent nectar droplets

form at the base of the male flowers in L. osbornei which

attract large numbers of Diptera during the day when the

air temperature reaches about 25°C. I have not observed

potential pollinators on L. teretifolium. Both Williams

(1972) and Rebelo (1995) note that L. teretifolium pro-

duces showers of pollen when shaken, indicating that it

is wind-pollinated.

Distribution and habitat

This species is restricted to mountains at the western

end of the Little Karoo in the southwestern Cape at ele-

vations between 700 and 1 500 m. Populations of L. os-

bornei occur on Elandsberg north of Sevenweeks Poort at

its northeasterly limits; the western end of the Klein

Swartberg at Wittepoort and Paardenfontein; the north-

eastern corner of Touwsberg; on Anysberg; on Matjies-

goedberg north of Anysberg at the western end of its range

and on Rooiberg near the Floriskraal dam. Leucadendron

osbornei occurs mainly in Dry Mountain Fynbos or in the

ecotone between Dry Mountain Fynbos and Karroid Bro-

ken Veld (Acocks 1988) or between Dry Mountain Fynbos

and Central Mountain Rhenosterveld (Moll et al. 1984).

Most of the known localities are on Witteberg Quartzite

and a few are on Table Mountain Sandstone (Figure 1 1 ).

These localities are generally extremely arid, receiving

a mean annual rainfall of between 150 and 200 mm.

Merxmuellera arundinacea and Erica spectabilis are com-

monly associated species.

In several populations many of the shrubs observed

were between 3 and 4 m in height. The females are almost

invariably slightly taller than the males. Vegetative growth

in such habitats is slow, and attempts to count annual

vegetative growth increments indicated that such speci-

mens were in excess of 50 years old. Leucadendron os-

bornei is strongly serotinous, retaining unopened cones

for up to seven and eight years (Figure 12). Flowering

takes place from early to late November depending on

site aspect.

Conservation status

The exact conservation status of this species is not

clear as the known populations have not yet been ade-

quately assessed. Most populations which I have observed

consist of about 100 or fewer mature individuals. It is

probably best described as naturally rare but under no

obvious man-made threat at present. The Matjiesgoedberg

and Anysberg stands are protected within the Anysberg

Nature Reserve administered by the Cape Dept of Nature

Conservation.

Specimens examined

WESTERN CAPE. — 3320 (Montagu): Matjiesgoedberg, above Mat-

Jieskloof on South side below trig, beacon 66, (-BC), Nov., Rourke

2114 (NBG); Anysberg, 3.8 km west of trig, beacon 66 on Farm Mat-

jieskloof, (-BC), March 1994, Osborne 94022302 (NBG); Wittepoort,

western end of Klein Swartberg above Kromkloof, (-BD), March

1995, Rourke 2071 (NBG); Rooiberg, Drielingskloof, westernmost part

of peak on Laingsburg-Prince Albert road, 7-11-1994, Osborne s.n.

(NBG). 3321 (Ladismith): Touwsberg, on Farm Basseur, (-CA), 23-

9-1994, A. September 94092301 (NBG); Elandsberg, on Ylandskloof

211, 0.8 km from spot height 1533, (-AD), 13-11-1995, K. Mars

95111303 (NBG).

ACKNOWLEDGEMENTS

I am most grateful to Mr David Osborne, Cape Dept

of Nature Conservation, Ladismith, who arranged several

BothaIia27,l (1997)

55

FIGURE 12. — Leucadendron osbor-

nei. A, female inflorescences in

flower, X 0.7; B, mature female

cone, X 0.7; C, seed, x 2.3; D,

section through leaf, x 5.4.

E-G, female flower, x 5.4: E, in

bud; F, open; G, gynoecium

with hypogynous scales. H,

shoot with male inflorescences,

X 0.7. I-K, male flower, x 5.4:

I, in bud; J, open; K, showing

pollen presenter and hypo-

gynous scales.

field excursions which enabled me to examine this species

at various localities within its distribution range and to Dr

Ion Williams who read and reviewed the manuscript.

REFERENCES

ACOCKS, J.PH. 1988. Veld types of South Africa, 3rd edn. Memoirs of

the Botanical Survey of South Africa No. 57.

MOLL, E.J., CAMPBELL, B.M., COWLING, R.M., BOSSI, L., JAR-

MAN, M L. & BOUCHER, C. 1984. A description of major

vegetation categories in and adjacent to the fynbos biome. South

African National Scientific Programmes Report No. 83.

REBELO, T. 1995. SASOL proteas. A field guide to the proteas of

southern Africa. Femwood Press.

WILLIAMS, I.J.M. 1972. A revision of the genus Leucadendron (Pro-

teaceae). Contributions from the Bolus Herbarium No. 3.

J.P. ROURKE*

* Compton Herbarium, National Botanical Institute, Kirstenbosch, Pri-

vate Bag X7, Claremont 7735, Cape Town.

MS. received: 1996-12-11.

BORAGINACEAE

THE TAXONOMIC STATUS OF LOBOSTEMON HORRID US

Levyns (1934) in describing L. horridus Levyns, lists

a Levyns 2881 collection. She did not, however, personally

collect this material as it is clear from the particular her-

barium label that J. Rennie was the collector. Levyns

based her description on two herbanum specimens avail-

able to her, namely Compton 2971 (BOL) and J. Rennie

s.n. sub Levyns 2881 (BOL, STE). Subsequent collections

have been few and far between. A total of four collections.

the last being Acocks 23698 (PRE) in 1965, had been

made at the onset of the current revision of the genus.

The aforementioned scarcity of collections resulted in L.

horridus being included in the Red Data List of southern

African plants (Hilton-Taylor 1996). All attempts to re-

collect the taxon at the type locality have failed. The most

recent collection close to the type locality has been M.H.

Buys 523 at Pienaarskloof, about 40 km northwest of Mat-

56

Bothalia27,l (1997)

jiesfontein. Examination of all available herbarium speci-

mens as well as field work led us to suspect that L. hor-

ridus and L. paniculatus are conspecific.

According to the diagram presented by Levyns (1934:

412), L. paniculatus and L. horddus are closely related.

Although Levyns provides no summary of differing char-

acters in her diagnosis of L. horddus, one can catch a

glimpse at what her thoughts might have been through

her choice of the specific epithet. The herbarium speci-

mens at her disposal both exhibit leaves with an extremely

spinous indumentum. The current revision has confirmed

that leaf characters based on the indumentum are unreli-

able for systematic purposes. Levyns (1934) warns against

the undue use of morphological characters in Lobostemon,

yet leaf shape and indumentum type distinguish L. pani-

culatus from L. horridus in her key.

A critical comparison of characters between L. pani-

culatus and L. horridus reveals no significant diagnostic

differences between them. The difference in hairiness of

the leaves, the pronounced midrib and thickened margins

are ascribed to arid habitats. The less spiny forms ( = L.

paniculatus) are generally to be found near water in and

amongst the larger mountain ranges of the Swartberg. The

spiny forms ( = L. horridus) become more abundant as

one proceeds further into the interior of South Africa. The

variation allowed here in the indumentum with the sinking

of L. horridus is no greater than that allowed for in L.

echioides for example.

L paniculatus and L horridus share the following char-

acter states to the exclusion of the other members of the

section Lobostemon: 1, adaxial surface of young leaves ap-

pearing glabrous, becoming hairy with age; 2, possession of

two distinct trichome lengths; 3, two corolla lobes slightly

larger than the rest; 4, hairs on the abaxial side of the corolla

lobes largely confined to the midveins (this differs markedly

from L. echioides where the hairs are confined to the distal

parts of the lobes); 5, identical fruit structure.

It is for the above reasons that L. horridus Levyns is

viewed to be conspecific with L. paniculatus. The no-

menclatural history of L. paniculatus therefore reads as

follows:

Lobostemon paniculatus (Tliunb.) Buck in Linnaea

11: 139 (1837); DC.: 8 (1846); C.H. Wright: 33 (1904);

Levyns: 418 (1934). Type: Cap. b. Spei, Thunberg s.n.

sub UPS 4109 (UPS, holo.!).

Echium paniculatum Thunb.: 33 (1794); Willd.: 784 (1798); Pers.; 163

(1805); Schrad.: 41 (1806); Poir.: 675 (1808); Thunb.; 9 (1811); Lehm.:

425, 473 (1818); Roem. & Schult: 11 (1819); Lehm.: t. 23 (1823);

Thunb.: 165 (1823).

Lobostemon horridus Levyns: 419 (1934). Type: Whitehill near Mat-

jiesfontein, J. Rennie s.n. sub Levyns 2881 (BOL, lecto.!; STE, iso.!).

REFERENCES

BUEK, H.W. 1 837. Echia capensia. Liwzflefl 11: 129-149.

DE CANDOLLE, A.P. 1846. Prodromus systematis naturalis regni vege-

tabilis 10: 8. Treuttel & Wiirtz, London.

GREUTER, W., BARRIE, F. R., BURDET, H. M., CHALONER, W. G.,

DEMOULIN, V., HAWKSWORTH, D. L., JORGENSEN, P. M„

NICOLSON, D. H., SELVA, P. C., TREHANE, P. & MCNEILL,

J. 1994. International Code of Botanical Nomenclature (Tokyo

Code) adopted by the Fifteenth International Botanical Congress,

Yokohama, August-September 1993. Regnum vegetabile 131.

HILTON-TAYLOR, C. 1996. Red Data List of southern African plants.

Strelitzia 4. National Botanical Institute, Pretoria.

LEHMANN, J.G.C. 1818. Plantae e familiae Asperifoliarum nuciferae:

425, 473. Diimmler, Berlin.

LEHMANN, J.G.C. 1823. leones et descriptiones novarum et numis

cognitarum stirpium auctore 1 : t. 23. Perthes & Besser, Hamburg.

LEVYNS, M.R. 1934. A revision of Lobostemon. Journal of the Linnean

Society 49: 393^51.

PERSOON, C.H. 1805. Synopsis plantarum, sen Enchiridium botanicum

complectens enumerationem systematicam specierum hucusque

cognitarum 1: 163. Cramerum, Paris.

POIRET, J.L.M. 1808. Encyclopedic methodique, Botanique 8: 675.

Agasse, Paris.

ROEMER, J.J. & SCHULTES, J.A. 1819. Caroli a Linne equitis systema

vegetubilium: 11. Cottae, Stuttgardt.

SCHRADER, H.A. 1806. N cues Journal f Hr die Botanik 1,3: 41.

THUNBERG, C.P. 1794. Prodromm: 33.

THUNBERG, C.P. m\. Flora capensis edn 1, 1,2: 9.

THUNBERG, C.P. 1823. Flora capensis edn 2; 165.

WILLDENOW, C.L. 179^. Species plantarum 1,2:784.

WRIGHT, C.H. 1904. Flora capensis 4,2: 27-44. Lovell Reeve, London.

M.H. BUYSUnd J.J.A. VAN DER WALT*

* Depailment of Botany, University of Stellenbosch, Private Bag XI,

Matieland, 7602 Stellenbosch.

MS. received: 1996-12-02.

Bothalia 27,1: 57-74(1997)

Composition and biogeography of forest patches on the inland mountains of

the southern Cape

C. J. GELDENHUYS*

Keywords: biogeography, corridors, evergreen forest, geomorphology, rare species, species-area relationship

ABSTRACT

Patterns in species richness of 23 small, isolated forests on the inland mountains of the southern Cape were studied. Species

richness of woody plants and vines of the Kouga-Baviaanskloof Forests was higher than in the western mountain complexes,

where species richness in the more southern Rooiberg and Kamanassie Mountains was higher than in the Swartberg range. The

Rooiberg, a dry mountain with small forests far away from the coastal source area, had more species than, and contained many

species which are absent from, the larger, moister forests of the Kamanassie which are closest to the coastal source areas. Neither

altitude nor distance from the source area, the forests south of the coastal mountains, nor long-distance dispersal, adequately

explained the variation in species richness. The variations are best explained in terms of dispersal corridors along the Gouritz

and Gamtoos River systems which connect the coastal forests with the inland mountains. The distribution patterns of four

species groups in relation to the geomorphological history of the two river systems provide relative dates for the expansion and

contraction of temperate forest, subtropical forest and subtropical transitional thicket in the southern Cape.

CONTENTS

Abstract 57

1 Introduction 57

2 Study area 59

3 Methods 60

4 Results 61

4.1 Species richness 61

4.2 Relationships between species richness and alti-

tude, forest area and distance from source ... 61

4.3 Distribution of taxa on different mountain com-

plexes 62

4.3.1 Widespread species on the inland mountains . 62

4.3.2 Species widespread in coastal forests but with

limited spread in study area 62

4.3.3 Species with limited/disjunct distribution in, or

absent from, the coastal forests 62

4.3.4 Endemic species of the region 62

4.4 Seed dispersal mechanisms 63

5 Discussion 63

5.1 Species richness 63

5.2 Habitat preferences of species 63

5.3 Species-area relationships and long-distance dis-

persal 64

5.4 Dispersal barriers and corridors 66

5.5 Forest migration in relation to climatic change . 66

5.5.1 Temperate forest 66

5.5.2 Subtropical forest 67

5.5.3 Subtropical transitional thicket and karroid

woodland 68

5.6 Endemic species 68

6 Conclusions 68

7 Acknowledgements 69

8 References 69

Appendix 70

* Division of Water, Environment and Forestry Technology, CSIR, R O.

Box 395, Pretoria 0001 .

MS. received: 1996-02-08.

1 INTRODUCTION

In the southern Cape several mountain complexes oc-

cur north (inland) of the coastal mountains. These inland

mountains are surrounded by semi-arid to arid valleys and

lowland. Two river systems, however, connect the inland

areas with the coast. A third river system almost breaks

through the coastal mountains near one of the inland

mountains. Forests on the inland mountains are very small

and isolated and are in sharp contrast to the large and

widespread forests along the coast (Anon. 1987). Their

distribution, composition and conservation status are

poorly known.

Axelrod & Raven (1978) and Deacon (1983) recon-

structed the palaeofloras of Africa. They speculated that

the temperate forest which covered the southern tip of

Africa during the Palaeocene (65 to 55 My BP) was re-

placed by subtropical forest during the Oligocene-Mio-

cene (37 to 5 My BP) and subsequently by fynbos and

arid shrublands during the Late Pleistocene (125 000 to

10 000 yBP). The general increasing aridity in the south-

ern Cape region since the beginning of the Miocene (24

My BP) (Deacon 1983) suggests further that over time

the inland forest patches became increasingly isolated.

Hypothetically the current flora of the isolated inland for-

ests would therefore consist of species which have sur-

vived in suitable refuge sites, species with tolerance

ranges which would enable them to survive in the changed

environment, and species which have colonized from suit-

able forest source areas. Geldenhuys (1986; 1992a, b;

1994) studied several aspects of the distribution, fragmen-

tation and biogeography of the South African forests in

general and the southern Cape coastal forests in particular.

The question was raised: how do the inland forests relate

to the coastal forests in the southern Cape? Several studies

have successfully used geomorphological evidence to re-

construct the biogeography of floras and taxa (Kaul et al.

1988; Moore 1988) and the phylogeny of families and

genera, e.g. of freshwater fishes (Skelton 1986). Smith

(1981) considered the geological and palaeogeographical

58

Bothalia27,l (1997)

FIGURE 1. — The study area of the southern Cape inland mountains in relation to the widespread forests south of the coastal ranges. The numbers represent the following study sites in the respective mountain complexes (the

forests are too small to be indicated on the map): Rooiberg:!, Bosrivier upper; 2, Bosrivier lower; 3, Assegaaibosrivier. Kamanassie: 4, Kluesrivier; 5, Meulrivier upper. Swartberg: 6, WaterkJoof; 7, Seweweekspoort;

8, Swartbergpoort; 9, Swartberg Pass; 10, Rust-en-Vrede; 11, Huisrivier upper; 12, Meiringspoort; 13, Tierkloof; 14, Cheridouwspoort. Kouga: 15, Sapreerivier. Baviaanskloof: 16, Doringkloof; 17, Diepnekkloof; 18,

Geelhoutbos; 19, Assegaaikloof; 20, Witrivierkloof upper; 21, Witrivierkloof lower; 22, Grootrivierpoort. Groendal: 23, Zungarivier.

Bothalia27,l (1997)

59

history of the landmasses around Tasmania as well as

colonizing ability and altitudinal range of species in a

study of the origin of the Tasmanian high mountain flora.

Taxa move from source areas along different pathways

(corridors, filters or sweepstake routes) to colonize new

sites (Brown & Gibson 1983). If they move along a cor-

ridor (which contains a wide variety of suitable sites or

habitats), the composition of the community in the new

site will be very similar to the community at the source.

If they move through a filter (which contains a limited

variety of suitable sites or habitats), the community at the

new site will contain a limited component of the source

community. If they move along a sweepstake route (which

cuts across areas with totally different environments), the

community in the new site will contain only species which

will survive long distance dispersal across unsuitable ar-

eas. Island biogeographic theory was evoked to explain

the extinction and colonization of animal species on oce-

anic islands of different size and distance away from the

source areas (MacArthur & Wilson 1967), and more re-

cently for plants in islands of fynbos in the southern Cape

forests (Bond et al. 1988). Dispersal is only of signifi-

cance if the organism can establish a viable population

upon arrival in the new area (Brown & Gibson 1983).

Physical and biological barriers to successful colonization

by forest plants are insufficient moisture (arid valleys),

extreme temperature (mountain tops, frost in valleys), dis-

turbance patterns (fire in fynbos) and absence of long dis-

tance dispersal agents (migrant frugivorous birds). How-

ever, man-induced changes in the vegetation and environ-

ment during historical times may eliminate or confuse the

evidence required to elucidate the biogeographical patterns.

The objectives of this study were:

• to determine the patterns in species composition and

richness of forest communities in different inland moun-

tain complexes of the southern Cape.

• to explain the observed patterns in terms of habitat

preferences of species; species-area and species-distance

relationships; long-distance dispersal; dispersal corridors

in relation to the geomorphological history of the land-

scapes in the region.

• to aid interpretation of vegetation changes in the

coastal forests from the dating of forest expansion and

contraction on the inland mountains.

2 STUDY AREA

The study covers the Cape Folded Belt in the southern

Cape of South Africa, between Ladismith and Riversdale

in the west and Humansdorp in the east (Figure 1 ). A site

in the Groendal Wilderness Area northwest of Port Eliza-

beth was included as a riverine site in the coastal moun-

tains which appeared to be similar to the sites in the inland

mountains. The dominant physiographic feature of the

main study area are the subparallel mountain ranges and

the intermontane lowland belts which run approximately

east- west (Lenz 1957). In the west the Swartberg range,

the highest range in the study area (up to 2 250 m) forms

the northern boundary. The Kouga-Baviaanskloof Moun-

tain complex forms the northern boundary to the east. It

consists of several parallel ranges with relatively narrow

east-trending valleys. The coastal Langeberg-Outeniqua-

Tsitsikamma ranges form the southern boundary. A

number of smaller, almost isolated mountains occur be-

tween the Swartberg and the Langeberg-Outeniqua ranges:

Rooiberg-Gamka Hill between Calitzdorp and Ladismith,

and Kamanassie Mountain between Oudtshoorn and Un-

iondale. The Oudtshoorn basin is a large semi-arid low-

land between the Swartberg, Rooiberg, Outeniqua and

Kamanassie Mountains which comprises hills of Creta-

ceous conglomerates (Lenz 1957). The relatively narrow

Langkloof valley separates the Tsitsikamma Mountains

from the Kouga Mountains. Groendal is a hilly landscape

between the Great Winterhoek and Elandsberg Mountains.

A relatively dry coastal plain occurs to the southeast of

Groendal.

Three river systems link, or partially link, the coastal

area with the inland mountains and are important for the

interpretation of the forest flora. The Gouritz River

breaches the Langeberg-Outeniqua ranges west of Mossel

Bay through the Gouritzpoort. Inland it is formed by the

confluence south of the Rooiberg of the Olifants River

which drains the Oudtshoorn basin, the Gamka River

which flows through the Swartberg north of Calitzdorp

and which drains the Karoo west of Beaufort West, and

the Groot River which drains the area to the west of the

Rooiberg. Tributaries of these rivers breach the Swartberg

range through several poorts such as, from west to east,

Seweweekspoort, Gamkapoort, Meiringspoort and Cheri-

douwspoort. A poort is a relatively narrow, steep-walled

opening, cutting almost perpendicularly across a topo-

graphic barrier, through which the open areas on either

side are connected, usually by a river (Lenz 1957). The

Keurbooms River which runs between the Kamanassie

Mountain and Plettenberg Bay, does not fully breach the

gap between the Outeniqua and Tsitsikamma ranges south

of the Kamanassie Mountain. A low, narrow ridge sepa-

rates its origin from the Kamanassie River which drains

the southern slopes of the Kamanassie Mountain and runs

in a northwesterly direction to join the Olifants River. The

Gamtoos River east of Humansdorp is formed by the con-

fluence of the Baviaanskloof and Groot Rivers. The

Witrivier is a small stream which joins the Groot River

north of the Grootrivierpoort at Cambria. The Zunga (or

Swartkops) River runs from Groendal towards the coast

at Port Elizabeth.

Geologically the mountains owe their existence to their

heavily folded structure and the resistance of the quartzitic

Table Mountain Sandstones to weathering. Softer sand-

stones and shales of the Bokkeveld Series eroded more

readily to form the syncline valleys (Lenz 1957; Theron

1962; Toerien 1979).

The few weather stations in the area all occur in the

lowlands. Data are available for short periods for a

number of rain gauges across the Swartberg (along the

Pass and in the east) and Kamanassie Mountains. Rainfall

increases linearly with increase in elevation, and rain

shadow effects are apparent on north slopes. In the Swart-

berg Pass area annual rainfall declines rapidly from 725

mm at 1 600 m to 210 mm at 884 m on the northern

foothills and to 182 mm at 686 m in Prince Albert on the

edge of the Great Karoo (Bond 1981). Annual rainfall on

the southern foothills is 411 mm at 640 m (Cango Caves)

60

Bothalia27,l (1997)

and 570 mm at 762 m (Rust-en-Vrede, forest site No. 10,

Figure 1, Table 1) (Weather Bureau 1986 unpubl.). In

Swartberg East, near Blesberg between the Tierkloof and

Cheridouwspoort sites (Nos 13 & 14 respectively), annual

rainfall ranged from 766 mm on the northern midslope,

to 847 mm on the crest, 798 mm on the upper south slope

and 572 mm on the southern foothills (unpublished data).

In the Kamanassie annual rainfall near the crest ranged

between 815 mm on the southern side and 682 mm on

the northern side, and declined to 239 mm on the southern

foothill and 169 mm on the northern foothill (Kamanassie

Policy Memorandum, Department of Nature Conserva-

tion). Rainfall data for Rooiberg are unreliable but appear

to be lower than for the Kamanassie, also as judged from

the appearance of the fynbos vegetation on the southern

slopes. Rainfall in the lowlands ranges between 220 mm

at Van Wyksdorp (305 m) on the southern footslopes of

the Rooiberg, 244 mm at Oudtshoom (335 m), 482 mm

at Joubertina (544 m) in the Langkloof, 321 mm at Studtis

(760 m) in the western end of Baviaanskloof, and 536

mm on the southern foothills of Groendal (229 m; forest

site No. 23) (Weather Bureau 1986 unpubl.).

Reliable information on temperature regimes is less

available. Diurnal and seasonal temperature variation is

large. Mean maximum temperature for the warmest month

is 31.8°C for Oudtshoom and 27.8°C for Uitenhage and

the mean minimum temperature for the coldest month is

3.5°C and 5.8°C for the two towns respectively. The mean

number of days per annum with frost is 7.3 and 1.4 for

the two towns (Weather Bureau 1986). Snow occurs five

or six times per annum on the Swartberg and may lie for

more than two weeks (Bond 1981).

The main vegetation types of the area have been de-

scribed by Acocks (1988), Taylor (1979), Bond (1981)

and Cowling (1984), and in unpublished reports. Moun-

tain fynbos and grassy fynbos cover most of the moun-

tains and are interspersed with small patches of evergreen

forest in protected gullies and valleys. Karroid broken veld

covers the low-lying valleys of the Little Karoo, and sub-

tropical transitional thicket extends south of Groendal and

occurs in parts of Baviaanskloof

3 METHODS

Twenty-three forested gorges, gullies and riverine sites

were visited to represent the variety of sites with forest

species encountered on the inland mountains (Figure 1;

Table 1). In this study the definition of a forest (Gelden-

huys et al. 1988) was extended to accommodate scattered

bush clumps and isolated trees of species which are usu-

ally associated with forest, as observed in Seweweek-

spoort, Meiringspoort and Cheridouwspoort. For each

study site a list was compiled of all plant species which

were associated with the forest communities. Emphasis

was placed on recording all tree and shrub species, but

taxa of other growth forms were also recorded as com-

pletely as possible.

