Bothalia
A JOURNAL OF BOTANICAL RESEARCH
Vol. 35,1
May 2005
TECHNICAL PUBLICATIONS OF THE SOUTH AFRICAN NATIONAL
BIODIVERSITY INSTITUTE, PRETORIA
Obtainable from the South African National Biodiversity Institute, Private Bag XI 01, Pretoria
0001, Republic of South Africa. A catalogue of all available publications will be issued on request.
BOTHALIA
Bothalia is named in honour of General Louis Botha, first Premier and Minister of Agriculture of
the Union of South Africa. This house journal of the South African National Biodiversity 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 sub-
jects are published annually.
Three booklets of the contents (a) to Vols 1-20, (b) to Vols 21-25 and (c) to Vols 26-30, are available.
STRELITZIA
A series of occasional publications on southern African flora and vegetation, replacing Memoirs of
the Botanical Sun’ey 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 artist at the South African National Biodiversity Institute. Many botanical artists
have contributed to the series, such as Fay Anderson, Peter Bally, Auriol Batten, Gillian Condy,
Betty Connell, Stella Gower, Rosemary Holcroft, Kathleen Lansdell, Cythna Letty (over 700
plates), Claire Linder-Smith and Ellaphie Ward-Hilhorst. The Editor is pleased to receive living
plants of general interest or of economic value for illustration.
From Vol. 55, twenty plates are published at irregular intervals.
An index to Vols 1 —49 is available.
FLORA OF SOUTHERN AFRICA (FSA)
A taxonomic treatise on the flora of the Republic of South Africa, Lesotho, Swaziland, Namibia
and Botswana. The FSA contains descriptions of families, genera, species, infraspecific taxa, keys
to genera and species, synonymy, literature and limited specimen citations, as well as taxonomic
and ecological notes.
Contributions to the FSA also appear in Bothalia.
PALAEOFLORA OF SOUTHERN AFRICA
A pal aeo flora on a pattern comparable to that of the Flora of southern Africa. Much of the informa-
tion is presented in the form of tables and photographic plates depicting fossil populations. Now
available:
Molteno Formation (Triassic) Vol. 1. Introduction. Dicroidium , 1983, by J.M. & H.M.
Anderson.
Molteno Formation (Triassic) Vol. 2. Gymnosperms (excluding Dicroidium), 1989, by J.M.
& H.M. Anderson.
Prodromus of South African Megafloras. Devonian to Lower Cretaceous, 1985, by J.M. &
H.M. Anderson. Obtainable from: A. A. Balkema Marketing, Box 317, Claremont 7735,
RSA.
Towards Gondwana Alive. Promoting biodiversity and stemming the Sixth Extinction, 1999,
by J.M. Anderson (ed.).
BOTHALIA
A JOURNAL OF BOTANICAL RESEARCH
Volume 35,1
Scientific Editor: G. Germishuizen
Technical Editor: B.A. Momberg
national
biodiversity
institute
S A N B I
2 Cussonia Avenue, Brummeria, Pretoria
Private Bag X101, Pretoria 0001
ISSN 0006 8241
May 2005
Editorial Board
D.F. Cutler
B.J. Huntley
P.H. Raven
M.J.A. Werger
Royal Botanic Gardens, Kew, UK
National Botanical Institute, Cape Town, RSA
Missouri Botanical Garden, St Louis, USA
University of Utrecht, Utrecht, The Netherlands
Acknowledgements to referees
Archer, Dr R.H. South African National Biodiversity Institute, Pretoria, RSA.
Archer, Mrs C. South African National Biodiversity Institute, Pretoria, RSA.
Bridson, Dr D. 21 Ferrymoor, Ham, Richmond, Surrey, England, UK.
Cope, Dr T. Royal Botanic Gardens, Kew, Richmond, England, UK.
Dold, A.P., Schonland Herbarium, Grahamstown, RSA.
Duncan, G.D. South African National Biodiversity Institute, Cape Town, RSA.
Edwards, Prof. TJ. University of KwaZulu-Natal, Pietermaritzburg, RSA.
Endress, Prof. P.K. University of Zurich, Switzerland.
Ferguson, Dr D.M. Louisiana State University, Baton Rouge, USA.
Geerinck, Dr D. Rue Charles Pas 4, Brussels, Belgium.
Goldblatt, Dr P. Missouri Botanical Garden, Missouri, USA.
Greuter, Prof. Dr W. Botanical Garden and Botanical Museum, Berlin, Germany.
Henderson, Ms L. Plant Protection Research Institute, Agricultural Research Council, Pretoria, RSA.
Herman, P.P.J. South African National Biodiversity Institute, Pretoria, RSA.
Jordaan, Mrs M. South African National Biodiversity Institute, Pretoria, RSA.
Killick, Dr D.J.B. 465 Sappers Contour, Lynnwood, Pretoria, RSA.
Langstrom, Dr E. University of Uppsala, Sweden.
Leistner, Dr O.A. South African National Biodiversity Institute, Pretoria, RSA.
Liebenberg, Mrs E.J.L. South African National Biodiversity Institute, Pretoria, RSA.
Manning, Dr J.C. South African National Biodiversity Institute, Cape Town, RSA.
Meerow, Dr A.W. National Germplasm Repository, Miami, USA.
Melot, Dr J. B.P.1599, IS-121 Reykjavik, Iceland.
Nordal, Prof. I. University of Oslo, Norway.
Nordenstam, Prof. R.B. Naturhistoriska Riksmuseet, Stockholm, Sweden.
Paterson-Jones, Dr D. South African National Biodiversity Institute, Cape Town, RSA.
Randle, Dr C.P. University of Kansas, Lawrence, USA.
Redeuilh, Dr G. 69 Bd des Fosses, Maule, France.
Retief, Ms E. South African National Biodiversity Institute, Pretoria, RSA.
Robbertse, Prof. P.J. University of Pretoria, RSA.
Ronse De Craene, Prof. Dr L. Catholic University of Leuven, Belgium.
Selvi, Prof. F. University of Florence, Italy.
Simon, B. Queensland Herbarium, Brisbane Botanic Gardens, Toowong, Australia.
Smithies, Mrs S.J. South African National Biodiversity Institute, Pretoria, RSA.
Thulin, Dr M.L. University of Uppsala, Sweden.
Van Wyk, Prof. A.E. University of Pretoria, RSA.
Victor, Ms J.E. South African National Biodiversity Institute, Pretoria, RSA.
Vorster, Dr P.J. University of Stellenbosch, RSA.
Welman, Ms W.G. South African National Biodiversity Institute, Pretoria, RSA.
Wilmot-Dear, Dr C.M. Royal Botanic Gardens, Kew, Richmond, England, UK.
CONTENTS
Volume 35,1
1. Three cryptic new species of Aristea (Iridaceae) from southern Africa. P. GOLDBLATT, A.P. DOLD and
J.C. MANNING 1
2. FSA contributions 18: Salicaceae s. str. M. JORDAAN 7
3. Three new species and a new synonym in Strumaria (Amaryllidaceae: Amaryllideae) from southern
Africa. D.A. SNIJMAN 21
4. Lamarck’s new species of Mesembryanthemaceae and the types of their names. P. CHESSELET and M.
PIGNAL 29
5. Cape plants: corrections and additions to the flora. 1. P. GOLDBLATT, J.C. MANNING and D. SNUMAN . . 35
6. Commiphora kaokoensis (Burseraceae), a new species from Namibia, with notes on C. dinteri and C.
namaensis. W. SWANEPOEL 47
7. Two new species of Asteraceae from Northern and Western Cape, South Africa and a new synonym.
J.C. MANNING and P. GOLDBLATT 55
8. Notes on African plants:
Agapanthaceae. Synonymy in Agapanthus. G.D. DUNCAN 87
Amaryllidaceae. A new variety in the genus Clivia. Z.H. SWANEVELDER, A.E. VAN WYK and
J.T. TRUTER 67
Boraginaceae. Codonoideae, a new subfamily based on Codon. E. RETIEF and A.E. VAN WYK . . 78
Iridaceae. Taxonomic notes on Babiana and F err aria in arid western southern Africa. P. GOLD-
BLATT and J. C. MANNING 71
Lycoperdaceae-Gasteromycetes. Bovista capensis , the correct name for Bovista promontorii.
J.C. COETZEE and A.E. VAN WYK 74
Lycoperdaceae-Gasteromycetes. The identity of Lycoperdon complanatum Desf. and its nomen-
clatural implication. J.C. COETZEE and A.E. VAN WYK 76
Poaceae. Name used in the FSA region for the Cymbopogon excavatus-caesius-giganteus Com-
plex. L. FISH 82
Poaceae. Notes on Eragrostis. L. FISH 80
Hyacinthaceae. Ornithogalum juncifolium var. emsii, a new cliff-dwelling Omithogalum from
Eastern Cape, South Africa. E.J. VAN JAARSVELD and A.E. VAN WYK 82
Orobanchaceae. A new species of Harveya from Western Cape, South Africa. J.C. MANNING
and P. GOLDBLATT 89
Proteaceae. A new Leucadendron (Proteeae) from Western Cape, South Africa. J.P ROURKE . . 63
Rubiaceae. Infraspecific taxa in a southern African Pavetta species. P.P.J. HERMAN 84
Salicaceae. Salix: the correct application of the name Salix mucronata, and a new combination.
R.H. ARCHER & M. JORDAAN 92
Scrophulariaceae. Nemesia zimbabwensis, a new record for the FSA region with notes on its phyto-
geographical significance. S.J. SIEBERT and A.E. VAN WYK 69
9. The systematic value of the leaf indumentum in Lobostemon (Boraginaceae). M.H. BUYS 93
10. Ovule-to-seed development in Dovyalis caffra (Salicaceae: Flacourtieae) with notes on the taxonomic
significance of the extranucellar embryo sac. E.M.A. STEYN, A.E. VAN WYK and G.F. SMITH . . 101
11. Miscellaneous notes:
Agapanthaceae. Chromosome counts in the genus Agapanthus. M. MUZILA and J.J. SPIES ... 109
12. Book reviews Ill
New subfamily, species, combinations, status and varieties in Bothalia 35,1 (2005)
Aristea cistiflora J.C. Manning & Goldblatt , sp. nov., 5
Aristea elliptica Goldblatt & A.P.Dold , sp. nov., 3
Aristea nana Goldblatt & J.C. Manning, sp. nov., 1
Babiana planifolia (G.J. Lewis) Goldblatt & J.C. Manning, comb, et stat. nov., 72
Bovista capensis (Fr.) J.C.Coetzee & A.E.van Wyk, comb, nov., 75
Brownanthus vaginatus (Lam.) Chesselet & M.Pignal, comb, nov., 33
Brownanthus vaginatus subsp. schenckii (Schinz) Chesselet & M.Pignal, comb, nov., 33
Chrysocoma hantamensis J.C. Manning & Goldblatt, sp. nov., 58
Clivia gardenii Hook, var citrina Z.H.Swanevelder, A.E.van Wyk & J.T.Truter, var. nov., 68
Codonoideae Retief & A.E.van Wyk, subfam. nov., 79
Commiphora kaokoensis W.Swanepoel, sp. nov., 47
Ferraria variabilis Goldblatt & J.C. Manning, sp. nov., 73
Harveya roseoalba J.C. Manning & Goldblatt, sp. nov., 89
Leucadendron immoderatum Rourke, sp. nov., 63
Ornithogalum juncifolium Jacq. var. emsii Van Jaarsv. & A.E.van Wyk, var. nov., 83
Pavetta zeyheri Sond. subsp. microlancea (K.Schum.) P.P.J. Herman, comb, et stat. nov., 86
Pavetta zeyheri Sond. subsp. middelburgensis (Bremek.) P.P.J. Herman, comb, et stat. nov., 86
Salix mucronata Thunb. subsp. subserrata (Willd.) R.H. Archer & Jordaan, stat. et comb, nov., 92
Senecio speciosissimus J.C. Manning & Goldblatt, sp. nov., 55
Strumaria luteoloba Snijman, sp. nov., 25
Strumaria prolifera Snijman, sp. nov., 21
Strumaria speciosa Snijman, sp. nov., 23
IV
Bothalia 35,1 : 1-6(2005)
Three cryptic new species of Aristea (Iridaceae) from southern Africa
P. GOLDBLATT*, A.P. DOLD** and J.C. MANNING***
Keywords: Aristea Aiton, Iridaceae, new species, southern Africa, systematics, taxonomy
Abstract
Field work in southern Africa over the past several years has resulted in the discovery of three new species of the sub-
Saharan African and Madagascan genus Aristea Aiton, which now comprises some 53 species. Aristea has a pronounced
centre in southern Africa and a centre of diversity in the winter rainfall zone of the subcontinent, where all three new species
occur, one extending eastward into the adjacent southern edge of the summer rainfall zone. All three novelties have been
collected in the past but were confused with related species. A. elliptica (subgenus Eucapsulares), confused in the past with
A. pusilla (Thunb.) Ker Gawl., has a more robust habit, usually with 4 or 5 flower clusters per flowering stem, pale blue
flowers, smooth ellipsoid seeds with flattened surface cells, and pollen shed as monads, whereas A. pusilla usually has 1-3
flower clusters per flowering stem, dark blue flowers, pollen shed as tetrads, and globose seeds with faint foveate sculptur-
ing and colliculate surface cells. A. nana (also subgenus Eucapsulares ), known from few collections, and also confused
with A. pusilla or A. anceps Eckl. ex Klatt, has the unbranched and leafless flowering stem of the latter but has large green
floral spathes, flowers borne on long pedicels, and lacks a leaf subtending the single terminal flower cluster in contrast to
nearly sessile flowers in A. pusilla and A. anceps , and in the latter, dry, rusty spathes. A. cistiflora (subgenus Pseudaristea)
closely resembles A. teretifolia Goldblatt & J.C. Manning but has linear to narrowly sword-shaped leaves and ± secund flow-
ers with the outer tepals only slightly smaller than the inner and with small, dark brown markings at the bases of all the
tepals. In contrast, A. teretifolia has narrower, sometimes terete leaves and flowers held upright with the outer tepals notice-
ably smaller than the inner and bearing dark markings covering the lower half, whereas the inner tepals are unmarked.
INTRODUCTION
In the course of field work in southern Africa, three
undescribed species of the sub-Saharan African and Mada-
gascan genus Aristea Aiton have come to light. All are
spring-flowering species native to the eastern half of the
southern African winter rainfall zone and adjacent south-
ern edge of the summer rainfall zone. All have been col-
lected before but have been confused with known species
similar in vegetative or floral morphology. Comparison
of the taxonomically critical features of the genus, in-
cluding seeds and pollen grains (Goldblatt & Le Thomas
1997; Goldblatt et al. 2004), has substantially aided in
distinguishing two of them. A. elliptica and A. nana, both
members of subgenus Eucapsulares: section Eucapsu-
lares (taxonomy following Goldblatt & Le Thomas
1997). In contrast, details of the flower have shown that
A. cistiflora, of subgenus Pseudaristea, differs from the
closely related A. teretifolia Goldblatt & J.C. Manning,
although its other vegetative and fruiting features accord
closely with the remaining members of the subgenus
(Goldblatt & Manning 1997a). All species have been
examined live in the field as well as in the herbarium.
With the addition of these three novelties, Aristea com-
prises an estimated 53 species. Seven species occur in
Madagascar (Goldblatt 1991, 1995a) and about 18 in
tropical and eastern southern Africa (Weimarck 1940;
Vincent 1985), one shared with Madagascar. There are
33 species in the southern African winter rainfall zone
(Goldblatt & Le Thomas 1997; Goldblatt & Manning
* B.A. Krukoff Curator of African Botany, Missouri Botanical Garden, RO.
Box 299, St. Louis, Missouri 63166, USA. peter.goldblatt@mobot.org
** Botany Department, Rhodes University, 6140 Grahamstown, South
Africa. t.dold@ru.ac.za
*** South African National Biodiversity Institute, Private Bag X7, 7735
Claremont, Cape Town, manning @nbi.ac.za
MS. received: 2004-07-22.
1997a, b), four of which are shared with eastern southern
Africa.
In the descriptions that follow, we ignore the homolo-
gies of the subtending foliar bracts of the inflorescence,
a binate rhipidium, and call the outer two bract members
spathes and those enclosed within them bracts, the latter
always smaller than the spathes. The individual inflores-
cence units, which vary in number and arrangement on
the flowering stem, are simply termed flower clusters.
Subgenus Eucapsulares section Eucapsulares
Aristea nana Goldblatt & J.C. Manning, sp. nov.
Plantae (50— )80— 1 50 mm altae, caule complanato
bialato 1.2-2. 2 mm lato. nodo terminale elongato, foliis
anguste ensiformibus vel linearibus caule pauciter exce-
dentibus, 1 .5—4.0 mm latis, marginibus hyalinis saepe
rubrescentibus, rhipidio binato unico terminale (l)2-flo-
rum, spathis inaequalibus viridibus marginibus siccis
hyalinis (1 3-) 15-25 mm longis, bracteis similaribus
6-12 mm longis spathis obtectis, floribus atrocaeruleis
.tepalis inaequalibus, extemis ± 10-16 x 3. 5-5.0 mm,
intemis 12-17 x 5.5-11.0 mm, filamentis 3-4 mm
longis, antheris ± 2.5 mm longis flavis, ovario ovoideo
4-6 mm longo, pedicelis 10-12 mm longis, stylo 3-loba-
to 5-6 mm longis, capsulis ovoideis (6-) 10- 18 mm
longis, seminibusque ignotis.
TYPE. — Eastern Cape, 3323 (Willowmore): hill slopes
immediately north of Joubertina (growing with A. pusil-
la), (-DD), 19 Sept. 2004, P. Goldblatt & L.J. Porter
12492 (NBG, holo.; K, MO, PRE, iso.).
Plants (50— )80— 1 50 mm high, sometimes in small tufts;
stem flattened and 2-winged, 1.2-2. 2 mm wide, one or
Bothalia 35,1 (2005)
2
two per plant, with elongate terminal intemode up to 90
mm long, occasionally with second short stem up to 30
mm long, stems each with single terminal flower cluster.
Leaves narrowly sword-shaped to linear, sometimes short-
ly exceeding stem, 1 .5—4.0 mm wide, with narrow translu-
cent margins sometimes flushed red. Flower clusters soli-
tary, terminal, (l)2-flowered; spathes unequal, green,
acute-attenuate, margins dry and hyaline, (1 3-) 15-25 mm
long; bracts similar, 6-12 mm long, enclosed within
spathes. Flowers on pedicels 10-20 mm long, dark blue,
outer tepals green on outside; tepals unequal, outer ±
10-16 x 3. 5-5.0 mm, inner 12-17 x 5.5-1 1.0 mm. Stamen
filaments 3-4 mm long; anthers ± 2.5 mm long, yellow;
pollen grains shed singly. Ovary ovoid, ± 4-6 mm long;
style 5-6 mm long, 3-lobed at apex, ± 6 mm diam., lobes
lightly fringed. Capsules ovoid, (6-) 10- 18 mm long, on
pedicels up to 20 mm long. Seeds unknown. Flowering
time: late July to September, rarely in early October.
Distribution and ecology : occurring in Western and
Eastern Cape, from Robinson’s Pass to the Baviaans-
kloof, on sandy and rocky sandstone slopes, in arid, mar-
ginal fynbos.
A dwarf species, mostly reaching less than 120 mm in
stature, Aristea nana has until recently rarely been col-
lected and then has usually been confused with similarly
low-growing A. pusilla (Thunb.) Ker Gawl. When we
came across the species in the wild in 2003, growing
together with A. pusilla near Joubertina, in Eastern Cape,
on a slope burned the previous summer, direct compari-
son of the two in full bloom made it clear that these were
different species, despite their similar dwarf growth form
and virtually identical flowers. They differ slightly in
their leaves, those of A. nana having a glaucous bloom
and wider translucent margins, whereas the pale green
leaves of A. pusilla are softer textured. The flowering
stem of A. nana is almost always unbranched and bears
leaves only near the base. The terminal intemode is sev-
eral times longer than the rest of the stem and terminates
in a single inflorescence of two or rarely three flowers.
Particularly striking are the well-developed pedicels of
the flowers, 10-12 mm long, and ovoid ovary, 4-6 mm
long, quite different to the short pedicels, ± 2 mm long,
of A. pusilla and triangular-columnar ovary, 12 mm long.
Subsequent examination of the pollen of the two
species from the Joubertina site showed that Aristea
pusilla has pollen shed in tetrads, the monads with oper-
culate apertures, as described by Goldblatt & Le Thomas
(1997). Pollen of A. nana, however, is shed singly, and
the grains are dizonasulculate, having two smooth, well-
defined apertures at opposite ends of the grain. Once we
had determined that A. nana was a distinct species, we
made a concerted effort in 2004 to establish its range. We
found it to be common in the Long Kloof and valleys to
the north from Avontuur to Joubertina. We were also able
to identify additional collections of A. nana in herbaria,
where the earliest collections that we have found are
those made by R.D. Bayliss in 1974 and mixed with A.
pusilla. Later collections in herbaria have consistently
been misidentified as A. pusilla.
Despite their superficial similarity, Aristea nana and
A. pusilla are probably not closely related. Instead, we
believe that A. nana is most closely allied to the eastern
southern African A. abyssinica Pax (currently including
A. cognata N.E.Br. (Goldblatt 1995b) and A. anceps
Eckl. ex Klatt. These two species also have unbranched,
flattened and broadly winged flowering stems with an
extended upper intemode, thus bearing leaves only near
the base (Weimarck 1940). In addition, A. nana, A. abys-
sinica, and A. anceps sometimes produce a short stem
held close to the base of the plant as well as normal
extended flowering stems, a feature not before reported.
Some examples are collections of A. abyssinica from
KwaZulu-Natal ( Goldblatt & Manning 9720, MO) and
Limpopo in South Africa ( Goldblatt <£ Porter 11954B,
MO) and Zimbabwe ( Chase 3650, MO) and A. anceps
(. Barker 6991 , NBG; Compton 20288, NBG).
Unlike Aristea nana, however, both A. abyssinica and
A. anceps have a small subterminal leaf subtending the
terminal flower cluster or terminal pair of flower clus-
ters, and rarely a second subterminal leaf 10-20 mm
below the flower clusters. A. anceps also differs in having
the inflorescence spathes and bracts ± dry at flowering
time, whereas those of A. nana are green. Spathes and
bracts of A. abyssinica are green with broad dry margins
at flowering time, later becoming entirely membranous.
They also have subsessile flowers (pedicels are 2-3 mm
long in A. anceps ), whereas those of A. nana have
pedicels up to 20 mm long at flowering time, extending
to 15 mm in fruit. Moreover, both A. abyssinica and A.
anceps have pollen grains with the apertures obscured by
masses of exine (Goldblatt & Le Thomas 1997). The
grains were thus termed sulculate; subsequent examina-
tion using transmission electron microscopy showed that
the apertures are either zonasulculate or disulculate, in
either case with a thick intine indicating the location of
the aperture (Goldblatt et al. 2004). Pollen grains of A.
nana differ significantly from those of its apparent rela-
tives in having smooth apertures, unique for a species of
section Eucapsulares. Whereas nearly mature capsules
of A. nana are known, mature seeds are not, making
impossible comparison of the latter character, important
in determining relationships in Aristea (Goldblatt et al.