The size of each forest (Table 1) was estimated in the

field. Most forests consisted of a long, narrow stand along

a stream or river. The length of the stream which was

covered in the survey was estimated from 1 : 50 000 topo-

graphic maps and the mean width of the forest was esti-

mated in the field. Altitude was read from the relevant

1 : 50 000 topographic maps.

Obvious or important disturbance factors affecting the

forest communities were recorded. These included fire,

flooding of rivers, landslides and wind.

TABLE 1 . — List of forested sites vjsited in the inland mountains of the southern Cape

Bothalia27,l (1997)

61

TABLE 2. — Number of species by growth forms for the forest sites on the inland mountains of the southern Cape

The relationship between the number of woody species

(plus vines) or the number of herbaceous species in a for-

est and the altitude, forest area and direct distance (km)

to the nearest source area was determined by means of

stepwise multiple regression (STSC 1986). Log transfor-

mations were used for all variables because of the highly

skew nature of the observations, a procedure usually fol-

lowed in such studies (Bond et al. 1988). The southern

Cape forests (Geldenhuys 1993a), marked in black as ‘in-

digenous forest’, and ‘island’ coastal forest west of Port

Elizabeth (Figure 1) were considered to be the nearest

source areas.

The distribution of taxa, mainly woody species, on the

inland mountains was represented by means of tables

which indicate the frequency of species on each particular

mountain range or in similar sites within a range. The

sites were grouped on the basis of assumed dispersal bar-

riers and corridors, as follows: Swartberg sites in three

subgroups, i.e. on the northern slopes, at high altitudes

and on the southern slopes; Rooiberg; Kamanassie;

Kouga-Baviaanskloof; Grootrivierpoort, including the

lower site of the Witrivierkloof; and the Groendal site (see

Tables 5-8). The most likely dispersal mechanisms for the

species are indicated and are based on Coates Palgrave

(1977) and my own observations (Geldenhuys 1993a).

The plant nomenclature follows Arnold & De Wet

(1993) and the plant author names are according to Brum-

mitt & Powell (1992).

4 RESULTS

4.1 Species richness

The species are listed in the Appendix and summarized

by growth forms for the different sites (Table 2). Woody

plants and vines form the bulk of the species.

Species numbers vary greatly between sites. In general

there is a decline in species richness from east to west

and from south to north. The mean number of species per

site in the Kouga-Baviaanskloof complex is double the

number in the Swartberg range. Note however that the

mean number of species, particularly woody species, in

the Rooiberg forests tends to be higher than the number

in the Kamanassie forests.

4.2 Relationship between species richness and altitude,

forest area and distance from source

The number of woody species (plus vines) is signifi-

cantly correlated with altitude and forest area (Table 3),

but not with distance to source area. Altitude alone ex-

62

Bothalia27,l (1997)

TABLE 3. — Analysis of variance for the stepwise multiple regression of the dependence of the (log) number of woody species and vines in a

forest on its altitude and area

Independent variable Coefficient Standard error t-value Significance level

Constant 3.132426 0.221281 14.1559 0.0000

Log Altitude, m -0.635723 0.078427 -8.1059 0.0000

Log Area, ha 0.195371 0.043653 4.4756 0.0002

(adjusted for Df) = 0.8582; Standard error of the estimate = 0.0975 1 .

plains 74% of the variation in number of woody species.

The number of herbaceous species is significantly corre-

lated only with forest size, but this regression model ex-

plains only 18% of the observed variation.

4.3 Distribution oftaxa on different mountain complexes

Only 10% of the species occur in more than 50% of

the sites and these are mostly tree and shrub species (Table

4). For example, 48 of the 118 tree and shrub species

occurred in one or two sites only, i.e. 10% or less of the

sites. The pattern of occurrence of species on the different

mountain ranges became clearer when the pattern was

considered in relation to the distribution range of species

in both the inland and coastal forests.

4.3.1 Widespread species on the inland mountains

Species which occur in more than 50% of the sites

occur in most sites of all the mountain groups (Table 5).

Site preferences of the species are based on their occur-

rence in the coastal forests (Geldenhuys 1993a). The

Kamanassie sites lack several species which have a pref-

erence for drier habitats although these species occur in

the high altitude sites of the Swartberg. Those which do

occur in the Kamanassie are confined to the lower end of

Kluesrivier.

4.3.2 Species widespread in coastal forests but with limited

spread in study area

Very few of the widespread species of the coastal for-

ests which have a limited spread on the inland mountains

do occur in the Swartberg range (Table 6). Those which

do occur there are confined to the southern sites. Note

again that very few of these species occur in the Kamanas-

sie. However, two of the Kamanassie species are absent

from the Rooiberg, i.e. Rapanea melanophloeos and

Diospyros whyteana. Ocotea bullata is confined to the

Kamanassie and one site in the Rooiberg. Many of the

Rooiberg species are absent from the Kamanassie. The

majority of the species occur in Groendal and Grootrivier-

poort and in one or more sites of the Kouga-Baviaanskloof,

and many of these are absent from the Kamanassie or

Rooiberg. Note the decrease in number of species from

Groendal through Grootrivierpoort to the Baviaanskloof.

4.3.3 Species with limited/disjunct distribution in, or absent

from, the coastal forests

These species fall in two groups: those which are con-

fined to the western ranges, i.e. Swartberg, Rooiberg and

Kamanassie; and those which are confined to the eastern

ranges (Table 7). Note again the decrease in number of

species from Groendal through Grootrivierpoort to the

Baviaanskloof.

4.3.4 Endemic species of the region

Nine woody species are endemic to the study area, or

have their main distribution in the southern Cape. Of these

only Virgilia divaricata is widespread in the study area

(Table 8). Laurophyllus capensis, which has a wide dis-

tribution in the coastal areas, is confined to a few sites in

the Groendal area where it grows in association with V.

divaricata. Strelitzia alba, not recorded during this study,

has since been recorded from one locality in the Kouga

Mountains (M. Yates pers. comm. 1991; see Geldenhuys

1992a for its distribution in the coastal forests).

TABLE 4. — Absolute and relative frequencies by which species of different growth forms occur in the 23 forested study sites in the inland moun-

tains of the southern Cape

Bothalia27,l (1997)

63

TABLE 5. — Distribution and fruit/seed dispersal of species widespread in forested sites on the inland mountain ranges of the southern Cape

' Fruit/seed dispersal mechanisms: •, bird/mammal; +, small/large dry seed not dispersed by wind; #, wind dispersal.

^ Mountain range; SB, Swartberg (see ^ below); RB, Rooiberg (sites 1-3); KN, Kamanassie (sites 4, 5); BK, Baviaanskloof (sites 15-20); GR,

Grootrivier (sites 21, 22); GD, Groendal (site 23).

^ N, north (sites 8, 13 & 14); U, upper (sites 9 & 11); S, south (sites 6, 7, 10 & 12).

* Maximum number of sites per range for comparison with number of sites in which a species is present.

4.4 Seed dispersal mechanisms

The majority of woody species (Tables 5-8) have

fleshy fruits or seeds which are dispersed by frugivorous

birds and/or mammals. Among the plants with dry

propagules, only the ferns, the bane Secamone alpini and

the two Brachylaena species are readily dispersed by

wind. Cunonia capensis, Nuxia floribunda, Buddleja

saligna and B. salviifolia produce small seeds in capsules

which may be blown over short distances in strong wind.

Gonioma kamassi, Ptaeroxylon obliquum and some of the

other species have winged seeds which are not well suited

for wind dispersal. Note that the majority of endemic spe-

cies (Table 8) have dry seeds which are not particularly

adapted for dispersal over longer distances.

5 DISCUSSION

5.1 Species richness

The forests on the inland mountains of the southern

Cape contain 118 tree and woody shrub species, 24 of

which are not included in the 140 tree and shrub species

of the coastal forests (Geldenhuys 1993a). The difference

in number of species of the herbaceous growth forms be-

tween the two areas is much larger. It has been assumed

that woody species are more persistent in suitable habitats

because the majority of them can resprout after fire, and

that they create the micro-habitat for forest understorey

plants. In this discussion interpretation of the patterns in

species richness and composition is almost confined to

the woody species.

5.2 Habitat preferences of species

Habitat preferences of species along an altitudinal gra-

dient do not explain the major differences in species com-

position of forests in different parts of the inland

mountains. The decrease in species richness of woody

plants and vines with increasing altitude suggests that

many species cannot grow at high altitude. The majority

of the widespread species of the eoastal forests which have

a limited distribution in the inland sites are indeed con-

fined to the low-lying mountain valleys and riverine sites.

However, these species are also absent from similar sites

on the northern side of the Swartberg range, and many

are absent from similar sites on the southern Swartberg,

the Kamanassie and the Rooiberg (Table 6).

64

BothaIia27,l (1997)

TABLE 6. — Distribution and fruit/seed dispersal of species widespread in the coastal forests of the southern Cape but with a limited spread in the

study area

‘ Fruit/seed dispersal mechanisms: •, bird/mammal; +, small/large dry seed not dispersed by wind.

^ Mountain range: SB, Swartberg (see ^ below); RB, Rooiberg (sites 1-3); KN, Kamanassie (sites 4, 5); BK, Baviaanskloof (sites 15-20); GR,

Grootrivier (sites 21, 22); GD, Groendal (site 23).

^ N, north (sites 8, 13 & 14); U, upper (sites 9 & 11); S, south (sites 6,7, 10 & 12).

Maximum number of sites per range for comptuison with number of sites in which a species is present.

’ Species which occur in only the GR & GD sites.

^ Species which occur only in Groendal.

Where habitat preferences have been attached to some

species (Table 5), their demonstrated wide tolerances do

not explain their absence from certain forests. The wide-

spread species associated with moist coastal sites, occur

on the inland sites both at high and low altitudes and in

sites which are relatively dry. In several sites they are not

confined to the moist, cool sites along the streams, but

grow on steep, exposed slopes and often shallow, rocky

sites far above the streams, for example in the Kamanas-

sie, the upper Bosrivier, Meiringspoort and Sapreerivier

sites. Only widespread inland species which are associated

with both dry and moist coastal sites, grow in the northern

Swartberg sites (Table 5), often where the stream courses

open up to the north into the arid Great Karoo.

5.3 Species-area relationships and long-distance dispersal

Area is a significant variable in the regression models;

this explains the richness of both woody and herbaceous

species but in both cases accounts for a relatively small

Bothalia27,l (1997)

65

TABLE 7. — Distribution and fruit/seed dispersal of species which have a limited or disjunct distribution in, or which are absent from the coastal

forests of the southern Cape

' Fruit/seed dispersal mechanisms: •, bird/mammal; +, small/large dry seed not dispersed by wind.

^ Mountain range; SB, Swartberg (see below); RB, Rooiberg (sites 1-3); KN, Kamanassie (sites 4, 5); BK, Baviaanskloof (sites 15-20); GR,

Grootrivier (sites 21, 22); GD, Groendal (site 23).

^ N, north (sites 8, 13 & 14); U, upper (sites 9 & 1 1); S, south (sites 6, 7, 10 & 12).

Maximum number of sites per range for comparison with number of sites in which a species is present.

66

Bothalia27,l (1997)

portion of the variation. Direct distance from the source

areas ranges from 30 to 80 km but is an insignificant

variable in the regression models. For example, the

Kamanassie Forests are some of the largest in the study

area, are the closest to the coastal forests, particularly the

large forests north of Knysna, and represent the moistest

sites; yet they contain fewer species than the Rooiberg

Forests and lack several species which are present in the

Rooiberg. The Rooiberg is a drier mountain, has small

forests and its closest source areas are scattered forests

between Mossel Bay and George to the southeast and

some Langeberg forests west of Riversdale.

The distribution patterns of very few species support

claims of long-distance dispersal despite the fact that the

majority of tree and shrub species have fleshy fruits or

seeds, many of which I have observed being eaten by

birds in the coastal forests.

• Rapanea melanophloeos and Ficus burtt-davyi occur

in one and two sites respectively in the Swartberg range

which could be attributed to long-distance dispersal.

Ocotea bullata may have been dispersed into the Rooiberg

and Kamanassie from relatively nearby forests with O.

bullata which exist on the northern slopes of the Lange-

berg (Garcia Pass near Riversdale) and Outeniqua Moun-

tains (Robinson Pass near Mossel Bay and near Noll west

of Uniondale).

• Most of the species which are present in the

Rooiberg but absent from the Kamanassie are generally

readily dispersed by birds. Their absence from the

Kamanassie and the limited distribution in the study area

of other similarly readily dispersed species of the coastal

forests such as Apodytes dimidiata, the two Podocarpus

species, the two Olea capensis subspecies, Psydrax

obovata and Canthium muiidianum casts doubt on the

relevance of long-distance dispersal in the study area.

• Several species which release minute dry seeds from

dry capsules, but which are not effectively dispersed by wind,

show distribution patterns which are similar to the patterns

of frugivorous species. Examples are Cunonia capensis,

Nuxia floribunda, Buddleja saligna and B. salviifolia. Their

dispersal by birds or mammals is very unlikely.

• The absence of suitable dispersal vectors may be the

reason for the insignificance of long-distance dispersal.

The sites were visited during different seasons over sev-

eral years but only one Rameron pigeon (Columba arqua-

trix) was seen in the Swartberg Pass Forest. No studies

exist to indicate the migration patterns of frugivorous birds

in the southern Cape. It has been suggested that the Ra-

meron pigeon migrates up and down the coast (Phillips

1927). That Rameron pigeons would fly from the coast

where food sources are more readily available to the in-

land mountain forests where their food sources are more

limited and possibly irregularly available seems highly un-

likely. Small flocks of Red-winged starlings {Onycho-

gnathus morio) were often seen in the vicinity of the

inland forests, but rarely seen along the coast.

5.4 Dispersal barriers and corridors

The parallel mountain ranges are obvious barriers to the

dispersal of forest taxa from the coast to the inland forests.

The mountain ridges exp>erience strong, cold winds and ex-

treme temperatures. Ice often forms in the sheltered gullies

near the top (pers. obs.). Frequent controlled and natural fires

in the fynbos on the mountain slopes prevent the estab-

lishment of forest species in the exposed sites and confine

or eliminate existing forests. This was observed in most

of the sites or was evident in the small sizes of trees near

the forest edge in more protected sites.

Dry lowlands and valleys of the Little Karoo minimize

the number of species which are able to cross them by

means of establishment in small bush clumps in a step-

ping-stone fashion.

The obvious dispersal corridors are the Zunga River,

the Gamtoos River through the Grootrivierpoort and Ba-

viaanskloof, the Keurbooms River and the Gouritzpoort.

The first two river systems are effective corridors for

stepping-stone dispersal. The sites along the Zunga River

the end constitute of the subtropical transitional thicket

and riverine forests which are connected with the Alex-

andria and other coastal forests. Many streams run into

the Baviaanskloof River from the mountain ridges to its

north and south and provide refuge sites for forest spe-

cies. The Baviaanskloof shares many of the species

which^ occur in the Grootrivierpoort and Witrivierkloof

and at Groendal. The remaining two rivers are not effec-

tive corridors. The Keurbooms River does not breach the

relatively low Outeniqua-Tsitsikamma mountain ridge to

provide direct and easy access for dispersal of forest spe-

cies from the large Knysna forests to the Kamanassie

Mountains. The Gouritzpoort contains no sheltered sites

for forest establishment and north of the poort is an arid

lowland. It is not connected with any nearby forests to

the south of the poort. It may have been an effective

dispersal corridor in earlier, moister periods, but not un-

der the present climate.

5.5 Forest migration in relation to climatic change

Following on from the earlier discussion, it is therefore

suggested that the variation in species richness on the in-

land mountains is mainly the result of different degrees

of intermingling during the contraction and expansion of

the different floras due to climatic and landscape changes

since the Palaeocene. Only tree and woody shrub species

have been considered for this interpretation because it has

been assumed they are the key elements which create the

specific micro-habitats for herbaceous elements of the un-

derstorey of particular vegetation units. Certain under-

storey species should therefore correlate with the

distribution pattern of particular groups of tree species.

Factors such as the altitudinal gradient, forest size, site

preferences and dispersal corridors and mechanisms are

merely contributing to this variation within a particular

mountain range. In this study at least four floras can be

recognized from the distribution patterns of the species as

listed in Tables 5 to 8, namely: temperate or austral forest

relicts; subtropical forest; subtropical transitional thicket

and karroid riverine woodland.

5.5.1 Temperate forest

It is suggested that most of the widespread inland spe-

cies characteristic of moist sites (Table 5) represent relicts

of the temperate austral forests (such as Cunonia capensis)

or high-altitude forests of tropical latitudes (such as Ilex

mitis, Halleria lucida and Kiggelaria africana) which cov-

Bothalia27,l (1997)

67

ered the southern tip of Africa during the Palaeocene,

rather than recent dispersal events. These temperate forests

were eliminated with changes toward warmer and more

humid climates associated with the northward drift of the

African continent (Axelrod & Raven 1978; Deacon 1983).

These constituent species grow in forests on all the moun-

tain complexes in the study area, and are the only species

which grow in sheltered sites on the northern slopes of the

Swartberg range. They are also the main species of the for-

ests in the cool, sheltered kloofs and gorges of the inland

mountain ranges in the southwestern Cape (pers. obs.). Many

of them occur in the isolated Affomontane forests of southern

and eastern Africa (Killick 1963; Chapman & White 1970;

Dowsett-Lemaire 1988). The poorts through the Swartberg

range had breached the ranges by the early Tertiary (Lenz

1957) and would have allowed dispersal of other readily

dispersed species through the poorts from south to north, if

they were present by that time. Even if it is argued that in

more recent times the fleshy-lhiited tree species may have

been dispersed to the northern side of the Swartberg by birds,

the argument does not account for the presence of Cunonia

capensis with its small, diy seeds. The wide habitat toler-

ances of these species enabled them to survive and to occur

widespread in the study area, and in southern Africa (Geld-

enhuys 1992b).

Species which could be added to the list of temperate

forest species are Pterocelastrus rostratus, Diospyros

glabra and Brachylaena neriifolia. Their eastern distribu-

tion limits in the study area (Table 7) coincide with the

longitude of their eastern limits in the coastal forests

(Geldenhuys 1992a). Their distribution patterns suggest a

more continuous distribution at some early period which

was later fragmented into their present pattern.

5.5.2 Subtropical forest

It is suggested that the widespread coastal forest spe-

cies with limited spread in the inland sites (Table 6) rep-

resent elements of the subtropical forest which replaced

the temperate forests since the Oligocene-Miocene (Axel-

rod & Raven 1978; Deacon 1983). Most of these species

occur also in the coastal forests of the southern and south-

western Cape (Geldenhuys 1993a; McKenzie 1978).

These forests have expanded from the east. The easterly

orientation of the Zunga and Gamtoos Rivers and the

Kouga-Baviaanskloof valleys suggests that they would

have been more readily colonized by the expanding sub-

tropical forests. The Gouritz River breached the Lange-

berg-Outeniqua range during the late Cretaceous as a

subsequent poort, i.e. it developed along a relatively

weaker part of the range by a headward eroding stream

(Lenz 1957). With widespread forest along the coast south

of the Outeniqua-Langeberg Mountains and with a more

humid climate (Hendey 1983) the Gouritz River could

have been a suitable dispersal corridor for some species

to enter the southern sides of the Rooiberg. The Rooiberg-

Gamka mountain range forms a loose connection to the

southeast with the Outeniqua range and may have had

sheltered sites on the southwestern side. Today several

species of the western fynbos element, e.g. Mimetes cu-

cullatus, that are characteristic of the wetter coastal

ranges, also occur on the Rooiberg-Gamka range and the

adjacent Outeniqua range (Taylor 1979; J.H.J. Vlok pers.

comm. 1988). They support the dispersal route suggested

for the forest species, although the fynbos migration re-

lates to Late Pleistocene times and different environmental

conditions which would not support forest.

The nature of the deposits in the Oudtshoom Basin

suggests that the climate during the Late Cretaceous was

similar to the present semi-arid climate and the Olifants

River portion of the breach between Rooiberg and Gamka

Hill occurred as late as Plio-Pleistocene (Lenz 1957). This

semi-arid climate and the late breach would have pre-

vented the spread of subtropical forest towards the Swart-

berg. The absence of a direct corridor between the

Keurbooms River and the Kamanassie accounts for the

absence from the Kamanassie of several Knysna forest

species, which are present in the Rooiberg (Table 6).

Some of the widespread as well as disjunctly distributed

coastal species are confined to the Groendal, Grootrivier-

poort and lower Witrivierkloof sites (Tables 6 & 7; Ap-

pendix). Two possible explanations for this pattern are: 1,

most of these species have relatively large fruits or seeds

which require specialized dispersal vectors. The seeds of

some species such as Podocarpus latifolius and Caloden-

drum capense loose viability fast when they dry out, and

those of other species such as Olea capensis subsp. macro-

carpa and Cassine papillosa have long germination peri-

ods due to woody seed coats (Geldenhuys 1975, 1996)

and are then liable to predation by rodents (pers. obs.).

However, others with similar seed types occur further into

the Kouga-Baviaanskloof complex such as Ekebergia cap-

ensis and Podocarpus falcatus (both are dispersed by bats;

see Geldenhuys 1993b for P falcatus); 2, a more likely

explanation is that the subtropical forests expanded in dif-

ferent waves and that each wave contained a different set

of species. The expansion and contraction could be related

to successive periods of high and low sea levels respec-

tively, which in turn were associated with humid and arid

periods respectively (Hendey 1983). During later periods

of forest expansion along the coast, some areas, particu-

larly the more arid inland areas, may not have been suit-

able for the colonization by forest species.

For example, Podocarpus falcatus occurs as far west

as Swellendam, and P latifolius as far west as the Cape

Peninsula (Von Breitenbach 1986). Both grow in small

forest patches near the southern exit of the Gouritz River

through the Langeberg-Outeniqua ranges. Both are readily

dispersed in the southern Cape coastal forests (Geldenhuys

1980, 1993a, b) but have a limited entry in the Baviaanskloof

and are absent from the Rooiberg and Kamanassie. It is

suggested that they represent a relatively late southwestern

expansion of the subtropical forests when barriers of semi-

arid lowlands inland of the coastal mountains prevented

their spread inland. This implies that the two Podocarpus

species have a tropical origin and are not part of the aus-

tral flora as has often been suggested (e.g. Levy ns 1964).

Their large fruit size indicates a tropical affinity (Givnish

1980), in contrast to the small fruit size of austral podo-

carps of Australia, New Zealand and Chili (pers. obs.). It

is suggested that the fossil podocarp pollens from some

southwestern Cape sites (e.g. Coetzee 1986) may repre-

sent austral podocarps which became extinct with the re-

gression of the early temperate forests and before the

present species arrived in the area.

Some species represent the spread of subtropical tran-

sitional thicket (Cowling 1984; Everard 1987). Species

68

Bothalia27,l (1997)

such as Ptaeroxylon obliquum, Diospyros scabrida and

Pavetta lanceolata have not yet reached the southern Cape

coastal forests. Others have reached the southern Cape but

were cut off by the Late Pleistocene-Holocene marine

transgression, e.g. Hippobromus pauciflorus, Schotia lad-

folia and Plumbago auriculata (Geldenhuys 1992a).

Some coastal dune forest species, such as Strychnos

decussata and Eugenia zeyheri, require specialized sites

which prevented their spread inland. For example, S.

decussata grows on a terrace along the Zunga River in a

site similar to that in which the species grows in Nature’s

Valley along the southern Cape coast (Geldenhuys 1986).

Strelitzia alba also falls in this category.

Some species, such as Trichocladus ellipticus, repre-

sent relicts of a retreating forest flora (Geldenhuys 1992a).

5.5.3 Subtropical transitional thicket and karroid wood-

land

The widespread inland species of drier sites (Table 5)

are generally associated with subtropical transitional

thicket (Cowling 1984; Everard 1987). They have prob-

ably become mixed with the more tolerant moist forest

elements with the increasing aridity since the beginning

of the Miocene-Pliocene (Deacon 1983). They occur in

few of the sites at higher altitudes. They are more promi-

nent in the bush clumps and subtropical transitional

thicket of the more arid lowlands and riverine sites in the

drier, open valleys of the Baviaanskloof and Karoo. They

occur in the drier parts of moist sites as they are found

where streams from the mountains open up into the dry

valleys, or on the drier slopes above the streams, or along

open valleys or gorges such as Meiringspoort. Seeds of

several of the species were found along the krantzes above

the study sites and it is assumed that Red-winged starlings

(O. mono) dispersed the seeds from the lowlands or other

nearby sites. Many of these species are, however, absent

from the Kamanassie sites although they are present in

the lowlands further away from the mountain. The higher

(Eo-Oligocene) and lower (Mio-Pliocene) surfaces on the

southern side of the Kamanassie (Lenz 1957) could ac-

count for this absence. At Boomplaas Cave, between the

Swartberg Pass and Rust-en-Vrede sites, charcoal assem-

blages indicate that A. karroo only became dominant in

the late Holocene during more mesic conditions (Scholtz

1986). This species was absent from charcoal layers older

than 12 000 years although it was apparently a much pre-

ferred firewood. It became a major component of wood-

land in the valley near the cave after 5 000 yBP, after an

initial spread into the valley during the early Holocene

(Deacon et al. 1983).

5.6 Endemic species

The endemic species represent two major groups; forest

margin species such as Virgilia divaricata (Phillips 1926),

Laurophyllus capensis (Phillips 1931; Geldenhuys 1993a)

and Widdringtonia schwarzii (Liickholf 1963); and species

of drier sites such as Lachnostylis bilocularis and

Loxostylis alata (Palmer & Pitman 1972). The sites in

which the forest margin species mature suggest that they

can only persist with less frequent fires than under which

fynbos persists. This has been shown for L. capensis (Vlok

& De Ronde 1989). Trees of V. divaricata (Phillips 1926)

and Widdringtonia schwarzii (Liickhoff 1963) are killed by

fires and depend on reseeding for regeneration. All except

Smelophyllum capense have dry seeds which do not appear

to be readily dispersed. It is assumed that these species

have evolved in this region. It is suspected that they formed

part of specific vegetation units but became separated and

isolated to a lesser or greater degree as a result of their

poor dispersability and sensitivity to frequent fires.

The distribution of Virgilia (Van Wyk 1986) corre-

sponds with the distribution pattern of the subtropical for-

est species which expanded during the Oligocene-

Miocene. It is suggested that V. divaricata was the parent

species from which the other species evolved because of

its presence in both the inland and coastal forests. Its ab-

sence from Rooiberg but presence in Kamanassie suggests

that the species became established at a relatively late

stage of expansion of subtropical forest. Its crossing of

the gap north of the Keurbooms River towards the

Kamanassie can perhaps be explained by dispersal of the

resistant seeds by primates, particularly the baboon {Papio

ursinus) and Vervet monkey (Cercopithecus aethiops).

Both these primates have been seen in stands of Acacia

karroo and the alien wattle A. mearnsii which have seeds

very similar to V. divaricata. Seeds of V. divaricata have

been found in the faeces of the baboon in the coastal

mountains. Van Wyk (1986) mentioned two forms of the

species: a form of drier localities such as in Seweweek-

spoort, Baviaanskloof and Groendal; and the form of the

coastal forests between Humansdorp and George. In the

context of the spread of the species I consider the form

of drier localities as the first stage and the form of the

coastal forest as the second stage.

Lachnostylis bilocularis was recorded in the Rooiberg

and Meiringspoort but it also occurs in various localities

between Ladismith and Uniondale (Palmer & Pitman

1972). In Meiringspoort the tree grows in the southern

part of the gorge up to a particularly narrow part of the

gorge but does not occur north of it. Its distribution sug-

gests that it was more widespread before. The related L.

hirta occurs in the coastal forests over a somewhat wider

range (Palmer & Pitman 1972). It is suggested that L

hirta spanned the Gouritz River valley during less arid

periods of the Oligocene-Miocene with a wider distribu-

tion in the dry, forested coastal areas. L. bilocularis

evolved inland as an adaptation to dry, open sites and

eventually became limited to the present localities when

the lowlands became even drier. Both species have dry

seeds in dry capsules and are poorly dispersed. Calpurnia

villosa has a similar but more restricted distribution than

L. bilocularis, which also centres on the Gouritz River

and its tributaries in the Oudtshoorn basin (Palmer & Pit-

man 1972). The record of C. villosa from Grootrivierpoort

could be based on my misidentification in the field of C.

aurea, a species which has a disjunct distribution along

the coast (Geldenhuys 1992a).

6 CONCLUSIONS

The systematic survey of forests in the inland mountain

ranges in relation to the geomorphological evolution of dis-

persal corridors which link them with the coastal forests has

provided a means to postulate relative dates for the expansion

and contraction of floristic elements of both the inland and

coastal forests. However, certain assumptions made in this

study require to be verified, such as the following;

Bothalia27,l (1997)

69

• Dispersal distances by wind of the small, dry seeds

of Cunonia capensis and Nuxia floribunda.

• Flight patterns of ffugivorous birds between the coastal

and inland forests, and the feeding behaviour of these birds.

• The phylogenetic relationships of species of genera

such as Lachnostylis and Virgilia, and of provenances of

several other species such as Ilex mitis and the Podocar-

pus species.

Disturbance regimes in the study area have changed

over time. The current man-induced disturbances of the

vegetation exert extreme pressures on the forests which

persisted in refuge sites in marginal environments. The

forest patches should be treated as rare ‘species’ to allow

the natural processes of population migration, settlement

and adaptation to continue. The following examples of

changed management could provide the required protec-

tion to these forests:

• Burning patterns during controlled block bums in

catchments which contain forest patches should be recon-

sidered. The forests should not be used as fire breaks as

has been done in several cases. Fires should be burnt down

the slopes as would occur with natural fires (Geldenhuys

1994) and not from the bottom of the valleys upwards.

• Smaller alluvial sites along the rivers and the area

surrounding the exit of streams from the mountains should

not be cultivated, grazed or burnt.

7 ACKNOWLEDGEMENTS

This study formed part of the activities of the Pro-

gramme for Natural Resources and Rural Development of

the Division of Water, Environment and Forestry Tech-

nology, CSIR. It was funded by the Department of Water

Affairs and Forestry (Forestry Branch). These activities

formed part of a Ph.D. study under supervision of Prof.

E.J. Moll, formerly of the Department of Botany, Univer-

sity of Cape Town. Several colleagues assisted during the

sampling of the forest patches in often very difficult ter-

rain: R. Bartholomew, M.J. Cameron, T. Hoekstra, H.J.

Homann, J.H. Koen, H. Kotze, A. Meyer, C.M. van den

Berg, C.J. van der Merwe, J.P.L. van der Walt, M. Viviers

and J.H.J. Vlok. Mr J. Dobson prepared the map of the

study area.

8 REFERENCES

ACOCKS, J.RH. 1988. Veld types of South Africa, 3rd edn. Memoirs of

the Botanical Survey of South Africa No. 57.

ANON. 1987. Forest map of southern Africa. Forest Biome Project Na-

tional Programme for Environmental Sciences, FRD, Pretoria.

ARNOLD, T.H. & DE WET, B.C. (eds) 1993. Plants of southern Africa:

names and distribution. Memoirs of the Botanical Survey of South

Africa No. 62. National Botanical Institute, Pretoria.

AXELROD, D.I. & RAVEN, PH. 1978. Late Cretaceous and Tertiary

vegetation history of Africa. In M.J. A. Werger, Biogeography and

ecology of southern Africa: 77-130. Junk, The Hague,

BOND, W. 1981. Vegetation gradients in southern Cape mountains.

M.Sc. thesis. University of Cape Town, Rondebosch.

BOND, W.J., MIDGLEY, J. & VLOK, J. 1988. When is an island not an

island? Insular effects and their causes in fynbos shrublands.

Oecologiall: 5l5-52\.

BROWN, J.H. & GIBSON, A.C. 1983. Biogeography. Mosby, St Louis.

BRUMMITT, R.K. & POWELL, C.E. (eds) 1992. Authors of plant names.

Royal Botanic Gardens, Kew.

CHAPMAN, J.D. & WHITE, F. 1970. The evergreen forests of Malawi.

Commonwealth Forestry Institute, Oxford.

COATES PALGRAVE, K. 1977. Trees of southern Africa. Stmik, Cape Town.

COETZEE, J.A. 1986. Palynological evidence for major vegetation and

climatic change in the Miocene and Pliocene of the southwestern

Cape. South African Jourrml of Science 82; 71, 72.

COWLING, R.M. 1984. A syntaxonomic and synecological study in the

Humansdorp region of the Fynbos Biome. Bothalia 15: 175-227.

DEACON, H.J. 1983. The comparative evolution of Mediterranean-type

ecosystems: a southern perspective. In E.J. Kruger, D.T. Mitchell

& J.U.M. Jarvis, Mediterranean-type ecosystems — the role of nu-

trients: 3-40. Ecological Studies 43. Springer- Verlag, Berlin.

DEACON, H.J., SCHOLTZ, A. & DAITZ, L.D. 1983. Fossil charcoals as

a source of palaeoecological information in the lynbos region. In

H.J. Deacon, Q.B. Hendy & J.J.N. Lambrechts, Fynbos palaeoe-

cology: a preliminary synthesis : 174-182. South African National

Scientific Programmes Report No. 74. CSIR, Pretoria.

DOWSETT-LEMAIRE, F. 1988. The forest vegetation of Mt Mulanje

(Malawi): a floristic and chorological study along an altitudinal

gradient (650-1950 m). Bulletin du Jardin botanique nationale de

Belgique 58: 77-107.

EVERARD, D.A. 1987. A classification of the subtropical transitional

thicket in the eastern Cape, based on syntaxonomic and structural

attributes. South African Journal of Botany 53: 329-340.

GELDENHUYS, C.J. 1975. Raising indigenous trees. Pamphlet No. 150.

Department of Forestry, Pretoria.

GELDENHUYS, C.J. 1980. The elfects of management for timber pro-

duction on floristics and growing stock in the southern Cape

indigenous forests. South African Forestry Journal 113: 6-15, 25.

GELDENHUYS, C.J. 1986. Nature’s valley: a refugium for rare southern

Cape forest species. Fa/aeoeco/ogy o/A/nca 17: 173-181.

GELDENHUYS, C.J. 1992a. Disjunctions and distribution limits of for-

est species in the southern Cape. South African Forestry Journal

161: 1-13.

GELDENHUYS, C.J. 1992b. Richness, composition and relationships of

the floras of selected forests in southern Africa. Bothalia 22:

205-233.

GELDENHUYS, C.J. 1993a. Floristic composition of the southern Cape

forest flora with an annotated checklist. South African Journal of

Botany 59: 26-44.

GELDENHUYS, C.J. 1993b. Reproductive biology and population

structures of Podocarpus falcatus and P. latifolius in southern

Cape forests. Botanical Journal ofLinnean Society 112: 59-74.

GELDENHUYS, C.J. 1994. Bergwind fires and the location pattern of

forest patches in the southern Cape landscape. South Africa.

Journal of Biogeography 21: 49-62.

GELDENHUYS, C.J. 1996. Fruit/seed characteristics and germimition

requirements of tree and shrub species of the southern Cape

forests. Report FOR-DEA 954, Division of Forest Science and

Technology. CSIR, Pretoria.

GELDENHUYS, C.J., KNIGHT, R.S., RUSSELL, S. & JARMAN, M.L,

(eds) 1988. Dictionary of fore. st structural terminology. South

African National Scientific Programmes Report No. 147. FRD,

Pretoria.

GIVNISH, T.J. 1980. Ecological constraints on the evolution of breeding

systems in seed plants: dioecy and dispersal in gymnosperms.

Evolution 34: 959-972.

HENDEY, Q.B. 1983. Cenozoic geology and palaeogeography of the

fynbos region. In H.J. Deacon, Q.B. Hendy & J.J.N. Lambrechts,

Fynbos palaeoecology: a prelimimtry synthesis: 35-60. South

African National Scientific Programmes Report No. 74. CSIR,

Pretoria.

KAUL, R.B., KANTAK, G.E. & CHURCHE.L, S.P 1988. The Niobrara

River Valley, a postglacial migration corridor and refugium of

forest plants and animals in the grasslands of central North Amer-

ica. The Botanical Review 54: 44—8 1 .

KILLICK, D.J.B. 1963. An account of the plant ecology of the Cathedral

Peak area of the Natal Drakensberg. Memoirs of the Botanical

Survey of South Africa No. 34.

LENZ, C.J. 1957. The river evolution and the remnants of the Tertiary

surfaces in the western Little Karoo. Annals of the University of

Stellenbosch No. 33; 197-234.

LEVYNS, M R. 1964. Migrations and origin of the Cape flora. Transac-

tions of the Royal Society of South Africa 37: 85-107.

LUCKHOFF, H.A. 1963. Die Baviaanskloof- of Willowmore-seder Wid-

dringtonia schwarzii (Marl.)Mast. Forestry in South Africa 3: 1-14.

MAC ARTHUR, R.H. & WILSON, E.O. 1967. The theory of island

biogeography. Princeton University Press, Princeton.

McKENZIE, B. 1978. A quantitative and qualitative study of the indige-

nous forests of the southwestern Cape. M.Sc. thesis. University of

Cape Town, Rondebosch.

70

Bothalia27,l (1997)

MOORE, A.E. 1988. Plant distribution and the evolution of the major

river systems in southern Africa. South African Journal of Geol-

ogy 91: 346-349.

PALMER, E. & PITMAN, N. 1972. Trees of southern Africa. Balkema,

Cape Town.

PHILLIPS, J.F.V. 1926. Virgilia capensis Lamb. (Keurboom): a contribu-

tion to its ecology and sylviculture. South African Journal of

Science 23; 435^54.

PHILLIPS, J.F.V. 1927. The role of the bush dove Columba arquatrix T

& K in fruit dispersal in the Knysna forests. South African Journal

of Science 24: 435—440.

PHILLIPS, J.F.V. 1931. Forest succession and ecology in the Knysna

region. Memoirs of the Botanical Survey of South Africa No. 14.

SCHOLTZ, A. 1986. Palynological and palaeobotanical studies in the

southern Cape. M.A. thesis. University of Stellenbosch, Stellen-

bosch.

SKELTON, P.H. 1986. Distribution patterns and biogeography of non-

tropical southern African freshwater fishes. Palaeoecology in Af-

rica 17:211-230.

SMITH, J.M. B 1981. Colonist ability, altitudinal range and origins of Mt

Field, Tasmania. Journal of Biogeography 8: 249-261.

STSC 1986. Statgraphics — statistical graphics system. Users guide. Sta-

tistical Graphics Corporation, Rockville, Maryland.

TAYLOR, H.C. 1979. Observations on the flora and phytogeography of

Rooiberg, a dry fynbos mountain in the southern Cape Province,

South Africa. Phytocoenologia 6: 524-531.

THERON, J.N. 1962. An analysis of the Cape Folding in the district of

Willowmore, C.P. Annals of the University of Stellenbosch 37A:

347-419.

TOERIEN, D.K. 1979. The geology of the Oudtshoom area. Explanation

to sheet 3322, scale 1 :250 000. Government Printer, Pretoria.

VAN WYK, B.-E. 1986. A revision of the genus Virgilia. South African

Journal of Botany 52: 347-353.

VLOK, J.H.J. & DE RONDE, C. 1989. The effect of low intensity fires

on forest floor vegetation in mature Pinus elliottii plantations in

the Tsitsikamma. South African Journal of Botany 55: 11-16.

VON BREITENBACH, F. 1986. National list of indigenous trees. Den-

drological Foundation, Pretoria.

WEATHER BUREAU 1986. Climate of South Africa. WB40, Depart-

ment of Environment Affairs, Pretoria.

APPENDIX. SPECIES LIST FOR FORESTS IN INLAND MOUNTAINS OF THE SOUTHERN CAPE

The forest sites are as follows;

Rooiberg: 1 , Bosrivier upper; 2, Bosrivier lower; 3, Assegaaibosrivier

Kamanassie: 4, Kluesrivier; 5, Meulrivier upper

Swartberg: 6, Waterkloof; 7, Seweweekspoort; 8, Swartbergpoort (Prins Albert); 9, Swartberg Pass (hotel site); 10, Rust-en-Vrede; 11, Huisrivier

upper; 12, Meiringspoort; 13, Tierkloof; 14, Cheridouwspoort

Kouga-Baviaanskloof: 15, Sapreerivier; 16, Doringkloof; 17, Bosrug; 18, Geelhoutbos; 19, Assegaaikloof; 20, Witrivier upper; 21, Witrivier lower;

22, Grootrivierpoort

Groendal: 23, Zungarivier (Chase’s kloof)

Forest site

Species Total

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Canopy and subcanopy tree species

Acacia karroo Hayne

Allophylus decipiens (Sond.) Radik.

Apodytes dimidiataE.A/ey. exArn.

Atalaya capensis R.A.Dyer

Brachylaena glabra (L.f.j Druce

Buddleja saligna Willd.

Calodendrum capense (L.f) Thunb.

Canthium

ciliatum (Klotzsch) Kuntze

inerme (L.f. ) Kuntze

mundianum Cham. & Schlechtd.

pauciflorum (Klotzsch) Kuntze

Cassine

aethiopica Thunb.

eucleiformis (Eckl. & Zeyh.) Kuntze

papillosa (Hochst.) Kuntze

peragua L.

Celtis africana Burm.f.

Chionanthus foveolata (E.Mey.) Steam

Clausena anisata (Willd.) Hook.f ex Benth.

Cunonia capensis L.

Curtisia dentata (Burm.f) C.A.Sm.

Cussonia

paniculata Eckl. & Zeyh.

spicata Thunb.

Diospyros

dichrophylla (Gand.) De Winter

whyteana (Hiern) F. White

Ekebergia capensis Sparrm.

Euclea

racemosa Murray

schimperi (A. DC.) Dandy var. schimperi

undulata Thunb.

6

9

4

1

4

7

2

2

6

2

I

3

1

1

9

7

2

3

16

7

4

14

• 13

• 6

• 3

2

1

7

Bothalia27,l (1997)

71

Species

Eugenia zeyheri Harv.

Ficus sur Forssk.

Gonioma kamassi E.Mey.

Halleria lucida L.

Heteromorphatrifoliata (Wendl.) Eckl. & Zeyh.

Hippobromus pauciflorus {L. f.) Radik.

Ilex mitis (L. j Radik.

Kiggelaria afiricana L.

Loxostylis alata Spreng.f. ex Rchb.

Maytenus

acuminata (L.f.) Loes.

nemorosa (Eckl. i& Zeyh.) Marais

oleoides (Lam.) Loes.

peduncularis (Sand.) Loes.

undata (Thunb.) Blake lock

Nuxia floribunda Benth.

Ochna arborea Burch, ex DC.

Ocotea bullata (Burch.) Baill.

Olea

capensis L.

subsp. capensis

subsp. macrocarpa ( C.H. Wr ) I. Verd.

europaea L. subsp. africana (Mill.) RS.Green

Olinia ventosa (L.) Cufod.

Pappea capensis Eckl. & Zeyh.

Pittosporum viridiflorum Sims

Pleurostylia capensis (Turcz.) Oliv.

Podocarpus

falcatus (Thunb.) R.Br. ex Mirb.

latifolius (Thunb.) R.Br. ex Mirb.

Psydrax obovata ( Eckl. & Zeyh. ) Bridson

Ptaeroxylon obliquum (Thunb.) Radik.

Pterocelastrus

rostratus Walp.

tricuspidatus (Lam.) Sond.

Rapanea melanophloeos (L.) Mez

Rhus

chirindensis Baker f.

lancea L.f.

lucida L.

Rothmannia capensis Thunb.

Schotia latifolia Jacq.

Scolopia

mundii (Eckl. &Zeyh.) Warb.

zeyheri (Nees) Harv.

Sideroxylon inerme L.

Smelophyllum capense (Sond.) Radik.

Sterculia alexandri Harv.

Strychnos decussata (Pappe) Gilg

Tarchonanthus camphoratus L.

Teclea natalensis (Sond.) Engl.

Vepris lanceolata (Lam.) G.Don

Virgilia divaricata Adamson

Widdringtonia schwarzii (Marloth) Mast.

Zanthoxylum capense (Thunb.) Harv.

Shrubs

Acokanthera oppositifolia (Lam.) Codd

Azima tetracantha Lam.

Brachylaena neriifolia (L.) R.Br.

Buddleja

salviifolia (L.) Lam.

sp.

Calpumia villosa Harv.

Capparis sepiaria L. var. citrifolia (Lam.) Toelken

Carissa

bispinosa (L.) Desf. ex Brenan

haematocarpa (Eckl.) A.DC.

Cassine

parvifolia Sond.

reticulata (Eckl. & Zeyh.) Codd

tetragona (L.f) Loes.

Chaetacme aristata Planch.

Chrysanthemoides monilifera (L.) Norl.

Colpoon compressum P.J.Bergius

Forest site

72

Bothalia27,l (1997)

Species

1 2 3 4 5 6 7

Forest site

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Total

Crotalaria capensis Jacq.

Diospyros

glabra (L.) De Winter

scabrida (Harv. ex Hiem) De Winter

Dovyalis rhamnoides (Burch, ex DC.) Harv.

Ehretia amoena Klotzsch

Euclea schimperi (A. DC.) Dandy var. daph-

noides (Hiem) De Winter

Ficus burtt-davyi Hutch.

Grewia occidentalis L.

Lachnostylis bilocularis R. A. Dyer

Laurophyllus capensis Thunb.

Maytenus

heterophylla (Eckl. & Zeyh.) N.Robson

mossambicensis (Klotzsch) Blake lock

Myrica serrata Lam.

Myrsine africana L.

Ochna serrulata (Hochst.) Walp.

Pavetta lanceolata Eckl.

Plumbago auriculata Lam.

Psoralea sp. cf. pinnata

Putterlickia pyracantha (L.) Szyszyl.

Rhamnus prinoides L’Her.

Rhus

glauca Thunb.

longispina Eckl. & Zeyh.

rehmanniana Engl.

sp. cf. refracta

Scutia myrtina (Burm.f.) Kurz

Trichocladus ellipticus Eckl. & Zeyh. ex Walp.

Widdringtonia nodiflora (L.) Powrie

Lianes and vines

Adenocline acuta (Thunb.) Baill.

Cissampelos torulosa E.Mey. ex Harv.

Clematis brachiata Thunb.

Coccinea palmata (Sond.) Cogn.

Cynanchum ellipticum (Harv.) R.A.Dyer

Dipogon lignosus (L.) Verde.

Kedrostis nana (Lam.) Cogn. var. schlechteri

( Cogn. } A.Meeuse

Microglossa mespilifolia (Less.) B.L.Rob.

Myrsiphyllum scandens (Thunb.) Oberm.

Oncinema lineare (L. f.) Bullock

Pelargonium sp.

Protasparagus

aethiopicus (L.) Oberm.

setaceus (Kunth) Oberm.

sp.

Rhoicissus

digitata (L.f.) Gilg & Brandt

tomentosa (Lam.) Wild & R.B.Drumm.

Rhynchosia caribaea (Jacq.) DC.

Rubus pinnatus Willd.

Rumex sagittatus Thunb.

Secamone

alpini Schult.

filiformis (L.f.) J.H.Ross

Senecio

angulatus L.f.

deltoideus Less.

quinquelobus (Thunb.) DC.

sp, 1

sp. 2

Thunbergia dregeana Nees

Tylophora lycioides (E.Mey.) Decne.

Zehneria scabra (L.f) Sond.

1

3

6

3

1

6

9

9

3

1

15

1

9

13

3

2

5

7

6

2

1

3

15

1

5

1

2

1

1

10

2

8

17

7

2

6

2

3

14

1

4

1

4

5

2

15

3

3

2

1

1

1

1

1

3

Ferns

Adiantum capillus-veneris L. --------- - 4

Asplenium

adiantum-nigrum L. var. solidum (Kunz.e)

J.P.Roux - 1

daWhopienm (Burm.f ) Bech. - g

lunulatum Sw. ----------------- • 4

BothaIia27,l (1997)

73

Species

Asplenium (cont.)

monanthes L,

protensum Schrad.

rutifolium (P.J.Bergius) Kunze

splendens Kunze

Blechnum

australe L

capense Burm.f.

attenuatum (Sw.) var. giganteum (Kaulf.) Bonap.

inflexum (Kunze) Kuhn

punctulatum Sw.

tabulare (Thunb.) Kuhn

Ceterach cordatum (Thunb.) Desv.

Cheilanthes

capensis (Thunb.) Sw.

hirta Sw.

multifida (Sw.) Sw.

viridis (Forssk.) Sw.

parviloba (Sw.) Sw.

Ctenitis lanuginosa (Willd. ex Kaulf.) Copel.

Cyathea capensis (L.f.)J.E.Sm.

Dryopteris callolepis C. Chr.

Elaphoglossum acrostichoides (Hook. & Grev.)