2004).
Additional material examined
EASTERN CAPE. — 3322 (Oudtshoom): sandstone slopes at top of
Nuwekloof Pass, near Farm Vaalwater, (-BC), 23 September 2004,
Goldblatt & Porter 12545 (MO, NBG, PRE); Kouga Mtns, Farm
Hoeree, ± 700 m, (-DB), 29 September 1986, Oelofsen 105 (PRE); 6.7
km E of Joubertina, sandy gravel slopes [growing with A. pusilla],
(-DD), 24 September 2003, Goldblatt & Porter 12361 (MO. NBG).
3324 (Steytlerville) Kouga Mtns, near Doringkloof, ± 900 m, (-CA),
30 July 1978, Bond 1421 (PRE). [Doubtful locality] 3326
(Grahamstown); Bathurst Dist., grassland (mixed collection, with A.
pusilla), (-DB), 1 August 1974, Bayliss 6255 (MO).
WESTERN CAPE. — 3321 (Ladismith) Gamka Mountain Reserve,
Zebra ridge near Oukraal, stony sandstone soil, (-CB), 1 5 August 1 983,
Cattell 286 (NBG). 3322 (Oudtshoom): Robinson’s Pass, N-facing
slopes in stony ground, (-CC), 22 September 2000, Goldblatt & Nanni
11583 (MO, NBG); Perdepoort N of Camfer, sandstone slopes burned
last summer, (-CD), 29 September 2004, Goldblatt & Porter 12572
(MO, NBG); upper Longkloof, rocky sandstone bank burned last sum-
mer, (-DD), 18 September 2004, Goldblatt & Porter 12486 (MO,
NBG, PRE). 3323 (Willowmore): Antoniesberg, north slopes, (-AD),
30 September 1989, Barker 679 (PRE); W end of the Kouga Mtns, Bo-
Kouga road, ± 5 km from Uniondale, (-CA), 24 September 2003 (fruit-
ing), Goldblatt & Porter 12369 (MO).
Bothalia 35.1 (2005)
3
Aristea elliptica Goldblatt & A.P.Dold, sp. nov.
Plantae 200-350(-500) mm altae, caule ovoideo
leviter bialato 2. 5-3.0 mm lato eramoso (2)3 vel 4 inflor-
escentibus sessilibus ferentibus (raro infima breviter
stipitato), foliis linearibus ad ensiformibus usitate (2.5-)
4. 0-6.0 mm latis coriaceis glaucis marginibus anguste
hyalinis, rhipidio binato 3-5 terminale 2-florum laterali-
bus usitate 1- vel 2-florum, spathis viridibus marginibus
siccis brunneis (25— )30— 38 mm longis, bracteis siccis
brunneis quam spathis brevibus, floribus pallide cae-
ruleis, tepalis extemis 16-18 x ± 7 mm, intemis ± 16 x ±
9 mm, filamentis ± 3 mm longis, antheris ± 3 mm longis
flavis, ovario triangulo-columnari 12-16 mm longo,
pedicelis ad 3 mm longis, stylo ± 6.5 mm longo ad
apicem 3-lobato fimbriatoque, capsulis triangulo-cylin-
dricis (20-)24—30 mm longis, seminibus ellipsoideis, in
capsulo obliquis.
TYPE. — Eastern Cape, 3326 (Grahamstown): Kariega
Park, between Kenton-on-Sea and Salem, stony quart-
zitic outcrops, (-DA), 21 August 2003, A.P. Dold 4604
(GRA, holo.; MO, NBG, iso.).
Plants 200-350(-500) cm high; stem oval in section
and prominently 2-winged, 2. 5-3.0 mm wide, normally
unbranched, with (2)3 or 4 lateral flower clusters, these
sessile or lowermost short-stalked. Leaves clustered at
base, with 2 cauline leaves, linear to narrowly sword-
shaped, (2.5-)4.0-6.0 mm wide, firm to leathery, glau-
cous, with narrow hyaline margins. Flower clusters 3-5,
terminal cluster 2-flowered, lateral clusters 1- or 2-flow-
ered; spathes green with membranous, hyaline or brown
margins, (25-)30-38 mm long; bracts green to dry, one-
third as long as spathes and concealed by them. Flowers
pale blue, outer tepals with broad green stripe on reverse;
tepals unequal, outer 16—18 x ± 7 mm, inner ± 16 x ± 9
mm. Stamen filaments ± 3 mm long; anthers ± 3 mm
long, smaller after dehiscing; pollen grains shed singly.
Ovary triangular-columnar, 12-16 mm long, on pedicels
up to 3 mm long; style 3-lobed and fringed, ± 6.5 mm
long. Capsules cylindric and three-lobed, (20-)24— 30
mm long; lobes angled. Seeds ellipsoid, many per locule,
often oriented obliquely to capsule axis in single row,
smooth, surface cells ± plane to weakly domed.
Flowering time: August to October. Figure 1 .
Distribution and ecology : occurring in Eastern Cape,
from the Zuurberg Mountains as far east as the Fish
River Mouth, mainly on sandstone slopes and often on
rock outcrops, in fynbos or grassy fynbos.
Aristea elliptica most closely resembles the diminu-
tive southern Cape species, A. pusilla, and was included
in what Weimarck (1940) called A. pusilla subsp. robus-
tior Weim. That taxon is based on a painting of a dark
blue-flowered plant called A. pusilla in Curtis’s
Botanical Magazine (Ker Gawler 1809). In the absence
of an associated specimen, however, it is impossible to
establish its identity with confidence. We believe it rep-
resents a well-grown specimen of A. pusilla but is not
distinct from that species, which has the dark blue flow-
ers clearly evident in the painting. However, some col-
lections from Eastern Cape referred to subsp. robustior
by Weimarck (1940) and later by Vincent (1985), appear
FIGURE 1 . — Aristea elliptica. A, flower; B, fruit. Photographed from
Dold 4604.
very different and, as Weimarck noted, it is difficult to
accept these as belonging to the same species.
Weimarck’s decision to treat the two as subspecies was
largely based on the presence of apparent intermediates,
which he did not enumerate, in the Uitenhage and Port
Elizabeth areas. Vincent (1985) who also recognized A.
pusilla subsp. robustior does list intermediates but some
of these specimens are subsp. pusilla and others are A.
spiralis (L.f.) Ker Gawl. (Marsh 660 , PRE, from
Franschhoek Pass) or A. aff. pauciflora Wolley Dod
(Oliver 5472, PRE, NBG, from Bailey’s Peak, Klein
Swartberg Mtns), both of which localities are outside the
range of both A. elliptica and A. pusilla. Measurements
4
Bothalia 35,1 (2005)
FIGURE 2. — SEM micrographs of seeds, Aristea elliptica : A, non-
raphal view, Dold 4604', B, with raphe across upper third, Dold
4618; C, SEM micrograph of seed of Aristea ensifolia,
Goldblatt & Porter 12353. Scale bars: 100 pm.
given by Vincent for subsp. robustior mostly do not
apply to A. elliptica.
While most of these so-called intermediates are mere-
ly more robust Aristea pusilla , the taller Eastern Cape
plants referred to A. pusilla subsp. robustior represent a
second species, recognized at first by its larger size and
associated broader stem and leaves, the latter firmer and
more leathery than those of A. pusilla. More careful
examination of these plants shows that they have ellip-
soid seeds, unique in Aristea , with a smooth surface and
surface cell outlines plane or weakly colliculate (Figure
2A, B). The seeds lie in a single row in each locule, as
they do in the closely related A. ecklonii Baker, A. ensi-
folia Muir and A. pusilla , but unlike the horizontally
packed seeds in these three species, those of A. elliptica
are loosely arranged and often oriented obliquely to the
long axis of the capsule in a zig-zag arrangement with
only their tips touching (Figure IB). In this arrangement
every alternate seed is more or less parallel. These plants
also have pollen grains shed singly (Table 1) and they are
bisulculate, with two large apertures covered with exine.
Both the seeds and pollen contrast starkly with those
of Aristea pusilla and its ally A. ensifolia. Seeds of these
species are depressed-globose to shortly columnar
(Figure 2C) (flattened dorsally and ventrally due to pres-
sure from adjacent seeds) with shallow foveate sculptur-
ing and surface cell outlines colliculate to tuberculate
(Goldblatt et al. 2004). The seeds are vertically stacked,
in capsules similar to those of A. elliptica but (14-)
20— 25(— 30) mm long. Pollen of A. pusilla is shed in
tetrads (Goldblatt & Fe Thomas 1997) and the monads
are operculate. We have now sampled six populations of
A. pusilla from across its entire range and confirm tetrads
in all of them (Table 1 ), whereas four populations of A.
elliptica examined for the feature have disulculate pollen
grains shed singly. The latter pollen type conforms to the
pattern in most members of section Eucapsulares
(Goldblatt & Le Thomas 1997; Goldblatt et al. 2004).
Typical Aristea pusilla is a dwarf plant, usually
80-150 mm high, with fairly soft-textured, pale green
leaves mostly 2-4 mm wide, and flowers slightly small-
er than those of A. elliptica , with outer tepals 12-14 x 7
mm, and inner 9-11 x ± 9 mm versus outer tepals 16-18
x ± 7 mm, and inner ± 16 x ± 9 mm in A. elliptica. While
A. pusilla occurs on both loamy clay and sandy slopes
TABLE 1 . — Pollen types in populations of Aristea pusilla and A. elliptica; all localities are in South Africa.
Vouchers are housed at GRA, MO, and NBG
* Pollen grains from Tyson 150, are described as sulculate because we cannot detect the aperture(s) beneath the exine
that covers the entire surface.
Bothalia 35,1 (2005)
5
and extends from Swellendam in the west along the coast
and in the Long Kloof to Grahamstown in the east, A.
elliptica favours rocky sandstone slopes, often occurring
in quartzite outcrops and is known from the Zuurberg at
Grahamstown and eastward to the Fish River Mouth.
The ranges of the two species overlap in the Grahams-
town area. Any confusion between the two is most like-
ly the result of depauperate specimens of A. elliptica
being mistaken for robust plants of A. pusilla. Apart from
the pollen and seed differences, A. elliptica can be sepa-
rated by the leaves, mostly 4—5 mm wide, leathery and
somewhat glaucous, the flowering stem usually bearing
at least four flower clusters, the lowermost sometimes
short-stalked, and the inflorescence spathes (25— )30— 38
mm long. Capsules of A. elliptica are elongate, (20-)
24-30 mm long, and contain ellipsoid seeds in a single
vertical row, with the long axis of the smooth seeds
oblique to the long axis of the locules as explained above.
The spathes of A. pusilla are 16-20 mm long and the
stem 1. 5-2.0 mm wide.
Additional material examined
EASTERN CAPE. — 3326 (Grahamstown): Grahamstown, Stones
Hill, (-BC), 18 October 1931, J.R. & B. Rennie 179 (GRA); Beggars
Bush State Forest, quartz outcrop in grassy fynbos overlooking
Beggars Bush Farm, 25 January 2004 (fr.), Dold 4625 (GRA); stony
sandstone-derived soil on slopes W of Fish River Mouth near
Sherwood turnoff, 20 January 1993 (fruiting), Goldblatt & Manning
9530A (MO); Kariega Park. 16 km N of Kenton on Sea, (-DA), 14
August 1997, H.H. Burrows 4658 (GRA), 17 September 1994, H.H.
Burrows 4113A (GRA); Port Alfred, grassy slopes, (-DB), Oct. 1916,
Tyson 150 (PRE); Kowie Nature Reserve. Port Alfred, (-DB), 26
October 2003 (fr.), Dold & Cocks 4607 (GRA, MO).
Subgenus Pseudaristea
Aristea cistiflora J.C. Manning & Goldblatt , sp. nov.
Plantae (100-)300-500 mm altae caespitose, caule
eramoso raro breviter 1-2 ramoso ovale, foliis linearibus
usitate 2. 5-3.0 mm latis leviter rotatis, rhipidio binato
( 1 — )4 — 6 usitate sessilibus 2-4 florum, spathis virides
marginibus translucentibus supra siccentibus 15-18 mm
longis, bracteis similibus pauce brevibus, floribus pallide
roseis ad pallide lilaceis vel cremeis, tepalis extemis
22-23 x 13-14 mm, intemis 23-25 x 15-17 mm cupreo-
brunneis in tertiis proximalibus, filamentis ± 4 mm
longis, antheris 5-6 mm longis flavis, ovario oblongo
± 10 mm longo, stylo ± 11 mm longo ad apicem late 3-
lobato fimbriato, capsulis elongato lignosis indehiscen-
tibus 35^10 mm longis, seminibus triangulato-colum-
naribus fimbriato-papillosis ad angulibus.
TYPE. — Western Cape, 3420 (Swellendam): Swellen-
dam, Marloth Reserve, Reservoir Hill, (-AB), 22 August
2003, J.C. Manning 2875 (NBG, holo.; MO, PRE, iso.).
Plants (100-)300-500 mm high, forming tufts up to
150 mm diam.; stem erect, occasionally with 1 or 2 short
branches, oval in section. Leaves linear to narrowly
sword-shaped, mostly 2. 5-3.0 mm wide, reaching to mid-
dle of stem, loosely twisted. Flower clusters (1— )4 — 6,
mostly sessile, each 2-4-flowered; spathes greenish with
translucent margins becoming dry above, 15-18 mm long;
bracts shorter than spathes. Flowers held at 45° to hori-
zontal, large, pale pink to pale lilac or cream-coloured,
with bases of tepals darker lilac, violet, or brown, streaked
dull brown on outside, outer tepals copper-brown in lower
third; tepals unequal, outer 22-23 x 13-14 mm, inner
23-25 x 15-17 mm. Stamen filaments ± 4 mm long;
anthers 5-6 mm long, yellow; pollen grains shed singly.
Ovary oblong, ±10 mm long, weakly curved outward;
style broadly 3-lobed and fringed, ± 1 1 mm long. Capsules
elongate, woody, indehiscent, 35^10 mm long. Seeds tri-
angular-columnar, 1 .5—1 .8 x ± 1.5 mm, reddish brown, flat
at apex and base, fimbriate-papillate along angles,
obscurely foveate on faces, surface cells domed.
Flowering time : August to mid September. Figure 3.
Distribution and ecology: Western Cape, on the lower
southern slopes of the Langeberg Mtns, in peaty sandstone
soil, flowering only after fire or clearing of the veld.
FIGURE 3. — Aristea cistiflora. Manning 2878. A, basal leaves and
flowering stem showing sessile lateral flower clusters and
secund flowers, B, flower, front view; C, fresh capsule; D, dry
capsule; E, seed. Scale bars: A-D, 10 mm; E, 1 mm. Artist:
John Manning.
6
Bothalia 35,1 (2005)
Evidently first collected by T.M. Wurts in 1952, and
only a few times since then, we found Aristea cistiflora
in 2003, in the spring after a wildfire on the Langeberg
near Swellendam, when it became clear that it was an
undescribed species. Although Wurts 326 is stunted and
only about 100 mm tall, the large, well-pressed flowers
are identical to those of the type collection. Interestingly,
in 1979 the South African botanist, Dr E.G.H. Oliver
noted on the Wurts specimen, "probably a new species,
not described as material inadequate’. His prediction has
proved correct.
Aristea cistiflora is apparently most closely related to
A. teretifolia Goldblatt & J.C. Manning (Goldblatt &
Manning 1997a), which also has unequal tepal whorls,
with oblique outer tepals smaller than the inner and bear-
ing dark basal marks. Aristea teretifolia is distinguished
by its linear to terete leaves, up to 2 mm wide, flowering
stems bearing at most two lateral inflorescences, and by
the slightly smaller flowers with the the outer tepals ± 20
mm long and the inner 24-28 mm long, thus substantially
longer than the outer, which are abruptly constricted at
the base. In A. cistiflora the leaves are narrowly sword-
shaped, mostly 2. 5-3.0 mm wide, the flowering stems
bear up to six lateral flower clusters and the inner tepals
are only 1-2 mm longer than the outer, which taper
gradually toward the base. Aristea cistiflora also has
larger capsules than A. teretifolia, 35-40 mm long versus
20-30 mm long. In general appearance A. cistiflora is
most likely to be confused with A. cantharophila
Goldblatt & J.C. Manning but this species has subequal,
symmetrical tepals with dark markings at the base of
both whorls, dark filaments and exceedingly long cap-
sules, 60-85 mm long.
As in other members of subgenus Pseudaristea
(Goldblatt & Le Thomas 1997), the pollen grains of
Aristea cistiflora are dizonasulculate and have reticulate
exine and smooth apertures. Recognition of A. cistiflora
brings the total number of species in subgenus Pseu-
daristea to eight.
Additional material examined
WESTERN CAPE. — 3420 (Swellendam): Swellendam, base of
Crown Mountain, clearing in pines, (-AB), 4 August 1952, T.M. Wurts
285 (NBG); Swellendam; 9 September 1952, T.M. Wurts 326 (NBG);
Swellendam. Marloth Nature Reserve, old rather sparse veld near path
on 'knol', 10 September 1969, R.A. Haynes H202 (NBG).
ACKNOWLEDGEMENTS
Support for this study by grants 7103-01 and 7316-02
from the National Geographic Society is gratefully
acknowledged. Collecting permits were provided by the
Nature Conservation authorities of Eastern and Western
Cape, South Africa. We thank Ingrid Nanni and Lendon
Porter for their assistance and companionship in the field.
REFERENCES
GOLBLATT, P. 1991. Iridaceae — Famille 45 (2me Edition). Flore de
Madagascar et des Comores : 1-45.
GOLBLATT, P. 1995a. Aristea ranomafana Goldblatt, a new species
from Madagascar. Adansonia 17: 159-162.
GOLBLATT, P. 1995b. Notes on Aristea Aiton (Iridaceae: Niveni-
oideae): taxonomy, chromosome cytology, and phylogeny.
Annals of the Missouri Botanical Garden 82: 139-145.
GOLBLATT, P. & LE THOMAS, A. 1997. Palynology, phylogenetic
reconstruction and the classification of the Afro-Madagascan
genus Aristea Aiton (Iridaceae). Annals of the Missouri Botani-
cal Garden 84: 263-284.
GOLBLATT, P„ LE THOMAS, A. & SUAREZ-CERVERA, M. 2004.
Phylogeny of the Afro-Madagascan Aristea (Iridaceae) revisit-
ed in light of new data for pollen morphology. Botanical
Journal of the Linnean Society 144: 41-68.
GOLBLATT, P. & MANNING, J.c' 1997a. New species of Aristea
section Pseudaristea (Iridaceae) from South Africa and notes
on the taxonomy and pollination biology of the section. Novon
7: 137-144.
GOLBLATT, P. & MANNING, J.C. 1997b. New species of Aristea
section Racemosae (Iridaceae) from the Cape Flora, South
Africa. Novon 7: 357-365.
KER GAWLER, J.B. 1809. Aristea pusilla. Curtis's Botanical
Magazine 30: t. 1231.
VINCENT, L.P.D. 1985. A partial revision of the genus Aristea
(Iridaceae) in South Africa, Swaziland, Lesotho, Transkei and
Ciskei. South African Journal of Botany 51: 209-252.
WEIMARCK, H. 1940. Monograph of the genus Aristea. Acta
Universitatis Lundensis (Lunds Universitets Arsskrift), new ser.
36: 1-140.
Bothalia 35,1: 7-20(2005)
FSA contributions 18: Salicaceae s. str.
M. JORDAANf
Deciduous trees or shrubs, sometimes creeping and
spreading by suckering, with scaly buds; bark bitter;
wood soft and light; plants dioecious. Leaves alternate,
simple, entire, toothed or occasionally lobed; stipules
foliaceous and subpersistent, or small and caducous. In-
florescences erect or pendulous spikes or catkins; bracts
membranous, entire, toothed or laciniate, fugacious or
persistent, subtending individual flowers, often appearing
before leaves. Flowers unisexual, regular; perianth
absent or represented by a cup-shaped or annular disc or
2 scales or glands at base; nectaries when present, vary-
ing in size, shape, colour and number. Male flowers with
2-many stamens; fdaments filiform, free or ± connate;
anthers small, ovate or oblong, opening by 2 longitudinal
slits; ovary rudimentary or absent. Female flowers with
ovary superior, sessile or shortly pedicellate, 1 -locular;
ovules 2-many on each parietal or basal placenta, anatro-
pous, crassinucellar, unitegmic; style short; stigmas 2^f,
short, thick, emarginate or 2-fid. Fruit a small 2-4-
valved capsule, ovoid or oblong, often acuminate. Seeds
many, very small, with large basal tuft of long, silky, tan-
gled, placental hairs, wind-dispersed, short-lived; endo-
sperm absent or very scanty and oily; embryo small and
straight.
A family of two genera and ± 490 species, chiefly of
moist or wet habitats, mainly in the cooler parts of the
northern temperate and subarctic zones, scarce, and
chiefly in highland areas in the tropics, absent from
Australasia and the Pacific Islands. Both genera are re-
presented in southern Africa: Populus (introduced spe-
cies only), and Salix (both introduced and indigenous
species). Populus is grown for matchwood, boxwood
and pulp, while both genera are cultivated as ornamental
trees and shrubs, for shade, fodder, for their sand-binding
properties and as windbreaks. Young twigs and leaves of
Salix are bitter and astringent, but yield salicin and are
the antecedent of modem aspirin. Baskets are made out
of the slender twigs of Salix species and the best cricket
bats are made from the timber of S. coerulea E.Wolf.
The taxonomic position of the Salicaceae and its rela-
tionship with the Flacourtiaceae are still in dispute, but in
this article the family concept is being treated in its nar-
row sense according to the classification of Cronquist
(1981). The flowers of Populus are wind-pollinated
(anemophilous), whereas those of Salix are predominant-
ly entomophilous (Fisher 1928). Meeuse (1975), in dis-
cussing the taxonomic position of the Salicaceae, argued
that the family underwent several, partly divergent evolu-
tionary processes through which Salix became predomi-
nantly insect-pollinated and developed nectaries. Fisher
t South African National Biodiversity Institute, Private Bag X101,
0001 Pretoria.
* Naturalized taxon.
MS. received: 1999-04-08.
(1928) was the first to describe the flower structure of the
Salicaceae 5. str. in great detail.