Schelpe

Gleichenia polypodioides (L.) J.E.Sm.

Histiopteris incisa (Thunb.) J.Sm.

Hypolepis sparsisora (Schrad.) Kuhn

Lycopodium gnidioides L.f.

Microsorium ensiforme (Thunb.) Schelpe

Osmunda regalis L.

Pellaea pteroides (L.) Brand

Pleopeltis macrocarpa (Bory ex Willd.) Kaulf.

Polystichum pungens (Kaulf.) Presl

Pteris dentata Eorssk.

Rumohra adiantiformis (G.Forst.) Ching

Thelypteris

bergiana (Schlechtd.) Ching

intemipta (W;7W.) K.lwats.

Todea barbara (L. ) T.Moore

1 2 3 4 5 6 7

Geophytes and graminoids

Agapanthus sp.

Aristea ensifolia J.Muir

Carex aethiopica Schkuhr

Chlorophytum comosum (Thunb.) Jacq.

Commelina africana L. var. lancispatha

C.B. Clarke

Cyperus albostriatus Schrad.

Dietes iridioides (L.) Sweet ex Klatt

Disperis lindleyana Rchb.f.

Ehrharta

sp. cf. ramosa/gigantea

sp. 1

sp. 2

Ficinia sylvatica Kunth

Gladiolus robustus Goldblatt

Haemanthus albiflos Jacq.

Oplismenus hirtellus (L.) PBeauv.

Omithogalum longibracteata Jacq.

Panicum deustum Thunb.

Scadoxus puniceus (L.) Friis & Nordal

Schoenoxiphium

lanceum (Thunb.) Kuk.

lehmannii (Nees) Steud.

Stipa dregeana Steud.

Zantedeschia aethiopica (L.) Spreng.

Soft shrubs and forbs

Abutilon sonneratianum (Cav.) Sweet

Aloe sp.

Asparagus macowanii Baker

Cardamine africana L.

Centella erianthe (Rich.) Drude

Ceropegia sp.

8 9

Forest site

10 11 12 13 14 15 16 17 18 19 20 21 22 23

Total

1

2

2

2

18

15

10

5

18

2

4

1

1

• 4

1

1

1

74

Bothalia27,l (1997)

Species

Clutia

natalensis Bernh.

pulchella L.

Crassula

nemorosa {Eckl. & Zeyh.) Endl. ex Walp.

pellucida L. subsp. marginalis (Dryand. in

Alton) Toelken

Cyrtorchis arcuata (Lindl.) Schltr.

Didymodoxa caffra (Thunb.) Friis & Wilmot-

Dear

Euphorbia kxaussiana Bernh.

Galium thunbergianum Eckl. & Zeyh.

Galopina circaeoides Thunb.

Gerbera cordata (Thunb.) Less.

Helichrysum petiolare Hilliard & B.L.Burtt

Hypoestes sp. cf. verticillatus

Isoglossa prolixa (Nees) Lindau

Jalropha capensis (L.f.) Sand.

Knowltonia vesicatoria (L.f. ) Sims subsp.

humilis H.Rasm.

Leidesia prochmbens (L.) Brain

Leonotis ocymifolia (Burm.f.) Iwarsson

Lobelia sp.

Nemesia mellissifolia Benth.

Oxalis incamata L.

Pavonia praemorsa (L.f. ) Cav.

Pelargonium

ribifolia Jacq.

scabroide Knuth

zonale (L.) L'Her.

Peperomia

retusa (L.f. ) A.Dietr.

tetraphylla (G.Forst.) Hook. & Arn.

Peucedanum sp.

Plectranthus verticillatus (L.f) Druce

Polygala myrtifolia L.

Solanum

giganteum Jacq.

retroflexum Dunal

tomentosum L.

Sp. cf. Acalypha

Stachys

aethiopicaL.

grandifolia E.Mey. ex Benth.

Streptocarpus rexii (Hook.) Lindl.

Zygophyllum morgsana L.

Forest site

Bothalia 27,1:75-82(1997)

Cytogenetic studies in some representatives of the subfamily Poddeae (Poaceae)

in South AMca. 3. The tribe Poeae

J.J. SPIES*, S.M.C. VAN WYK*, I.C. NIEMAN* and EJ.L. LIEBENBERG**

Keywords: basic chromosome numbers, chromosomes, meiosis, Poaceae, Poeae, polyploidy, Pooideae, South Africa

ABSTRACT

This is a report on chromosome numbers for the tribe Poeae, which is represented in South Africa mainly by naturalized

exotics. Chromosome numbers of 67 specimens, representing 26 species and 1 1 genera, are presented. These numbers include

the first reports on Poa binata Nees (n = 3x = 21 and n = 4x = 28), Puccinellia acroxantha C.A.Sm. & C.E.Hubb. (n = 3x = 21)

and P angusta (Nees) C.A.Sm. & C.E.Hubb. (n = x = 7). New ploidy levels are reported for Catapodium rigidum (L.) C.E.Hubb.

(n = 2x = 14), Festuca caprina Nees (n = 2x = 14) and F. scabra Vahl (n = x = 7).

INTRODUCTION

The first paper in this series indicated the importance

of determining the ploidy levels and basic chromosome

numbers of naturalized and endemic flora in South Africa

(Spies et al. 1996a). In the second paper chromosome

numbers of the rest of the tribe Aveninae were presented

(Spies et al. 1996b). This third paper in the series on chro-

mosome numbers of representatives of the subfamily

Pooideae in South Africa, is restricted to the tribe Poeae.

During this study we followed the classification system

of Gibbs Russell et al. (1990) for the tribal separation. There-

fore we did not recognize the tribe Hainardeae Greuter, and

all the species usually belonging to this small tribe are in-

cluded in the Poeae. The tribe Poeae consists, therefore, of

approximately 55 genera and more than 5000 species (Clay-

ton & Renvoize 1986). Most local species belonging to this

tribe are naturalized exotics. The Poeae are represented in

South Alfica by the genera Brizja. L., Catapodium Link,

Colpodium Trin., Cynosunis L., Dactylis L., Festuca L., Hai-

nardia Greuter, Lamarckia Moench, Folium L., Parapholis

C.E.Hubb., Poa L., Puccinellia Pari., Sphenopus Trin. and

Vulpia C.C.Gmel. (Gibbs Russell et al. 1990).

The aim of this study is to determine the chromosome

numbers, polyploid levels and meiotic chromosome be-

haviour of the South African representatives of the tribe

Poeae. These results will, eventually, be compared with

the results obtained from indigenous and endemic taxa to

compare the frequency of polyploidy between indigenous

and introduced grasses.

MATERIALS AND METHODS

For the purpose of this study, cytogenetic material was

collected in two dilferent ways. The material was either

collected and fixed in the field, or living material was

collected in the field and transplanted in the nurseries of

* Department of Botany and Genetics, University of the Orange Free

State, RO. Box 339, Bloemfontein 9300.

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

MS. received: 1996-08-07.

either the Department of Botany and Genetics, University

of the Orange Free State (Bloemfontein) or the National

Botanical Institute (Pretoria), where the cytogenetic ma-

terial was collected and fixed. The specimens used and

their localities are listed in Table 1. Voucher specimens

are housed either in the Geo Potts Herbarium, Department

of Botany and Genetics, University of the Orange Free

State, Bloemfontein (BLFU) or in the National Herbarium,

Pretoria (PRE).

Anthers were squashed in aceto-carmine and meioti-

cally analysed (Spies et al. 1996a). Chromosome numbers

are presented as haploid chromosome numbers to conform

to previous papers on chromosome numbers in this journal

(Spies & Du Plessis 1986a). Genome homology in some

tetraploid specimens was determined according to the

models described by Kimber & Alonso (1981).

RESULTS AND DISCUSSION

Briza consists of 16 species, with three (B. maxima L.,

B. minor L. and B. subaristatum Lam.) being naturalized

in South Africa. Two of these species were studied. Both

were diploid with B. maxima having the haploid chromo-

some number seven (n = x = 7) and B. minor five (n -

X = 5) (Table 1). These findings confirm previous reports

on chromosome numbers for these species by Fedorov

1969; Ornduff 1967-1969; Moore 1970, 1971, 1972,

1974, 1977; Goldblatt 1981, 1983, 1985, 1988; Goldblatt

& Johnson 1990, 1991, 1994, who observed tetraploid

Briza specimens in several species. However, polyploidy

seems to be absent in both B. maxima and B. minor speci-

mens studied here and abroad. Meiosis was regular in both

species, and bivalents were formed (Figure lA & B).

TTiese results, in combination with the available chromo-

some numbers given in the literature consulted, indicate

that Briza has two basic chromosome numbers, namely

five and seven.

Catapodium consists of two species, one of which is

naturalized in South Africa [C. rigidum (L.) C.E.Hubb.].

Three of the studied specimens were diploid (n = x = 7),

whereas the other specimen was tetraploid (n = 2x = 14)

(Table 1). The diploid chromosome number observed sup-

76

Bothalia27,l (1997)

TABLE 1. — Haploid chromosome numbers of representatives of the tribe Poeae (Poaceae, Pooideae) in southern Africa with their voucher specimen

numbers and specific localities. Species are listed alphabetically, voucher numbers within a species numerically and the locality is presented

according to the system described by Edwards & Leistner (1971)

LocaUty

WESTERN CAPE.— 3218 (Clanwilliam): Versveld Pass, (-DC)

WESTERN CAPE. — 3421 (Riversdale): 4 km from Kweekkraal to Droerivier, (-AA)

WESTERN CAPE. — 3318 (Cape Town): 19 km from Darling to Malmesbury, (-AD)

NORTHERN CAPE. — 3229 (Sutherland); 10 km from Sutherland to Matjiesfontein, (-BC)

WESTERN CAPE. — 3418 (Simonstown); Redhill, (-AB)

WESTERN CAPE. — 3318 (Cape Town): Bothmaskloof, (-BC)

WESTERN CAPE. — 3420 (Bredasdorp); 4 km N of De Hoop Nature Reserve, (-AD)

WESTERN CAPE. — 3420 (Bredasdorp): 20 km from Bredasdorp to Spitskop, (-CA)

WESTERN CAPE. — 3420 (Bredasdorp): 6 km from Ouplaas to De Hoop Nature Reserve,

(-AD)

WESTERN CAPE. — 3319 (Worcester): 1 km south of old toll house in Mitchell’s Pass, (-AD)

WESTERN CAPE.— 3418 (Simonstown): Redhill, (-AB)

EASTERN CAPE. — 3228 (Butterworth); on beach at Bonza Bay, (-CC)

MPUMALANGA. — ^2530 (Lydenburg): 18 km from Lydenburg to Weltevreden, (-AB)

MPUMALANGA. — 2530 (Lydenburg): 6 km from Dullstroom to Goede Hoop, (-AC)

EASTERN CAPE. — 3028 (Matatiele): 47 km from Rhodes in Naude’s Neck, (-CC)

EASTERN CAPE. — 3027 (Barkly East): Beestekraal se loop, (-DC)

MPUMALANGA. — 2530 (Lydenburg): 16 km from Lydenburg to Sabie, (-BA)

MPUMALANGA. — 2530 (Lydenburg): 17 km from Lydenburg to Roossenekal, (-AB)

EASTERN CAPE. — 3424 (Humansdorp); 10 km from Humansdorp to Cape St Francis,

(-BB)

WESTERN CAPE. — 3421 (Riversdale): 25 km from Droerivier to Vermaaklikheid via

Oudemuragie, (-AC)

WESTERN CAPE. — 3319 (Worcester): 7 km from Villiersdorp to Franschhoek, (-CC)

EASTERN CAPE. — 3127 (Lady Frere): 9 km from Dordrecht to Barkly East, (-AC)

WESTERN CAPE. — 3319 (Worcester): 6 km from Franschhoek to Villiersdorp, (-CC)

WESTERN CAPE.— 3318 (Cape Town): 1 km east of Mamre Road, (-BC)

MPUMALANGA. — 2530 (Lydenburg): 14 km from Dullstroom to Lydenburg, (-AC)

EASTERN CAPE. — 3027 (Lady Grey): near Barkly East, (-DC)

FREE STATE.— 2826 (Brandfort): Glen, (-CD)

WESTERN CAPE.— 3318 (Cape Town): 1 km east of Mamre Road, (-BC)

WESTERN CAPE. — 3319 (Worcester): 8 km from Wellington to Worcester in Bains-

kloof, (-CA)

EASTERN CAPE. — 3325 (Port Elizabeth): King Neptune Beach, (-DC)

WESTERN CAPE. — 3319 (Worcester): Katbakkies tum-off on road between Ceres and

Citrusdal, (-AB)

WESTERN CAPE. — 3217 (Vredenburg): on beach outside Cape Columbine Nature

Reserve, (-DD)

WESTERN CAPE. — 3318 (Cape Town): 1 km east of Mamre Road, (-BC)

WESTERN CAPE. — 3017 (Hondeklipbaai): dunes at Groenrivier Mouth, (-DC)

WESTERN CAPE. — 3318 (Cape Town): 7 km from Yzerfontein to Darling, (-AC)

WESTERN CAPE. — 33 1 8 (Cape Town); 5 km from Langebaan to Langebaanweg, (-DC)

WESTERN CAPE. — 3320 (Montagu): 22 km from Villiersdorp to Worcester via Koppies,

(-AD)

WESTERN CAPE. — 3420 (Bredasdorp); De Hoop Nature Reserve, (-AD)

EASTERN CAPE. — 3027 (Lady Grey): 34 km from Rhodes to Lundean’s Neck, (-DD)

FREE STATE — 2729 (Volksrust): Verkykerskop, (-CC)

WESTERN CAPE. — 3218 (Clanwilliam); 5 km south of Elandsbaai, (-AB)

WESTERN CAPE. — 3318 (Cape Town ): 1 km north of Uilekraal on road between

Hopefield and Darling, (-AB)

EASTERN CAPE. — 3424 (Humansdorp): 10 km from Humansdorp to Cape St Francis, (-BB)

WESTERN CAPE. — 3320 (Montagu): 22 km from Villiersdorp to Worcester via Koppies, (-AD)

WESTERN CAPE. — 3118 (Vanrhynsdorp): 2 km from Doombaai to Donkinbaai, (-CB)

MPUMALANGA. — 2530 (Lydenburg): 39 km from Lydenburg to Roossenekal, (-AA)

WESTERN CAPE.— 3318 (Cape Town): 1 km east of Mamre Road, (-BC)

MPUMALANGA. — 2530 (Lydenburg): 6 km from Goede Hoop to Dullstroom, (-AC)

EASTERN CAPE. — 3028 (Matatiele): 12 km from Rhodes in Naudesnek, (-CC)

NORTHERN CAPE. — 3017 (Hondeklipbaai); 6 km from Kamieskroon in Kamiesberg Pass,

(-BB)

WESTERN CAPE. — 3318 (Cape Town): Afrikaanse taal memorial, (-DD)

NORTHERN CAPE. — 3119 (Calvinia): 20 km from Calvinia to Loeriesfontein, (-AB)

NORTHERN CAPE. — 3220 (Sutherland); 2 km from Sutherland to Calvinia, (-BC)

WESTERN CAPE. — 3318 (Cape Town): Tinie Versveld Nature Reserve, (-AD)

WESTERN CAPE. — 3218 (Clanwilliam): 5 km south of Bobbejaanberg Point, Elands-

baai, (-AD)

WESTERN CAPE. — 3318 (Cape Town): 10 km from Wellington to Porterville, (-DB)

WESTERN CAPE. — 3319 (Worcester): Katbakkies tum-off on road between Ceres and

Citrusdal, (-AB)

WESTERN CAPE. — 3218 (Clanwilliam): 5 km south of Eland’s Bay, (-AB)

NORTHERN CAPE. — 3018 (Kamiesberg): 16 km east of Kamieskroon, (-AC)

WESTERN CAPE. — 3319 (Worcester): 4 km to Franschhoek from tum-off on Villiersdorp-

Grabouw road, (-CC)

WESTERN CAPE. — 3418 (Simonstown): Silvermine Nature Reserve, (-AB)

EASTERN CAPE. — 3027 (Lady Grey): 45 km from Barkly East to Rhodes, (-DD)

WESTERN CAPE. — 3319 (Worcester): Katbakkies tum-off on road between Ceres and

Citmsdal, (-AB)

NORTHERN CAPE. — 3018 (Kamiesberg): 21 km from Kamieskroon to Gamoep, (-AC)

Bothalia27,l (1997)

77

ports the previous numbers recorded by various authors

(TFedorov 1969; Omduff 1967-1969; Moore 1970, 1971,

1972, 1974, 1977; Goldblatt 1981, 1983, 1985, 1988;

Goldblatt & Johnson 1990, 1991, 1994). To the best of

our knowledge, the tetraploid number observed is a new

ploidy level for C. rigidum. Tetraploid specimens have,

however, been observed in the other Catapodium species.

Both ploidy levels studied, formed bivalents only (Figure

1C & D) and meiosis was normal. This indicates that the

tetraploid specimen may be of alloploid origin.

FIGURE 1. — Photomicrographs of

meiotic chromosomes in the

genera Briza and Catapodium.

A, B. maxima. Spies 4420,

diakinesis with 7n; B, B. minor.

Spies 4425, diakinesis with 5ii;

C, C. rigidum. Spies 3854,

diakinesis with 7n; D, C.

rigidum. Spies 3451, metaphase

I with 14n. Scale bars: 10 |itm.

In this study the alloploid origin of the tetraploid C.

rigidum specimen is supported by its genome constitution

analysis, in which the observed chromosome associations

concurred best with the associations expected for the 2:2

model of Kimber & Alonso (1981) (Table 2). This model

indicates that two sets of genomes are present (two

genomes per set) and the relative similarity between the

genomes within a set is 0.5. The relative similarity be-

tween sets of genomes is expressed by an x-value that

may vary between 0.5 (differences between sets are simi-

TABLE 2. — Genomic relationships in some tetraploid representatives of the tribe Poeae (Poaceae, Pooideae) in southern Africa according to the

models of Kimber & Alonso (1981). The number in square brackets indicates the relative affinity of the different genomes

78

Bothalia27,l (1997)

lar to the differences within a set, therefore the genomes

may be presented in the form of AAAA) and 1 (sets dif-

fering greatly, therefore the genomes may be presented as

AABB). The x-value for the tetraploid Cl rigidum speci-

men is almost 1 (Table 2) and, therefore, very little to no

homology exists between the two chromosome sets

(genomes may be presented by AAA’ A, where A differs

greatly from A’).

Colpodium contains three species, with C. hedbergii

(Melderis) Tzvelev being indigenous to this country. This

species was not included in this study. However, chromo-

some numbers of 2n = 4, 8, 14 and 28 have been reported

for the genus Colpodium (Ornduff 1967-1969; Moore

1970, 1971, 1972, 1974, 1977; Goldblatt 1981, 1983,

1985, 1988; Goldblatt & Johnson 1990, 1991, 1994), and

2n = 8 for C. hedbergii (Hedberg & Hedberg 1977). Fur-

ther studies are necessary to determine the basic chromo-

some number of this genus and also to determine the

phylogenetic development of chromosome numbers.

Cvnosurus has eight species worldwide, with C. col-

oratu^ Lehm. ex Nees being a very rare indigenous spe-

cies and C. echinatus L. a naturalized species. Neither of

these species was included in this study. Chromosome

number reports indicate that all species studied are diploid

and C. echinatus was included in some of those studies

(Ornduff 1967-1969; Moore 1970, 1971, 1972, 1974,

1977; Goldblatt 1981, 1983, 1985, 1988; Goldblatt &

Johnson 1990, 1991, 1994).

Dactylis is represented in South Africa by one natural-

ized species {D. glomerata L.). The specimen of D. glom-

erata studied, was diploid (Table 1). This finding supports

the various chromosome number reports of 2n = 14 or 28

for D. glomerata, as well as for the genus Dactylis in

general (Fedorov 1969; Ornduff 1967-1969; Moore 1970,

1971, 1972, 1974, 1977; Goldblatt 1981, 1983, 1985,

1988; Goldblatt & Johnson 1990, 1991, 1994).

Festuca has approximately 360 species worldwide and

is represented by eight indigenous species: F. africana

(Hack.) Clayton, F caprina Nees, F costata Nees, F. dra-

comontana H.P.Linder, F. killickii Kenn. -O’Byrne, F

longipes Stapf, F scabra Vahl and F vulpioides Steud.,

and one naturalized species: F elatior L. ( = F arundi-

nacea). Four Festuca species were included in this study.

They were all found to be tetraploid except for one of the

seven specimens of F scabra, which was diploid (Table

1 ). In this study the tetraploid chromosome numbers (n =

2x = 14) were observed for F caprina; this represents a

new ploidy level for this species. Previous reports gave

an octoploid number (n = 4x = 28) for this species (Spies

& Du Plessis 1986a, b). This study confirms a tetraploid

chromosome number (n = 2x = 14) for F costata (De Wet

1958). The tetraploid number (n = 2x = 14) for F elatior

is in support of one of the ploidy levels previously de-

scribed for this species, namely 2n = 14, 28, 42, 56 and

70 (Fedorov 1969; Ornduff 1967-1969; Moore 1970,

1971, 1972, 1974, 1977; Goldblatt 1981, 1983, 1985,

1988; Goldblatt & Johnson 1990, 1991, 1994). Festuca

longipes was not included in this study but Hill (1965)

reported a hexaploid number for this species. This is the

first study to report a diploid chromosome number for F.

scabra. The tetraploid number observed confirms the level

previously reported with nonaploid and decaploid num-

bers (De Wet & Anderson 1956; De Wet 1958; Spies &

Du Plessis 1986a, b). The chromosome numbers of four

South African species of Festuca are still unknown, i.e.

F. africana, F dracomontana, F killickii and F. vulpioides.

Meiosis was normal in most specimens, with only bi-

valents being formed (Figure 2), excepting a telophase I

cell with four micronuclei in F. costata (Figure 2C) and

a telophase II cell in one F. scabra specimen. Spies 3962,

with a micronucleus and the possible remnants of an ana-

phase II bridge (Figure 2E). This low frequency of abnor-

malities (less than one cell in the forty cells studied per

specimen), may be attributed to accidental misdivisions.

The presence of only bivalents found in the tetraploid

specimens of F. caprina (Figure 2A), F. costata (Figure

2B), F. elatior and F. scabra (Figure 2D), suggests that

these species have alloploid origins.

FIGURE 2. — Photomicrographs of

meiotic chromosomes in the ge-

nus Festuca. A, F. caprina,

Saayman 1 16, diakinesis with

14ii. B, C, F. costata. Spies

4692: B, metaphase 1 with 14n;

C, telophase I with four micro-

nuclei. D, E, F. scabra: D, Saay-

man 99, diakinesis with 14ii; E,

Spies 3962, telophase II with

micronucleus and possible rem-

nant of anaphase II bridge.

Scale bars: 10 pm.

BothaJia27,l (1997)

79

FIGURE 3. — Photomicrographs of meiotic chromosomes in the genus Lolium. A, L rigidum. Spies 3190, diakinesis with 7ii; B, L. perenne x L.

multiflorum. Spies 2463, metaphase I with 7ii; C, L. multiflorum. Spies 4428, anaphase 1 with 4 laggards. D-G, L. temulentum: D, Spies 4567,

metaphase I with 7n + IB; E, Spies 4637, diplotene with 7n; F, Spies 4637, diakinesis with 14n; G, Spies 4569, anaphase 1 with chromatin

strand connecting two segregating chromosomes. H, I, Lolium sp.. Spies 5062: H, anaphase II with four laggards; I, telophase 11 with two

micronuclei. Scale bars; 10 pm.

Genome analyses of all tetraploid Festuca specimens

indicated that, in each case, the observed chromosome

associations concurred best with the expected associations

for the 2:2 model (Table 2). The x-values varied from

0.95 to 1 and the analysed specimens of all four species

suggest an alloploid origin (Table 2).

Hainardia is a monotypic genus, and H. cylindrica

(Willd.) Greuter is naturalized in this country. We ob-

served a diploid chromosome number of n = x = 7 in this

species (Table 1). This number deviates from the 2n = 26

given in the literature (Scrugli & Bocchieri 1977, men-

tioned in Goldblatt 1981). Further studies of this ex-

tremely rare species are necessary to determine its basic

chromosome number and the possible evolutionary

change from x = 7 to x = 13.

Lamarckia aurea (L.) Moench represents another

monotypic naturalized genus in South Africa. No suitable

material of this species could be obtained for this study.

All the published chromosome numbers given agree that

the species is diploid (2n = 2\ - 14) (Fedorov 1969;

Moore 1970, 1971, 1972, 1974, 1977; Goldblatt 1988;

Goldblatt & Johnson 1994).

Four species of the genus Lolium are naturalized in

South Africa, i.e. L. multiflorum Lam., L. perenne L., L.

rigidum Gaudin and L. temulentum L. All four species

were studied and were all diploid (Figure 3A, B, D & E),

thus confirming previous chromosome number reports

(Fedorov 1969; Ornduff 1967-1969; Moore 1970, 1971,

1972, 1974, 1977; Goldblatt 1981, 1983, 1985, 1988;

Goldblatt & Johnson 1990, 1991, 1994). A single

tetraploid cell was observed in a L. temulentum specimen

(Figure 3F). This tetraploid cell may be attributed to cell

fusion (Spies & Van Wyk 1995). A few meiotic abnor-

malities were observed. They include anaphase I laggards

(Figure 3C), the presence of B-chromosomes (Figure 3D),

an anaphase I bridge (Figure 3G), anaphase II laggards

(Figure 3H) and telophase II micronuclei (Figure 31).

The genus Parapholis comprises six species, but only

P. incurva (L.) C.E.Hubb. is locally naturalized. Eive

specimens were included in this study (Table 1). With a

80

Bothalia27,l (1997)

FIGURE 4. — Photomicrographs of

meiotic chromosomes in Para-

pholis incurva. A, Spies 5349,

metaphase I with 19n. B-E,

Spies 4596: B, anaphase I with

18-18 segregation of chromo-

somes; C, anaphase I with chro-

mosome laggards; D, anaphase

I with chromatid bridge; E, te-

lophase I with four micronuclei.

Scale bars: 10 pm.

basic chromosome number of 7, two specimens were dip-

loid (n = X = 7), two aneuploid (n = 18 and n = 19)

(Figure 4A & B), and one hexaploid (n = 3x = 21). These

conflicting chromosome numbers are accentuated by

the literature consulted, where 2n = 24 (Goldblatt &

Johnson 1994), 28 (Moore 1977), 36 (Fedorov 1969;

Moore 1972; Goldblatt 1981), 38 (Fedorov 1969; Moore

1972; Goldblatt 1981) and 42 (Fedorov 1969) are found

for P. incurva. Other Parapholis species are either diploid

(2n = 14) (Fedorov 1969) or tetraploid (2n = 28) (Moore

1972). Further studies in P incurva are needed to deter-

mine the evolution of chromosome numbers.

Meiosis was relatively normal and abnormalities were only

observed in the aneuploid specimens (Figure 4). These abnor-

malities include chromosome laggards during anaphase I

(Figure 4C), a chromatid bridge during anaphase I (Figure

4D) and micronuclei during telophase I (Figure 4E).

FIGURE 5. — Photomicrographs of meiotic chromosomes in the genus Poa. A, B, P. binata, Saayman 1 17, diakinesis with 21n. C-G, P. pratensis:

C, Spies 4670, diakinesis; D, Spies 4670, metaphase I with several univalents; E, Spies 3196, anaphase 1 with laggards; F, Spies 4720, anaphase

1 with laggards; G, Spies 4720, telophase 1 with micronucleus. Scale bars: 10 pm.

Bothalia27,l (1997)

FIGURE 6. — Photomicrographs of meiotic chromo-

somes in the genus Puccinellia. A, B, P.

acroxantha: A, Spies 3126, diakinesis with

21ii; B, Spies 3134, metaphase I with various

univalents. C, P. angusta. Spies 3157a, diaki-

nesis with 7ii; D, Puccinnellia sp., Spies 3154,

diakinesis with 1\\ + IB; E, P. angusta. Spies

3157a, anaphase I with a chromosome bridge

and either a B-chromosome or fragment in

upper pole (see arrows). Scale bars: 10 pm.

The largest genus in the Poeae, Poa, consists of approxi-

mately 500 species, with three indigenous species {P. biiiata

Nees, P. bulbosa L. and P leptoclada A.Rich.) and three

naturalized species (P annua L., P pratensis L. and P. tri-

valvis L.). Four of these sptecies were studied. The P. annua

specimens were tetraploid, P. binata had one hexaploid (Fig-

ure 5A & B) and one octoploid sptecimen, whereas all the

P. bulbosa and P pratensis sftecimens were hexaploid (Figure

5C). To the best of our knowledge, this is the first report on

chromosome numbers for P. binata. The rest of our chro-

mosome number reports support the previous counts made

for P. annua, P bulbosa, P. pratensis and P. trivalvis.

Meiosis was abnormal in the P pratensis specimen

(n = 21). These abnormalities included numerous univa-

lents during metaphase I (Figure 5D), chromosome lag-

gards during anaphase I (Figure 5E & F) and micronuclei

during telophase I (Figure 5G).

There are ± 80 species in the genus Puccinellia, with

three indigenous species, P. acroxantha C.A.Sm. &

C.E.Hubb., P. angusta (Nees) C.A.Sm. & C.E.Hubb. and

P fasciculata (Torr.) C.Bicknell; and a naturalized species

P. distans (L.) Pari. The Puccinellia acroxantha specimens

were hexaploid (Figure 6A) and the P angusta specimen

was diploid (Figure 6C & D). These are thought to be the

first reports on chromosome numbers for both species. A

metaphase I cell with many univalents was observed (Fig-

ure 6B) in one P. acroxantha specimen. One B-chromo-

some was present in some cells of P angusta (Figure 6D)

and a chromosome bridge was observed in one cell of

this species (Figure 6E).

1 ^ *

B

D

i

10^

82

Bothalia27,l (1997)

The genus Sphenopus contains two species, with only

S. divaricatus being naturalized in South Africa. Both S.

divaricatus specimens were diploid with n = x = 7. This

finding contradicts the previous numbers based on six as

listed in Moore 1972 & 1974 and Goldblatt 1981, and

supports the 2n = 28 listed in Moore 1972 & 1974.

Four of the 23 species of the genus Vidpia are natu-

ralized in South Africa, i.e. V. bwmoides (L.) Gray, V.

fasciculata (Forssk.) Samp., V. muralis (Kunth) Nees and

V myuros (L.) C.C.Gmel, One Vidpia bwmoides specimen

was diploid, the other was hexaploid (Figure 7 A), V. fas-

ciculata was tetraploid and V. muralis (Figure 7C) and V.

myuros (Figure 7D) were hexaploid. These numbers con-

firm the previous reported findings (Fedorov 1969; Orn-

duff 1967-1969; Moore 1970, 1971, 1972, 1974, 1977;

Goldblatt 1981, 1983, 1985, 1988; Goldblatt & Johnson

1990, 1991, 1994). Meiosis was normal in most cells with

occasional chromosome laggards during anaphase I (Fig-

ure 7B & E).

This study confirms a basic chromosome number of

seven for the tribe Poeae. Further studies are, however,

necessary to determine the origin of the other basic chro-

mosome numbers present in the tribe. In this regard Briza

(x- 5 & 7), Colpodium (x = 7 and an aneuploid reduction

series exists, or x = 2), Hainardia (x = 7 or 13), Para-

pliolis (x = 7, 18 or 19) and Sphenopus (x = 6 or 7) should

receive special attention.

ACKNOWLEDGEMENTS

The University of the Orange Free State and the Foun-

dation for Research and Development are thanked for fi-

nancial assistance during this study. The National

Botanical Institute is thanked for providing some of the

meiotic material used during this study.

REFERENCES

CLAYTON, W.D. & RENVOIZE, S.A. 1986. Genera graminum—

grasses of the world. Kew Bulletin Additioml Series 13: 1-389.

DE WET, J.M.J. 1958. Additional chromosome numbers in Transvaal

grasses. Cytologia 23: 113-118.

DE WET, J.M.J. & ANDERSON, L.J. 1956. Chromosome numbers in

Transvaal grasses. Cytologia 21: 1-10.

EDWARDS, D. & LEISTNER, O. A. 1 97 1 . A degree reference system for

citing biological records in southern Africa. Mitteilungen der

Botanischen Staatssammlung, MUnchen 10: 501-509.

FEDOROV, A. A. 1969. Chromosome numbers of flowering plants. Acad-

emy of Sciences of the U.S.S.R., Leningrad.

GIBBS RUSSELL, G.E., WATSON, M., KOEKEMOER, M., SMOOK,

L., BARKER, N.P., ANDERSON, H.M. & DALLWITZ, M.J,

1990. Grasses of southern Africa. Memoirs of the Botanical Sur-

vey of South Africa No. 58.

GOLDBLATT, P 1981 . Index to plant chromosome numbers 1975-1978.

Monographs in Systematic Botany 5.

GOLDBLATT, P. 1983. Index to plant chromosome numbers 1979-1981.

Monographs in Systematic Botany 8.

GOLDBLATT, P. 1985. Index to plant chromosome numbers 1982-1983.

Monographs in Systematic Botany 13.

GOLDBLATT, P 1988. Index to plant chromosome numbers 1984—1985.

Monographs in Systematic Botany 23.

GOLDBLATT, P, & JOHNSON, D.E. 1990. Index to plant chromosome

numbers 1986-1987. Monographs in Systematic Botany 30.

GOLDBLATT, P. & JOHNSON, D.E. 1991. Index to plant chromosome

numbers 1988-1989. Monographs in Systematic Botany 40.

GOLDBLATT, P. & JOHNSON, D.E. 1994. Index to plant chromosome

numbers 1990-1991. Monographs in Systematic Botany 51.

HEDBERG, 1. & HEDBERG, O. 1977. Chromosome numbers of afro-al-

pine and afromontane angiosperms. Botaniska Notiser 130: 1-24.

HILL, H.D. 1965. Karyology of species of Bromus, Festuca and Ar-

rhenatherum (Gramineae). Bulletin of the Torrey Botanical Club

92: 192-197.

KIMBER, G. & ALONSO, L.C. 1981. The analysis of meiosis in hybrids.

HI. Tetraploid hybrids. Canadian Journal of Genetics and Cytol-

ogy 23: 235-254.

MOORE, R.J. 1970. Index to plant chromosome numbers for 1968.

Regnurn Vegetabile 68.

MOORE, R.J. 1971. Index to plant chromosome numbers for 1969.

Regnurn Vegetabile 77.

MOORE, R.J. 1972. Index to plant chromosome numbers for 1970.

Regnurn Vegetabile 84.

MOORE, R.J. 1974. Index to plant chromosome numbers for 1972.

Regnurn Vegetabile 91.

MOORE, R.J. 1977. Index to plant chromosome numbers for 1973/74.

Regnurn Vegetabile 96.

ORNDUFF, R. 1967. Index to plant chromosome numbers for 1965.

Regnurn Vegetabile 50.

ORNDUFF, R. 1968. Index to plant chromosome numbers for 1966.

Regnurn Vegetabile 55.

ORNDUFF, R. 1969. Index to plant chromosome numbers for 1967.

Regnurn Vegetabile 59.

SPIES, J.J. & DU PLESSIS, H. 1986a. Chromosome studies on African

plants. 1. Bothalia 16: 87, 88.

SPIES, J.J. & DU PLESSIS, H. 1986b. Chromosome studies on African

plants. 2. Bothalia 16: 269, 270.

SPIES, J.J. & VAN WYK, S.M.C. 1995. Cell fusion: a possible mecha-

nism for the origin of polyploidy. South African Journal of Botany

61:60-65.

SPIES, J.J., SPIES, S.K., VAN WYK, S.M.C, MALAN, A.F. & LIE-

BENBERG, E.J.L, 1996a. Cytogenetic studies of the subfamily

Pooideae (Poaceae) in South Africa. 1. The tribe Aveneae, sub-

tribe Aveninae. Botiuilia 26: 53-61 .

SPIES, J.J., SPIES, S.K., VAN WYK, S.M.C, M/U-AN, A.F. & LIE-

BENBERG, E.J.L. 1996b. Cytogenetic studies of the subfamily

Pooideae (Poaceae) in South Africa. 2. The tribe Aveneae, sub-

tribes Phalaridinae and Alopecurinae. Bothalia 26: 63-67.

Bothalia27,l: 83-89 (1997)

Comparative field performance of three different gas exchange systems

G.F. MIDGLEY*, M. VESTE**, DJ. VON WILLERT**, G.W. DAVIS*, M. STEINBERG** and L.W. POWRIE*

Keywords: ecophysiology, gas exchange measurement, photosynthesis, transpiration

ABSTRACT

We compared portable and continuously monitoring gas exchange systems under field conditions, using Protea glabra

Thunb. as a test species. The aim was to determine if the same patterns of gas exchange and ancillary parameters could be

obtained with rather different measurement systems, and whether the same interpretation and conclusions about environmental

control of gas exchange could be drawn. The following systems were compared: 1, a ‘closed’ portable IRGA manufactured by

Ll-Cor (LI-6200); 2, an ‘open’ portable porometer manufactured by Walz; and 3, a continuously monitoring minicuvette system

with temperature control facility, also manufactured by Walz.

All three systems yielded similar diurnal curves for CO2 uptake, although absolute flux values for the minicuvette system

were lower than those obtained for the portable systems. This was likely due to stem respiration and self-shading of leaves on

the shoot enclosed in the minicuvette. Differences in sampling technique between the two portable systems, primarily with

regard to changes in leaf orientation, resulted in some differences in absolute values of gas fluxes and ancillary parameters such

as leaf temperature and leaf to air vapour pressure difference. However, data from all three systems allowed similar interpreta-

tions to be made about the environmental dependencies of gas exchange patterns. It appears that each system has certain

drawbacks associated with widely varying field conditions. A combination of portable and continuous monitoring techniques

would seem to be the most powerful approach to investigating the gas exchange patterns of terrestrial plants in their natural

environment.

INTRODUCTION

In the last decade, a number of advances in the simul-

taneous measurement of water vapour and CO2 exchange

by plant organs have been made. The present availability

of several commercially produced systems provides a

healthy competitive environment which is to the benefit

of researchers and the quality of their science. However,

the different approaches to gas exchange measurement

used by gas exchange systems make them suitable for

different purposes, and may also have important implica-

tions for the interpretation of results obtained when using

them. In essence, no perfect all-purpose gas exchange sys-

tem exists (Field et al. 1989). The choice of instrument

for any particular task involves two fundamental trade-

offs— between portability and the facility for environ-

mental control, and between replication and resolution.

Field et al. (1989) suggest that a combination of instru-

ments with complementary strengths is a good solution

to this dilemma. This view assumes that different gas ex-

change systems yield similar results, but this assumption

should be tested (Reich et al. 1988). Apart from inade-

quate calibration protocol (Reich & Middendorf 1990),

concern has been expressed about the accuracy of gas flux

and leaf temperature data obtained with leaf cuvettes, and

the need for correction procedures (Rochette et al. 1990;

Idso 1992). On the other hand, Monteith (1990) suggests

that correction of these data is not necessary, as long as

correct sampling procedures are followed.

Considering that sampling approach and technique may

differ a great deal between gas exchange systems, can one

* Stress Ecology Research Programme, National Botanical Institute,

Private Bag X7, Claremont, 7735 Cape Town.

** Department of Applied Botany, University of Munster, D-4400 Munster,

Federal Republic of Germany.

MS. received: 1996-10-18.

expect results yielded by them to be directly comparable?

Only one published study (Winner et al. 1989) that we

know of has addressed this important question, by com-

paring a closed portable photosynthesis system (LI-6200,

LICOR, Lincoln, Nebraska) with an open system (ADC

LCA-2, Analytical Development Corporation, Hoddesdon,

England). Although the results of this study suggested that

the systems gave comparable gas flux values, problems

with experimental protocol prevented a conclusive result.

In this paper, we provide a direct comparison of gas

exchange data obtained by three different gas exchange

systems in a highly variable field environment. This is an

important form of data control in a field where different

research groups become more or less committed to one

make or type of instrument. Our comparison is prelimi-

nary in that we do not address subtle and complex ques-

tions of cuvette design differences between instruments.

We also attempt to draw attention to the advantages of

combining the use of different approaches to gas exchange

measurement in an ecophysiological study.

MATERIALS AND METHODS

Study site and measurement protocol

The study was carried out on the Farm Papkuilsfontein,

near Niewoudtville, Cape Province, South Africa, in an

area of natural vegetation comprising arid Fynbos and

some Karoo elements. The study site was situated near

the edge of an escarpment of the Bokkeveld Mountains

(altitude 800 m, 33°30'S 19°05'E).

For our primary comparison, we present the results of

gas exchange measurements on Protea glabra Thunb., an

evergreen, broad-leaved, sclerophyllous shrub, made on

84

Bothalia27,l (1997)

31 January 1991 (i.e. midsummer). Measurements were

made with a LICOR LI-6200 portable IRGA (LICOR,

Lincoln, Nebraska, USA), a Walz CO2/H2O portable

porometer (Walz, Effeltrich, Germany), and a continuous-

monitoring minicuvette system with temperature control

facility (Walz, Effeltrich, Germany). We also provide the

results of a comparison between only the minicuvette sys-

tem and the Walz porometer, using data collected on 29

September 1990 (spring).

In January, measurements were carried out on adult

individuals which were approximately 1.5 to 2 m tall. To

minimize disturbance to the individual being continuously

monitored by the Walz minicuvette system for other pur-

poses, we used the portable gas exchange systems to sam-

ple an individual of matching size and water status less

than 150 m away (water potentials were measured before

dawn and through the day using a pressure chamber). All

gas exchange instruments were unmodified, and were

used according to their instruction manuals. We attempted

to synchronize sampling with the portable systems as far

as possible. One leaf (that most recently fully expanded

on the shoot) on each of five shoots was measured with

each portable instrument at each sampling time. We re-

turned to measure the same leaves throughout the day,

except that a small number of leaves measured with the

Walz porometer became detached from the plant stem;

these were replaced by leaves of a comparable age and

position on an adjacent stem. Towards the end of the day

(after 16h00), permanently marked leaves sampled by the

LI-6200 became shaded; after this occurred, well-irradi-

ated leaves of a similar age and stem position were sam-

pled and removed at each sampling event. Sampling of

leaves with the LI-6200 was carried out with as little dis-

turbance to the natural leaf angle as possible, although

some disturbance was usually unavoidable. With the Walz

porometer, the upper leaf surface was turned to face the

sun after being enclosed in the cuvette, this being neces-

sary to prevent shading of the leaf by the cuvette lid.

In September 1990, the Walz porometer was compared

only with the minicuvette system. An identical procedure

was followed, except that three leaves were sampled per

sampling period with the Walz porometer, from a smaller

shrub situated not more than 50 m from the continuously

sampled individual.

Leaf areas were measured with a LI-3000 belt system,

or a digitized CAD system (Summa Sketch II, programme

from the Department of Plant Physiology, University of

Wien, Austria).

Instrumentation

Ll-Cor 6200 (Ll-Cor Inc., Lincoln, Nebraska, USA)

We used this battery-powered instrument in its normal

configuration, i.e. a closed system (Welles 1986). After

the leaf is placed in the cuvette, air circulates between the

cuvette and the gas analyser, and the CO2 exchange rate

is computed from the rate of change of CO2 concentration

due to net CO2 uptake or loss by the leaf. Air vapour

pressure can be held constant by manually adjusting a

valve which allows a portion of the circulating air to pass

through a magnesium perchlorate dessicant column — in

this way the effect of leaf transpiration on the vapour pres-

sure of the enclosed air volume can be countered. Leaf

temperature is measured by a chromel-constantan thermo-

couple which makes contact with the underside of the leaf

when the hinged lid of the cuvette is closed. Air tempera-

ture in the cuvette is measured by a shielded thermistor.

Relative humidity is measured by a capacitance sensor

(Vaisala Humicap), which is situated beneath the radiation

shield in the cuvette. The gas analyser (LI-6250) is a non-

dispersive, infrared type which is tuned to the 4.26 mi-

crometer band, providing rejection of IR absorption by

gases other than CO2. The analyser uses as a reference

gas, a closed loop of air that is continuously scrubbed of

C02- Any drift in this zero reference was checked roughly

every two hours during the field work, by switching the

measurement air loop through a soda lime scrubber, with-

out a leaf in the sample cuvette. We used a standard 0.251

chamber (LI-6000- 13), which is constructed of polycar-

bonate, and has a teflon-coated inner surface to minimize

adsorption and desorption effects. The cuvette contains a

small fan which minimizes boundary layer resistance. In-

cident photosynthetic photon flux density is measured by

a LI-190S-1 quantum sensor, which is mounted parallel

with the sampled leaf surface.

Walz CO2/H2O porometer (Walz, Effeltrich, Germany)

This instrument is best operated with the aid of gen-

erator-supplied power. A 12V motor car battery may be

used under field conditions, but this results in a poor IRGA

temperature stabilization, and subsequent drifts in the CO2

zero point. The instrument is normally configured as an

open system (Schulze et al. 1982). Gas exchange rate is

calculated from the difference in concentration between a

reference gas line which samples ambient air, and a sam-

ple gas line which is passed through a cuvette containing

the sampled leaf. Water vapour and CO2 concentrations

are measured by a BINOS I differential infrared gas an-

alyser (Leybold Heraeus, Hanau, Germany). The zero

point of the H2O and CO2 of the BINOS is recorded after

every five measurements for later calculation correction.

This is carried out by making a measurement in the normal

way, but without a leaf in the cuvette. The flowrate in the

measurement line is controlled by a flowmeter (Tylan,

Carson, California, USA).

The sample cuvette is cylindrical (inner diameter 42

mm, height 130 mm) and has a nickel-plated inner surface.

A hinged lid covered by polyethylene foil is used to seal

the sampled leaf at the top of the cuvette. The cuvette has

a circular radiation shield, and a fan ventilates the space

between this and the cuvette to maximize heat transfer.

Leaf temperature is measured by a chromel-alumel ther-

mocouple which presses on the underside of the leaf when

it is enclosed in the cuvette. Cuvette air temperature is

measured by a thermistor. Cuvette humidity is measured

by a capacitance sensor (Vaisala Humicap). The cuvette

contains a small fan which minimizes boundary layer re-

sistance. Incident photosynthetic photon flux density is

measured by a LI- 1 90S quantum sensor, which is mounted

parallel with the sampled leaf surface.

Bothalia 27,1 (1997)

85

Walz minicuvette system (Walz, Effeltrich, Germany)

This system can be operated in the field only with the

aid of power supplied by at least a 0.6 kW generator. The

instrument is configured as an open system, but with a

continuous zero reference, obviating the requirement for

removing the leaf from the sample cuvette to check the

IRGA zero. The humidity of air in the sample and refer-

ence paths may be manipulated by a dewpoint controller.

Differences in H2O and CO2 concentrations between

measurement and reference paths are measured by a dif-

ferential infrared gas analyser (BINOS I, Leybold

Heraeus, Hanau, Germany). Two dewpoint mirrors (MTS

MKl, Walz, Effeltrich, Germany) are mounted in the

flowpath of the measurement gas, one measuring the dew-

point of air entering the cuvette, and the other the exiting

air. This allows calculation of air vapour pressure and tran-

spiration rate which is independent of the reading pro-

vided by the BINOS. This is especially important when

high daytime transpiration rates exceed the range of the

BINOS water vapour channel.

The minicuvette (GK 022) consists of two parts: an

environmental control system of mainly nickel construc-

tion mounted inside a polyethylene shield, and a plexiglass

leaf chamber. Chamber air temperature is controlled by

Peltier elements which are thermally connected to a heat

sink ventilated by a small fan. Cuvette air temperature

can be set to track that of ambient air (measured by a

ventilated PTIOO resistance temperature sensor), or can

be set to maintain a user-defined constant temperature.

Under field conditions the former option is most com-

monly used. Chamber vapour pressure deficit can also be

controlled, or set to track that of the ambient air. The

instrument setup can utilize more than one sample cuvette.

The cuvette design allows a whole shoot of the target

plant to be sampled in its natural position (different cham-

ber types can be constructed which provide great flexibil-

ity in sampling). Leaf temperature is measured by a

nickel-chromel thermocouple which is pressed to the un-

derside of a representative leaf. Chamber air temperature

is measured by a radiation-shielded thermistor. Photosyn-

thetic photon flux density is measured by a LI- 1 90S quan-

tum sensor. A data logger stores data from relevant

channels, and controls the timing of the IRGA zeroing

sequence. The data can be transferred to a personal com-

puter for further computation. The IRGA, pumps and data

logging facilities are best mounted in a medium-sized ve-

hicle (such as a minibus) for mobility, and to alleviate the

harsh conditions often encountered in the field.

Calculation of gas fluxes and conductances

All gas exchange parameters were calculated after Von

Caemmerer & Farquhar (1981) for all three measurement

systems. All fluxes are expressed on a total leaf area basis

(i.e. the total of the upper plus lower leaf surfaces), as

leaves are amphistomatous in this species.