Terminal buds present, rarely lacking; winter buds with several
clearly visible unequal outer scales present; leaves
broad, deltoid or broadly ovate, broader than long; peti-
ole longer than 15 mm; catkins pendulous; floral bracts
apically serrate or laciniate; male disc cup-shaped or
annular * *Populus
Terminal buds lacking; winter buds with only 1 calyptrate scale
present; leaves narrow, ovate to linear-lanceolate or
elliptic, longer than wide; petiole shorter than 10 mm;
catkins usually upright; floral bracts entire; male disc of
1 or 2 small distinct glands at base Salix
1872000 POPULUS*
*Populus L., Species plantarum edn 2: 1034 (1753);
L.: 456 (1754); Willd.: 802 (1806); Wesm.: 323 (1868);
Benth.: 412 (1880); Pax: 35 (1889); C.S.Hubb.: 340
(1926); Chalmers Smith: 275 (1943); Adamson: 311
(1950); Franco: 54 (1964); Willis: 937 (1973); R.A.Dyer:
30 (1975); Jalas & Suominen: 48 (1976); Coates Palgrave:
91 (1977); Wilmot-Dear: 4 (1985); Wilmot-Dear: 121
(1991); Jordaan: 500 (2000). Type species: P. alba L.
[lecto. fide Britton & Brown: 587 (1913)].
Small or large trees with pale furrowed bark and soft
white wood; branches terete or angled, with terminal
buds; winter buds often resinous and aromatic, with several
unequal outer scales. Leaves mostly broadly ovate to
rhombic, long-stalked, entire or dentate; stipules membra-
nous, small, narrow. Inflorescence a pendulous, droop-
ing, odourless catkin, appearing before leaves. Flowers
unisexual, wind-pollinated, borne in axil of a serrate or
laciniate bract; perianth reduced to a cup-like disc. Male
flowers with 4—30 or more stamens; fdaments free;
anthers 2-thecous, oblong to ovate, red. Female flowers
with ovary sessile or subsessile, 1 -locular; ovules many;
style very short, 2^1-branched, each branch entire or 2-
or 3-fid. Capsule 2^1-valved. Seeds many, brown, small,
ovoid or obovoid, with a tuft of long silky hairs from
base.
A genus of ± 40 species (Wilmot-Dear 1985) and con-
fined to northern temperate and subtropical regions apart
from the single East African species Populus ilicifolia
(Engl.) Rouleau. Natural hybrids are common in the
genus, and some have been described as species, e.g. P. x
canescens (Aiton) Sm. It is possible that other named
species are in fact of hybrid origin. In southern Africa a
few species have been widely cultivated, principally for
wood pulp, matches and boxes, e.g. P deltoides
W.Bartram ex Marshall (match poplar), which occasion-
ally escapes from cultivation. Populus species are often
planted to stabilize dongas but they can spread to adjacent
streambeds. P alba L. (white poplar) and P. x canescens
(grey poplar) become naturalized, especially in marshy
Bothalia 35,1 (2005)
areas and in river valleys, where they spread by sucker-
ing. The latter is now one of the most prominent riverine
invaders in southern Africa (Henderson & Musil 1984;
Henderson 1989, 1991a, 1992, 1998). P. nigra L. var. it ali-
en (Moench) Koehne also escapes from cultivation, and
has invaded rivers in various parts of southern Africa.
Key to species
la Leaves tomentose or puberulous underneath, margins lobed:
2a Leaves, at least near tip of long shoots, palmately 1-5-
lobed and with thick, persistent, distinctly white
tomentum below, margin irregularly dentate; flowers
sessile 1 . *P. alba
2b Leaves obscurely lobed and thinly greyish tomentose
below, becoming almost hairless with age below, mar-
gin coarsely serrate; flowers pedicellate . . 2. *P. x canescens
lb Leaves glabrous, margins serrate or dentate:
3a Branches spreading or ascending; crown broad (Figure
4B); leaves 70-110 mm long, base truncate; glands
present at base of lamina; capsules slender-pedicellate
3 . *P. deltoides subsp. deltoides
3b Branches steeply ascending; habit columnar (Figure 6B);
leaves 20-60 mm long, base cuneate or rounded;
glands absent at base of lamina; capsules ± sessile . . .
4, *P. nigra var. italica
1. *Populus alba L., Species plantarum edn 2: 1034
(1753); Boiss.: 1193 (1879); Femald: 522(1950); D.R.Maire:
39 (1961); Franco: 54 (1964); A. Neumann; 12 (1969);
L.H. Bailey & E.Z.Bailey: 901 (1976); Jalas & Suominen:
48 (1976); Browicz & Yalt.: 717 (1982); Meikle: 1490
(1985); L.Hend.: 188 (2001). Type: ‘Habitat in Europa
temperatiori’. Herb. Burser XXIII: 19 [UPS, lecto., desig-
nated by Jonsell: 78 (1993)].
Tree up to 30 m tall, with trunk up to 1 m or more
diam. and branches spreading to form a wide, rounded
crown; spreading by root sprouts. Bark greyish green to
whitish grey and smooth on upper part of trunk and
branches, rough and fissured on basal part of old trunks;
branchlets at first densely white tomentose, becoming
dull grey-brown with age; buds ovoid, blunt, densely
white-tomentose. Leaves of two kinds: those on short
lateral shoots and at base of long leading shoots up to 50
mm long, broadly and bluntly ovate, irregularly bluntly
serrate-lobed and thinly tomentose or glabrous; those
towards apex of long shoots often deeply palmately 3-5-
lobed, up to 125 mm long, persistently white-tomentose
below, apex acute, base rounded or subcordate, some-
times with 2 glands, margin toothed, teeth triangular;
petioles terete, tomentose, shorter than lamina, 20-42
mm long. Male catkins not found in southern Africa.
Female catkins twice as long as male catkins; pedicels
1-2 mm long; discs ± 1.5 mm diam.; floral bracts den-
tate, margins with long white hairs. Female flowers :
ovary 3-5 mm long, tomentose; stigmas with 2 terete,
horizontally divergent branches. Capsules ovoid, 3-5
mm long, 2- or 3-valved. Seeds mostly abortive.
A native of central, E and SE Europe and Asia; now
naturalized from plantations in Gauteng, Mpumalanga,
Free State and Lesotho (Figure 1 ). Not as widespread as
R. x canescens. Spreads by suckering, as only female
plants are found in southern Africa (Hubbard 1926). Very
difficult to distinguish from P. x canescens, of which it is
one of the putative parents. See the differences under P.
x canescens.
12 14 16 16 20 22 24 26 26 30 32
FIGURE 1. — Distribution of Populus alba in southern Africa, drawn
from SAPIA Database, PPRI, 1979-2005.
Vouchers: Du Plessis 427 (PRE, PRU); Henderson 628 (PRE,
PRU), 961 , 1010 (PRE); Kluge 888 (PRE, PRU); Theron 1908 (PRU).
2. *Populus x canescens (Aiton) Sm. in Flora Brit-
tanica 3: 1080 (1804); Burtt Davy: 340 (1921); Burtt
Davy: 432 (1932); Femald: 522 (1950); Butcher: 966
(1961); Jalas & Suominen: 49 (1976); Coates Palgrave:
91 (1977); L.Hend.: 189 (2001). Type from Europe.
Tree up to 25 m tall, with moderately spreading
crown, often suckering. Bark on young stems white,
smooth, becoming rough on old trunks. Leaves of long
shoots deltoid-ovate, cordate, up to 120 mm long, grey-
tomentose beneath, becoming almost hairless with age;
leaves on short branches, smaller, suborbicular to ovate,
obtuse, subcordate, 30-60 mm long; not ciliate, with a
narrow translucent border; apex acute, base cordate,
sometimes with 2 glands at base, margin irregularly
glandular-serrate, ciliate; petiole terete, 15-35 mm
long. Male catkins 60-100 mm long; stamens 8-15.
Figure 2.
A native of Europe and W Asia which has spread to
many parts of the world and is now found as an escape
from cultivation and established along streams and rivers in
the Limpopo Province, North-West, Mpumalanga, Gauteng,
Free State, KwaZulu-Natal and Northern, Western and East-
ern Cape (Henderson 1991a) (Figure 3).
Commonly known as the grey poplar, it was original-
ly introduced into South Africa to supply the match indus-
try but in the summer rainfall area at least, it has since
been replaced by P. deltoides. Although originally named
as a species, it is apparently a hybrid between P. alba and
P. tremula L. (Browicz & Yaltirik 1982). Only male plants
are found in southern Africa, and the plants spread vege-
tatively by root suckers. A semi-evergreen variety, ‘Rossii’,
also occurs in the area and was originally described as a
taxonomic variety by Hubbard (1926), from a specimen
collected at the Apies River, Gauteng. It is fast-growing,
very frost-resistant and fairly drought-hardy. It differs
from P. alba mainly in leaf characters: the leaves are
Bothalia 35,1 (2005)
9
FIGURE 2. — Populus x canescens: A, terminal shoot, x 1; B, male inflorescence, x 1; C, floral bract, x 8. A, Jordaan 356 (PRE); B, C, Jordaan
3528 (PRE). Artist: G. Condy.
usually smaller, shorter than 40 mm, not lobed, with four
or five coarse, broadly rounded teeth on each side and
becoming glabrescent with age below. The leaves of P
alba are usually longer than 45 mm, are 1-5-lobed and
have more than five, small, irregular, sharp, triangular
teeth on each side, with the white tomentum persistent
below.
Vouchers: Duggan & Henderson 15 (PRE); Meadows s.n. (GRA);
Olivier 1432 (NBG); Parker 3823 (NBG); Potts 4968 (UOVS).
3. *Populus deltoides W.Bartram ex Marshall,
Arbustrum americanum: 103 (1785); Britton & Brown:
591 (1913); Femald: 522 (1950); Franco: 55 (1964);
L.H. Bailey & E.Z.Bailey: 901 (1976); Eckenw.: 203
(1977); L. Fiend.: 190 (2001). Type: United States of
America, Carolina and Florida, Bartram s.n. [(BM?,
holo., not traced, fide Eckenwalder (1977)].
subsp. deltoides
Tree up to 30 m tall, with large, erect trunk, ± 2 m
diam.; branches ascending to spreading, forming a broad
crown. Bark ashy grey, thick, deeply fissured to form
broad rounded ridges; branchlets usually stout, glabrous,
strongly angular, pale yellowish green to brownish or
grey when young, becoming greyish brown with age;
buds large, ovoid to ellipsoid, acuminate, resinous, with
6 or 7 scales, outer scales puberulent at base, bright red-
dish brown, glabrous, 12-30 mm long, crenate-dentate,
ciliolate, with 2 or 3 basal glands. Leaves deltoid or sub-
orbicular-ovate, 70-180 mm long, apex abruptly triangu-
lar-acuminate, base truncate, coriaceous, bright green
and glossy above, paler beneath, glabrous, turning yel-
low in autumn, margin irregularly glandular-serrate, with
2 or more large conspicuous glands at point of attach-
ment with petiole; petiole flattened at summit, slender.
10
Bothalia 35,1 (2005)
12 14 16 18 20 22
26 28 30 32
FIGURE 3. — Distribution of Populus x canescens in southern Africa, FIGURE 5. — Distribution of Populus deltoides subsp. deltoides in south-
drawn from SAPIA Database, PPRI, 1979-2005. em Africa, drawn from SAPIA Database, PPRI, 1979- 2005.
pilose becoming glabrous, yellowish or tinged with red. Commonly called match poplar, cottonwood or neck-
up to 1 50 mm long. Male catkins densely flowered, lace poplar. A native of the United States of America, but
cylindrical, 70-100 mm long, red and yellow, appearing is cultivated and occasionally naturalized along water-
in spring; pedicels 8-10 mm long. Male flowers: stamens courses in South Africa (Henderson & Musil 1984;
30-60 or more; anthers ± 1 mm long. Figure 4.
Henderson 1989, 1991b, 1992, 1998) (Figure 5).
/a’sjt&s/e sj&fuv-
V (7
FIGURE 4 — Populus deltoides subsp. deltoides. A, terminal shoot, x 2/3; B, tree outline; C, male catkin, x 2/3. Artist: Marietjie Steyn. Taken from
Henderson (2001), with permission.
Bothalia 35,1 (2005)
11
FIGURE 6. — Populus nigra var. italica. A, terminal shoot, x 2/3; B. tree outline; C, male catkin, x 1 . Artist: Marietjie Steyn. Taken from Henderson
(2001 ), with permission.
Vouchers: Gubb KMG10755 (PRE); Henderson 1088 (PRE); Siebert
& Siebert 2012 (PRU).
4. *Populus nigra L., Species plantarum edn 2:
1034 (1753); Jacot Guill.: 161 (1971). Type: ‘Habitat in
Europae temperatiore’ (type not designated).
var. italica (Moench) Koehne in Deutsche Dendro-
logie: 81 (1893), non Du Roi: 2141 (1772), nom. illeg.;
Franco: 55 (1964); Bugala: 45 (1967); A.Neumann: 8
(1969); L.H. Bailey & E.Z. Bailey: 901 (1976); Browicz
& Yalt.: 719 (1982); Meikle: 1490 (1985); L.Hend.: 191
(2001). P. italica Moench: 79 (1785). P. nigra subsp. ital-
ica (Moench) Seemen: 41 (1908); D.R.Maire: 45 (1961).
Type: from Lombardy (Italy).
Tree up to 30 m tall, with short trunks up to ± 2 m
diam.; branches steeply ascending and brittle, giving tree
a columnar outline; rarely suckering. Bark dark grey and
deeply fissured; branchlets terete, glabrous, pale yellow-
ish brown, becoming grey; buds ellipsoid-conic, re-
curved at apex, resinous, reddish, glabrous, scales with
entire margins. Leaves firm, glabrous, yellowish green;
lamina rhombic to triangular-ovate, 20-60 mm long,
apex abruptly acute to acuminate, base rounded to trun-
cate, margins finely crenate-serrate, eglandular at base;
petiole slender, 30-50 mm long, flattened laterally. Male
catkins 50-70 mm long. Male flowers: stamens 8-20, red.
Figure 6.
Commonly known as the Lombardy poplar, a cultivar
from the original species, which is a native of Europe.
Infertile and often spreads by means of suckers along
watercourses, particularly in the eastern Free State
(Henderson 1991a), Lesotho, Western Cape and Eastern
Cape (Henderson 1992) (Figure 7).
Vouchers: Du Preez 1929 (PRU); Gibbs Russell, Robinson, Herman
& Downing 169 (PRE); Zambatis 169 (PRE).
12
Bothalia 35,1 (2005)
12 14 16 18 20 22 24 26 28 30 32
FIGURE 7. — Distribution of Populus nigra var. italica in southern
Africa, drawn from SAPIA Database, PPRI, 1979-2005.
1873000 SALIX
Africa in the north, from Syria and Egypt, the Arabian
Peninsula southwards through East and West tropical
Africa following the Nile and the great lakes and along
the Congo-Zambezi watershed as far south as the
Kunene, Zambezi and Chobe Rivers in Namibia and
Botswana. No attempt is made so far to propose an infra-
specific classification of Salix north of the Limpopo
River for tropical Africa, because of very inadequate
material as pointed out by Meikle (1958) and Friis
(1992). Further south in the FSA region, only three taxa
occur, all geographically correlated to their own river
groups and drainage lines. S. mucronata subsp. mucrona-
ta (= subsp. capensis) occurs in the Orange and Vaal
Rivers and their tributaries (North-West, Free State,
Lesotho, Northern Cape and southern Namibia) and
Western Cape from the Breede River Valley northwards
to the Eastern Cape rivers, mainly the Great Fish River
Valley, and as far north as the Umzimkulwana River in
southern KwaZulu-Natal. S. mucronata subsp. hirsuta is
confined to the Olifants River and probably the Berg
River in Western Cape. S. mucronata subsp. woodii
occurs in the northern provinces of South Africa, in
Swaziland, KwaZulu-Natal and Lesotho.
Salix L., Species plantarum, edn 2: 1015 (1753);
L.: 456 (1754); Willd.: 655 (1806); Thunb.: 30 (1823);
Andersson: 1 (1867); Andersson: 190 (1868); Benth.:
411 (1880); Pax: 36 (1889); Boiss.: 1181 (1879);
Marloth: 130 (1913); Engl.: 7 (1915); Burtt Davy: 62
(1922); Skan: 575 (1925); M.J.Fischer: 307 (1928); Burtt
Davy: 431 (1932); Adamson: 310 (1950); Ball: 369
(1961); Rech.f.: 43 (1964); Friedr.-Holzh.: 1 (1967);
Willis: 1021 (1973); Argus: 2 (1973); Argus: 1613
(1974); R. A. Dyer: 30 (1975); Coates Palgrave: 91
(1977); Wilmot-Dear: 1 (1985); Immelman: 171 (1987);
Meikle: 258 (1989); Wilmot-Dear: 120 (1991); McKean:
83 (1996); Jordaan: 500 (2000); Coates Palgrave: 121
(2002). Type species: S. alba L. [lecto., designated by
Britton & Shafer (1908)].
Deciduous trees or shrubs with usually terete branches
lacking terminal buds; winter buds each protected by a
single calyptrate scale, buds sometimes glutinous; branch-
es ± flexible. Leaves alternate, petiolate or sessile, oblong,
lanceolate or linear, entire or serrulate; stipules present or
absent. Inflorescences dense, usually erect, firm catkins,
appearing before or with leaves. Flowers unisexual, chiefly
insect-pollinated, borne in axil of an entire bract, with 1 or
2 small nectariferous glands at base. Male flowers : stamens
2, or in a few species, 3-many; filaments slender, free or
sometimes connate, exceeding scale. Female flowers : a
single ovary, composed of 2 carpels, sessile or stipitate;
ovules often 4-8, arranged on 2 placentas; style often short,
with 2 short, retuse or bifid stigmas. Capsule many-seeded,
dehiscing by 2 recurving valves. Seeds many, minute, nar-
rowed at ends, dark chestnut-brown or nearly black,
enveloped in silky wool; testa white, translucent, surface
rough; cotyledons oblong.
A large genus of ± 450 species (Argus 1997), mostly
in temperate parts of the northern hemisphere. One
indigenous species with four subspecies occurs in south-
ern Africa and is widespread along rivers; three intro-
duced species have become naturalized. S. mucronata
subsp. subserrata (the safsaf willow) occurs in tropical
The indigenous Salix species in southern and tropical
Africa are sometimes difficult to identify because of the
dimorphic character of their leaves, which is not always
the case with the introduced and naturalized species. Two
types of leaves are produced in different seasons. The
first spring leaves are usually smaller, broader in propor-
tion to their length, often obovate instead of ovate or
lanceolate, with margins entire instead of toothed, apex
often rounded instead of acute. The summer leaves
which follow, often remain on the branches until pushed
off by the swelling buds in their axils, hence these trees
are sometimes almost evergreen (Newsholme 2002).
Owing to this overlap of leaf characters (leaf shape and
proportion), the complete geographical separation, with
subspecies restricted to certain drainage basins, is there-
fore the most distinctive character to use for the infra-
specific classification of Salix mucronata.
Camus & Camus (1904) classified the European species
of Salix into sections, mainly on account of the number of
nectaries and number of stamens. S. babylonica and S.
fragilis with two nectaries and 4— 12 stamens and pubescent
filaments belong to section Fragilis Koch, and S. caprea
with one nectary and two stamens belongs to section
Capreae. Argus (1997) published a new infrageneric clas-
sification of Salix and placed S. fragilis in Salix subgenus
Salix sect. Salix. S. babylonica belongs to Salix subgenus
Salix sect. Subalbae Koidzumi and S. caprae to Salix sub-
genus Vetrix (Dumort.) Dumort. sect. Cinerella Seringe.
There has never been any attempt made to place the
African species S. mucronata in any section. With its nec-
tary represented by an irregularly lobed ring, or sometimes
reduced to one abaxial and one adaxial gland and 5-8 sta-
mens, it may constitute a section of its own.
Some species of Salix, e.g. S. babylonica , S. fragilis
and S. mucronata, are classified as rheophytes, which in
nature are confined to the beds of swift-running streams
and rivers where they grow up to flood level, but not
beyond the reach of regularly occurring flash floods
(Van Steenis 1981).
Bothalia 35,1 (2005)
13
Key to species and subspecies (see Figures 8 & 9)
la Branches hanging ± vertically; leaves tapering to a long-acuminate apex; stipules persistent, at least 8 mm long;
female flowers sessile 6. *S. babylonica
lb Branches ascending, spreading or droopy but not hanging vertically; leaves rounded, acute or acuminate but not
tapering to a long apex; stipules varying; female flowers and fruits usually pedicellate:
2a Stipules usually ± present, foliaceous or linear, at least 3 mm long:
3a Catkins pedicellate; mature leaves usually not wider than 15 mm and longer than 45 mm 7. *S. fragilis
3b Catkins subsessile; mature leaves usually wider than 20 mm and not longer than 40 mm 8. *S. caprea
2b Stipules absent or very soon caducous, minute, only up to 0.3 mm long:
4a Leaves and branches densely silver-hirsute; found along the Olifants River and probably the Berg River
(Western Cape) and their tributaries 5b. S. mucronata subsp. hirsuta
4b Leaves and branches glabrous or grey-canescent; not found along the Olifants River and probably the Berg
River (Western Cape):
5a Leaves (10 — )15 — 40 mm wide; found in N Namibia and Botswana 5d. S. mucronata subsp. subserrata
5b Leaves 4 — 10(— 15) mm wide; absent in N Namibia and Botswana:
6a Summer leaves usually shorter than 55 mm; petioles 2-5 mm long; twigs always glabrous; found mainly
along Orange and Vaal Rivers and their tributaries and rivers in Western and Eastern Cape as far
north as southern KwaZulu-Natal 5a. S. mucronata subsp. mucronata
6b Summer leaves usually longer than 60 mm; petioles 4—14 mm long; twigs grey-canescent to puberulous,
sometimes glabrous; found mainly along the Limpopo, Olifants, Maputo, Komati, Umbuluzi, Tugela
and Black and White Umfolozi Rivers and their tributaries 5c. S. mucronata subsp. woodii
FIGURE 8. — Leaves of Salix mucronata. A, spring leaf of subsp.
mucronata , x 1; B, summer leaf of subsp. mucronata, x 1; C,
spring leaf of subsp. subserrata, x 1; D, summer leaf of subsp.
subserrata, x 1;E, summer leaf of subsp. woodii, x 1 ; F, spring
leaf of subsp. hirsuta, x 1; G, summer leaf of subsp. hirsuta, x
1.A, Berry KMG13218 (PRE); B, Gubb KMG10847 (PRE); C,
Merxmuller & Giess 30494 (PRE); D, Roux 332 (NBG); E,
Dieterlen 6718 (SAM); F, Middlemost 1733 (NBG); G, Van
Jaarsveld 4496 (NBG). Artist: G. Condy.
5. Salix mucronata Thunb., Prodromus plantarum
capensium: 6 (1794); Willd.: 685 (1806); Thunb.: 31
(1807); Andersson: 14 (1867); Burtt Davy: 70 (1922);
Friis: 154 (1992); McKean: 84 (1996); Coates Palgrave:
121 (2002); R.H. Archer & Jordaan: 92 (2005). Type:
Cape, Thunberg s.n. UPS23065 [lecto., designated by
Immelman (1987), IDC microfiche 1063/968!].
Tree with branches ascending or drooping, slender or
stout. Bark dark grey or brown, becoming deeply verti-
cally fissured with age; branches glabrous or with dense
grey canescence, becoming glabrous with age, often red-
dish when young. Leaves dimorphic, small, entire spring
leaves followed by much larger toothed ones, broadly
lanceolate to elliptic, 25-160 x 3-15 mm, apex long
acuminate to acute to obtuse, base cuneate, usually dis-
colorous, silvery hairy to puberulous to glabrous, soon
glabrous on both surfaces or only above, margin entire,
subentire, denticulate or serrate, reticulate venation
slightly visible above, hardly visible beneath; petiole
slender or stout, 3-15 mm long, glabrous or pubescent,
often reddish; stipules small, falling soon, serrulate,
glandular on adaxial surface, often absent. Inflorescence :
flowers arranged in dense spikes or catkins. Male catkins
34-70 mm long; bracts usually pubescent, sometimes
densely so, or sometimes glabrous; gland an irregularly
lobed ring, sometimes reduced to one abaxial and one
adaxial gland, 0. 3-1.0 mm long. Male flowers appearing
from midwinter to spring and again in late summer; sta-
mens 5-8. Female catkins 20-35 mm long. Female flow-
ers on pedicels 2^1 mm long. Capsules 4—6 mm long,
dehiscent to release tufted, woolly seed; fruiting stipe
1.5-3. 5 mm long. Seeds ovoid, ± 1 mm long.
5a. subsp. mucronata. Immelman: 173 (1987).
S. capensis Thunb. var. mucronata (Thunb.) Andersson: 14 (1867);
Andersson: 198 (1868); Sim: 329 (1907); Skan: 577 (1925).
S. capensis Thunb.: 31 (1807); Harv.: 309 (1838); Andersson; 197
(1868) excl. vars. mucronata, hirsuta ; Sim: 328 (1907); Marloth: 130
(1913); Skan: 576 (1925) excl. var. mucronata & syn. S. aegyptiaca\
Burtt Davy: 432 (1932); Von Breitenbach: 72 (1965); Friedr.-Holz.: 14
(1967); Jacot Guill.: 161 (1971); Palmer & Pitman: 413 (1972);
Newsholme: 59 (2002). 5. mucronata Thunb. subsp. capensis (Thunb.)
Immelman: 173 (1987); Jordaan: 255 (2002a); Jordaan: 122 (2002b).
14
Bothalia 35,1 (2005)
FIGURE 9. — Leaves and stipules of introduced species of Salix. A, B,
S. caprea, Keet STEU13012 (NBG), x I ; C, D, S. babylonica,
Haugh 517 (NH) & Henderson 671 (NH), x 1; E, S. fragilis
subsp. fragilis, Henderson 786 (PRE), x 1. Artist: G. Condy.
Type: Northern Cape, near rivers in mountains near Hantam, Thunberg
s.n. ( UPS22958 , lecto., designated here, IDC microfiche 1063/963!).
S. gariepina Burch.: 317, t. 6 (1824); Burtt Davy: 338 (1921); Von
Breitenbach: 73 (1965). S. capensis var. gariepina (Burch.) Andersson:
13 (1867); Andersson: 197 (1868); Sim: 329 (1907); Skan: 579 (1925);
Burtt Davy: 432 (1932); Newsholme: 118 (2002). Syntypes: Northern
Cape: Prieska District, banks of the Orange River, Burchell 1637 (K!)
and Orange River, near Petrusville, Burchell 2669 (K!).
S. mucronata var. caffra Burtt Davy: 71 (1922). Type: Eastern
Cape, Eastern Districts, Cooper 48 (PRE!, holo.; BM!, K!, iso.).
Tree with dense to sparse drooping crown, up to 10 m
tall, occasionally attaining 15-20 m in height, with stem
diam. of 300-600 mm; twigs always slender, glabrous,
yellow or red. Bark becoming rough, deeply furrowed
with age; branchlets terete, slender, ± pubescent when
young, quite glabrous, often shining and reddish to dark
brown when older; wood white to purplish white, satiny,
soft, light and brittle. Leaves thin, becoming leathery,
few reddish hairs when young, otherwise glabrous, green
on both sides (Figure 8A); summer leaves (30-)40-55
(-75) x 5-11 mm (Figure 8B), apex acuminate or acute,
base cuneate, margin closely or remotely serrulate, rarely
entire; petiole glabrous, 2-5 (-7) mm long.
In this treatment, subsp. mucronata is considered to
be the same as subsp. capensis and partially as subsp.
mucronata of Immelman (1987). This view is different
from Jordaan (2002a) and Coates Palgrave (2002), where
subsp. mucronata was applied to the tropical plants and
is considered in this treatment as subsp. subserrata, be-
cause of the correct application of the type of Salix
mucronata (see elsewhere in this publication). It has the
smallest summer leaves of all the subspecies, usually
shorter than 55 mm and narrower than 11 mm, and is
commonly known as the small-leaved willow.
It is the most widespread wild willow and is found main-
ly in the drainage basins of the Vaal and Orange Rivers and
their tributaries in North-West, Free State, Lesotho,
Northern Cape and southern Namibia, and rivers in Western
and Eastern Cape as far north as the Umzimkulwana River
in KwaZulu-Natal (Figure 10). It occurs on islands or near
and on the banks of streams and rivers in bushveld and
grassland at altitudes of 600-2 000 m.
Vouchers: Burtt Davy 1503 (BOL, PRE); Dieterlen 314A (NH, PRE,
SAM); Jacot Guillarmod 7384 (GRA, PRE); McDonald 914 (NBG,
PRE); Merxmiiller & Giess 2270 (WIND); Van Wyk BSA325 (PRU).
5b. subsp. hirsuta ( Thunb .) Immelman in Bothalia
17: 173 (1987). Type: Cape, Thunberg 23038 [UPS, lecto.,
incorrectly designated by Immelman (1987) as UPS23028,
microfiche in PRE!].
5. hirsuta Thunb.: 6 (1794); Willd.: 695 (1806); Thunb.: 31 (1823);
Skan: 579 (1925); Von Breitenbach: 74 (1965); Palmer & Pitman: 416
(1972); Coates Palgrave: 91 (1977). S. capensis var. hirsuta (Thunb.)
Andersson: 14 (1867); Andersson: 198 (1868); Sim: 329 (1907).
12 14 16 18 20 22 24 26 28 30 32
FIGURE 10. — Distribution of Salix mucronata subsp. mucronata in
southern Africa.
Bothalia 35,1 (2005)
15
Shrub or small tree 2-4 m tall, with yellow winter buds;
branches rather stout, densely silver-hirsute. Bark smooth,
grey; branches at first densely whitish villous, becoming
glabrescent, red-brown and somewhat rugose; branchlets
angular. Leaves oblong to lanceolate, 20-70 x (5—) 10— 15
mm (Figure 8F, G), apex mucronate, acute or acuminate,
base round to cuneate, thinly covered above and densely
beneath with grey silky hairs, becoming glabrous, margin
entire or sometimes remotely and obscurely serrulate; peti-
ole short, 1-6 mm long, densely hirsute; stipules brown,
membranous, obliquely ovate, ± 1.5 mm long, silky-hairy,
soon deciduous. Flowering tune : September to October.
Fruiting time: October to November.
S. mucronata subsp. hirsuta is the most easily disting-
uished subspecies because of its grey silky indumentum
on branchlets, stipules and leaves. Commonly known as
the silver willow and confined to the Western Cape along
the Olifants River and probably the Berg River and their
tributaries (Figure 11).
Krauss (1845), Skan (1925) and Adamson (1950)
cited the distribution of S. hirsuta as the Cape Peninsula,
at Hout Bay, Bergvliet and rivulets near Constantia and
Stellenbosch. The natural habitat in the Cape Peninsula
and surroundings has been largely destroyed and the dis-
tribution of subsp. hirsuta given by these early authors
has raised the suspicion that it might have been much
wider until early in the last century or they must have
confused these trees with subsp. mucronata.
Vouchers: Lewis 3524 (SAM); Marloth 11035 (NBG, PRE); Pillans
9831 (PRE); Van Jaarsveld 4496 (NBG, PRE); Wagener 217 (NBG).
5c. subsp. woodii ( Seemen ) Immelman in Bothalia
17: 176 (1987). Type; Natal [KwaZulu-Natal], Upper
Tugela River, near Colenso, Wood 4970 (NH, holo.!).
S. woodii Seemen: 53 (1896); J.M.Wood: 121 (1907); Marloth: 130
(1913); Bunt Davy: 339 (1921); Bews: 79 (1921); Skan: 577 (1925);
Bum Davy; 432 (1932); Von Breitenbach: 70 (1965): Jacot Guill.: 161
(1971); Palmer & Pitman: 415 (1972); Compton: 172 (1976).
S. natalensis Wimm. ex Andersson: 14 (1867); Andersson: 198
(1868). Type: Natal [KwaZulu-Natal], Port Natal (Herb. Vindob.
Gueinzius 136).
FIGURE 11. — Distribution of Salix mucronata subsp. hirsuta, ▲;
subsp. subserrata. •; and subsp. woodii, ■. in southern Africa.
S. wilmsii Seemen: 9 (1900); Marloth: 130 (1913); Burtt Davy: 340
(1921); Bum Davy: 432 (1932). S. woodii var. wilmsii (Seemen) Skan:
578 (1925). S. mucronata Thunb. subsp. wilmsii (Seemen) Immelman:
176 (1987), syn. nov. Type: Eastern Transvaal [Mpumalanga],
Lydenburg District, Wilms 1350 [PRE, lecto.!. designated by Immel-
man (1987): BOL, isolecto.!].
S. wilmsii x safsaf Burtt Davy: 432 (1932). Type: Mpumalanga,
Lydenburg. Grootfontein River, foot of Burgers Pass, Burtt Davy
HI 559 (PRE, holo.!). Burtt Davy also cites: Eastern Transvaal
[Mpumalanga], Barberton, Pole Evans H2965 (PRE!).
S. wilmsii x woodii Burtt Davy: 432 (1932). Type: Eastern Trans-
vaal [Mpumalanga], Barberton, Galpin 1278 (GRA!, PRE!).
An arborescent shrub or small tree 2-10 m tall, with
long, lax, somewhat drooping branches, much-branched;
branches always stout, usually sparsely to densely grey-
canescent, rarely puberulous to glabrous. Bark dark
brown, deeply fissured; young twigs red. Leaves narrow-
ly lanceolate, (50-)60-160 x 5— 1 5(— 22 ) mm (Figure 8E),
apex long-acuminate, tapering at base, smooth to slight-
ly hairy, pale green above, with greyish bloom below,
with dense, grey, silky indumentum, sometimes hairy,
becoming glabrescent and glaucescent with age, margin
entire or serrulate; petiole 2-14 mm long. Flowering
time : August to September. Fruiting time: October to
April. Figure 12.
The width of the summer leaves of this subspecies
might overlap with that of the other subspecies but it has
the longest leaves, up to 160 mm. Commonly known as
the flute willow because suitable lengths of the branches
are used for making toy whistles (Smith 1966). Occurs in
the drainage basins of the Limpopo, Crocodile, Olifants,
Komati, Umbuluzi and Maputo Rivers and adjacent
streams in Limpopo, North-West, Gauteng and Mpuma-
langa, and of the Umbuluzi River in Swaziland. Also in
the Tugela and Black and White Umfolozi River basins
in the eastern Free State, Lesotho and KwaZulu-Natal
(Figure 1 1 ).
There are no constant distinguishing characters that
separate S. mucronata subsp. woodii and subsp. wilmsii
(Coates-Palgrave 1977). Specimens have entire to serrate
leaves and the young twigs are all hirsute, becoming
glabrous with age. Therefore only one taxon occurs in
the eastern parts of the Drakensberg Escarpment in the
drainage lines of the rivers that run into the Indian
Ocean, from the Umzimkulu in the south to the Komati
and Maputo Rivers in the north. The subspecific epithet
woodii is the oldest name and has therefore been used.
Vouchers: Balkwill & Cadman 2472 (J, PRU); Bayer & McClean
101 (BOL, PRE); Compton 31168 (NBG, NH, PRE); Dieterlen 314B
(NH, PRE): Galpin 1278 (BOL, SAM); Theron 3569 (GRA. PRU).
5d. subsp. subserrata ( Willd. ) R.H. Archer & Jordaan
in Bothalia 35: 92 (2005). Type; Egypt, near Cairo,
Bulak, Herb. Willd. 18137 (B-WILLD, holo., fragm.,
IDC microfiche 7440-30/1313!).
S. subserrata Willd.: 671 (1806); Milne-Redh.: 474 (1936); Meikle:
588 (1958); D.R.Maire: 50 (1961); Von Breitenbach: 70 (1965);
Friedr.-Holzh.: 14 (1967); Leonard & Geerinck: 2 (1967); Palmer &
Pitman: 413 (1972); Wilmot-Dear: 1 (1985); Meikle: 258 (1989);
Wilmot-Dear: 121 (1991).
S. safsaf Forssk.: LXXVI (1775) as S. safsaf baelledi nom. nud. ex
Trautv.: 6, tab. 2 (1836); Andersson: 196 (1868); Boiss.: 1183 (1879);
16
Bothalia 35,1 (2005)
FIGURE 12. — Salix mucronata
subsp. woodii. A, terminal
shoot, x 0.9; B, female flow-
er, x 8.8; C, female catkin, x
8.8; D, dehiscent capsule
with seeds, x 4.4; E, male
flower, x 8.8. A, B, C, E,
Poynton 1H10464\ D, Ober-
meyer TM31027. Artist; G.
Condy.
Skan: 318 (1917); Newsholme: 1 18 (2002). Type: Egypt, Herb. Sieber
(?LE, holo., K? iso.).
S. aegyptiaca sensu Thunb. non Willd.: 30 ( 1806).
Tree up to 12 m tall; branches drooping, slender or
stout, glabrous or with dense grey canescence, becoming
glabrous and often reddish with age. Leaves broadly
lanceolate to elliptic, 40-100(-150) x (10-)15-40 mm
(Figure 8C, D), apex acute to obtuse, base cuneate, olive-
green and glossy above, glaucous below, soon glabrous
on both surfaces, margin subentire or denticulate, reticu-
late venation slightly visible above, hardly visible be-
neath; petiole slender or stout, 3-15 mm long, glabrous
or pubescent. Flowering time : early spring. Fruiting time :
July to December.
S. mucronata subsp. subserrata has the broadest
leaves, up to 40 mm wide, and is widespread in Africa,
from North Africa (Egypt and Arabian Peninsula) south-
wards through tropical Africa to southern Africa in
Namibia along the Kunene and Zambezi Rivers and
abundant along the Chobe River in Botswana. A wide-
spread tree in fringing bush, on sandy soil along rivers
and streams, and on islands in places likely to become
inundated for at least part of the year (Figure 1 1 ).
Vouchers; Maguire 1707 (NBG); Merxmiiller & Giess 30494 (PRE,
WIND); Roux 332 (NBG, PRE); Tinley 1493 (WIND); Van Wyk BSA52
(PRE, PRU); Ward. Ward & Seely 10438 (PRE, WIND).
6. *Salix babylonica L., Species plantarum edn 2:
1017 (1753); Andersson: 212 (1868); Boiss.: 1185 (1879);
A.Camus & E.-G.Camus: 246 (1904); E.F.Linton: 21
(1913); Britton & Brown; (1913); Burtt Davy: 338 (1921);
Burtt Davy: 81 (1922); M.J.Fischer: 311 (1928); Burtt
Davy: 431 (1932); Femald: 506 (1950); D.R.Maire: 57
(1961); A. K. Skvortsov: 28 (1969); Jacot Guill.: 161
(1971 ); Coates Palgrave: 91 (1977); Meikle: 1488 (1985);
Wilmot-Dear: 124 (1991); A.E.van Wyk & P.van Wyk:
154 (1997); L.Hend.: 161 (2001). Type: Orient, Tourne-
fort (Herb. LINNI 158.20, lecto., designated here, micro-
fiche in PRE!).
A weeping tree up to 10 m tall with long, slender,
dense, spreading crown and branches hanging ± vertical-
ly and reaching the ground; stems yellow-brown, terete,
glabrous. Leaves narrowly ovate-lanceolate to linear-
lanceolate, 60-85 x 8-17 mm, tapering into a long fine
acumen (Figure 9C, D), apex asymmetrical, base cuneate,
bright green above, glaucous beneath, glabrous, margin
regularly serrulate; petiole 3-6 mm long, glabrous or
Bothalia 35,1 (2005)
17
12 14 16 18 20 22 24 26 28 30 32
FIGURE 13. — Distribution of Saltx babylonica in southern Africa,
drawn from SAPIA Database, PPRI. 1979-2005.
thinly pubescent; stipules foliaceous, 4-12 mm long,
ovate-lanceolate, dentate, caducous. Inflorescences on
short axillary branches, .appearing with the leaves or
soon after; bracts ovate-lanceolate, glabrescent, axis,
basa! part and margins with long white hairs. Female
catkins 15-25 mm long; female flowers with 1 nectary
gland; ovary glabrous, sessile or subsessile; style very
short, ± 2 mm long; stigmas 2-lobed.
The weeping willow was introduced from Europe,
but originally came from central and southern China and
is planted extensively in southern Africa, along rivers
and around dams. It is one of the most widespread
invaders of watercourses in South Africa and is particu-
larly abundant in the Grassland Biome, where in places,
it forms pure stands along whole river reaches (Hen-
derson 1991a). Only female trees are known to exist in
southern Africa (Poynton 1973) — the University of the
Free State, Bloemfontein, has one male tree — see Hen-
derson (1991c).
Naturalized in southern Africa since at least the time of
Ecklon (± 1823-1832). Now recorded from most of south-
ern Africa where there is permanent water (Figure 13). A
widespread weed in many parts of the world, spreading in
marshy places and along streambanks by means of branch-
es breaking off and taking root. Place of origin uncertain,
probably China. Despite the specific epithet, it does not
occur naturally in the Middle East.
Vouchers: Bum Davy 1887 (NH, PRE); Du Preez 1910 (PRU);
Gubb KMG10838 (PRE); Henderson 671 (NH, PRE); Jacot Guillar-
mod 2970 (GRA, PRE); Oliver 3121 (PRE, NBG).
7. *Salix fragilis L., Species plantarum edn 2;
1017 (1753); Andersson: 209 (1868); Boiss.: 1184 (1879);
A.Camus & E.-G.Camus: 257 (1904); E.F.Finton: 14
(1913); Britton & Brown: (1913); Femald: 505 (1950);
D.R.Maire: 51 (1961); Rech.f.: 45 (1964); Jalas &
Suominen: 14 (1976); A.K.Skvortsov & J.R.Edm.: 707
(1982); F.Hend.: 162 (2001). Type; ‘Habitat in Europae
borealibus’ (Herb. LINN1158.19, lecto., designated here,
microfiche in PRE!).
var. fragilis
A robust tree up to 15 m tall, with a broad rounded
crown; branches ascending; trunk deeply furrowed. Bark
grey; branchlets at first thinly pubescent, becoming
glabrous and brittle with age. Leaves narrowly lanceolate
to slightly rhomboid-lanceolate, 60-160 x 15-40 mm
(Figure 9E), acute at both ends, glabrous, dark glossy
above, glaucous below, margin coarsely serrulate; petiole
5-15 mm long, with 2 glands at base of leaf; stipules
small, sublinear, 4—7 mm long, caducous. Inflorescences
axillary, short-stalked, appearing with the leaves, droop-
ing; axis hairy; bracts green, covered with many long,
silky hairs. Female catkins very slender, cylindrical,
60-110 mm long. Female flowers with 2 nectaries; ovary
sessile or shortly pedicellate, narrowly acuminate, slen-
der; style very short, ± 1 mm long; stigmas 2-lobed. No
male plants found in southern Africa.
Commonly known as the crack or brittle willow and
originally from Western Europe. An invasive alien
species less widely distributed than S. babylonica, but in
places forming pure stands along rivers (Henderson
1991a) at high altitudes in Free State, KwaZulu-Natal
and Eastern Cape, with one record from the Cape Penin-
sula (Figure 14).
Vouchers: Compton 21071 (NBG); Fuls 183 (PRE, PRU); Hen-
derson 930 (PRE); Hilliard & Bunt 13432 (NU, PRE); Scott 32 (NH);
Wolley Dod 2517 (BOL).
8. *SaIix caprea L., Species plantarum edn 2: 1020
(1753); Andersson: 222 (1868); Boiss.: 1188 (1879);
A.Camus & E.-G.Camus: 102 (1905); E.F.Linton: 47
(1913); Britton & Brown: (1913); Femald: 517 (1950);
Rech.f.: 50 (1964); Newsholme: 59 (2002). Type:
‘Habitat in Europae siccis’ [ Herb LINN 1158.88, lecto.,
designated by Jonsell & Jarvis: 151 (1994)].
Tree up to 6 m or occasionally up to 9 m tall, with
spreading branches forming a broad crown; branchlets
pubescent, becoming glabrous, reddish or dark brown.
Bark grey, fissured. Leaves broadly ovate to ovate-
oblong, 60-160 x (8— )20— 50(— 80) mm (Figure 9A, B),
FIGURE 14. — Distribution of Salix fragilis var. fragilis in southern
Africa, drawn from SAPIA Database, PPRI, 1979-2005.
18
Bothalia 35,1 (2005)
12 14 16 18 20 22 24 26 28 30 32
FIGURE 15. — Distribution of Salix caprea in southern Africa, drawn
from SAPIA Database, PPRI, 1979-2005.
apex acute with a twisted tip, base rounded or cordate,
glabrescent and green above, persistently softly pubes-
cent beneath, margin undulate, crenate or entire; petiole
up to 6 mm long; stipules semicordate, acuminate, often
dentate, persistent. Inflorescences axillary, catkins, silky,
silver-grey, appearing before leaves in late winter, sub-
sessile; bracts narrowly ovate, subacute, with long silky
hairs, brownish below black upper portion. Male catkins
rather stout, 30-45 x 15-23 mm. Male flowers : stamens
2, long-exserted, ± 10 mm long; nectary cylindrical,
short. Female catkins very slender, 20-30 mm long,
cylindrical. Female flowers: nectary 1; ovary hairy, pedi-
cellate, pedicel much longer than nectary; style very
shorter; stigma small, undivided. Capsules ovoid, up to
10 mm long, subsessile on axis. Seeds 1.0 x 0.4 mm,
cylindrical, with fine vertical ridges, grey.
Commonly known as the goat willow or great sallow.
It was introduced from Europe and has been recorded as
an occasional escape from cultivation in KwaZulu-Natal
and the northeastern parts of Eastern Cape (Henderson
1991c) (Figure 15).