The LI-6200 and the two Walz systems differ slightly

in their approach to calculating leaf conductance to water

vapour. Tdie LI-6200 software computes stomatal conduc-

tance (gs) from leaf conductance to water vapour (gH2o)

by correcting for leaf boundary layer conductance (gb),

according to the equation

1/gs = l/gH20 - 1/gb

The boundary layer conductance value should be experi-

mentally verified for different leaf shapes and sizes, and

may vary according to leaf position in the cuvette. We

used a nominal figure for g^ of 1.7 mol m-2 s->, which

was obtained by using a wet filter paper replica of a sam-

pled leaf in a standard position in the cuvette. The Walz

systems compute total leaf conductance to water vapour

(gH2o)> and do not derive stomatal conductance.

Converting gs computed by the LI-6200 to gH20

duced the conductance value by roughly 0.6% per 10

mmol m-2 s-* (i.e. a gs of 100 mmol m-2 s-' is equal to

gH20 of 94 mmol m-2 s-2), which, within the conductance

range of the species used in this study, is a trivial correction

in relation to other possible sources of error. Therefore, in

this paper we treat gs derived by the LI-6200 as equivalent

to gH20 from the Walz instmments, as would be the situation

when comparing separately published values.

RESULTS

Carbon dioxide exchange

The diurnal pattern of CO2 exchange (Figure lA)

yielded hy the three instruments was qualitatively similar,

with a clear mid-morning peak, followed by a rapid de-

crease (less rapid for the LI-6200) towards midday, and a

steady but less marked decrease towards the evening. The

daily maximum CO2 uptake rate was recorded during the

same period for all systems; maximum rate yielded by the

minicuvette system (3 pmol m-2 $-*) was lower than the

mean recorded by the LI-6200 (3.8 pmol m-2 s'^) and the

Walz porometer (4.9 pmol m-2 s-^- Data variability for

the portable systems (i.e. the coefficient of variation for

each sample period mean expressed as a percentage), was

considerably greater for the LI-6200 (typically 37%) than

for the Walz (typically 27%) during the light period. Nei-

ther portable system gave a realistic value for respiration

rate at low light levels in the early morning, but the LI-

6200 measured a mean respiration rate comparable to that

given by the minicuvette system at the end of the day.

Integrated CO2 uptake for the light period on the Janu-

ary sampling date for the Walz porometer was 120 mmol

CO2 m-2, double that of the minicuvette system (57 mmol

CO2 m-2), with the value for the LI-6200 between these

(99 mmol CO2 m-2).

Water vapour exchange and stomatal conductance

All three systems gave qualitatively matching patterns

for transpiration, and comparable maximum values (Fig-

ure IB). However, peak transpiration was measured ear-

liest by the Walz porometer (around lOhOO) later by the

minicuvette system (12h00), and latest in the day by the

LI-6200 (I3h00). Data variability for the two portable sys-

tems was similar (coefficient of variation around 30% of the

mean for each sample time during the light period). It is

likely that large transpirational water loss rates by leaves in

the minicuvette system resulted in condensation in the meas-

86

Bothalia 27,1 (1997)

FIGURE 1. — Diurnal trends, using Protea glabra as test species, of

parameters measured by three different gas exchange systems on

31 January 1991. A, COj flux; B, transpiration rate; C, cuvette

temperature and dewpoint temperatures of air entering and exiting

sample cuvette on Walz minicuvette system; D, stomatal conduc-

tance for Walz and LICOR portable systems. A, B, D: Walz

porometer, •; L1-62CX), ■; C, Walz minicuvette system, — ; dew-

point out, ; dewpoint in — — - . Vertical bars represent standard

deviations.

uring gas flowpath in the region of the cuvette. This is

reflected in the parallel changes in ambient cuvette tem-

perature and the dewpoint temperature of exiting air be-

fore 12h00 in the minicuvette system (Figure 1C), and it

is unlikely that the transpiration values as measured by

this instrument under these conditions are biologically

meaningful.

The pattern of stomatal conductance as measured by

the two portable systems, as well as the absolute values

and data variability, agree well for most of the day, except

for a marked divergence between the two systems before

09h00 (Figure ID). The transpiration rates measured and

conductances calculated by the LI-6200 for the last sam-

pling period were highly variable, and sometimes nega-

tive, and are not given.

Physical parameters

Each measurement system was applied in a slightly

different way in the field, and this led to some differences

in measured physical parameters such as PPFD (Figure

2A). The Walz porometer measured higher values from

earlier in the day than the LI-6200, due to the need to

orientate the enclosed leaf towards the sun (unfortunately

the Walz porometer used by us did not have the facility

to provide readings greater than 2000 |imol m-2 s’l). Meas-

ured leaf temperature in the Walz minicuvette system was

much lower through the day than that measured by either

portable system, and the Walz porometer yielded higher

leaf temperatures than the LI-6200 (Figure 2B). These tem-

perature differences led also to different leaf to air vapour

pressure differences (AW) in each system (Figure 2C).

Secondary comparison

The comparison of transpiration rate measured by the

Walz porometer and minicuvette system in September

1990 (Figure 3 A) show better agreement in qualitative pat-

tern, and give the expected lower maximum (due to lower

AW) measured by the minicuvette system. Also, conductance

patterns for these two systems were quantitatively and

qualitatively comparable on that day (Figure 3B).

Summary relationships

Linear regressions fitted to plots of CO2 exchange rate

against stomatal conductance (Figure 4A, B) were signifi-

cantly positive. The slopes of this relationship compared

well with the Walz systems in September 1990, but the

LI-6200 gave a somewhat reduced slope value than the

Walz porometer in January 1991, and the lowest correla-

tion coefficient for the regression.

Linear regressions fitted to plots of stomatal conduc-

tance against leaf to air vapour pressure difference were

significantly negative (Figure 5A, B), and the slope of this

relationship given by the portable instruments was com-

parable on both dates. The Walz minicuvette system gave

by far the highest correlation coefficient for this regres-

sion, and the slope of the relationship was slightly steeper

than for the portable machines.

Bothalia27,l (1997)

87

FIGURE 2. — Diurnal trends, using Protea glabra as test species, of

physical parameters measured by three different gas exchange

systems on 31 January 1991. A, photosynthetic photon flux den-

sity; B, leaf temperature; C, leaf to air vapour pressure difference.

Walz porometer, •; LI-6200, ■; Walz minicuvette system, — .

Vertical bars represent standard deviations.

DISCUSSION

Our simultaneous use of three different gas exchange

systems highlighted some heartening similarities in their

data outputs. In contrast to the study of Winner et al.

(1989), the test species used in this study had a clear di-

urnal pattern of water vapour and CO2 exchange which

was revealed by all three systems. However, the data also

suggest some important differences between systems

which may to a large extent be minimized by improve-

ment in system design or standardization of sampling

technique. The biggest difference was found between the

portable machines and the Walz minicuvette system, in

terms of absolute values of CO2 and water vapour flux,

and subsequently calculated conductances. This may be

explained by the inclusion of a whole shoot in the mini-

cuvette system, an approach which had three obvious im-

plications: 1, the gas exchange of leaves with a range of

ages was sampled. It has been established that young

leaves of Protea species tend to have lower photosynthetic

rates than developing and mature leaves (Von Willert et

al. 1989; Van der Heyden & Lewis 1990). The contribu-

tion of all leaves in the cuvette to net CO2 and water

vapour exchange is equally weighted as these fluxes are

calculated on a leaf area basis. This would lead to an

underestimation of these fluxes relative to those measured

by the portable systems, which were used to sample only

mature leaves; 2, leaves on the shoot were self-shaded to

a greater or lesser extent or obliquely positioned relative

to the sun’s rays depending on the position of the sun (and

more closely representing the real situation in the field);

3, stem material enclosed in the Walz minicuvette sample

chamber would have contributed respired CO2.

Because sampling with the LI-6200 system involved

minimum disturbance to leaf orientation, while the Walz

cuvette was aimed directly at the sun for some time before

taking the measurement, the expectation was that Walz

values of CO2 flux during the light period would be

greater than LI-6200 values. This proved to be the case,

especially at midday, when leaf orientation led to the

greatest difference in sampled leaf orientation between the

two portable systems. Which is the correct way to sample,

or indeed, is there a correct way? It can be argued that

enclosing a leaf in any sampling chamber constitutes a

disturbance to the leaf environment. This is especially true

of gas exchange cuvettes, which use turbulent airflow gen-

erated by an internal fan to reduce the leaf boundary layer

resistance. If sampling occurs rapidly enough, it is as-

sumed that stomata do not have time to respond to this

disruption, but it is likely that this assumption is violated

FIGURE 3. — Diurnal trends, using Protea glabra as test species, of

parameters measured by two different gas exchange systems on

29 September 1990. A, transpiration rate; B, stomatal conduc-

tance. Walz porometer, O ; Walz minicuvette system, — . Vertical

bars represent standard deviations.

Bothalia27,l (1997)

FIGURE 4. — The relationship for Protea glabra between stomatal con-

ductance and CO2 flux as measured by three different gas ex-

change systems. A, 31 January 1991;B,29 September 1990. Walz

porometer, O; LI-6200, ■; Walz minicuvette system, •. Statistics

are as follows: A, Ll-6200, ^ 0.65, df = 5 1 , Y = 0.05X -1- 0.44.

Walz porometer, r^ = 0.80, df = 46, Y = 0.08X -t 0.15. B, Walz

porometer, r^ = 0.88, df = 45, Y = 0.08X - 0.22. Walz minicuvette

system, = 0.82, df = 90, Y = 0.08X 0.24.

under certain circumstances, and is species-specific. The

energy balance of a leaf is also altered after being enclosed

in a cuvette, but this is also assumed to have limited im-

mediate effect on leaf function during sampling. McDer-

mitt (1990) provides a brief summary of important

considerations in this regard. Changes in leaf orientation

during sampling constitute a disruption to the function of

the leaf which may have immediate or delayed impacts

on leaf energy balance, stomatal movements and the ve-

locity of leaf photo- and biochemical reactions. Leaf pho-

tochemical reactions may be rapid in response to changes

in light energy, but stomatal responses tend to be rather

slower (Gross & Chabot 1979). If leaf orientation is to be

altered for a sample, it seems prudent to establish first the

rapidity of stomatal and biochemical changes in the spe-

cies under study.

By changing sampled leaf orientation it is possible to

control, to some extent, the PPFD incident on a sampled

leaf surface using a portable system. This can be a useful

technique, for example, for standardizing light conditions

for different samples. In this study, the effect of stand-

ardizing light conditions (i.e. by aiming the leaf directly

at the sun), using the Walz porometer, appeared to result

in less noisy data, as can be seen in the correlation coef-

ficients for the relationship between gH20 ^^d A for the

portable instruments. This method could shortcut more

comprehensive but time-consuming sampling strategies

for portable systems such as stratifying sampling accord-

ing to leaf orientation or angle classes, but depends on

knowledge of the rapidity of the physiological response.

Gas flux values obtained with a continuous monitoring

system, such as the minicuvette system, may be the most

accurate means of estimating diurnal carbon and water

budgets, but the technique is limited by low potential for

replication. Certainly, this type of system offers a level of

data resolution which may improve interpretation of the

effects of changing environmental conditions on gas ex-

change processes. This can be seen clearly in the relation-

ship between AW and gH20> which reveals a remarkably

close relationship between these parameters that is masked

by considerable variability in the data from the portable

systems.

Technical limitations

Each system we used revealed shortcomings in the

widely varying field environment. For the minicuvette

system, the main problem seemed to be adsorption and

desorption processes, especially under conditions of high

air dew point temperature (i.e. in the morning on the Janu-

ary sampling date). The problem was not apparent on the

September sampling date, when relative humidity was

relatively low during the morning.

FIGURE 5. — The relationship for Protea glabra between leaf to air

vapour pressure difference and stomatal conductance as meas-

ured by three different gas exchange systems. A, 31 January 1991;

B, 29 September 1990. Walz porometer, o; LI-6200, ■; Walz

minicuvette system, •. Statistics are as follows: A, LI-6200, r^ =

0.47, df = 53, Y = -0.88X -1- 62.43. Walz porometer, r^ = 0.32, df

= 32, Y = - 0.83X -I- 68.48. B, Walz porometer, r^ = 0.33, df =

36, Y = - 0.79X -I- 52.32. Walz minicuvette system, r^ = 0.93, df

= 68, Y = -1.12X-h 39.38.

Bothalia27,l (1997)

89

Accurate measurement of water vapour concentration

in the sample cuvette appeared to be a major problem for

the LI-6200 early and late in the day, when humidity was

high. The LI-6200 relies on an accurate measurement of

this parameter for calculating transpiration rate and hence

stomatal conduetance. It is well documented that the ac-

curacy of the Vaisala Humicap sensor is strongly affected

above about 80% relative humidity (McDermitt 1990),

and that fairly small errors in humidity measurement can

lead to large errors in conductance when ambient humidity

is either very low or very high (Welles 1986; McDermitt

1990) ; this may explain the deviation of conductance val-

ues between the portable systems early in the morning

and at the end of the light period. However, the Vaisala

Humicap is relatively robust within the effective range

(10%-80%, McDermitt 1990), as is clear from the simi-

larity between measurements of water vapour flux be-

tween this system and the Walz porometer through the

day. The combination CO2/H2O IRGA used by the Walz

porometer appeared to be a superior system under high

humidity conditions.

Sampling with the portable systems was plagued pri-

marily by cuvette heating problems, which were of two

types: firstly, enclosed leaves heated up rapidly during

measurement, and secondly, the euvettes themselves

heated up during a sampling run. Apart from direct effects

on leaf function, this affects compound parameters such

as leaf to air vapour pressure difference, a parameter

which is thought to be of considerable importance in sto-

matal movements (Aphalo & Jarvis 1991), possibly

through its effect on transpiration rate (Mott & Parkhurst

1991) . Schulze et al. (1982) suggested shading the head

of portable porometers between measurements to avoid

heating, yet Tyree & Wilmot ( 1 990) showed how a shaded

LICOR Li- 1600 porometer cuvette rapidly reduced the

temperature of irradiated sugar maple leaves, leading to

considerable modification of water vapour flux and cal-

culated conductance. Recently developed portable systems

which use Peltier cooling systems to allow chamber tem-

perature to track ambient temperature may remove this

limitation. This is a positive step in reducing the intru-

siveness of sampling with a portable system.

CONCLUSIONS

All sampling techniques used by us yielded equivalent

results, and therefore appear to be directly comparable.

However, we urge users of portable systems to describe

the procedure followed when clamping cuvettes onto

leaves; this will contribute to more effective assessment

and cross-comparison of data.

In general, matching interpretations about complex en-

vironmental and stomatal determinants of gas exchange

patterns could be made using the data obtained from all

three systems, which can be seen clearly in the relation-

ships obtained between stomatal conductance and daytime

CO2 fluxes, and between AW and stomatal conductanee.

Therefore, we eoncur with the suggestion of Field et al.

(1989: 239) that the combination of a continuous moni-

toring technique with a well designed stratified sampling

strategy using a portable system, may be the most pow-

erful way to investigate gas exchange patterns in the field.

It remains to be seen whether more recently developed

portable systems with peltier-eooled cuvettes will increase

the effectiveness of clamp-on gas exchange systems.

ACKNOWLEDGEMENTS

The support of the Deutscher Akademischer Austausch-

dienst and the Deutsche Forschungsgemeinschaft is ac-

knowledged (MV, DJvW, MS). Volkswagen of SA Ltd

(Uitenhage, South Africa) supplied two vehicles. We thank

Willem van Wyk, owner of Papkuilsfontein at Nie-

woudtville, for permission to work on his property, and

for his helpfulness and hospitality during our stay.

REFERENCES

APHALO, P.J. & JARVIS, P.G. 1991. Do stomata respond to relative

humidity? P/ant, Cell and Environment 14: 127-132.

FELD, C.B., BALL, J.T. & BERRY, J.A. 1989. Photosynthesis: princi-

ples and field techniques. In R.W. Pearcy, J.R. Ehleringer, H.A.

Mooney & P.W. Rundel, Plant physiological ecology, field meth-

ods and instrumentation: 209-253. Chapman & Hall, London.

GROSS, L.J. & CHABOT, B E. 1979. Time course of photosynthetic

responses to changes in incident light energy. Plant Physiology

63: 1033-1038.

IDSO, S.B. 1992. Net photosynthesis: corrections required of leaf chamber

measurements. Agriculniral and Forest Meteorology 58: 35-42.

MCDERMITT, D.K. 1990. Sources of error in the estimation of stomatal

conductance and transpiration from porometer data. Hortscience

25: 1538-1548.

MONTEITH, J.L. 1990. Porometry and baseline analysis: the case for

compatability. Agricultural and Forest Meteorology 49: 1 55-167.

MOTT, K.A. & PARKHURST, D.F. 1991. Stomatal responses to humid-

ity in air and helox. Plant, Cell and Environment 14: 509-515.

REICH, P.B. & MIDDENDORF, L. 1990. Sources of variation in

porometry data. Plant, Cell and Environment 13: 879.

REICH, P.B., WALTERS, M B. & TABONE, T.J. 1988. Variation in

response of five identical steady-state porometers. Plant, Cell and

Environment 1 1 : 785, 786.

ROCHETTE. P, PATTEY, E., DESJARDINS. R.L. & DWYER, L.M.

1990. Adjustment of leaf temperature measurements in LI-COR

6200 assimilation chamber using energy balance calculations.

Agricultural and Forest Meteorology 53: 149-156.

SCHULZE, E.-D., HALL, A.E., LANGE, O.L. & WALZ. H. 1982. A

portable steady-state porometer for measuring the carbon dioxide

and water vapour exchange of leaves under natural conditions.

Oecologia 53: 141-145.

TYREE, M.T. & WILMOT, TR. 1990. Errors in the calculation of evapo-

ration and leaf conductance in steady-state porometry: the impor-

tance of accurate measurement of leaf temperature. Canadian

Journal of Forest Research 20: 103 1-1035.

VAN DER HEYDEN, F. & LEWIS. O.A.M. 1990. Environmental control

of photosynthetic gas exchange characteristics of lynbos species

representing three growth forms. South African Journal of Botany

56; 654-658.

VON CAEMMERER, S. & FARQUHAR, G.D. 1981. Some relation-

ships between the biochemistiy of photosynthesis and the gas

exchange of leaves under natural conditions. Planta 153:

376-387.

VON WILLERT, D.J., HERPPICH, M. & MILLER, J.M. 1989. Photo-

synthetic characteristics and leaf water relations of mountain fyn-

bos vegetation in the Cedarberg area (South Africa). South African

Journal of Botany 55; 288-298.

WELLES, J. 1986. A portable photosynthesis system. In W.G. Gensler,

Advanced agricultural instrumentation, design and use: 21-38.

Nijhoff, Dordrecht.

WINNER, W.E., PARKINSON, K.J., MCDERMITT, D.K., ROEMER,

S.C, & CLINE, S.P. 1989. Measurement and comparison of gas

exchange of Pinus halepensis and Flaveria brownii using com-

mercial photosynthesis systems. Photosynthetica 23: 585-602.

Bothalia 27,1: 91-96(1997)

OBITUARY

LESLIE CHARLES LEACH (1909-1996)

Leslie Charles Leach (Figure 1), known to friends as

Larry, was bom at Southend-on-Sea in Essex, England on

18 November 1909. After leaving school he passed the

entrance examination of the British Army Technical

School where he completed a course in electrical technol-

ogy. Serving as an Army technician, he gained wide ex-

perience in electrical engineering. He applied successfully

for a post in Southern Rhodesia [now Zimbabwe] and

arrived in Salisbury [now Harare] in January 1938. Here

he met and married Ann (often called Nan) in 1939. In

1944 he established his own business, following the ac-

quisition of premises on Sinoia Street in Harare. Concern-

ing himself mainly with the supply of electrical equipment

for vehicles and aircraft, he traded productively under the

name of L.C. Leach.

Larry’s interest in succulents started in 1950 after he

had purchased a 10 acre plot near Harare, which he named

‘Farview’. Initially he cultivated a variety of plants, but

gradually his attention centred on succulents. Larry en-

deavoured to learn more about succulents than could be

gleaned from available literature or expertise. Soon he be-

gan to concentrate on the systematic collection of Stape-

lieae, Euphorbieae and species of Aloe.

In 1956, after selling his business, he visited the eastern

Transvaal [now Mpumalanga] and southern Mozambique

to compare Euphorbia confinalis R.A.Dyer with closely

related plants that he had studied in Zimbabwe. At that

time he purchased a 25 morgen plot south of Nelspmit.

Satisfied that the Zimbabwean plants were distinct from

E. confinalis, he soon embarked on several, mainly self-

financed expeditions to investigate and collect the species

of the Flora zambesiaca area, and to see as many succu-

lent euphorbias as possible at their type localities. After

the death of his wife in 1958, following a long illness, he

intensified his research efforts. The journeys encompassed

most of southern Africa, including Zambia, Angola,

Mozambique, South Africa and Namibia, as well as Tan-

zania and Kenya. He was at times accompanied by R.D.A.

Bayliss, I.C. Cannell, H. Hall, A.R.A. Noel, A. Percy-

Lancaster, R.O.B. Rutherford-Smith, E.A.C.L.E. Schelpe,

J.A. Whellan, G. Williamson, and H. Wild. His extensive

expeditions are summarized in Codd (1966), Schelpe

(1969) and Gomes e Sousa (1971).

The Angola trip in 1967 was partly financed by grants

from the CSIR. On this occasion he collected several new

records of ferns, written up by Schelpe (1968). These col-

lecting trips were often fraught with difficulties and dan-

gers. Two attempts to reach Goa Island, northern

Mozambique, to see E. angularis, failed because of an

allergy to mosquito bites, which he had developed on a

previous excursion. However, at the third attempt, with

the assistance of the Portuguese Naval Authorities, he

landed on Goa Island only to find that E. angularis was

neither flowering nor fruiting. Three years later, he was

back on Goa Island, collecting flowers and fruits. In 1962

Larry was accompanied by Schelpe to collect ferns and

Euphorbia spp. in northern Mozambique. However, a

near-fatal viral dysentery grounded the expedition for a

week in Blantyre.

From 1972 to 1981, Larry worked as Honorary Bota-

nist on the staff of Zimbabwe’s National Herbarium. Here

he described himself as ‘probably Rhodesia’s only unpaid

civil servant’. Larry persuaded the Aloe, Cactus and Suc-

culent Society of Zimbabwe to publish a taxonomic series

supplementary to Excelsa. Four volumes, containing,

among others, monographs of the Stapelieae taxa Orbea,

Stapelia, Huemia and Tridentea, appeared between 1978

and 1988 with Larry as the sole author. In December 1981

he emigrated to South Africa and settled first at the Bo-

tanical Research Institute in Pretoria (Figure 2) and was

then employed at the National Botanic Garden at Worces-

ter from 1982 to 1989. Until the time of his death he was

employed as Honorary Research Fellow in the Depart-

ment of Botany of the University of the North. Larry

passed away on 18 July 1996 at the age of 86. Sadly he

could not complete the treatment of the succulent Euphor-

biaceae for Flora zambesiaca which would have culmi-

FIGURE 1.— Leslie (Larry) Charles Leach (1909-1996).

92

Bothalia 27,1 (1997)

FIGURE 2. — Larry Leach at his desk at the Botanical Research Institute,

Pretoria, December 1981 . Photo.: courtesy of The Pretoria News.

nated forty years of immensely dedicated research on

Euphorbieae and Stapelieae.

In later years Larry spent ‘far too much’ time on no-

menclatural disputes. The identity of the controversial

name, Euphorbia candelabrum is such an example (Carter

1982, 1985, 1988; editor’s notes in The Euphorbia Journal

4: 4 (1987); Leach 1974c, 1981a, 1985b, 1986f, 1992a).

Lairy presented convincing arguments that E. candela-

brum is the correct name for an arborescent species in

Angola (letters to the editor of The Euphorbia Journal

1987). He also fought against the apparently misapplied

lectotypification of Stapelia variegata (Leach 198 Id).

Convinced that the International Code of Botanical No-

menclature could be improved, a number of proposals for

amendation of parts of the Code were unsuccessfully sub-

mitted.

Larry was a prolific collector, often in inaccessible lo-

calities in the Flora zambesiaca and Flora of southern

Africa regions with ± 10 000 personal accessions being

taken up in various southern African and European her-

baria, although current collecting numbers are in the re-

gion of 18 000 through the inclusion of specimens sent

to him for study.