Vouchers: Bester 1001 (PRU); Henderson 668, 670 (PRE); Jacobs
8572 (PRE); Keet STEU 13012 (NBG); Loock NBG18870 (NBG).
SPECIES INSUFFICIENTLY KNOWN
Salix crateradenia Seemen: 9 (1900); Skan: 578 (1925).
Type: Botswana, Passarge 41 of 1896 (not found). From
the description and locality this is probably S. mucrona-
ta subsp. subserrata.
Salix mucronata var. integra Burtt Davy: 70 (1922).
Type; Eastern Cape, Camdeboo, on the flats and at the
river near the Camdebooberg, 2000-3000 ft [615-923
m], Drege s.n. (K!, S!). The Stockholm specimen is S.
mucronata subsp. mucronata, but the Kew specimen is
subsp. hirsuta, with large entire leaves and dense silvery
pubescence on the young parts. No other specimen of
subsp. hirsuta from the Camdeboo has been seen, and
neither did Drege collect it at the Olifants River, where
subsp. hirsuta occurs. It is possible that the Kew speci-
men has been mislabelled.
Salix woodii var. grandifolia Burtt Davy: 432 (1932)
nom. nud. Specimens cited: Burtt Davy 10614\ Legat
H433F, Robertson 1474. Burtt Davy speculated that this
could be a hybrid between S. woodii and S. wilmsii.
Salix woodii x safsafl Burtt Davy: 432 (1932). Type:
northern Transvaal [Limpopo], Louis Trichardt, 3100 ft
[954 m], Rogers 21690 (not found). This is S. mucrona-
ta subsp. woodii.
SPECIMENS EXAMINED
Abbott 2696 (5a) PRU. Acocks H828 (5a); 10120 (5c) NH, PRE.
Adamson 1002 (5a) BOL, PRE. Archibald 3967 (5a) GRA; 5238 (5a) PRE.
Balkwill <£ Cadman 2472 (5c) J, PRU. Barker 4828. 6257 (5a) NBG.
Barnard SAM36109 (5a) SAM. Bayer 1873 (5c) NH. Bayer &
McClean 101 (5c) BOL. PRE. Bayliss BRI.B.87 (5a) GRA, PRE;
BRI.B. 159, BRI.B.1358 (7) PRE; BRI.B.6160 (6) PRE. Berry
KMG13219 (5a) PRE. Biggs 5 (5c) PRE. Bohnen 8896 (5a) NBG.
Bolus 468 (5a) BOL; A7767 (5a) GRA. Botha 793, 1447 (5c) PRE.
PRU. Botha & Ubbink 1033 (5a) PRE. Boucher 1985 (5b) NBG, PRE;
3509 (7) NBG, PRE. Bourquin 888 (5a) PRE. Brink s.n. (5a) PRE.
Britten 2439 (5a) GRA, PRE. Brown 557 (5b) NBG. Buckle 13980 (5c)
BOL. Buitendag 616. 997 (5c) NBG, PRE; 1082 (5c) NBG. Burke 330
(5c) BOL. Burrows 3785 (5a) GRA. Burtt Davy 605 (5c) PRE; 1266
(5c) BOL; 1503 (5a) BOL, PRE; 1559, 1584 (5c) PRE; 1887 (6) NH,
PRE; 5185 (5c) GRA.
Codd 8246, 10095 (5c) PRE. Codd & De Winter 5558 (5c) PRE.
Coetzee 600 (6) PRE. Compton 4960, 6913 (5b) NBG; 7405, 18349,
21058 (5a) NBG; 19743 (5c) NBG; 21071 (7) NBG; 27697, 30094 (5c)
NBG, PRE; 30731 (5c) NBG, NH; 31168 (5c) NBG, NH, PRE. Cooper
223 (5a) BOL. Culverwell 0183 (5c) PRE.
Devenish 889, 1016, 1464 (5c) PRE. De Winter 7685 (5c) PRE.
Dieterlen 314A (5a) NBG, NH, PRE, SAM; 314B (5c) NH. PRE; 6718
(5c) SAM. Dinter 5132 (5a) PRE, SAM. Dlamini s.n. (5c) PRE.
Duggan & Henderson 15 (2) PRE; 24 (5c) PRE; 52 (6) PRE. Du
Plessis 387 (6) PRE; 427 (1) PRE. Duthie STEU10587, 17701 (5a)
NBG.
Ecklon & Zeyher 147 (5a) BOL, NBG. Edwards 2033 (5c) PRE; 2054
(7) PRE; 2059 (5c) PRE. Ellery 267 (5c) PRE. Erasmus KMG12112
(5a) PRE. Esterhuyse 424 (5d) WIND. Eyles 706 (5d) SAM.
Ferrar 1873 (5a) NH. Flanagan 1417 (5a) NBG. Fourcade 3105 (5a)
BOL, NBG. Fuls 72 (5a) PRE; 181 (5a) PRU; 183 (7) PRE; 196 (5c)
PRE.
Galpin 1278 (5c) BOL, SAM; 7873 (5a) PRE; 9157, 12169 (5c) PRE;
BOL13733, BOL21497 (5c) BOL. Geldenhuys 276, 343 (5a) NBG; 344
(5a) NBG, PRE. Gemmell 6474 (5a) PRE. Germishuizen 6093 (5c)
PRE. Gerstner 3852 (5c) NH; 5492 (5c) PRE. Gertenbach &
Groenewald 9221 (1) PRE. Giess, Volk & Bleissner 5399 (5a) PRE.
Gibbs Russell et al. 169 (4) PRE. Giffen 249 (5a) PRE; G940 (5a)
GRA. Gillett 1946 (5a) NBG; 2955 (5c) BOL. Gilmore 428 (5c) PRE.
Glen 2473, 3092 (5c) NH; 3640 (5a) PRE. Goldblatt 3278 (5b) NBG.
Goossens 1144 (5a) PRE. Gubb KMG10755 (3) PRE; KMG10838,
KMG10840, KMG11175 (6) PRE; KMG10845, KMG10847, KMG12165
(5a) PRE.
Hafstrom H961 (5a) PRE. Haugh 517 (6) NH. Hanekom 1272 ( 5b)
NBG, PRE. Hardy 401 , 975 (5c) PRE. Hemm 452 (5c) PRE, PRU.
Henderson 628, 643, 653 (1) PRE; 595, 679, 794, 906, 996, 1017,
1090, 1102 (2) PRE; 1088 (3) PRE; 989 (5a) PRE; 823 (5c) PRE; 671
(6) NH; 681, 782, 807, 833, 897, 903, 919, 928, 931, 976, 988, 995 (6)
PRE; 784, 786, 788, 798, 814, 824, 828, 898, 930, 992, 999, 1004 (7)
PRE; 668, 670 (8) PRE. Henrici 3090 (5a) PRE. Herbs! 21 (5a) PRU.
Herre STEU12072 (5a) NBG; STEU20350 (5d) NBG; STEU24283 (2)
NBG. Heson NH26940 (5c) NH. Hilliard & Burtt 13257 (5a) NU,
PRE; 13432 (7) NU, PRE. Hugo 746 ( 5b) NBG, PRE. Hutchinson,
Forbes <6 Verdoorn 139 (5c) NH.
Immelman PRE60975 ( 1 ) PRE.
Jacobs 8572 (8) PRE. Jacot Guillarmod 2970 (6) PRE; 2972, 2973,
7384 (5a) GRA, PRE; 8691 (5c) GRA. Jarman 22 ( 1 ) PRE. Jenkins s.n.
Bothalia 35,1 (2005)
19
(5b) PRE; JMH8I68 (5c) BOL. Jordaan 356, 3528 (2) PRE. Jurgens
22419 (5a) PRE.
Keet 1463, 1567 (5c) NBG: STEU13011 (5a) NBG: STEU13012 (8)
NBG; STEU13015, STEU24285 (2) NBG. Killick 510 (6) PRE; 4309
(7) PRE; 4383 (5a) PRE. Killick & Marais 2130 (5c) PRE. Kinges 1797
(6) PRE. Kluge 48 (2) NBG; 888 (1) PRE. Krynauw 281 (5c) PRE.
Leendertz 4272 (5c) PRE. Legat 161 (5c) PRE. Leistner 1342 (5a) PRE.
Leistner et al. 110 (5d) PRE. Le Roux & Ramsey 219 (5a) NBG, PRU;
811 (5a) NBG. PRE. Letty 483 (5c) PRE. Lewis 1511 (5c) SAM; 2922
(5a) SAM; 3534 (5b) SAM. Long 769 (5a) PRE. Loock NBG18870 (8)
NBG. Louw 1467 (5a) PRE.
MacMurtry 1659 (5c) PRE. MacOwan 1645 (5a) SAM. Maguire 1707
(5d) NBG; 1974 (5a) NBG. Marais 343 (7) PRE. Marloth 817 , 6188
(5a) NBG; 11035 (5b) PRE, NBG; 11493 (5b) NBG; 8941, 11843,
12387 (5a) NBG, PRE. McDonald 914 (5a) NBG. PRE. Merxmiiller &
Giess 2270 (5a) PRE, WIND; 3661 (5a) PRE; 30494 (5d) PRE, WIND.
Metelerkamp 380 (5a) BOL. Meyer 987 (5c) PRE. Middlemost 1733
(5b) NBG; 2168 (5a) NBG. Moffett 627 (6) NBG, PRE. Mogg 7543.
14134, 19529, 37129 (5c) PRE. Moll 5278 (5c) NH; 5290 (5c) NH.
PRE. Muir 3857 (6) PRE. Muller 1099 (5a) PRE. Munro s.n. (5c) PRE.
Noel RU11610 (5a) GRA.
Oates 319 (5a) PRE. Obermeyer 576, TM31027 (5c) PRE. Oliver s.n.
(5b) NBG; 3121 (6) NBG, PRE; STE30264 (5a) NBG. Olivier 1432 (2)
NBG. Onderstall 1173, 1346 (5c) PRE. Ortendahl 294 (5a) PRE.
Pearson 3111, 3255 (5a) SAM. Pegler 529, 937 (5a) BOL, PRE.
Penzhom 5805 (5a) PRE. Phillips MOSS1323 (5b) PRE; 1403 (5a)
PRE. Pillans 5101 (5a) BOL; 9831 (5b) PRE. Player 70 (5c) PRE.
Powrie 107 (5a) PRE. Poynton 17889 (5a) PRE. Pringle 14790 (5a)
PRE. Prior 33 (5c) PRE. Prosser 1036 (5c) PRE.
Ramsay 1600 (5a) GRA. Range 595, 1551 (5a) SAM. Rattray 121 (5a)
PRE. Rehmann 6509 (5c) BOL, NBG. Repton 3485 (5c) PRE. Retief &
Herman 114 (5c) PRE. Roberts STEU 17438 (2) NBG. Rodger 3858 (6)
GRA. Rodin 3687 (5a) BOL, PRE. Rogers 2736 ( 5c) BOL, GRA;
TM4759, TM4834 (5c) BOL, PRE. Ross 2032 (5c) NH. Roux 332 (5d)
NBG, PRE; 1212 (5a) NBG; 1231 (5c) NBG, PRE. Rudatis 1243 (5c)
NBG.
Salisbury 443 (5a) GRA. Scharf 1491 (5a) PRE. Scheepers 1196 (5c)
PRE. Schelpe 8121 (5a) BOL. Schlechter 3938 (5c) BOL; 9016, 9017
(5a) NBG. PRE. Schmitz 8844, 9355 (7) PRE. Scott 32 (7) NH.
Shearing 79, 654 (5a) PRE. Sim 1 (7) PRE; 1503 (5a) PRE. Smith 6054
(5c) PRE. Smook 7268 (5a) PRE; 7317 (7) PRE; 8023 (5a) PRE. Smuts
PRE51969 (5c) PRE. Smuts & Pole Evans 857 (5c) NBG, PRE. Stayt
5 (5c) PRE. Steynberg 714 (5c) PRE. Strey 5866 (5a) NH; 7695 (6) NH;
10995 (5c) NH. PRE. Sutton 295 (5a), 907 (5c) PRE.
Taylor 370, 557 (5a) NBG. PRE; 11799 (5b) NBG. Teague 125 (5d)
SAM. Theron 1249, 1312 (5a) PRE; 1402 (5c) PRE, PRU; 3569 (5c)
GRA, PRU. Thode 8175 (5c) NBG. Thompson 1545 (5b) NBG. PRE.
Thomcroft 586 (5c) NH; 2150, 3007 (5c) PRE. Tinley 1493 (5d)
WIND. Trauseld 623 (5c) PRE; 818 (6) PRE. Troughton 226 (5a) PRE.
Tyson 2504 (5a) SAM.
Ubbink 988 (5c) PRE.
Van der Merwe 29 (6) PRE. Van der Schijff 51, 357, 812, 4750 (5c)
PRE. PRU. Van der Westhuizen 44/78 (5a) NBG, PRE; 97/80 (5a)
NBG. Van Graan & Hardy 459 (5c) PRE. Van Jaarsveld 2410, 2411
(5b) NBG; 2596 (5a) NBG, PRE; 2640, 2647, 2651, 2711 (5a) NBG;
4496 (5b) NBG, PRE; 4621, 5741, 6853A (5a) NBG. Van Jaarsveld &
Bean 5788 (5a) NBG, PRE. Van Jaarsveld, Forrester & Jacobs 8448
(5a) NBG. Van Rensburg 27 (6) PRE. Van Wvk 1843 (5a) PRU. Van Wvk
BSA52 (5d) PRE, PRU; BSA3102 (5d) PRU; 169 (6) PRE. Van Wyk &
Kok 5781 (5a) PRE. Van Wvk, Retief & Herman 6737 (5a) PRE. Venter
8124 (5a) PRE; 11834 (5c)’ PRE. Viviers 903 (5a) NBG.
Wagener 217 (5b) NBG. Walgate BOL33831 (5a) BOL. Ward 2224 (5c)
GRA; 2225 (5c) NH. Ward, Ward & Seely 10438 (5d) PRE, WIND. West
496 (5c) PRE; 1329 (7) PRE. Westphal TM25 (5c) PRE. Wilman 2600
(5a) BOL. Wdms 1350 (5c) BOL, PRE. Wolley Dod 2517 (7) BOL.
Wood 6669 (5c) NH; 9769 (5c) SAM.
Zambatis 169 (4) PRE; 1028 (5c) PRE. Zietsman & Zietsman 1247 (5a)
PRE.
REFERENCES
ADAMSON, R.S. 1950. Salicaceae Lindl. In R.S. Adamson & T.M.
Salter, Flora of the Cape Peninsula. Juta, Cape Town.
ANDERSSON, N.J. 1867. Monographia Salicum hucusque cognitarum.
Kungliga Svenska vetenskapsakademiens handlingar 6,1: 1-180.
ANDERSSON, N.J. 1868. Salicineae. Salix. In A.P de Candolle, Pro-
dromus systematis naturalis regni vegetabilis 16,2: 191-323.
Masson, Paris.
ARCHER, R.H. & JORDAAN, M. 2005. Salix: the correct application
of the name Salix mucronata, and a new combination. Bothalia
35: 92.
ARGUS, G.W. 1973. The genus Salix in Alaska and the Yukon.
National Museums of Canada, Publications in Botany 2: 1-279.
ARGUS, G.W. 1974. An experimental study of hybridization and pol-
lination in Salix (willow). Canadian Journal of Botany 52:
1613-1619.
ARGUS, G.W. 1997. Infrageneric classification of Salix (Salicaceae) in
the New World. Systematic Botany Monographs, vol. 52. The
American Society of Plant Taxonomists, USA.
BAILEY, L.H. & BAILEY, E.Z. 1976. Popidus. Hortus third, a concise
dictionary of plants cultivated in the United States and Canada.
Macmillan, New York.
BALL, C.R. 1961. Salix. In C.L. Lundell, Flora of Texas 3: 369-372.
Texas Research Foundation. Texas.
BENTHAM. G. 1 880. Salicineae. In G. Bentham & J.D. Hooker, Genera
plantarum 3. Reeve, London.
BEWS, J.W. 1921. Salicaceae. The flora of Natal & Zululand. City
Printing Works, Pietermaritzburg.
BOISSIER, PE. 1879. Salicineae. Flora onentalis 4, 2: 1181-1194. Georg,
Geneve.
BRITTON, N.L. & BROWN, A. 1913. An illustrated flora of the north-
ern United States, Canada and the British possessions, edn 2,
1: 587, 591. Scribner, New York.
BRITTON, N.L. & SHAFER. J.A. 1908. North American trees. Henry
Holt, New York.
BROWICZ, K. & YALTIRIK, F. 1982. Salicaceae. Populus. In PH.
Davis, Flora of Turkey 7; 716-720. Edinburgh University Press,
Edinburgh.
BUGALA, W. 1967. Systematyka euroazjatyckich topoli z grupy
Populus nigra L. Arboretum Komickio 12: 45-219.
BURCHELL, W.J. 1824. Travels in the interior of southern Africa, vol.
2. Longman. London.
BURTT DAVY, J. 1921. New or noteworthy South African plants IV.
Kew Bulletin 1921: 335-343.
BURTT DAVY, J. 1922. The distribution and origin of Salix in South
Africa. Journal of Ecology 10: 62-86.
BURTT DAVY, J. 1932. Salicaceae. A manual of the flowering plants
and ferns of the Transvaal with Swaziland 1 . Longmans &
Green, London.
BUTCHER, R.W. 1961. Salicaceae. A new illustrated British flora 1:
966-987. Leonard Hill, London.
CAMUS, A. & CAMUS, E.-G. 1904. Classification des Saules d’Europe
et monographic des Saules de France. Journal de Botanique 18:
245-296.
CAMUS, A. & CAMUS, E.-G. 1905. Classification des Saules d’Europe
et monographic des Saules de France. Journal de Botanique 19:
87-144.
CHALMERS SMITH, E. 1943. A study of cytology and speciation in
the genus Populus. Journal of the Arnold Arboretum 24: 275-305.
COATES PALGRAVE, K. 1977. Trees of southern Africa. Struik, Cape
Town.
COATES PALGRAVE, M. 2002. Keith Coates Palgrave Trees of south-
ern Africa, edn 3. Struik, Cape Town.
COMPTON. R.H. 1976. Flora of Swaziland. Journal of South African
Botany, Suppl. 11: 172.
CRONQUIST, A. 1981. An integrated system of classification of flower-
ing plants. Columbia University Press, New York.
DU ROI, J.P. 1772. Die Harbkesche wilde Baumzucht. Waisenhaus
Buchhandlung, Braunschweig.
DYER, R.A. 1975. Salicaceae. The genera of southern African flower-
ing plants, vol. 1 . Department of Agricultural Technical Services,
Pretoria.
ECKENWALDER, J.E. 1977. North American cottonwoods ( Populus ,
Salicaceae) of sections Abaso and Aigeiros. Journal of the
Arnold Arboretum 58: 193-208.
ENGLER, H.G.A. 1915. Salicaceae. Die Pflanzenwelt Afrikas 3, 1:6, 7.
Engelmann, Leipzig.
FERNALD, M.L. 1950. Salicaceae. In A. Gray, Manual of botany, edn
8: 487-523. American Book Company, New York.
20
Bothalia 35,1 (2005)
FISHER. M.J. 1928. The morphology and anatomy of the flowers
of the Salicaceae. American Journal of Botany 15: 307-326,
372-394.
FORSSKAL, P. 1775. Flora aegyptiaco-arabica: LXXVI, 170. Moller,
Copenhagen.
FRANCO, J.M.A. do A. 1964. Salicaceae. Populus. In T.G. Turin el al..
Flora europaea 1: 53-55. Cambridge University Press, Cam-
bridge.
FRIEDRICH-HOLZHAMMER, M. 1967. Salicaceae. Prodromus einer
Flora von Siidwestafrika 14: 1, 2.
FRIIS, I. 1992. Salix. Forests and forest trees of northeast tropical
Africa. Her Majesty’s Stationery Office, London.
HARVEY, W.H. 1838. Salix. The genera of South African plants.
Robertson, Cape Town.
HENDERSON, L. 1989. Invasive alien woody plants of Natal and the
north-eastern Orange Free State. Bothalia 19: 237-261.
HENDERSON, L. 1991a. Invasive alien plants of the Orange Free
State. Bothalia 21: 73-89.
HENDERSON, L. 1991b. Invasive alien plants of the northern Cape.
Bothalia 21: 177-189.
HENDERSON, L. 1991c. Alien invasive Salix spp. (willows) in the
Grassland Biome of South Africa. South African Forestry
Journal 157: 91-95.
HENDERSON. L. 1992. Invasive alien plants of the eastern Cape.
Bothalia 22: 119-143.
HENDERSON, L. 1998. Invasive alien woody plants of the southern
and southwestern Cape region. South Africa. Bothalia 28:
91-112.
HENDERSON, L. 2001. Alien weeds and invasive plants. Agricultural
Research Council, Pretoria.
HENDERSON, L. & MUSIL, K.J. 1984. Exotic woody plant invaders
of the Transvaal. Bothalia 15: 297-313.
HUBBARD, C.S. 1926. A review into the species of Populus intro-
duced into South Africa. South African Journal of Science 23:
340-360.
IMMELMAN, K.L. 1987. Synopsis of the genus Salix (Salicaceae) in
southern Africa. Bothalia 17: 171-177.
JACOT GUfLLARMOD, A. 1971. Salicaceae. Flora of Lesotho ( Basuto-
land). Cramer, Lehre.
JALAS, J. & SUOM1NEN, J. 1976. Salicaceae. Atlas florae europeae
3: 13-51.
JONSELL, B. 1993. Populus. In C.E. Jarvis, List of Linnean generic
names and their types. Regnum Vegetabile 127. Koeltz Scien-
tific Books, Konigstein.
JONSELL, B. & JARVIS, C.E. 1994. Lectotypification of Linnaean
names for Flora Nordica, vol. 1 (Lycopodiaceae-Papaveraceae).
Nordic Journal of Botany 14: 145-164.
JORDAAN, M. 2000. Salicaceae. In O.A. Leistner, Seed plants of
southern Africa: families and genera. Strelitzia 10: 499, 500.
National Botanical Institute, Pretoria.
JORDAAN, M. 2002a. Abstract: The infraspecific classification of
Salix mucronata Thunb. (Salicaceae) in southern Africa. South
African Journal of Botany 68: 255.
JORDAAN, M. 2002b. Salix. In M. Coates Palgrave, Keith Coates
Palgrave Trees of southern Africa, edn 3: 121, 122. Struik, Cape
Town.
KOEHNE, B.A.E. 1893. Deutsche Dendrologie. Ferdinand Enke, Stuttgart.
KRAUSS, C.F.F. VON. 1845. Beitrage zur Flora des Cap- und
Natallandes: 88. Regensburg.
LEONARD, J. & GEERINCK, D. 1967. Salicaceae. Flore du Congo
du Rwanda et du Burundi, Spermatophytes, Fasc.: I -A. Bruxelles.