Larry discovered and described four genera and 150

species and infraspecific taxa in the Euphorbieae, Stape-

lieae and the genus Aloe. Yet his most valuable contri-

FIGURE 3. — Prof. J.N. Eloff presenting the Golden Cactus Award of the International Organization for Succulent Plant Study to Larry Leach in

1990. Photo.; courtesy of The Pretoria News.

Bothalia27,; '997)

93

bution probably was the tracking down of imperfectly

known species to their type localities and establishing

their correct identity. His wide interest in plants resulted

in his being commemorated in the following taxa: Aloe

leachii Reynolds, Cheilanthes leachii (Schelpe) Schelpe,

Crassula leachii R.Fern., Dombeya leachii Wild, Echid-

nopsis leachii Lavranos, Eulophia leachii Greatrex ex

A.V.Hall, Huernia leachii Lavranos, x Orbeostemon

leachii RV.Heath, Leachia Plowes, Leachiella Plowes

and Larryleachea Plowes.

In 1990 Leach was the first recipient in Africa of the

Golden Cactus Award of the International Organization

for Succulent Plant Study based in Zurich (Figure 3). He

was also honoured with the Harry Bolus Medal by the

Botanical Society of South Africa in 1968 (Schelpe 1969),

the Gold Medal of the Rhodesian Scientific Association

in 1977 (Kimberley 1977), the Certificate of Merit of the

South African Association of Botanists in 1981 and the

Allen Dyer medal of the Succulent Society of Southern

Africa (Anon. 1994).

Larry was elected a Fellow of the Aloe and Succulent

Society of Zimbabwe in 1975 and the Cactus and Succu-

lent Society of America in 1983 (Mitich 1983). He was

a vice president of the Rhodesian Aloe & Succulent So-

ciety and the African Succulent Plant Society of England.

Various tributes to Larry Leach appeared in Gomes e

Sousa (1971), Schelpe (1969), Kimberley (1988), Mitich

(1983) and Downs (1996).

South African botany has profited immensely from the

work of highly motivated, well-versed amateurs. Larry

Leach was one of the greatest among them.

ACKNOWLEDGEMENTS

I am grateful to Dr H.F. Glen (NBI) for providing some

information and pictures.

REFERENCES

ANON. 1994. Allen Dyer Award. Aloe 31 : 79.

CARTER, S. 1982. New succulent spiny euphorbias from East Africa.

Hooker's leones Plmtarum 39, 3: 1-1 17.

CARTER, S. 1985. Euphorbia candelabrum, a question of validity

(Euphorbiaceae). Taxon 34: 699-701.

CARTER, S. 1988. Euphorbia candelabrum: a controversial name. The

Euphorbia Journal 5: 105-107.

CODD, L.E. 1966. Contemporary botanists and botanical collectors. 2.

Leslie Charles Leach. Bothalia 8. Suppl. 1: 61-64.

DOWNS, PE. 1996. Larry Leach 1909-1996. British Cactus and Succu-

lent Journal 14: 146, 147.

GOMES E SOUSA, A. 1971. Exploradores e naturalistas da Flora de

Mogambique. Boletim da Sociedade de Estudos de Mogambique

40: 99-115.

KIMBERLEY, M.J. 1977. Gold Medal Award [Including excerpts of the

Financial Gazette 14-10-1977; Rhodesia Herald 14-10-1977].

Quarterly Newsletter No. 36. of the Aloe, Cactus and Succulent

Society of Rhodesia.

KIMBERLEY, M.J. 1988. Leslie Charles Leach: Euphorbiarum botani-

cus insignis. The Euphorbia Journal 5: 4, 5.

MITICH, L.W. 1983. Leach, Barthlott and Cole are named fellows in

C.S.S.A. Cactus & Succulent Journal (US) 55: 250, 251.

SCHELPE, E.A.C.L.E. 1968. Pteridophyta collected in Angola by Messrs

Leach and Cannell. Boletim da Sociedade Broteriana 42:

249-261.

SCHELPE, E.A.C.L.E. 1969. Mr L.C. Leach— Bolus Medalist, 1968.

Journal of the Botanical Society of South Africa 55: 21-23.

PUBLICATIONS BY L.C. LEACH

BALLY PR O. & LEACH, L.C. 1962. 13. A new Euphorbia from South

central Africa. In PR O. Bally, Miscellaneous notes on the flora of

tropical East Africa including description of new taxa: 8-15.

Caitdollea 18: 347-351.

DE WINTER, B., GIBBS RUSSELL, G.E., KILLICK, D.J.B.,

ROURKE, J.P., OLIVER, E.G.H. & LEACH, L.C. 1992. (93)

Proposal to limit the retro-activity of Article 8.3. Taxon 41: 359.

LEACH, L.C. 1963a. Euphorbia johnsonii, its rediscovery and an ampli-

fied description. Kirkia 3: 34—36.

-1964a. Euphorbia species from the Flora zambesiaca area. Journal of

South African Botany 30: 1-12.

-1964b. Euphorbia species from the Flora zambesiaca area: 2. Kirkia 4:

15-23.

-1964c. Euphorbia species from the Flora zambesiaca area: III. Journal

of South African Botany 30: 209-217.

-1965a. Stapelieae from south tropical Africa, I. Journal of South African

Botany 31: 241-249.

-1965b. Euphorbia species from the Flora zambesiaca area: IV. Journal

of South African Botany 3 1 : 25 1-257.

-1966a. Euphorbia species from the Flora zambesiaca area: V. Journal of

South African Botany 32: 173-182.

-1966b. Euphorbia species from the Flora zambesiaca area: VI. Kirkia 6:

133-147.

-1967. Euphorbia species from the Flora zambesiaca area; VII. Journal

of South African Botany 33: 247-262.

-1968a. Two new stapeliads from Cape Province. Journal of South Afri-

can Botany 34: 135-142.

-1968b. A new Euphorbia from Tanzania. Journal of South African

Botany 34: 289-294.

-1968c. A new Aloe from Rhodesia. Journal of South African Botany 34:

363-370.

-1968d. Euphorbiae succulentae angolenses — I. Boletim da Sociedade

Broteriana 42: 161-179.

-1969a. Euphorbiae succulentae angolenses: II. Boletim da Sociedade

Broteriana 43: 163-182.

-1969b. Euphorbia species from the Flora zambesiaca area: VIII. Jour-

nal of South African Botany 35: 13-23.

-1969c. Stapelieae from south tropical Africa. V. Bothalia 10: 45-54.

-1969d. Huemia urceolata. The Flowering Plants of Africa 39: t. 1550.

-1970a. Euphorbia species from the Flora zambesiaca area: IX. Journal

of South African Botany 36: 13-52.

-1970b. The identity of Aloe vituensis Baker. Journal of South African

Botany 36: 57-62.

-1970c. Stapelieae (Asclepiadaceae) from south tropical Africa: VI. Jour-

nal of South African Botany 36: 157-189.

-1970d. Euphorbiae succulentae angolenses: III. Boletim da Sociedade

Broteriana 44: 185-206.

-1971a. Euphorbia virosa Willd., with a new synonym and two new taxa

from Angola. Boletim da Sociedade Broteriana 45: 349-362.

-1971b. Two new species of Aloe (Liliaceae) from tropical Africa. Jour-

nal of South African Botany 37: 41-52.

94

-1971c. Two new species of Aloe (Liliaceae) from south tropical Africa.

Journal of South African Botany 37: 249-266.

-1972. Two new species of Aloe (Liliaceae) from Zambia. Journal of

South African Botany 38: 185-194.

-1973a. Euphorbia species from the Flora zambesiaca area: 10. Journal

of South African Botany 39: 3-32.

-1973b. Euphorbia tirucalli L.: its typification, synonymy and relation-

ships with notes on ‘Almeidina’ and ‘Cassoneira’ . Kirkia 9:

69-86.

-1973c. New and interesting taxa of the tribe Euphorbieae (Euphor-

biaceae) from Portuguese Africa. Garcia de Orta, ser. Bot. I:

31-42.

-1973d. Stapelieae (Asclepiadaceae) from south tropical Africa: VII.

Bothalia 11: 133-137.

-1974a. Stapelieae (Asclepiadaceae) from south tropical Africa: 8. Jour-

nal of South African Botany 40: 1 5-25.

-1974b. Notes on the aloes of south tropical Africa with four new species

and a new variety. Journal of South African Botany 40: 101-122.

-1974c. Euphorbiae succulentae angolenses: IV. Garcia de Orta, ser. Bot.

2: 31-54.

-1974d. Stapelieae from south tropical Africa: IX. Kirkia 9: 349-358.

-1975a. Euphorbia gummifera, E. gregaria and a new species from

Damaraland. Bothalia 1 1 : 495-503.

-1975b. Notes on Euphorbia mauritanica, E. gossypina and some related

species with an amplified description of E. berotica. Bothalia 11:

505-510.

-1975c. The lectotype species of Stapelia L. and the reinstatement of

Orbea Haw. (Asclepiadaceae). Kirkia 10: 287-291.

-1975d. A new species of Euphorbia (Euphorbiaceae) from Zambia and a

new name for an Angolan species. Bulletin du Jardin botanique

national de Belgique 45: 205-209.

-1975e. Euphorbiae succulentae angolenses: V. Garcia de Orta, ser. Bot.

2: 111-116.

-1975T Euphorbia species from the Flora zambesiaca area: 1 1 . Kirkia 1 0:

293.

-1976a. Euphorbia (Tetracanthae) in Angola and northern Kaokoveld.

Dinteria 12: 1-35.

-1976b. Distributional and morphological studies of the tribe Euphor-

bieae (Euphorbiaceae) and their relevance to its classification and

possible evolution. Excelsa 6: 3-19.

-1976c. A preliminary review of the prominently papillose Huemia spe-

cies (Asclepiadaceae). Journal of South African Botany 42:

439-487.

-1976d. A review of the Euphorbia brevis-imitata-decidua complex

with descriptions of five new species. Bulletin du Jardin bo-

tanique national de Belgique 46: 241-263.

-1976e. The Huernia kirkii-H. hislopii confusion. Bulletin of the African

Succulent Plant Society 11:1 8-20.

-1977a. Notes on Aloe (Liliaceae) species of the Flora zambesiaca area.

Kirkia 10: 385-389.

-1977b. Euphorbia species from the Flora zambesiaca area: 12. Kirkia

10: 391-400.

-1977c. Euphorbiae succulentae angolenses: VI. Garcia de Orta, ser. Bot.

3: 99-102.

-1978a. A contribution towards a new classification of Stapelieae

(Euphorbiaceae) with a preliminary review of Orbea Haw. and

description of three new genera. Excelsa Taxonomic ser. 1 : 1-75.

-1978b. On the classification of Stapelieae and the re-instatement of

Tridentia Haw. Proceedings and Transactions of the Rhodesia

Scientific Association 59: 1-5.

-1979a. Euphorbia proballyana Leach and Euphorbia waterbergensis

R. A. Dyer. Cactus & Succulent Journal (U.S.) 51: 57-59.

-1979b. The correct spelling of Euphorbia monteiri Hook.f and the

operation of Art. 73.10, ICBN 1978. Timm 28: 602-604.

-1980a. A new species of Euphorbia (Section Tetracanthae) from Trans-

vaal. Journal of South African Botany 46: 207-2 1 1 .

-1980b. Euphorbia quadrangularis Pax and a closely related new species

from Tanzania. Journal of South African Botany 46: 3 1 3-323.

-1980c. A review of Tridentia Haw. Excelsa Taxonomic ser. 2: 1-68.

-1980d. Miscellaneous notes on Stapelia L. Excelsa Taxonomic ser. 2:

69-73.

-1980c. A taxonomic review of Euphorbia gariepina Boiss., its distribu-

tion, variability and synonymy. Excelsa Taxonomic ser. 2: 74— 81 .

-1980f Some comments on intergeneric hybrids. Asclepiadaceae 20: 11, 12.

-1980g. On the classification of Tridentia Haw. Asclepiadaceae 21 : 20, 21.

Bothalia 27,1 (1997)

-1981a. Euphorbia candelabrum Welw. versus Euphorbia candelabrum

Kotschy. Taxon 30: 483-485.

-1981b. A new Euphorbia from the Karoo. Journal of South African

47: 103-107.

-1981c. A new Euphorbia from South West Africa. Journal of South

African Botany 47: 807-8 1 1 .

-198 Id. On the lectotypification of Stapelia L. and a proposal for the

clarification of Article 8 of the ICBN. Taxon 30: 227-229.

-1982a. A new species of Euphorbia from the Namib, South West Africa.

Dinteria 16: 27-31.

-1982b. The re-instatement of Tromotriche Haw. (Stapelieae). Journal of

South African Botany 48: 425, 426.

-1983a. A new Euphorbia from South West Africa. Journal of South

African Botany 49: 189-192.

-1983b. A new species of Synadenium from Mozambique. Garcia de

Orta, ser. Bot. 6: 47-50.

-1983c. Tridentea pachyrrhiza. The Elowering Plants of Africa 47: t.

1871.

-1983d. On the classification of the Stapelieae. Bradleya 1: 79, 80.

-1983e. A discussion of some aspects of Eluernia taxonomy. Bradleya 1:

81-83.

-1984a. A new Euphorbia from South Africa. Jourtuil of South African

Botany 50: 341-345.

-1984b. A revision of Tromotriche Haw. (Asclepiadaceae). Journal of

South African Botany 50: 549-562.

-1984c. A new Euphorbia from the Richtersveld. Journal of South Afri-

can Botany 50: 563-568.

-1984d. New taxa of Stapelia L. (Asclepiadaceae) from the Cape Prov-

ince. South African Journal of Botany 3: 169-178.

-1985a. A new species of Euphorbia (Euphorbiaceae). South African

Journal of Botany 51: 281-283.

-1985b. Euphorbia candelabrum Welw. and other tree-like species. The

Euphorbia Journal 3: 91-94.

-1985c. A revision of Stapelia L. (Asclepiadaceae). Excelsa Taxonomic

ser. 3: 1-157.

-1986a. The identity of Huernia concinna N.E. Br. and two new species

from the Somali Republic. Excelsa 12: 93-98.

-1986b. Euphorbia fdiflora. The Elowering Plants of Africa 49: t. 1927.

-1986c. Euphorbia hcdipedicola. The Elowering Plants of Africa 49: t.

1928.

-1986d. A new Euphorbia (Euphorbiaceae) from the western Cape Prov-

ince. South African Journal of Botany 52: 10-12.

-1986e. Anew Euphorbia from the western Knersvlakte. South African

Journal of Botany 52: 369-371.

-1986L Euphorbia candelabrum Welw. and related matters. Ttmm 35:

711-714.

-1987. To ‘lump’ or not to ‘lump’, are infraspecific concepts always

useful or desirable? Veld cfe Elora 12: 123.

-1988a. A revision of Huernia R.Br. (Asclepiadaceae). Excelsa Taxo-

nomic ser. 4: 1-197.

-1988b. A new species of Euphorbia (Euphorbiaceae) from the south-

western Cape. South African Journal of Botany 54: 501-503.

-1988c. The Euphorbia juttae-gentilis complex, with a new species and

a new subspecies. South African Journal of Botany 54: 534-538.

-1988d. A new species of Euphorbia (Euphorbiaceae) from the Mossel

Bay area. South African Journal of Botany 54: 539, 540.

-1989. Validation of the combination Duvalia immaculata (Luckhoff)

Bayer. South African Journal of Botany 55: 268.

-1990. Euphorbia species from the Flora zambesiaca area: Xlll. A new

species from the Chimanimani Mountains. Kirkia 13: 319-322.

- 1 99 1 a. Euphorbia species from the Flora zambesiaca area XIV. New and

imperfectly known species of section Tetracanthae Pax from

Zambia and Malawi with discussion of two ‘key’ species from

Tanzania and a diagnosis of section Tetracanthae. Excelsa 15:

7-22.

-1991b. Euphorbia griseola Pax: its subspecies and relationships. The

Euphorbia Journal 1: 131-137.

-1991c. Comments on a review of the revision of Huernia R.Br. Ask-

lepios 52: 83, 84.

-1992a. Euphorbia candelabrum auctt. Collectanea Botanica2\: 91-95.

-1992b. Euphorbia knuthii Pax: its distribution and other matters. The

Euphorbia Journal 8: 72, 73.

-1993. Tavaresia versus Decabelone (Asclepiadaceae). Ttmm 42:

665-667.

Bothalia27,l (997)

95

-1995. The preparation of good (useful) herbarium specimens of succu-

lent plants especially of the spiny euphorbias and of the stapeliads

and aloes. Asklepios 64: 24—26.

LEACH, L.C. & LAVRANOS, J.J. 1963. A new species of Hiiernia from

Mozambique. Kirkia 3: 38^0.

LEACH, L.C. & PLOWES, D.C.H. 1966a. Stapelieae from south tropical

Africa; 11. Journal of South African Botany 32: 41-60. Reprinted

in Asklepios 28: 68-80 (1983).

-1966b. Stapelieae from south tropical Africa: 111. Journal of South

African Botany 32: 299-303.

-1967. Stapelieae from south tropical Africa; IV. Journal of South African

Botany 33: 99-106.

LEACH, L.C. & WILLIAMSON, G. 1990. The identities of two con-

fused species of Euphorbia (Euphorbiaceae) with descriptions of

two closely related new species from Namaqualand. South Afri-

can Journal of Botany 56; 71—78.

LIST OF TAXA DESCRIBED BY L.C. LEACH

Sections and subsections of genera not included. Subsequent combinations are not indicated.

EUPHORBIACEAE

Endadenium

Euphorbia

albipollinifera

ainbroseae

ambroseae var. spinosa

atrocarmesina

atrocarmesina subsp. arborea

bayeri

baylissii

bougheyi

bruynsii

cannellii

carunculifera

carunculifera subsp. subfastigiata

congestiflora

contorta

confmalis subsp. rhodesiaca

cooperi N.E.Br. var. calidicola

cuneniana

cuneniana subsp. rhizomatosa

curocana

daniarana

debilispina

decidua Bally & Leach

decliviticola

dedzana

demissa

dispersa

distinctissima

dissitispina

eduardoi

ephedroides E.Mey. ex Boiss.

var. debilis

var. imminuta Leach & Williamson

exilis

fanshawei

faucicola

fortissimo

fragiliramulosa

francescae

gentilis N.E.Br. subsp. tanquana

giessii

glandularis Leach & Williamson

gracilicaulis

grandicomis Goeb. ex N.E.Br. subsp. ejuncta

graniticola

griseola subsp. mashonica

griseola subsp. zambiensis

halipedicola

indurescens

ingenticapsa

inundaticola

jubata

lavrani

linearibracteata

lividiflora

louwii

luapulana

lumbricalis

malevola

malevola subsp. bechuanica

mira

miscella

mlanjeana

monteiri Hook.f. subsp. ramosa

mwinilungensis

namuskluftensis

nubigena

nubigena var. rutilans

oligoclada

otjipernbana

papillosicapsa

parviceps

pedemontana

perplexa

perplexa var. kasamana

persistentifolia

platyrrhiza

proballyana

pteroclada

quadrilatera

radiifera

ramulosa

richardsiae

richardsiae subsp. robusta

rugosiflora

schmitzii

scitula

semperflorens

sereti De Wild, subsp. variantissima

speciosa

subsalsa Hiem subsp. fluvialis

strangulata N.E.Br. subsp. deminuens

teixeirae

tholicola

vaalputsiana

vallaris

viduiflora

virosa Willd. subsp. arenicola

whellanii

wildii

williamsonii

Monadeniurn

cannellii

torrei

Synadenium halipedicola

ALOACEAE

Aloe

andongensis Baker, var. repens

bicomiticum

cannellii

enotata

esculenta

excelsa Berger var. breviflora

inamara

lepida

luapulana

procera

scorpioides

tauri

trigonantha

vallaris

96

Bothalia27,l (1997)

ASCLEPIADACEAE Orbeopsis

Caralluma caudata N.E.Br. subsp. rhodesiaca Orbeanthus

Huemia

archeri

bayeri

brevirostris N.E.Br. subsp. baviaana

erectiloba Leach & Lavranos

formosa

guttata (Masson) Haw. subsp. calitzdorpensis

hislopii Tuirill subsp. robusta Leach & Plowes

hystrix (Hook.f.) N.E.Br. var. parvula

lavrani

longituba N.E.Br. subsp. cashelensis Leach & Plowes

occulta Leach Sc Plowes

pillansii N.E.Br subsp. echidnopsioides

plowesii

quinta (Phillips) White & Sloane var. blyderiverensis

thudichumii

urceolata

verekeri Stent var, angolensis

verekeri Stent var. pauciflora

Orbea

halipedicola

halipedicola subsp. septentrionalis

speciosa

Pachycymbium

Stapelia

baylissii

erectiflora N.E.Br. var. prostratiflora

kougabergensis

montana

montana var. grossa

obducta

praetermissa

praetermissa var. luteola

scitula

Stapelianthus baylissii

Trichocaulon mossamedense

Tridentea baylissii (Leach) Leach var. ciliata

R.H. ARCHER*

* National Botanical Institute, Private Bag XlOl, Pretoria (X)01.

Bothalia 27,1:97-99 (1997)

Book Reviews

TREE ATLAS OF SOUTHERN AFRICA/BOOMATLAS VAN SUIDER-

AFRIKA. Section/Seksie 1, compiled by PTUED & JUTTA VON BREITEN-

BACH. 1992. Dendrological Foundation, RO. Box 104, Pretoria 0001.

Pp. 226. Hard cover: ISBN 0620105046, price R3 12.00.

....‘in every way a larger than life character’. This is how Hugh Glen

and Mienkie Welman describe Dr Friedrich von Breitenbach in his obitu-

ary published in Bothalia 25: 260-264 (1995). The scope of the many

achievements of this man of global vision and unquenchable determina-

tion, enthusiasm and single-mindedness equally tended to be beyond what

could or would be reasonably expected. The present work, judging by

its modest title, could be assumed to be little more than a collection of

distribution maps of the 1 000-odd species of trees found on the sub-

continent. Yet it turns out to be conceived and designed as the standard

reference work on the local tree flora — the encyclopaedia of southern

African trees.

The first step towards realization of this grand project was accom-

plished during the lifetime of Dr Fried: section 1 of a projected total of

24 sections, dealing with 1 077 tree species in 98 famiUes and 374 genera,

was published in 1992. It is dedicated to T.R. Sim, author of the first

standard work on South African trees. Its 226 pages cover the 31 tree

species among the ferns and gymnosperms recorded for the subcontinent.

For each species the following is provided (where relevant and/or

known): National Tree number, scientific name, synonyms, common

name (in up to 15 languages), description (detailed but in non-technical

language), wood, taxonomy, distribution, ecology, reproduction biology,

epiphytes, fungi, insects and mites, conservation, national list of Big

Trees, cultivation and names (explanation of meaning/origin of common

name). In addition, the following are given for each species: one or more

line drawings by Jutta von Breitenbach, two or more black-and-white

photographs, a colour photograph (only one species is not so illustrated

in the present section), a distribution map covering half a page and de-

picting distribution in terms of quarter-degree squares (one-sixteenth of

a one-degree square) as well as a table listing a single record, ‘usually

the earliest’, for each quarter-degree square. For widely occurring species

these lists occupy more than a page and one shudders to think what

space the list of Acacia karroo will take up. There are detailed family

and genus accounts with literature references and keys to species. No

indication is given of plans to provide keys, or other aids to the identi-

fication of families, once the whole work has been completed. Colour

photos are grouped together in the relevant genus account but are not

specifically referred to in the text.

The book is well bound in a hard cover and is provided with a white

dust cover with green print displaying title and contents, as well as the

globally significant logo of the Dendrological Foundation: arborum sil-

varumque conservatio salus mundi est — the conservation of trees and

forests is/will lead to, the prosperity/happiness/welfare/salvation/redemp-

tion (the German word ‘Heil’, and similarly, the Afrikaans ‘heil’, say it

much more succinctly) of the earth/mankind. The book, in A4 format,

was produced with the aid of DTP, with English text in the left-hand

column and Afrikaans on the right. Use of paper and general layout must

be described as lavish, especially in view of the vast volume of the

subject matter still to be dealt with and the high price of the first section.

The maps could be reduced in size and produced with the aid of the

MAPPIT computer program (available from the National Botanical In-

stitute, Pretoria), the tables of distribution records much condensed, if

not omitted or relegated to an appendix, the number of illustrations de-

creased and the layout of the work redesigned without loss of information

and, if expertly handled, without detracting from the appearance of the

work.

At a price of R3 12.00 (likely to be higher for future sections) one

feels that the market of the work is largely restricted to institutions, to

the most enthusiastic dendro-friends and to book collectors. But the pur-

chaser can be assured that he acquires a work packed with thoroughly

researched, comprehensive information on the dendroflora of the sub-

continent. If the project can be brought to a successful conclusion it

should indeed constitute the standard encyclopaedia of southern African

trees. We wish Jutta and her collaborators the resources, of various kind,

the energy and the time to bring this grand project to fruition.