LINNAEUS, C. 1753. Species plantarum, edn 2. Laurentius Salvius,
Stockholm.
LINNAEUS, C. 1754. Genera plantarum, edn 5. Laurentius Salvius,
Stockholm.
LINTON, E.F. 1913. The British willows. The Journal of Botany (London)
51, Suppl.: 1-92.
MAIRE, D.R. 1961. Salicales. Flore de I’Afrique du Nord 7: 37-71.
Lechevalier, Paris.
MARLOTH, R. 1913. Salicaceae. The flora of South Africa, vol. 1:
130-131. Darter, Cape Town.
MARSHALL, H. 1 785. Arbustrum americanum. Cruikshank, Philadel-
phia.
MCKEAN, D.R. 1996. Salicaceae. In A.G. Miller & T.A. Cope, Flora
of the Arabian Peninsula and Socotra, vol 1 : 83-85. Edinburgh
University Press, Edinburgh.
MEIKLE, R.D. 1958. Salicaceae. In J. Hitchinson & J.M. Dalziel,
Flora of West tropical Africa, edn 2, 1,2: 588, 589. Crown Agents,
London.
MEIKLE, R.D. 1985. Salicaceae. Flora of Cyprus 2: 1486-1491 . Betham-
Moxon Trust, Royal Botanic Gardens, Kew.
MEIKLE, R.D. 1989. Salicaceae. In I. Hedberg & S. Edwards, Flora of
Ethiopia 3: 258-260.
MEEUSE, A.D.J. 1975. Taxonomic relationships of Salicaceae and
Flacourtiaceae: their bearing on interpretative floral morpholo-
gy and dilleniid phylogeny. Acta Botanica Neerlandica 24:
437-M-57.
MILNE-REDHEAD, E. 1936. Tropical African plants XIV. Kew Bulletin
1936: 469—489.
MOENCH. C. 1785. Verzeichniss auslandischer Bdume und Stauden
des Luftschlosses Weissenstein. Fleischerischen Buchhandlung,
Leipzig.
NEUMANN, A. 1969. Salicaceae. Populus. In K.H. Rechinger, Flora
iranica 65: 1-12. Akademische Druck, Graz-Austria.
NEWSHOLME, C. 2002. Willows, the genus Salix. First paperback edn.
Batsford, London.
PALMER, E. & PITMAN, N. 1972. The willow family (Salicaceae).
Trees of southern Africa, vol. 1 . Balkema, Cape Town.
PAX, F. 1889. Salicaceae. In A. Engler & K. PrantI, Die nattirlichen
Pflanzenfamilien 3,1: 29-37. Engelmann, Leipzig.
POYNTON, R.J. 1973. Two hundred selected indigenous and exotic
species: how to recognise and grow them. In W.F.E. Immelman,
C.L. Wicht & D.P Ackerman, Our green heritage. Tafelberg,
Cape Town.
RECHINGER, K.H. 1964. Salicaceae. Salix. In T.G. Tutin et al.. Flora
europaea, vol. 1: 43-55. Cambridge, University Press, Cambridge.
SEEMEN, O. VON. 1896. Neue Weidenarten in dem Herbar des Konig-
lichen botanischen Museums zu Berlin II. Botanische Jahr-
biicher 21, Beibldtter 53: 50-58.
SEEMEN, O. VON. 1900. Zwei neue Weidenarten aus Siid-Afrika.
Botanische Jahrbiicher 27, Beibldtter 64: 9, 10.
SEEMEN, O. VON. 1908-1909. Salicaceae. In P.F.A. Ascherson & P.
Graebner, Synopsis der mitteleuropdischen Flora 4: 13-350.
Engelman, Leipzig.
SIM, T.R. 1907. Salicineae. The forests and forest flora of the Colony
of the Cape of Good Hope. Taylor & Henderson, Aberdeen.
SKAN, S.A. 1917. Salicineae. In D. Prain, Flora of tropical Africa 6,2:
316-326. Reeve, London.
SKAN, S.A. 1925. Salicineae. In W.T. Thiselton-Dyer, Flora capensis
5,2: 574-579. Reeve, London.
SKVORTSOV, A.K. 1969. Salicaceae. Salix. In K.H. Rechinger, Flora
iranica 65: 12. Akademische Druck, Graz-Austria.
SKVORTSOV. A.K. & EDMONDSON, J.R. 1982. Salicaceae. In PH.
Davis, Flora of Turkey 1: 694—720. Edinburgh University Press,
Edinburgh.
SMITH, C.A. 1966. Common names of South African plants. Memoirs
of the Botanical Survey of South Africa No. 35.
SMITH, J.E. 1804. Salicineae. Flora brittanica 3: 1080. Davis,
London.
THUNBERG, C.P 1794. Prodromus plantarum capensium. Edman,
Uppsala.
THUNBERG, C.P. 1807. Salix. Flora capensis : 30-31. Edman, Uppsala.
THUNBERG, C.P 1823. Flora capensis, edn Schultes. Cottae, Stutt-
gardt.
TRAUTVETTER. C.R. 1 836. Salicetum sive Salicum formae quae hodie
innotuere, descriptae et systematice dispositae: 1-30, t. 1-4.
Academiae Ceasareae Scientiarum, St Petersburg.
VAN STEENIS, C.G.G.J. 1981. Rheophytes of the world. Sijthoff &
Noordhoff, Alphen aan den Rijn, The Netherlands.
VAN WYK, B. & VAN WYK, P. 1997. Field guide to trees of southern
Africa. Struik, Cape Town.
VON BREITENBACH, F. 1965. Salicaceae. The indigenous trees of
southern Africa 2: 68-74. Government Printer, Pretoria.
WESMAEL, A. 1868. Salicineae. Populus. In A.P. de Candolle, Prodromus
systematis naturalis regni vegetabilis 16,1 : 323-331. Masson, Paris.
WILLDENOW, C.L. 1806. Salix. Caroli a Linne, Species plantarum,
edn 4, 1. Nauk, Berlin.
WILLIS, J.C. 1973. Populus, Salix. Dictionary of flowering plants and
ferns, edn 8. University Press, Cambridge.
WILMOT-DEAR, C.M. 1985. Salicaceae. Flora of tropical East Africa:
1-7. Balkema, Rotterdam.
WILMOT-DEAR, C.M. 1991. Salicaceae. In E. Launert & G.V. Pope,
Flora zambesiaca 9,6: 120-124. Flora Zambeziaca Managing
Committee, London.
WOOD, J.M. 1 907. Salicineae. Handbook of the flora of Natal. Bennett
& Davis, Durban.
Bothalia 35,1:21-27 (2005)
Three new species and a new synonym in Strumaria
(Amaryllidaceae: Amaryllideae) from southern Africa
D.A. SNIJMAN*
Keywords: Amaryllidaceae, Amaryllideae, key, new rare species, southern Africa, Strumaria Jacq.
ABSTRACT
Newly described are three species of Strumaria Jacq. subgenus Strumaria. S. prolifera Snijman from the
Kourkammaberg in Namaqualand, South Africa, is a rare species closely allied to S. barbarae Oberm. S. speciosa Snijman
from the Sonberg, southern Namibia and S. luteoloba Snijman from Namuskluft, southern Namibia and the Richtersveld in
Northern Cape, South Africa, are rare species closely related to S. phonolithica Dinter. S. gigantea D.Mull.-Doblies &
U.Miill.-Doblies is formally presented as a new synonym of S. phonolithica. A key to the species in Strumaria subgenus
Strumaria is given.
INTRODUCTION
Strumaria Jacq., a genus belonging to the tribe
Amaryllideae subtribe Strumariinae is endemic to the
semi-arid areas of southern Africa. Most species are
found in the winter rainfall region, whereas only two
taxa extend into the summer rainfall karroid areas.
Among southern African Amaryllidaceae, Strumaria is
second only to Cyrtanthus Aiton (tribe Cyrtantheae) in
terms of the natural rarity of individual species, and as
many as 80% of Strumaria species have been assessed as
rare and threatened (Snijman & Victor 2002).
When last revised, Strumaria encompassed 25 species
(Snijman 1994), unlike the earlier classification of
Miiller-Doblies & Miiller-Doblies (1985), which pre-
sented the group as comprising four small genera:
Strumaria sensu stricto, Gemmaria Salisb., Bokkeveldia
D.Mull.-Doblies & U.Miill.-Doblies, and Tedingea
D.Mull.-Doblies & U.Miill.-Doblies. The results of a
recent phylogenetic study of the Amaryllideae, using
morphology and molecular data from internal tran-
scribed spacer (ITS) sequences (Meerow & Snijman
2001), however, have supported the broad classification
of Strumaria by Snijman (1994), which recognizes sub-
genus Strumaria, subgenus Gemmaria and subgenus
Tedingea.
More recently, several new collections from under-
explored areas of Namaqualand and southern Namibia
have come to hand. Some of these were found to repre-
sent undescribed species that are described here for the
first time. In addition, S. gigantea D.Miill.-Doblies &
U.Miill.-Doblies, which was published after Strumaria
was last revised in 1994, is formally presented as a new
synonym of S. phonolithica Dinter.
Strumaria phonolithica and the three new species all
belong to subgenus Strumaria, bringing the number of
species in the subgenus to eight. Like other members of
* Compton Herbarium, South African National Biodiversity Institute,
Private Bag X7, 7735 Claremont, Cape Town.
MS. received: 2004-07-22.
subgenus Strumaria they have strap-shaped, glabrous
leaves, dorsifixed anthers and a persistent infractes-
cence. Other features that are not consistently present in
all representatives of the subgenus, but which help to
characterise the group, are the three or more leaves ar-
ranged in a spreading fan, a sheathing cataphyll, and the
apparently tubular to hypocrateriform or campanulate
flowers. In addition, the pedicels often approximate the
length of the flowers and the filaments are mostly fused
into a basal tube divided into three nectar wells. A key to
the eight species currently placed in subgenus Strumaria
is given.
Strumaria prolifera Snijman, sp. nov., ex affini-
tate S. barbarae Oberm., sed bulbo prolifero, foliis
oppositis et humo patentibus, perigonio 18-20 mm dif-
fert. Figura 1.
TYPE. — Northern Cape, 2917 (Springbok): Kourkam-
maberg, (-CD), 11-7-1989, Bruyns 3883 (NBG, holo.;
K, PRE).
Deciduous, bulbous herb, 120-200 mm tall in flower.
Bulbs clumped, producing bulblets, subglobose, ± 25
diam.; outer tunics pale brown, parchment-like; neck
slender, up to 20 mm long. Leaves emerging with inflo-
rescence, distichous, 2 or 3, suberect to recurved at
first, spreading flat on ground when mature, narrowly
lorate, 80-180 x 4—8 mm, plane, thin-textured, pale
green, glabrous; apex subacute; outermost 2 leaves
sheathing at base; cataphyll not exserted above ground;
seedlings glabrous. Inflorescence 2-4-flowered, unilat-
erally clustered, 15-30 mm across; scape more or less
erect, roundish in cross section, 120-200 x 2 mm, pale
pink to pale green, glabrous, withering but remaining
attached to bulb when dry; spathe valves 2, narrowly
lanceolate, 10-20 x 3 mm, pink, becoming papery and
spreading, persistent; bracteoles linear, ± 5 mm long;
pedicels lax and pendulous at anthesis, 6-15 x 1 mm,
pale pink to pale green. Perigone actinomorphic, fun-
nel-shaped, 10-15 mm wide at mouth, delicate shell-
pink, drooping, faintly narcissus-scented, collapsing
and turning dark pink when old; tepals 6, free to base,
both whorls adnate to staminal tube for ± 1 mm.
22
Bothalia 35,1 (2005)
FIGURE I . — Strumaria prolifera : A, clumped bulbs and young leaves; B, inflorescence; C, whole flower; D, flower with tepals removed; E, one
inner stamen removed showing fusion of outer stamens with style; F, anther, lateral view; G, anther, dorsal view. Scale bar: A, B, 8.25 mm;
C, 6 mm; D, E, 3 mm; F, G, 1 mm. Drawn from Bruyns 3883. Artist: John Manning.
oblanceolate, 18-20 x 4. 0-4. 5 mm, narrowing to 1.5
mm at base, spreading slightly. Stamens 6, slightly
shorter than tepals, spreading distally; filaments ± 15
mm long, connate proximally into a 3-5 mm long tube,
outer whorl proximally adnate to 3-winged style for ± 3
mm, inner whorl free from style forming ± 3 mm deep
nectar wells; anthers dorsifixed, 2 mm long and cream-
coloured before opening; pollen pale lemon-coloured.
Ovary ovoidal, 2-3 mm diam., trilocular with 2 ovules
per locule, pinkish green; nectaries septal; style 14 mm
long, slightly shorter than stamens, ± 3-winged proxi-
mally; stigma trifid, papillate. Capsule papery, small.
dehiscing loculicidally. Seeds green, fleshy, ± 5 mm
diam.; embryo green.
Phenology, in cultivation, S. prolifera flowers
simultaneously with the bulb’s newly emerging leaves
in the middle of May. Several attempts have been
made to study the species flowering in nature, but
despite visits to the Kourkammaberg over three con-
secutive years in late April and May, the species has
not been seen flowering in the wild. This may suggest
that the bulbs flower infrequently in nature, possibly
only after good autumn rains. The soft, delicate leaves
Bothalia 35,1 (2005)
23
FIGURE 2. — Known distribution of Strumaria prolifera.
remain green throughout winter but die off with the
onset of hot, dry summer conditions at the end of
spring.
Diagnostic features'. Strumaria prolifera is easily rec-
ognized by its pink, funnel-shaped flowers, which enclose
the stamens. Only two other species have similarly
shaped flowers with short, included stamens, namely S.
barbarae Oberm. from southern Namibia and the eastern
Richtersveld, and S. pubescens W.F.Barker from the
Roggeveld escarpment and Laingsburg District. Based on
the stamen and style morphology, S. prolifera shows
greatest affinity with S. barbarae, which belongs to sub-
genus Strumaria. In both species the filaments form a
proximal tube up to 3-5 mm long and through the fusion
of the outer stamens to the three-winged style base, the
tube is divided into three separate hollows, where nectar
collects from the peristylar pores on the ovary dome
(Figure IE). Thus insects seeking nectar are forced to
probe each hollow consecutively. Moreover, both species
have drooping, scented flowers. The perigone in 5. bar-
barae, however, is longer (28-38 mm) and mostly white
when fresh, unlike the shorter (18-20 mm long), distinct-
ly pink perigone in S. prolifera. More divergent is S.
pubescens, belonging to subgenus Gemmaria, that has
spreading to erect flowers in which all six stamens are
adnate to the narrowly trigonous style base. Nectar there-
fore discharges and collects as three small droplets in the
sinuses between the inner filaments and style.
A noteworthy feature is that all the species belonging
to subgenus Strumaria (S. bidentcita Schinz, S. barbarae,
S. hardy ana D.Miill.-Doblies & U.Mtill.-Doblies, S. luteo-
loba Snijman, S. phonolithica, S. speciosa Snijman and
S. truncata Jacq.), except S. prolifera, have erect to fal-
cate leaves arranged in a single fan. Strumaria prolifera
is unique in the subgenus in having straight, soft, oppo-
site leaves, which ultimately spread flat on the ground.
The species is furthermore distinguished by the forma-
tion of bulblets and extremely thin-textured, pale green
leaves.
Distribution and habitat'. Strumaria prolifera is
known only from the Kourkammaberg, an isolated, pre-
dominantly quartzite mountain on the coastal forelands
of Namaqualand (Figure 2). The bulbs grow on southern
slopes near the mountain’s summit, where they are con-
fined to partial shade between large rocks (R Desmet
pers. comm.). The population consists of highly local-
ized colonies of up to 100 plants. Low succulent shrubs
dominate the surrounding vegetation.
Etymology, the species is named prolifera for its char-
acteristic proliferation by offsets, which slowly give rise
to dense colonies of bulbs and apparently compensate for
the bulbs’ infrequent flowering and fruiting.
Strumaria speciosa Snijman, sp. nov., S. phono-
lithicae Dinter affinis, sed floribus nutantibus, pedicellis
30-50 mm longis, perigonio campanulato et staminibus
±10 mm exsertis differt. Figura 3.
TYPE. — Namibia, 2817 (Vioolsdrif): Sonberg, E of
Lorelei and Rosh Pinah, (-AA), 7-7-1997, Bruyns 7233
(NBG, holo.; PRE, WIND).
Deciduous, bulbous herb, 230-330 mm tall in flower.
Bulbs clumped, subglobose, 30-35 mm diam.; outer tunics
tough and pale brown; neck stout, 10-15 mm across,
reaching up to 100 mm above ground. Leaves emerging at
flowering, 4-6, erect to falcate, arranged in a single plane,
broadly lorate. 90-150 x 17-25 mm, plane, pale green,
glabrous; apex subacute; cataphyll unknown. Inflorescence
11-14-flowered, widely and evenly spreading, 80-110
mm across; scape straight, recurved at apex during anthe-
sis, becoming erect in fruit, 260-400 x 3. 0-4.5 mm, pale
green, glaucous, withering and collapsing after fruiting;
spathe valves 2, lanceolate, 15-28 x 2-6 mm, papery,
spreading, pale pink, persistent; bracteoles linear, up to 15
mm long; pedicels firm, mostly straight, curved down-
wards at apex when flowering but completely straight
when fruiting, 30-50 x 1.5 mm, pale green. Perigone
actinomorphic, campanulate and ± 1 1 mm across in lower
half, recurved above, nodding, pure white, honey-scented;
tepals 6, free to base, linear-oblong, 11-12 x 2 mm, nar-
rowing to 0.5 mm at base, closely imbricate in lower half,
channelled, strongly reflexed about midway along their
length, outer whorl reflexing ± 1 mm below inner whorl.
Stamens 6, well exserted, slightly spreading; fdaments
15-18 mm long, connate proximally into a tube, ± 2 mm
long, outer whorl fused to 3-winged style base for ± 3 mm,
inner whorl free from style, forming ± 2 mm deep nectar
wells; anthers dorsifixed, ± 3 mm long and maroon before
opening; pollen cream-coloured. Ovary ovoidal, ± 3 mm
diam., trilocular with 3-5 ovules per locule; nectaries sep-
tal; style ±19 mm long, slightly longer than stamens, 3-
winged in proximal third, slender and columnar distally;
stigma trifid, papillate. Capsule small, papery, ± 7.5 mm
diam., dehiscing loculicidally. Seeds green, fleshy, ± 5 mm
across.
Phenology : in cultivation the species flowers in the middle
of May when the new leaves are present. The leaves remain
green during winter and die back sometime in late spring at
the end of the rainfall season. As yet, the species has not
been recorded flowering in nature.
Diagnostic features : in its vegetative state, Strumaria
speciosa is almost indistinguishable from the southern
24
Bothalia 35,1 (2005)
FIGURE 3. — Strumaria speciosa : A,
bulb and new leaves; B, in-
florescence; C, whole flower;
D, flower with tepals and one
inner stamen removed show-
ing fusion of outer stamens
with style. E-G, anther: E,
ventral view; F, lateral view;
G, dorsal view. Scale bar: A,
B, 8.25 mm; C, 3 mm; D, 1.5
mm; E-G, 1.5 mm. Drawn
from Bruyns 7233. Artist:
John Manning.
Namibian endemic, S. phonolithica Dinter. Both species
have a stout bulb neck that extends well above ground level
(up to 1 00 mm long in S. speciosa and 1 60 mm long in S.
phonolithica). The leaves are broad, 17-25 mm wide in S.
speciosa and 12-50 mm wide in S. phonolithica, and they
curve laterally to form an outspread, erect fan which is
arranged in a single plane. S. speciosa, however, is easily
distinguished when flowering. The scape is recurved api-
cally and the 30-50 mm long pedicels are firm, outspread
and straight, but deflexed distal ly so that the campanulate
flowers are nodding. Furthermore, the tepals are strongly
recurved in the distal half so that the stamens protrude by
almost 10 mm. In contrast, the firm, outspread pedicels in
S. phonolithica never exceed 25 mm long so that the flower
cluster remains compact. The tepals also remain imbricate
for about three-quarters of their length and only recurve
near the tips, thus only the tips of the stamens protrude
from the perigone throat. Although S. phonolithica and S.
speciosa are easily distinguished, their strong morphologi-
cal similarities, nevertheless, emphasize a close alliance.
Distribution and habitat : Strumaria speciosa is
known from only two collections on the Sonberg, a
mountain range in the semi-arid, winter rainfall region of
southern Namibia (Figure 4). The small, localized popu-
lations are found on south-facing slopes below dolomite
outcrops at about 900 m. The bulbs grow in soft, loamy
soil among stones and low, predominantly succulent
shrubs (P.V. Bruyns pers. comm.).
Etymology, this new species is named speciosa since
its splendid head of flowers is unrivalled in the genus.
Bothalia 35,1 (2005)
25
FIGURE 4. — Known distribution of Strumaria speciosa.
Other specimen examined
NAMIBIA. — 2716 (Witputz): Sonberg, (-DD), 3-9-2001, Bruyns
8856 (NBG).
Phenology : in cultivation the bulbs flower in May,
before the new leaves appear. The leaves emerge shortly
after flowering and remain green throughout winter until
September, when they die back and the bulbs become
dormant over the hot, dry summer.
Diagnostic features : the erect, fan-shaped leaf
arrangement and dorsifixed anthers of S. luteoloba are
sufficiently distinctive to place the species in subgenus
Strumaria , but unlike most species in the subgenus, the
filaments of S. luteoloba are not proximally fused into a
tube. Like S. phonolithica the new species has narrow,
closely imbricate tepals. However, S. luteoloba is easily
recognized by the tepals being deeply channelled and
sharply reflexed from ± halfway along their length, and
by the outer tepal whorl reflexing 2-4 mm below the
inner whorl so that the perigone has a distinctive bise-
riate appearance. The flower colour is also unique in the
genus. The perigone is pinkish red near the base and pale
lemon- to cream-coloured in the distal half. S. gemmata
Ker Gawl., belonging to subgenus Gemmaria, is the only
other species with lemon- or cream-coloured flowers but
these are stellate and concolorous.
Strumaria luteoloba Snijman, sp. nov., speciei
subgeneris Strumariae affinis, sed tepalis angustis
canaliculatis biseriatis roseis et citrinis, filis libris differt.
Figura 5.
TYPE. — Namibia, 2716 (Witputz): Namuskluft just
SE of Rosh Pinah, (-DD), collecting date unknown,
Lavranos s.n. (i NBG167717 , holo.; PRE, WIND).
Deciduous, bulbous herb, 180-280 mm tall in flower.