O.A. LEISTNER

TROPICAL ALPINE ENVIRONMENTS— PLANT FORM AND

FUNCTION, edited by RE. RUNDELL, A.P. SMITH and FC. MEIN-

ZER. 1994. Cambridge University Press, The Edinburgh Building, Cam-

bridge CB2 2RU, UK. pp 376. Price, hard cover: £65.00, $100.00. ISBN

0-521-42089-X.

The tropic-alpine environments, with winter every night and summer

every day, are radically different climatically from any other environment

on our beloved, but abused planet. This appears to have resulted in some

rather peculiar plant growth forms, which imply intriguing ecophysiologi-

cal relationships, and consequently has excited scientific interest. The

isolated occurrence of these habitats on four widely scattered land masses:

Hawaii (rather marginally). South America, Africa and New Guinea, ef-

fectively provides four replicates of the natural climatic experiment.

The book aims explicitly to promote our understanding of the plant

form and function in the tropic-alpine regions. It is therefore not a de-

scriptive account, and there is little attempt at a complete regional cover

(there is only a single paper on work in the New Guinean alpine region),

or even a complete cover of the diversity of growth forms. The tropic-

alpine setting is almost incidental to the central issue of the relationships

between form and function in a peculiar environment.

The first chapters in the volume describe the tropic-alpine environ-

ments. Alan Smith’s excellent review of the tropic-alpine systems con-

stitutes a most readable chapter that stimulates a strong desire to travel

to these wonderful mountains. This is followed by a rather dry chapter

on the climates, full of useful information. The remaining 18 chapters

fall into two groups: the first deals with general problems and adaptations

(drought, temperature, anatomy), and the second constitutes reports of

case studies of individual tropic-alpine taxa: Polylepis, Isoetes, Senecio,

Espeletia, giant Lobelia, Argyroxiphium. These cover a range of eco-

physiological subjects, as well as a few studies of herbivory, and two

papers on reproductive biology. There is a single chapter on the tropic-

alpine region of New Guinea: clearly this area needs more attention,

especially if there is to be an attempt at arguing for convergence.

The main problems are thermoregulation and diurnal drought. Meinzer

et al. indicate how Espeletia species may control minimum temperatures

through a combination of growth form variation and leaf anatomy and

morphology by comparing species from different altitudes, and correlat-

ing morphological and anatomical differences with altitudinal-climatic

differences. In a second chapter they also deal with the drought issue,

again using Espeletia, showing that a mechanism of major importance

is capacitance, whereby plants store water to deal with a large transpi-

ration load in the early morning, when ground temperatures may be below

freezing. This is in agreement with Hedberg’s findings on the Den-

drosenecio species in Africa. However, Beck shows that the process is

more complex, and that many plant tissues in Dendrosenecio are freeze-

tolerant. He suggests that variables linked to freezing might be more

important in regulating species distribution patterns.

Goldstein et al. show that Polylepis deals with its environment by a

complex combination of ‘adaptations’ (implying morphology or anatomy

that differs from the common or general situation), allowing the plants

to utilize the early morning sun for photosynthesis, and to deal with low

night temperatures. It does not store water as the giant rosette plants do,

but tends to occupy more protected microsites.

98

Bothalia 27,1 (1997)

Draba also has a diversity of survival mechanisms. Some species,

like D. chionochloa, store enough water to transpire in the morning sun

while the ground is still frozen, whereas other species, like D. bellardii,

do not have this storage and are restricted to microsites where soil tem-

peratures do not drop below zero. This paper is based on a comparison

between the more buffered rocky habitats and the open paramo habitats.

Keeley et al. indulge in the bizarre when they document the extent

to which tropic-alpine species of Isoetes derive their carbon from sedi-

ments, rather than from atmospheric CO2. These species also completely

lack stomata, and have a CAM photosynthetic pathway capable of func-

tioning at or near freezing.

Miller investigated the role of pubescence in the inflorescences of

Puya, and in a series of careful comparative and experimental studies

shows that (a) the pubescence kept the flowers warmer and (b) warmer

flowers set more seed. This is in contrast to the situation of leaf pubes-

cence, which appears also to be related to increasing reflectance.

In one of the two papers on reproductive biology of tropic-alpine

plants. Berry and Calvo document a gradual shift from bees to hum-

mingbirds and anemophily with increasing altitude. Anemophily tends to

be associated with some degree of mast-flowering. Hybridization also

appears to be quite common, and Berry and Calvo suggest that habitat

specificity may maintain species integrity, rather than breeding barriers.

Smith & Young demonstrate many similarities in the reproductive biology

between African Dendrosenecio and South American Espeletia species,

including a seedling recruitment system that appears to have two ‘hur-

dles’ : the first massive bout of mortality occurs during the seedling’s first

dry season, and the second is the failure of the seedlings to become

caulescent in the presence of adults (is this a peculiar form of chemical

control?). Plants tend to take decades to become reproductive.

Forty year’s worth of data show that the ranges of the two giant

Lobelia species in the Teleki Valley of Mt Kenya have undergone some

dramatic changes, and Young, presenting these data, argues that occa-

sional extreme events may have profound effects on the demography of

these large plants. The fate of the giant Lobelia and Dendrosenecio on

Mt Kenya is also strongly influenced by the patterns of hyrax predation:

this is elegantly demonstrated by Young and Smith. This susceptibility

to extreme events reminds us of the Karoo.

This book is an excellent survey of the current understanding of the

ecophysiology of tropic-alpine plants. These bizarre plants provide a type

of null hypothesis against which other, more typical, growth forms can

be compared. I have a number of quibbles, which are with the approach,

rather than the content.

A sad phenomenon is the absence of any systematic-evolutionary

analysis of the evolution and origins of these weird floras. There are

frequent references to the evolutionary background of the floras, yet no

attempt is made to trace their origins and the route by which they acquired

these amazing adaptations. The advances made in this direction by the

work on the evolution of heterostyly are not utilized here. The book is

very much in the non-evolutionary mould of old ecophysiological studies,

and shows no signs of the new biology yet. This means that the route

by which these weird and wonderful adaptations were attained is not

traced.

The adaptive nature of some of the structures is not always evident

from the papers. Correlation of structure and environmental variables is

not a sound methodology. Much more convincing are the papers that

report the results of experimental manipulations. It is very evident from

this book that the study of tropic-alpine ecophysiology is still at an early

stage. Different plants appear to show different methods of dealing with

the same problems. Many of the presumed adaptations have not been

tested rigorously, and there has been rather little experimental work.

Striking is the lack of work by indigenous African botanists. Con-

sidering the proximity of Mt Kenya to Nairobi, with its four universities

and a large herbarium, this is a sad indictment, which strongly underlines

the need for capacity building on our fabulous continent, especially since

the pioneering work on tropic-alpine environments was done by Hedberg

in Africa,

Tropic-alpine environments arc often inaccessible, and evoke rather

romantic images, like ‘The mountains of the Moon’, the Andean Paramo,

and Mt Wilhelm in Papua New Guinea, which appeal to both scientists

and the general public. Few people have worked in all four areas, arid

this is the first book that brings together diverse papers from all of them.

It is therefore essential reading for anybody interested in the tropic-alpine

flora. Many of the papers are reviews by scientists who had previously

published several papers on their research, and as such it is a timely

book, providing an overview of the fascinating adaptations of tropic-al-

pine plants.

H.P. LINDER*

* Bolus Herbarium, University of Cape Town, Rondebosch 7700, South

Africa,

BIODIVERSITY AND CONSERVATION OF NEOTROPICAL MON-

TANE FORESTS, edited by S.P. CHURCHILL, H. BALSLEV, E.

FORERO & L. LUTEYN. 1995. The New York Botanical Garden, Bronx,

New York 10458-5126, USA. Pp xiv + 702. Price, hard cover: $85.00.

ISBN 0-89327-400-3.

The New York Botanical Gardens hosted a symposium in June 1993 at

which the 52 papers published in this book were presented. The authors

are all intimately involved with the montane regions of the Neotropics

in a variety of ways. The ecosystems in focus: the montane forests of

the Neotropics, are under severe pressure and this is amply spelled out

in this volume. In terms of research, emphasis has generally been placed

on biodiversity and conservation of the lowland rainforests, whereas the

montane region has been neglected. Large areas of mountain forests have

been destroyed, resulting in dramatic fragmentation of the montane re-

gion. This is not only regrettable because of the simple loss of biodiver-

sity but also because the genetic base of many potentially important

medicinal and crop plants is lost or eroding rapidly. These inadequacies

are highlighted in the book and point to the need for much greater effort

to study and conserve the biodiversity of the montane Neotropics.

The book is dedicated to the memory of two outstanding neotropical

botanists, Alwyn Gentry and Linda Albert de Escobar. They were two

of four biologists killed in a light plane crash on 3 August 1993 in Ecua-

dor while involved with Conservation International’s Rapid Assessment

Programme. It is fitting tribute that a compendium of papers such as this

should honour their memory because they were so active in promoting

the cause of conservation of the tropical forests. Alwyn Gentry was par-

ticularly concerned with floristic inventory and ecological diversity of

tropical forests and his paper delivered at the symposium and published

posthumously, bears testimony to the energy and enthusiasm he displayed

in his chosen task. Other authors are no less dedicated and a continuous

thread throughout the different papers is the severe impact the activities

of man have had on the floral richness of the neotropical mountains over

a period of about five centuries. The most startling statistic is, that

whereas approximately 12% of the Amazonian rainforests have been de-

stroyed, about 12% is all that remains of the tropical montane ecosystems.

It is predicted that the future survival or demise of tropical highland

vegetation in South America will be determined during the next decade,

depending on whether effective conservation measures are instated or

not.

The work is divided into six sections, each with a brief editorial

summary. The first section sets the scene in that it explores the vegeta-

tional history from the Cretaceous to the Holocene.

Looking at this tableau as a palaeobotanist, you ask yourself: why

the extraordinary plant diversity in neotropical montane forests? And truly

extraordinary it is, with around one in every five species on Earth being

found there. How did this unique diversity arise? What do we know of

the history of the vegetation of this small and dynamic comer of our

planet’s surface? And what can we learn from all this?

Our answers to all these questions are patchy and embryonic. As

stated in the introduction to this volume: ‘An immense potential lies

ahead for the contributions that can be provided by evolutionary biolo-

gists’. Just four of 52 papers published in this admirable volume of 702

pages are devoted to the floral evolution of the Neotropics. These four

papers, covering the last 130 million years or so, from the Cretaceous

to the present, provide the merest glimpse at some of the most important

of all possible questions.

The first two contributions focus on the broad pattern of the floral

origins of this mountainous cornucopia; the second pair of papers home

in on the specific picture revealed in the sediments of the high plains of

Bogota (Colombia) over the past five million years. It is largely through

BothaIia27,l (1997)

99

the study of the fossil pollen record that the outline of the vegetation

history is told. Understanding the complex biogeographic processes — the

webs of cause and effect — will require modelling the myriad tectonic,

climatic and biological factors involved in the system. Between chaos

theory on the one hand and the solar constant!?) on the other, emerges,

through many millions of years, the extant reality of diversity and dis-

tribution.

The palaeovegetation has suffered frequent shifts in floristic com-

position. The Andean forests were subject to considerable change, as

shown particularly clearly for the past five million years — the interval

during which man was evolving from ape in Africa. Mountain building

has been episodic over the 130 million years discussed: with the centre

of Andean uplift shifting from south to north and west to east. The

southern Andes first arose around 100 million years back, whilst the

history of the northern stretch of the chain is confined to the latter

half of this period. The breakup of Gondwana and the shifting of

continental plates has played a critical role: ‘considerable interchange

[of floras] was possible between South America and Antartica until

the early Tertiary, between South America and Africa until the mid-

Cretaceous, and between South America and Meso- and North America

during the Late Tertiary’. Both mountain building and continental drift

dramatically affect climatic patterns. Environmental change drives

evolution, brings extinction and diversification.

The incredible floral richness of the tropical northern Andes com-

prises many immigrants from both north and south as well as many

families (17 of 40 with a known fossil record) that are centred there.

The mountainous terrain is both refugium for immigrants and centre of

diversification and radiation. The four papers reviewed have provided a

fine and stimulating glimpse — the merest glimpse — at the floral history

leading to this richness. There remains much to be learned why the Neo-

tropics are so very much richer than the tropics of Africa and southeast

Asia. Knowledge and understanding usually lead to love and devotion.

This volume and these papers must surely shift our global will closer to

preserving such a wonderful focal point of diversity within the biosphere

and to our reaping its full benefits.

The second section of the book documents a series of studies con-

cerning the inventory of the remarkable biological diversity. It includes

17 papers, only one of which deals with birds rather than plants. These

papers cover a wide range of topics pertaining to the common goal of

describing the patterns of species abundance and diversity as well as the

structure of the plant communities in question. Useful information about

patterns of endemism is also provided; the serious student of diversity

would do well to pay attention to these papers; they provide a good

source for comparison with other species-rich ecosystems in the world.

It may come as a surprise to some to find a whole section (section

three) on the taxonomic diversity of cryptogamic plants. However, the

extraordinary high diversity of these plants in the montane forests of the

Neotropics, and the important ecological role they play in montane eco-

systems, fully warrant such an inclusion. The first two papers describe

diversity in selected genera of Agaricales (fungi) in Quercus forests

(Mueller & Hailing), and in lichens of the Colombian montane forests

(Sipman). Although the neotropical fungi, and to a lesser extent lichens,

are very incompletely known, the studies show that the potential diversity

is high. An analysis of distribution patterns in selected groups of fungi

reveals great potential endemism, while almost half of the lichens studied

have a wide distribution throughout the tropics.

Bryophytes are better known and diversity for the tropical Andes is

exceedingly high, estimated at 1 500-1 700 species distributed among 343

genera and 75 families for mosses (ChurchiU, Griffin & Lewis), and

800-900 species in 1 35 genera and 42 families for liverworts (Gradstein).

The Andes may be four times as rich in liverworts, and eight times as rich

in mosses as the surrounding tropical lowland areas, and the highest species

diversity is in the upper montane forest, roughly between 2 500 and 3 300

m. The reason for this high diversity is thought to be topographic relief

which provides favourable habitats and climatic conditions for bryophytes.

This section gives a good overview of cryptogamic plant diversity in the

montane Neotropics and emphasizes the need to accelerate exploration

and taxonomic revision in this fascinating group of plants.

The amazing taxonomic diversity of the Neotropics is well described

in the fourth section. One paper deals with ferns, one with the Lycopo-

diaceae and the remaining 1 3 with flowering plants in a variety of families:

Theaceae, Bonnetiaceae, Symplocaceae, Rosaceae, Onagraceae, Araliaceae,

Asclepiadaceae, Rubiaceae, Asteraceae, Arecaceae, Poacae (Bambuseae)

and HeUconiaceae. The editorial commentary on this section particularly

highlights the plight of plant systematics in the Neotropics, where it is

estimated that it will take almost another 500 years to complete the taxo-

nomic treatment of flowering plants in the Flora neotropica. It is clear that

among plants many unknown species, perhaps some unknown genera and

even some unknown families, may become extinct before even being dis-

covered. This potentially great loss underlies the plaintive cry to train and

encourage more young taxonomists and comparative biologists who could

change this negative condition and foster knowledge of tropical biology and

more particularly of the tropical highland flora.

The fifth section concerns the human impact and utilization of the mon-

tane environments. Disturbance of the montane ecosysteins goes a long way

back in history. However, the human need for physiological stimulants

(drugs!) in various forms has produced the modem ever-increasing demand

for such substances as caffeine, cocaine and morphine. Illegal opium plan-

tations cover 20 000 ha in Colombia and coca cultivation results in the

annual destruction of 1 000-2 (XX) ha of forest. In Colombia, what was

once a major centre of plant diversity in the 900-2 000 m elevation zone,

is the 'coffee zone'. Add to this the global desire for hardwoods, and

the result is a tragic tale of rapid destruction of montane forests, particu-

larly in Colombia, Bolivia and Ecuador. Only a decreased demand for

habit-forming substances in the developed world will reduce the contin-

ued transformation of montane forests.

The final section appropriately starts with a paper by Thomas van

der Hammen entitled ‘Global change, biodiversity, and conservation of

neotropical montane forests’. At a time when global changes are occur-

ring in climate, geology and sociology, it is fitting that a paper such as

this succinctly focuses on the response by both plants and man to these

changes. Using the example of the destruction of Andean montane forests,

suggestions are made as to what should be done to conserve what little

is left after the historical depredations of man and the impact of future

far-reaching changes of global change. The lessons learned apply globally

and it is up to us to recognize the warning signs and respond positively

in order to stem the tide of global degradation and attrition of biodiversity.

Ethnobotany is of common concern at a global scale and it is true

that in many places many useful plants that are known and many plants

that are potentially useful have been lost through habitat transformation

and overexploitation. This is true in Africa, Asia and no less in South

America. The paper by AnneMarie Sprensen and Inge Schjellerup, shows

clearly how dependent the peasant community of Chachapoyas people

in the district of Chuquibamba, is on natural products from the montane

forests. The prognosis is not good — as the forest disappears so too do

the natural resources upon which the peasants depend. The potential

pointers to valuable products for modem medicine and improved food

crops can also be lost in this process.

The main aim of this book is to encourage the conservation of the

neotropical montane region. It is well edited and admirably succeeds in

not only introducing the Neotropics to the uninitiated but also in provid-

ing a wealth of information about the region. It has a useful index even

though the scientific names are not included. The book heightens the

awareness of the need for concerted global effort and responsibility if

there is to be any hope of survival for the neotropical vegetation with

its great diversity.

D.J. MCDONALD*, J.H. ANDERSON** (palaeobotany)

and J. VAN ROOY** (cryptogams)

* Ecology & Conservation, National Botanical Institute, Private Bag X7,

7735 Claremont, Cape Town.

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

r.0

I - .*,^. - fr^-..- »

JIMM

<||M» -

ffegyiaW I' i 1

.■••i^-Mi».,-r :a ..xi*,i8wj#i . ...

■> w '

- . 1.4. •n*a»

'4^

.■ll■|1 ti*i^ 1 1 : ^,,. ,. „ u ..-'-Jt <»■

~4 n#* ^ ' ‘c; - . .-wa^.-iki, ;*^'

.,

v.y r. ..w»y

:vT"s<fcVi£ .

.■'A-Wfl'.l?';

? •ifcifV-

^k''-

2. ti.r-fr. -■ f’4*«i|i||pik4i^'t

fi»^ -(tj;

^ ^ »4^^^iplt|Jllli" »;

r-'v.f^,/< .,, > vk../ H(p»^ Mrif

f, •e^f'i'Vtir <

T -r;

.- V(j* '> >•)»

».• ■» ■'V* ..iMii'*' -.1.

'*-''■■* fjjWii*/ (■••• , •-ak',

r> - -<(,w!n''

MICWi

JW

'-1— ■'■ *' V

»- -t#

■ ..I- T < - '»^iyiMr.’

lit . -:.<- f^^m.

- !- 'fc ^ * ■ '’

U<

V'Viv ,

'«' Jl 4#.r ■ ’l^v''lt**fTI^ P— — a;: ■

w*^t ' - " ' u' ’ '

■ Ai

'\l 4\JS0t ' ^ i

- '■■

* ». . jjv I^njjni ^

• ■'“ ■'V 4^. -iifr^ 0^-

tv ,. t .■ r •m)iim0mti ' ' > -sf' «tq» 'r«.tp>

* ' -'Ai. -**'rt’.--AS!' ' ' ’ifVr^,

V .t - ■~im- :f '*iV"V ♦' • "'■■i™’

.r-* '..- ♦>

...V,. .-• .^:

^ f*'V .ft *t! Mf.f

■* V9’'^>'€yiv#w

^ i’ VJI/ ./ Iirf

l«- <aHi in4,,.'W. nJ* ' “ •■■; ■■

., - .J-^'": LJ^- .

#■ i5j rs

■'’ IjW*'* ^ j t

■ rat* ’*• ''■ ♦• •"piw"

.:■ - » »»r. »* , : i. ..^ *»- f i*ff0 ■

- V*, !, • . -f-

i »* r * «1

^■' * '

♦fi ’■'^•►..J; I** tSl^iS'V

./t

f

' ^ii .;-JUfT|fwS.

-r- .

Vi ir;««ir ^-. . -

- .. i -K'.^ u;.c »l|^.

i»«K V* «lk'*ioS'.ir-i^

.'4\‘

. -, Off* < -

^ -jr .,-•*« is ■•■sjwwr » ,

*■' ^ .IS. Ij^^i

NOW AVAILABLC

BOTHALIA — Contents to vols 21-25

by B.A. Momberg &l J.M. Mulvenna

comprising:

List of papers alphabetically arranged according to senior author and dates and including all co-authors

in alphabetical listing.

Subject index compiled from keywords and titles, with reference to individual articles.

Price: R10,00 (inch VAT, excl. postage)

BOTHALIA — Contents to vols 1-20

by H.F. Glen, B.A. Momberg &c E. Potgieter (1991)

comprising:

a brief history of Bothalia; a list of all papers published; a list of all authors, co-authors, keywords and

titles; and tables with publication dates, major subjects covered and some information on authors.

Price: R9,10 (inch VAT, excl. postage)

Available from the Bookshop

National Botanical Institute

Private Bag XlOl

Pretoria 0001

RSA

Tel. (012) 804-3200 Fax. (012) 804-3211

BOTHALIA

Volume 27,1

May 1997

CONTENTS

1. Notes on Plectranthus (Lamiaceae) from southern Africa. E.J. VAN JAARSVELD and TJ. ED-

WARDS 1

2. Five new species of Lachenalia (Hyacinthaceae) from arid areas of South Africa. G.D. DUNCAN 7

3. Studies in the liverwort genus Fossombronia (Metzgeriales) from southern Africa. 1. Three new

species from Northern Province, Gauteng and Mpumalanga. S.M. PEROLD 17

4. Studies in the liverwort genus Fossombronia (Metzgeriales) from southern Africa. 2. An amendment

to three species from Western Cape, described by S.W. Arnell. S.M. PEROLD 29

5. Studies in the liverwort genus Fossombronia (Metzgeriales) from southern Africa. 3. An amendment

io F spinifolia. S.M. PEROLD 39

6. Notes of African plants:

Apiaceae (Umbelliferae). A new name for a South African Peucedanum. B.L. BURTT 51

Asteraceae. New combination in Dicoma. S. ORTIZ, J. RODRIGUEZ-OUBINA and I. PUL-

GAR 48

Boraginaceae. The taxonomic status of Lobostemon horridus. M.H. BUYS and J.J.A. VAN DER

WALT 55

Fabaceae. A survey of antipodals in the gametophyte of the tribes Podalyrieae and Liparieae.

A.L. SCHUTTE 43

Proteaceae. A new species of Leucadendron from the western Little Karoo. J.P. ROURKE. . . 52

Rubiaceae. A new species of Vangueria from the Soutpansberg. N. HAHN 45

Thymelaeaceae. New combinations in Lac/tntz^fl. J.B.P. BEYERS 45

Vitaceae. A new species of Rhoicissus from the Eastern Cape. E. RETIEF and E. J. VAN

JAARSVELD 49

7. Composition and biogeography of forest patches on the inland mountains of the southern Cape. C.J.

GELDENHUYS 57

8. Cytogenetic studies in some representatives of the subfamily Pooideae (Poaceae) in South Africa. 3.

The tribe Poeae. J.J. SPIES, S.M.C. VAN WYK, I.C. NEMAN and E.J.L. LIEBENBERG . . 75

9. Comparative field performance of three different gas exchange systems. G.F. MIDGLEY, M. VESTE,

D.J. VON WILLERT, G.W. DAVIS, M. STEINBERG and L.W. POWRIE 83

10. Obituary: LeslieCharles Leach (1909-1996). R.H. ARCHER 91

11. Book reviews 97

Ab.stracted, indexed or li.sted in • AETFAT Index • AGRICOLA • BIOSIS: Biological Ab,stract.s/RRM • CAB: Herbage Ab.stract.<i, Field Crop

Abstract. 1 • CAB.S : Current Advances in Plant Science • LSI : Current Contents, Scisearch, Re.search Aleti • Kew Record of Taxonomic Literature • Taxon :

Reviews and notices.

ISSN 0006 8241

© Published by and obtainable from: National Botanical Institute, Private Bag XlOl , Pretoria 0001, South Africa. Typesetting and page layout:

S.S. Brink (NBI). Reproduction & printing: Afri.scot Litho (Pty) Ltd, P.O. Box 23663, Innesdale, 0031 Pretoria. Tel. (012) 331-3698/9. Fax (012) 331-1747.