Bulb solitary, subglobose, ± 17.5 mm diam., tunics
parchment-like, brown; neck of short, loose, dry, broken
tunics. Leaves emerging shortly after flowering, 2 or
3(4), erect to slightly falcate, outermost sheathing at
base, ± arranged in a single plane, 254-0 x 4-9 mm,
oblong, glabrous, dark green, glaucous or shiny; apex ±
obtuse; cataphyll unknown. Inflorescence 4-7-flowered,
secund, drooping, slightly spreading, 30-40 mm across;
scape ± straight, recurved near apex while flowering,
200-280 x 1. 5-2.0 mm, glaucous, green, withering and
collapsing after fruiting; spathe valves 2, narrowly lanceo-
late, 15-20 x 2-3 mm, papery, parchment-coloured,
tinged with pink, spreading, persistent; bracteoles few,
linear, up to 9 mm long; pedicels lax, straight to slightly
curved, 8-25 x 1 mm, green. Perigone actinomorphic,
narrow and rose-pink in lower half, ± 2.5 mm across,
recurved and pale lemon- to cream-coloured above, nod-
ding, becoming increasingly pink with age, heavily
scented of fresh coriander; tepals 6, free to base, linear-
oblong, 16-19 x 1. 5-2.0 mm, narrowing to 1 mm at base,
imbricate in lower half, strongly recurved and channelled
in distal half, outer whorl curving backwards 24 mm
before inner whorl. Stamens 6, well exserted, slightly
spreading; filaments 22-25 mm long, free to base, free
from style, outer and inner approximately equally long at
anthesis; anthers dorsifixed, ± 3 mm long and maroon
before opening; pollen whitish. Ovary ovoidal, ± 3 mm
diam., trilocular with 4 or 5 ovules per locule; nectaries
septal; style up to 28 mm long, slightly exceeding sta-
mens, slender throughout; stigma trifid, inner surface
papillate. Capsule unknown. Seeds unknown.
Distribution and habitat : Strumaria luteoloba is
known only from the winter rainfall region of southern
Namibia and the Richtersveld, South Africa (Figure 6).
At Namuskluft, southern Namibia the plants are found in
dolomite-derived soil, whereas the only other known
population is on the Rooiberg (D. Hannon pers. comm.),
a granite mountain massif in the Richtersveld.
Other specimen examined
NORTHERN CAPE. — 2817 (Vioolsdrif): Richtersveld. Rooiberg,
about 6 km NE of Eksteenfontein, (-CD), collecting date unknown,
Lavranos s.n (NBG barcode 0197778. picture only).
A new synonym
Described in 1994, S. gigantea was distinguished
from S. phonolithica by several quantitative characters:
the width of the leaves, the number of flowers per inflores-
cence, the length of the perigone and the size of the
unopened anthers (Miiller-Doblies & Muller- Doblies
FIGURE 6. — Known distribution of Strumaria luteoloba.
26
Bothalia 35,1 (2005)
FIGURE 5. — Strumaria luteoloba: A, bulb and mature leaves; B, inflorescence; C, whole flower; D, one inner stamen removed showing that outer
stamens are free from style. E-G, anther; E, dorsal view; F, lateral view; G, ventral view. Scale bar: A, B, 8.25 mm; C, D, 2.5 mm; E-G,
1.5 mm. Drawn from Lavranos s.n. (NBGI677I7). Artist: John Manning.
Bothalia 35,1 (2005)
27
TABLE 1 . — Diagnostic features of Strumaria plants from the Klinghardt and Aurus Mountains, southern Namibia
* Data for S. phonolithica and S. gigantea given by Miiller-Doblies & Miiller-Doblies (1994).
+ Data for Bruyns 3081 (NBG) not referred to by Miiller-Doblies & Miiller-Doblies (1994).
t Data for 5. phonolithica as amplified in this study.
1994). In all other respects, however, the two species
were regarded as the same, being distinguished from all
other Strumaria species by their flowers, in which the
tepals remain erect and overlapping for most of their
length, and only recurve near the apex. When S. gigantea
was first described, the two species were considered to
be allopatric, each being known from a separate insel-
berg in southern Namibia. Strumaria gigantea was
described from the Aurus Mountains (Miiller-Doblies &
Miiller-Doblies 1994), whereas S. phonolithica was
described from the Klinghardt Mountains (Dinter 1923),
± 60 km to the northwest. In the apparent absence of
intermediate material. S. gigantea was thus separated
from S. phonolithica on size alone. However, Miiller-
Doblies & Miiller-Doblies overlooked a specimen at
NBG (Bruyns 3081), collected on the Aurus Mountains
in 1988, which indicates that the two taxa overlap geo-
graphically and intergrade morphologically. Strumaria
gigantea is thus formally placed into synonymy under S.
phonolithica based on the data given in Table 1 .
Strumaria phonolithica Dinter in Feddes Reper-
torium 19: 178 (1923). Type: Klinghardtgebirge, Aug. 1922,
Dinter s.n. (B, lecto.!), designated by Miiller-Doblies &
Miiller-Doblies (1985).
Strumaria gigantea D.Miill.-Doblies & U.Miill.-Doblies: 346 (1994).
Type: Namibia, Aurus Mountains, 18-9-1988, Miiller-Doblies 88144c
(WIND, holo.; B. BOL, BR. BTU, K, M, MO. PRE, S, STE), not yet
deposited at BOL, PRE, STE, WIND; syn. nov.
Key to species of Strumaria subgenus Strumaria
la Flowers ascending; tepals less than 8 mm long; style broad
in lower half, abruptly narrowed into a slender column
above S. bidentata
lb Flowers pendulous to spreading, rarely ascending; tepals
longer than 8 mm; style at most 3-angled below but
slender throughout:
2a Tepals linear-oblong, less than 5 mm wide, channelled and
strongly recurved in upper parts:
3a Tepals closely imbricate for at least two thirds of their
length, recurved towards their tips; stamens exserted
from perigone throat for less than a quarter of their
length S. phonolithica
3b Tepals closely imbricate in lower half, recurved in upper
half; stamens exserted from perigone throat for ±
half their length:
4a Perigone narrow and rose pink in lower half, ± 2.5 mm
across at throat; tepals pale lemon- to cream-
coloured above at anthesis S. luteoloba
4b Perigone slightly expanded in lower half, ± 11 mm
across at throat; tepals usually pure white at anthe-
sis S. speciosa
2b Tepals oblanceolate, more than 5 mm wide, ± plane, erect
to spreading, not strongly recurved:
5a Stamens included in perigone at anthesis, shorter than
tepals by 3 mm or more:
6a Leaves firm and shiny green; perigone more than 25
mm long, white at anthesis, flushing delicate pink
when old S. barbarae
6b Leaves soft and pale green; perigone less than 25 mm
long, shell pink at anthesis, turning dark pink
when old S. prolifera
5b Stamens exserted from perigone at anthesis, exceeding
tepals by up to 3 mm or more:
7a Leaves plane or slightly undulate, margin hyaline, apex
emarginate; cataphyll subterranean S. hardyatia
7b Leaves twisted or rarely plane, apex entire; cataphyll
reddened and exserted above ground S. truncata
ACKNOWLEDGEMENTS
I am most grateful to Dr RV. Bruyns, Dr P. Desmet
and Mr D. Hannon for contributing data on the. new
species, to Mr G. Duncan for carefully tending the living
bulbs and Dr J.C. Manning for kindly drawing the botani-
cal figures. Mr J. Lavranos is acknowledged for having
sent bulbs of Strumaria luteoloba to Kirstenbosch
Botanical Garden.
REFERENCES
DINTER, K. 1923. Beitrage zur Flora von Siidwestafrika. II. Feddes
Repertorium 19: 177-186.
MEEROW. A. & SNIJMAN. D.A. 2001. Phylogeny of Amaryllidaceae
tribe Amaryllideae based on nrDNA ITS sequences and mor-
phology. American Journal of Botany 88: 2321-2330.
MULLER-DOBLIES, D. & MULLER-DOBLIES, U. 1985. De Lilii-
floris notulae 2. De taxonomia subtribus Strumariinae (Amaryl-
lidaceae). Botanische Jahrbiicher 107: 17-47.
MULLER-DOBLIES, D. & MULLER-DOBLIES, U. 1994. De Lilii-
floris notulae 5. Some new taxa and combinations in the Amaryl-
lidaceae tribe Amaryllideae from arid Southern Africa. Feddes
Repertorium 105: 331-363.
SNIJMAN, D.A. 1994. Systematics of Hessea, Strumaria and Carpolyza
(Amaryllideae: Amaryllidaceae). Contributions from the Bolus
Herbarium 16: 1-162.
SNIJMAN, D.A. & VICTOR, J. 2002. Amaryllidaceae. In J.S. Golding,
Southern African plant Red Data Lists. SABONET Report No.
14: 96, 97, 101, 107, 108.
■
'
.
.
■
Bothalia 35,1: 29-33 (2005)
Lamarck’s new species of Mesembryanthemaceae and the types of
their names
P. CHESSELET* and M. PIGNAL**
Keywords: Lamarck Herbarium, Mesembryanthemaceae, new names, nomenclatural changes, types
ABSTRACT
The holotype of Mesembryanthemum echinatum Lam. is an unpublished, sketchy drawing kept in the Lamarck herbarium
(P-LAM). The holotype of the name M. vaginatum Lam., which has not been taken up again since its original publication,
is also in P-LAM. That name has priority over M. ciliatum Aiton, pertaining to a species currently placed in Brownanthus.
Two new combinations are proposed: Brownanthus vaginatus (Lam.) Chesselet & M.Pignal and Brownanthus vaginatus
subsp. schenckii (Schinz) Chesselet & M.Pignal.
INTRODUCTION
The French naturalist Jean Baptiste de Monnet de
Lamarck (1744—1829) established the first evolutionary
synthesis of modem biology and participated in the great
debates about living species and evolution of life that
took place in late 18th century Europe (Corsi 2001).
Lamarck’s interests covered diverse topics including
botany, chemistry, meteorology, and notably zoology,
where he made the fundamental distinction between ver-
tebrates and invertebrates, his contribution to botany is
voluminous. In the Flore frangoise (1779), Lamarck
used the principle of dichotomous sorting for the identi-
fication of all taxa, thereby enabling identification to
species level, a technique now widely used by botanists.
In the Encyclopedic methodique , he described numerous
new species, but in the Mesembryanthemaceae only two.
These are the topic of this contribution.
The Encyclopedic methodique, published between
1783 and 1808, comprises eight volumes plus supple-
ments. The two first volumes, and the third up to the
letter P, are entirely the work of Lamarck. Many of the
plants described there correspond to specimens kept in
the Lamarck Herbarium (P-LAM). This valuable historic
collection is rich in types and comprises ± 19 000 speci-
mens. An Internet site dedicated to the works and contri-
butions of Lamarck (www.lamarck.net) provides digi-
tized images of the first 7 000 specimens, and the others
too, are being made progressively available there and
through the SONNERAT database (http://www.mnhn.
fr/base/sonnerat.html). The Lamarck herbarium changed
hands several times: it was sold, towards the end of
Lamarck’s life, to the German Botanist Johannes Roeper
(the first to use floral diagrams), then bought by the
University of Rostock in Germany and finally acquired,
in 1886, by the Museum of Paris. The collection is now
housed not far from its origin in the house of Buffon
* Compton Herbarium, South African National Biodiversity Institute,
Private Bag X7, 7735 Claremont, Cape Town.
E-mail: chesselet@sanbi.org
** Herbier National Paris (P), USM 0602 Taxonomie et collections,
Departement Systematique et Evolution, Museum National d'Histoire
Naturelle, 16 Rue Buffon, F-75005 Paris. E-mail: pignal@mnhn.fr
MS. received: 2004-10-26.
where Lamarck had his office at the Jardin des Plantes
(Aymonin 1980, 1981).
Lamarck, in the section on Ficoide, described two
new mesemb species in the Encyclopedic (1788: 478):
Mesembryanthemum echinatum Lam., a distinctive
species, and the oldest name in the genus Delosperma
N.E.Br., and Mesembryanthemum vaginatum Lam. the
first validly published name that applies to a species cur-
rently in the genus Brownanthus Schwantes.
Mesembryanthemum echinatum Lam.
In the case of M. echinatum, a mix-up due to the re-
assembly of the Lamarck collection has engendered con-
fusion. In addition to being moved several times, the
Lamarck herbarium was included into and later separat-
ed from the general herbarium at Rostock (the latter
process taking no less than five years: see Aymonin 1981).
Lamarck did not wish specimens to be glued as he pre-
ferred to observe the plants from all angles. His original
herbarium consisted of unmounted specimens of dried
plants with loose labels in species covers. They were
mounted subsequently when Edmond Bonnet, curator of
the herbarium at the Museum, was given the task of re-
arranging the Lamarck collection in 1900. Bonnet first
had the specimens attached with paper strips and, for its
arrangement, adopted the order of Durant (that largely
follows the system of Bentham and Hooker). Lamarck
preferred an arrangement that reflected natural relation-
ships to an alphabetical system such as that of Linnaeus.
The consequences of keeping the plants unmounted
would not have been too serious had the herbarium been
kept in Paris, such as those of the Jussieu (P-JU), acquired
by the Museum in 1857 after the death of Adrien de
Jussieu, and that of Michel Adanson (1727-1806), donat-
ed to the Paris herbarium in 1924. Fixing of specimens of
the Jussieu herbarium with pins and strips is thought to
have taken place in the mid- 19th century but the fixing
of specimens with sticky strips took place much later in
the history of this herbarium.
The name Mesembryanthemum echinatum was pub-
lished by Lamarck (1788) where the species was de-
scribed and qualified as v.v. ( vu vivant, or vidi vivum ), to
30
Bothalia 35,1 (2005)
indicate that the description was based on live material
grown at the Jardin des Plantes. The species, currently in
Delosperma N.E.Br., has been the subject of past debate
(Taylor & Eggli 1986). The combination in Delosperma
was made by Schwantes (1927) who gave the basionym
as 'M. echinatum Ait.’ (Aiton 1789). However,
Lamarck’s name was published one year earlier in 1788
[not 1786 as stated by Taylor & Eggli (1986), see Stafleu
& Cowan (1979)]. In his book on succulent plants,
Candolle (1799) suggested that the material from which
Lamarck’s description was made, originated from the
Cape of Good Hope and was brought to Britain by
Francis Masson in 1774, from where it came to the
Jardin des Plantes. It is possible that Aiton’s concept of
M. echinatum was based on material of the same origin
as Lamarck’s plant, although this cannot be shown
unequivocally.
What was still a debatable issue in 1986 has since
been clarified by a new provision in the ICBN (Greuter
et al. 2000, Art. 33.6(a)): even though Schwantes, in
publishing the combination Delosperma echinatum, refer-
FIGURE 1 . — Herbarium sheet from
the Lamarck Herbarium of
Mesembryanthemum scabrum
showing appended descrip-
tion and drawing of Delo-
spemui ( Mesembryanthemum )
echinatum.
Bothalia 35,1 (2005)
31
V . v' . ,i \* #T***vfe ' »w ** **’ **v '* } ^1 1 % * * a •
, >^V ill.
2y£v « V* * '**>*
ft1/ |t:»Jtfl‘»2')' ^ —V^'T'
FIGURE 2. — Holotype of Mesembryanthemum echinatum Lam. including description and sketch. ‘ mesembrianth . echinatum floribus sessilibus. mes.
foliis obovatis, tereti-triquetris. echinato-hispidis. Caules ramosissimi diffusi cespitosi decumbentes, subpedales, teretes, punctis albis ver-
niculosi, et setis aliquot spinulosiformibus, aspersis et retroversis subechinati. Folia opposita, obovato-teretia subtriquetra, viridia,
incequaliter echinato-hispida. punctis elevatis sed non coloratis undique scabra, camosa, semi-pollicaria. / lores albi sessiles, cal. 5-fidus,
cor. paulo brevior. laciniis incequalibus, quorum dux opp. majores camosiores, tereti-papulosce, foliiformae. Petala linearia, erecta vix-
patentia. now spec, africana, ex hort. Reg. 1786. floruit augusto. lesfl. viennent dans les dichot. et au sommet des petits rani des cotes. Elies sont
solitaires.' Translates as: 'Mesembrianthemum echinatum with sessile flowers. Mesembrianthemum with leaves obovate, terete-triquetrous,
echinate-hispid. Stems highly branched diffuse caespitose decumbent, of about 1 foot long [30 cm] terete, subechinate white verruculose
dots, and sometimes setae spinulose, rugose sparse reflexed. Leaves opposite, obovate-terete almost triquetrous green, unequally echinate-
hispid with elevated dots but not coloured everywhere, scabrid, fleshy, half an inch long [1.25 cm]. Flowers white, sessile, calyx 5-partite,
a little shorter than the corolla, the lobes unequal of which two opposite ones are longer and fleshy, terete papillose and leaf-shaped. Petals
linear, erect, hardly spreading. New African species from the royal garden. 1786, Flowered in August. The flowers come in the dicots and
at the summit of small branches on the sides. They are solitary.’
red to the wrong basionym author, Aiton; and even
though Aiton did not mention Lamarck’s earlier valida-
tion of Mesembryanthemum echinatum, of which he may
have been unaware; still Lamarck is to be considered as
the basionym author, and the correct author citation for
the combination in Delosperma is D. echinatum (Lam.)
Schwantes.
As no original specimens of M. echinatum are extant,
Taylor & Eggli (1986) designated a photograph of
Schwantes (1927) as neotype, and this neotypification
was accepted by Hartmann (2001a). However, an origi-
nal element is extant in the Lamarck herbarium, an illus-
tration which (in the absence of a preserved specimen of
the living plant) must be accepted as the holotype.
Therefore, the designated neotype has no standing.
Although Taylor & Eggli (1986) consulted a micro-
fiche of the herbarium sheet represented in Figure 1, they
failed to realize the true significance of the relevant slip
of paper (Figure 2). What Taylor & Eggli (1986) inter-
preted as a misidentification is in fact the result of the
complex history of Lamarck’s herbarium. When the col-
lection was finally pasted to herbarium sheets, a small
piece of paper with the description and sketch of M. echi-
natum, by Lamarck, became associated with the wrong
herbarium specimen, of Lampranthus scaber (L.) N.E.Br.
(= Mesembryanthemum scabrum L.) (Figure 1). It is pos-
sible that the word scabra in the description of M. echi-
natum prompted the association of the note with that
specimen.
Lamarck’s slip obviously dates back to the time when
the protologue was written. Along with the description, it
includes a sketch of the plant itself, which is original
material for the name as defined in the ICBN (Art. 9
Note 2). Although very simple, this illustration agrees
with D. echinatum as currently understood, a highly dis-
tinctive mesemb with echinate leaves and sessile flow-
ers, unique and atypical in the genus as pointed out by
Koutnik & O’Connor-Fenton (1985). As no other origi-
nal material is known to exist, we consider Lamarck’s
original sketch of M. echinatum as the holotype of the
name (Figure 2). The associated description begins with
the phrase-name provided by Lamarck in the protologue.
It reads: ‘ Mesembryanthemum echinatum. Mesem-
bryanthemum foliis obovatis tereti-triquetris verruculo-
sis echinato-hispidis, floribus sessilibus' . The type illus-
32
Bothalia35,l (2005)
FIGURE 3. — Holotype of Mesembry-
anthemum vaginatum Lam.
tration, in the Lamarck herbarium in Paris (P-LAM), is
attached to the specimen of M. scabrum bar-coded as
P00307737.
Mesembryanthemum vaginatum Lam.
The second new mesemb species described by
Lamarck (1788) is Mesembryanthemum vaginatum, with
the accompanying phrase-name: ‘'Mesembryanthemum
foliis oppositis basi connato-vaginantibus, vaginis per-
sistentibus crebris inferne barbatis, floribus corymbo-
sis' . This validating diagnosis is followed by the indica-
tion ‘(v..s.)’ [vu sec, or vidi siccum ], meaning that herbar-
ium material was available for study. The corresponding
specimen in P-LAM (Figure 3) was brought to Paris
from the Cape of Good Hope by Pierre Sonnerat, the
French naturalist and draughtsman who visited the Cape
in the late 18th century, but is known for his work on the
Seychelles, India, the East Indies and China (Gunn &
Codd 1981). His specimens were mainly presented to
Jussieu and Lamarck, in Paris.
The specimen held in the Lamarck herbarium (P-
LAM, bar code P00307735) is considered to be the holo-
type of M. vaginatum Lam. The re-discovery of this type
specimen has nomenclatural implications. In modem lit-
erature the species represented by that type is known as
Brownanthus ciliatus (Aiton) Schwantes, with the
basionym, M. ciliatum, validated by Aiton (Gerbaulet &
Pierce in Hartmann 2001a). Aiton lists the ‘Ciliated Fig
Marigold’ in Hortus kewensis (1789) with the phrase-
name: ‘M. foliis oppositis connatis semiteretibus, stipulis
membranaceis reflexis laceris ciliiformibus' . Aiton’s
material was introduced to Britain by Francis Masson in
1774. A drawing by Masson (BM) has been designated
by Gerbaulet & Pierce (in Hartmann 2001a) as the lecto-
type of the name M. ciliatum. In De Candolle’s (1828)
treatment of the ‘Ficoideae’, M. ciliatum Aiton is cited
under species non satis notce.
There is a second, later Mesembryanthemum vagina-
tum, by Haworth (1803), which in the original volume of
Index kewensis is unaccountably treated as if it were but
a mere later re-use of M. vaginatum Lam. In reality, it is
an independent, illegitimate later homonym that belongs
to a completely different species. In De Candolle (1828),
Mesembryanthemum vaginatum Haw. is correctly listed
among the taxa currently placed in the genus Ruschia
Schwantes, with a reference to ‘Haw. misc. 95. syn. 284.
excl. var. (3 rev. 127. non. Lam.’. It bears the legitimate
name Ruschia vaginata Schwantes (with priority dating
from 1927, not 1789: see ICBN Art. 58). The neotype
designation by Hartmann (2001b) is effective for both
Mesembryanthemum vaginatum Haw. and the homotypic
Ruschia vaginata Schwantes.
Bothalia 35,1 (2005)
33
Lamarck’s name is not accounted for elsewhere in De
Candolle’s work, and is now utterly forgotten. We have
considered the option of submitting a proposal to con-
serve the name M. ciliatum Aiton against it, in the inter-
est of stability of botanical nomenclature. We are, how-
ever, discarding that option and proposing a name
change for nomenclatural reasons. Although the contin-
ued use of the name Brownanthus ciliatus (Aiton)
Schwantes may be preferred for practical reasons, it is
not proposed here as the species is little known, unim-
portant in horticultural trade, only occurs in Northern
and Western Cape, and the name is scantly used in the lit-
erature. The application of the rule of priority requires
the following nomenclatural changes:
Brownanthus vaginatus (Lam.) Chesselet & M.Pig-
nal , comb. nov.
Mesembryanthemum vaginatum Lam.: 478 (1788). Type: Sormerat
s.n. (P-LAM!, holo. P00307735). M. ciliatum Aiton: 179 (1789).
Trichocyclus ciliatus (Aiton) N.E.Br.: 151 (1923). Brownanthus cilia-
tus (Aiton) Schwantes: 21 (1927). Psilocaulon ciliatum (Aiton)
Friedrich: 216 (1968).
Brownanthus vaginatus subsp. schenckii (Schinz)
Chesselet & M.Pignal , comb. nov.
Mesembryanthemum schenckii Schinz: 80 (1897). Trichocyclus
simplex N.E.Br. ex Maass: 234 (1928). T. schenckii (Schinz) Dinter &
Schwantes ex Range: 18 (1934). Brownanthus simplex (N.E.Br. ex
Maass) Bullock: 494 (1937). B. ciliatus subsp. schenckii (Schinz)
Ihlenf. & Bittrich: 316 (1985). B. schenckii (Schinz) Schwantes: 21
(1927). Lectotype: (following Gerbaulet & Pierce in Hartmann 2001a)
Schenck 174 (K).
ACKNOWLEDGEMENTS
We gratefully acknowledge advice and useful discus-
sion with Gerard G. Aymonin, Joel Jeremie (P), Denis
Lamy (PC), Justyna Wiland-Szymariska from the A.
Mickiewicz University, Poland and Estelle Potgieter,
Mary Gunn Library, National Herbarium, South Africa.
We are extremely grateful to Werner Greuter and an
anonymous referee for their valuable comments on the
manuscript. The first author thanks staff at the Paris
Herbarium for their kind hospitality.
REFERENCES
AITON, W. 1789. Hortus kewensis edn. 1, 2: 179. Nicol, London.
AYMONIN, G.G. 1980. L’herbier de Lamarck: un patrimoine scien-
tifique national, temoin de Thistoire. Bulletin de la societe
botanique de France , 127, Lettres botaniques 4: 393-401.
AYMONIN, G.G. 1981. L’herbier de Lamarck. Revue d'histoire des
sciences 34: 25-58.
BROWN, N.E. 1923. The genera Aloe and Mesembryanthemum as pre-
sented in Thunberg’s herbarium. Bothalia 1: 139-169.
BROWN, N.E. 1925. Mesembryanthemum and some new genera sepa-
rated from it. Gardeners' Chronicle 78: 412, 413.
BULLOCK, A. A. 1 937. Notes on Mesembryanthemaceae 1 . The genus
Brownanthus Schwantes. Kew Bulletin 1937: 494.
CORSI, P. 2001. Lamarck. Genese et enjeux du transformisme.
1770-1830. CNRS Editions, Paris.
DE CANDOLLE. A.P. 1828. Ficoidete. Prodromus systematis natu-
ralis regni vegetabilis 3: 4 1 5 — 456. Treuttel et Wiirtz, Paris.
DE CANDOLLE, A.P. 1799. Plantarum historia succulentarum.
Dugour & Durand, Paris.
FRIEDRICH, H.-C. 1968. Mesembrianthemenstudien III. Mitteilungen
der Botanischen Staatssammlung Miinchen 7: 216.
GREUTER, W.. MCNEILL. J„ BARRIE. F.R., BURDET. H.M., DE-
MOULIN, V., FILGUEIRAS, T.S., NICOLSON, D.H.. SILVA,
PC., SKOG, J.E., TREHANE, P„ TURLAND, N.J. &
HAWKSWORTH, D.L. (eds). 2000. International Code of
Botanical Nomenclature (Saint Louis code). Koeltz Scientific
Books. Konigstein.
GUNN, M. & CODD, L.E. 1981. Botanical exploration of southern
Africa. Balkema, Cape Town.
HARTMANN. H.E.K. 2001a. Illustrated handbook of succulent plants.
Aizoaceae A-E. Springer Verlag, Berlin. Heidelberg, New York.
HARTMANN, H.E.K. 2001b. Illustrated handbook of succulent plants.
Aizoaceae F-Z. Springer Verlag, Berlin, Heidelberg, New York.
HAWORTH, A. H. 1821. Miscellania naturalia: 126.
HERRE. H. 1971. The genera of the Mesembryanthemaceae. Tafelberg.
Cape Town & Johannesburg.
IHLENFELDT, H.-D. & BITTRICH, V. 1985. Morphologie, Gliede-
rung und Abgrenzung der gattung Psilocaulon N.E.Br. s.l. (Mesem-
bryanthemaceae). Botanische Jahrbucher 105: 316.
KOUTNIK. D.L. & O’CONNOR-FENTON, M. 1985. Lectotypifica-
tion of the genus Delosperma (Mesembryanthemaceae). South
African Journal of Botany 51: 197, 198.
LAMARCK, J.B.P.A. 1778 [1779]. Flore franqoise ou description suc-
cincte de toutes les plantes qui croissent mturellement en France,
disposee selon une nouvelle methode d’ analyse, & a laquelle on a
joint la citation de leurs vertus les mains equivoques en medecine,
& utilisees dans les arts, 3 volumes. Imprimerie Royale, Paris.
LAMARCK, J.B.P.A. 1786. Encyclopedic Methodique, Botanique II:
1—400. Panckoucke, Paris.
LAMARCK, J.B.P.A. 1788. Encyclopedic Methodique, Botanique II:
401-end. Panckoucke, Paris.
MAASS, C.A. 1928. Mesembryanthemum. Genus Trichocyclus N.E.Br.
Zeitschrift fur Sukkulentenkunde 3: 233, 234.
RANGE, P. 1934. Die Flora des Namalandes 4. Aizoaceae. Feddes Reper-
torium 36: 10-19.
SCHINZ, H. 1897. Die Pflanzenwelt Deutsch-Siidwest-Afrikas (mit
Einschlu]) der westlichen Calahari). Bulletin de VHerbier Bois-
sier 5, Appendix 3: 5-101.
SCHWANTES, G. 1927. System der Mesembryanthemen. Zeitschrift
fur Sukkulentenkunde 3: 14—21.
STAFLEU, F.A. & COWAN, R.S. 1979. Taxonomic Literature II: H-Le.
Junk, The Hague.
TAYLOR, N. & EGGLI, U. 1986. The lectotype of Delosperma NE.
Brown (Aizoaceae). Taxon 35: 709-711.
.
Bothalia 35,1: 35 — 46 (2005)
Cape plants: corrections and additions to the flora. 1
P. GOLDBLATT*, J.C. MANNING** and D. SNIJMAN**
Keywords: Cape Floristic Region, floristics, phytogeography, plant diversity, southern Africa, speciation
ABSTRACT
Comprising an area of ± 90 000 km2, less than 5% of the land surface of the southern African subcontinent, the Cape
Floristic Region (CFR) is one of the world's richest areas for plant species diversity. A recent synoptic flora for the Region
has established a new base line for an accurate assessment of the flora. Here we document corrections and additions to the
flora at family, genus and species ranks. As treated in Cape plants, which was completed in 1999, the flora comprised 173
families (five endemic), 988 genera ( 160 endemic: 16.2%), and 9 004 species (6 192 endemic: 68.8%). Just four years later,
a revised count resulting from changes in the circumscriptions of families and genera, and the discovery of new species or
range extensions of species, yields an estimate of 172 families (four endemic), 992 genera (162 endemic: 16.3%) and 9 086
species (6 226: 68.5% endemic). Of these, 948 genera and 8 971 species are seed plants. The number of species packed into
so small an area is remarkable for the temperate zone and compares favourably with species richness for areas of compa-
rable size in the wet tropics. The degree of endemism is also remarkable for a continental area. An unusual family compo-
sition includes, in descending order of size, based on species number. Asteraceae, Fabaceae. Iridaceae, Ericaceae,
Aizoaceae, Scrophulariaceae. Proteaceae, Restionaceae, Rutaceae, and Orchidaceae. Disproportionate radiation has result-
ed in 59.1% of the species falling in the 10 largest families and 74.6% in the largest 20 families. Thirteen genera have more
than 100 species and the 20 largest genera contribute some 31.5% of the total species number.
INTRODUCTION
Published in September 2000, Cape plants (Goldblatt
& Manning 2000) is a synoptic account of the vascular
plant flora of the Cape Floristic Region (CFR) of southern
Africa. As with such endeavours, it contained its share of
errors and omissions. Moreover, changes to the systematics
of plant taxa of the Cape flora have accumulated at a
steady pace. Thus, despite our best efforts to produce an
accurate and lasting account, we find after just four years
since publication, and about five years since completed
copy was handed to the publisher, that a fair number of
corrections and additions are necessary. Most of these con-
cern species: 9 004 species were recognized in 2000, and
we now include 9 086 species in the flora region. Some
988 genera were recognized in the flora in 2000, but with
additions and taxonomic changes there are now 992 genera.
We hope to continue to publish lists of additions periodi-
cally when such compilations seem useful. Changes to the
account are discussed in detail below under the headings
Families, Genera , and Species.
Families. Changes in familial classification, resulting
from the continuing molecular research in angiosperm
phylogeny, have compelled adjustments to the generic
constitution of several families [Angiosperm Phylogeny
Group (APG) 1998, 2002], The rule of priority has also
made necessary some name changes. Most notable for the
CFR is the work of Olmstead and his collaborators in the
Lamiales-Scrophulariales which has shown that Veronica
L. and its allies, previously Scrophulariaceae, should be
included in an expanded Plantaginaceae (Olmstead et al.
2001). In the Cape flora the genera affected include
Limosella L. (two spp.) and Ilysanthes Raf. (one sp.).
* B.A. Krukoff Curator of African Botany, Missouri Botanical Garden,
RO. Box 299, St. Louis, Missouri 63166, USA. peter.goldblatt@mobot.org
** Compton Herbarium, South African National Biodiversity Institute,
Private Bag. X7, 7735 Claremont, Cape Town. manning@nbict.nbi.ac.za
MS. received: 2004-07-22.
which are accordingly transferred to Plantaginaceae.
These studies also show that several African genera of
Scrophulariaceae and Loganiaceae comprise a clade with
the Cape family Stilbaceae. In order to retain monophylet-
ic family circumscriptions, these genera have been trans-
ferred to Stilbaceae (alternatively Stilbaceae and several
more families would have to be included in Scro-
phulariaceae). Thus Halleria L. (three spp.) and Ixianthes
Benth. (one sp.) are removed to Stilbaceae, which ceases
to be endemic to the CFR. Nnxia Comm, ex Lam. (pre-
viously Loganiaceae) also belongs in an expanded
Stilbaceae while Buddleja L., previously Loganiaceae and
more recently Buddlejaceae, is now included in Scro-
phulariaceae. This leaves Scrophulariaceae with 3 1 genera
and 409 species, and Loganiaceae with one genus
( Strychnos L.) and two species in the CFR.
Another important familial change is the union of the
endemic southern African family Achariaceae with
Kiggelariaceae (Savolainen et al. 2000). The conserved
name Achariaceae is used for the expanded family. The
endemic Cape tree Hyaenanche Lamb. & Vahl (one sp.)
is now known to be nested in Picrodendraceae (= Pseu-
danthaceae) (Savolainen et al. 2000) and is transferred to
this family from Euphorbiaceae. The status of the mono-
typic Curtisia Aiton, usually included in Comaceae, has
varied with time. Molecular data show that the genus is
sister to the Cape endemic family, Grubbiaceae, and
recognition of a monogeneric Curtisiaceae is recommend-
ed by APG (2003). The family Comaceae is thus no
longer represented in the Cape flora. In the monocots, the
South African and near Cape endemic, Prioniaceae (with
the monotypic Prionium E.Mey.) is sister to the South
American Thumia Hook.f. and has been referred to
Thumiaceae by Chase et al. (2000). Tamaricaceae, repre-
sented by one species of Tamarix L. in the Cape flora,
was omitted in error from Cape plants.
Convallariaceae, the family that includes Dracaena
L„ Eriospermum Jacq. ex Willd. and Sansevieria Thunb.,
36
Bothalia 35,1 (2005)
must now be known by the earlier name, Ruscaceae. The
APG (2003) recommend that Ruscaceae as well as
Anthericaceae and Hyacinthaceae be included in an
enlarged monophyletic Asparagaceae. They also favour
union of Agapanthaceae and Amaryllidaceae in Alliaceae
and Asphodelaceae and Hemerocallidaceae in Xanthorr-
hoeaceae. Alternative treatment is permitted and for the
present we do not follow these recommendations and
maintain the narrower family circumscriptions in Cape
plants.
Another family realignment is the inclusion of
Centella L. in Apiaceae (Araliaceae in Cape plants) as a
result of extensive molecular study (Lowry et al. 2004).
Also likely as a result of sequence studies, is the removal
of Ceraria H. Pearson & Stephens and Portulacaria Jacq.
from Portulacaceae to Didieriaceae. This previously
endemic family of Madagascar forms a sister clade with
Portulacaceae s.s. (Applequist & Wallace 2001). The
changes involving Portulacaceae have not yet been for-
malized in the literature (APG 2003), and are not taken
into account here.
At family level it is most notable that the flora loses
its second largest endemic family, Stilbaceae, which in
its expanded circumscription extends through tropical
Africa and Madagascar to Arabia ( Halleria , Nuxia). This
leaves only four endemic families in the CFR: Penae-
aceae in Myrtales (21 species), Grubbiaceae in Cornales
(three species), Roridulaceae in Ericales (two species),
and Geissolomataceae in Saxifragales (one species)
(classification following APG 2003). Based on a molec-
ular clock calibrated using 135 mya for the divergence of
the eudicot lineage (Savolainen et al. 2000; Wikstrom et
al. 2001; V. Savolainen unpubl. ms), Penaeaceae may
have diverged 20 mya from its sister clade, the African
Oliniaceae plus the Neotropical Alzateaceae, whereas
Roridulaceae diverged from Ericaceae, its closest relative
(Savolainen et al. 2000), ± 48 mya. Geissolomataceae
appears to be older, having diverged perhaps 55 mya
from Ixerbaceae plus Strasburgeraceae. Grubbiaceae may
have diverged from Curtisiaceae in the early Tertiary, 63
mya.
Bruniaceae, one of the distinctive families of the Cape
flora, has an estimated 64 species in 1 1 genera (Goldblatt
& Manning 2000). Just three species in two genera
extend outside the confines of the Cape Region, two
locally, and one as far east as southern KwaZulu-Natal.
Bruniaceae may be the sister group to the order
Dipsacales (Savolainen et al. 2000), perhaps meriting
recognition at ordinal rank. The discovery of pollen
matching modern Bruniaceae in early Tertiary and late
Cretaceous (?Senonian) deposits in northern Namaqua-
land (S.E. de Villiers pers. comm.), well to the north of
the CFR, attests to considerable age for the family in
southern Africa. The pollen record also accords with a
preliminary early Tertiary dating of the divergence
between Bruniaceae and Dipsacales at about 57 mya (V.
Savolainen unpubl. ms.). In the later APG (2003) classi-
fication, however, Brunicaeae are not assigned to any
order in the Euasterids II group.
With the above changes, there are now 149 families of
seed plants, and 23 families of ferns and other vascular
cryptogams, for a total of 172 families of vascular plants
in the CFR. This is one less than was recognized in Cape
plants. Anticipated transfer of two genera of
Portulacaceae to Didieriaceae would bring the total num-
ber of seed plant families to 150. The CFR is character-
ized by an unusual family composition that includes, in
descending order of size (species number) following
Asteraceae and Fabaceae, the families Iridaceae, Eri-
caceae, Aizoaceae, Scrophulariaceae, Proteaceae, Restio-
naceae, Rutaceae, and Orchidaceae, among the 10 most
species-rich families in the flora. Asteraceae alone, with
1 048 species, contributes 11.5% of the total species in
the flora. Disproportionate radiation in the 20 largest
families (Table 1) has resulted in over 59% of the species
falling in the 10 largest families and more than 74% in
the largest 20 families.
TABLE 1. — Ranking of the 20 largest families in the Cape flora as
indicated by species number from Goldblatt & Manning (2000,
2002a) combined with changes in this paper. These families
contribute 7 022 species to the flora, or 75% of the total 9 087
species. Family circumscriptions reflect the recommendations
of the Angiosperm Phylogeny Group (1998, 2003)
Genera. A total of 942 genera of seed plants (or 988
genera of vascular plants) were included in the Cape
flora by Goldblatt & Manning (2000), comprising about
half of all those occurring in southern Africa. Of those,
some 160 genera (all of them seed plants), were endem-
ic, constituting 16.2% of the total vascular plant flora.
The inclusion of Tamaricaceae in the flora adds one more
genus, Tamarix, and the recent discovery of a species of
Clivia Lindl. (Amaryllidaceae) in the Cape flora (Rourke
2002) is a second generic addition. The genus Pilularia
L. (Marsileaceae) has also recently been recorded for the
first time in the flora (Roux 2002). Carpolyza Salisb.
(Amaryllidaceae) has been found by molecular analysis
to be nested in Strumaria Jacq. ex Willd., in which it is
now included (Meerow & Snijman 2001). Also in
Amaryllidaceae, the monotypic Cybistetes Milne-Redh.
& Schweick. is now included in Ammocharis Herb.
(Snijman & Archer 2003). In Aizoaceae, some species
previously assigned to Lampranthus N.E.Br. and Ruschia
Schwantes are now referred to the new endemic genera
Bothalia 35. 1 (2005)
37
Brianhuntleya Chesselet, S. A. Hammer & I. Oliver
(monotypic) (Chesselet et al. 2003), and Phiambolia Klak
(7 species) (Klak 2003).
Old herbarium records, until now overlooked, show
that Calystegia R.Br. (Convolvulaceae) and Chaetacme
Planch. (Celtidaceae) occur naturally in the Cape flora,
adding another two genera to the total. Another generic
omission is Gomphostigma Turcz., now added to Scro-
phulariaceae. In Asteraceae, a new monospecific genus
Roodebergia B.Nord. has been described (Nordenstam
2002b) and two species of Dicoma Cass, have been
transferred to Macledium Cass., thus adding two more
genera to the flora. The endemic genus Alciope DC.
(Asteraceae). has been found to be nomenclaturally ille-
gitimate, and is now called Capelio B.Nord.
(Nordenstam 2002a, 2003a). Lastly, in Hyacinthaceae
the endemic species Scilla plumbea has been shown by
molecular data to be misplaced generically and has been
assigned to the new and endemic genus Spetaea
Wetschnig & Pfosser, as S. lachenaliiflora Wetschnig &
Pfosser (2003). Also in Hyacinthaceae, Albuca L.,
Dipcadi Medik. and Neopatersonia Schonland have been
sunk in Ornithogalum L., Whiteheadia Harv. in
Massonia Thunb. ex Houtt., and Polyxena Kunth is
included in Lachenalia J.Jacq. ex Murray (Manning et
al. 2004). This reduces the number of genera but signifi-
cantly enlarges Ornithogalum , which now has 72 species
(previously 40 species) in the Cape flora, while
Lachenalia now has 69 species in the flora.
Recent molecular studies on Zygophylloideae (Beier
et al. 2003) have resulted in substantial restructuring of
the genera in the subfamily. Zygophyllum L. is now
understood to be restricted to Asia and the majority of the
southern African species of Zygophyllum have been
referred to the genus Roepera A.Juss. The species of
Zygophyllum subg. Agrophyllum Endl. are placed in the
genus Tetraena.
With these changes there are currently 948 genera of
seed plants, or 992 genera of vascular plants recognized
in the CFR. This is an increase of four genera to the pre-
vious total for the flora (Goldblatt & Manning 2000). Of
these, 162 genera (16.3%) are endemic, representing an
insignificant increase in generic endemism over that
recorded in Cape plants. Thirteen genera have more than
100 species and the 20 largest genera contribute some
2 858 species, or 31.5% of the total in the flora (Table 2).
TABLE 2. — Ranking by numerical size of the twenty largest genera in
the CFR (endemic species number) from Goldblatt & Manning
(2000, 2002a) with changes following this paper
Species. The number of species added to the flora is
considerable and we list changes below by family alpha-
betically. New inclusions are provided with treatments
comparable to those in Cape plants, including brief
descriptions, and notes on flowering time, distribution,
habitat, and phytogeographic centre(s) within the Cape
Region. Corrections to species names, descriptions, or
ranges are included in the list that follows. The sign *
indicates endemic to the CFR and the sign ! indicates
introduced species. With the additions and corrections,
the Cape flora now includes 9 086 species of vascular
plants, 6 226 endemic (8 971 species of seed plants, 6
217 endemic) with a percentage endemism of 68.5 %.
This represents an increase of 80 species since the publi-
cation of Cape plants and a reduction in endemism from
68.8%. In Table 1 we list the ten largest families in the
flora with their number of species and degree of
endemism.
Six phytogeographical subcentres are recognized in
the Cape Region and their abbreviations used here are:
NW (Northwest Centre); SW (Southwest Centre); AP
(Agulhas Plain); KM (Karoo Mountain Centre); FB
(Fangeberg Centre); and SE (Southeast Centre) (see
front endpaper of Cape plants).
AIZOACEAE
Acrodon deminutus Klak Spreading, tufted succulent shrublet to 10
cm. Leaves trigonous, free almost to base, margins toothed.
Flowers with petals magenta at base and tips, white between, sta-
minodes white with magenta tips, 15-20 mm diam. Capsule 5-loc-
ular, 7-8 mm diam. Sept. Stony quartz outcrops on clay, SW
(Swellendam to Bredasdorp)* (Klak 2003).
Amphibolia hutchinsonii (L.Bolus) H.E.K. Hartmann is a synonym of
A. laevis (Aiton) H.E.K.Hartmann (Hartmann 2001).
Antimima aristulata (Sond.) Chesselet & Gideon F.Sm. is a new com-
bination for Ruschia aristulata (Sond.) Schwantes (Chesselet &
Smith 2001 ).
Antimima insidens (L.Bolus) Chesselet is a new combination for
Ruschia insidens L.Bolus (Chesselet 2001).
Antimima viatorum (L.Bolus) Klak is a new combination for A. bina
(L.Bolus) H.E.K.Hartmann. NW, SW (Namaqualand to Malmes-
bury). Revised distribution, not endemic (Klak 2003).
Brianhuntleya Chesselet, S. A. Hammer & I. Oliver is a new genus for
the flora (Chesselet et al. 2003).
Brianhuntleya intrusa (Kensit) Chesselet, S.A.Hammer & I.Oliver
[= Ruschia intrusa (Kensit) L.Bolus] Tufted succulent, 7-10 cm,
with spreading branches, bearing persistent dry leaves. Leaves
swollen-trigonous, obtuse, entire, ± 55 x 8 mm, grey-green.
Flowers solitary, to 35 mm diam., with an intrusive calyx tube, pale
rose-purple. Fruits 5-locular. June-July. Shale slopes, NW
(Worcester Karoo)*. Revised description.
Drosanthemum asperulum (Salm-Dyck) Schwantes is a new combina-
tion for Delosperma asperulum (Salm-Dyck) L.Bolus (Klak 2003).
Drosanthemum quadratum Klak Like D. asperulum but plants
smaller, to 15 cm high, with thicker leaves 2-4 mm wide, and wider
capsules 5-10 mm diam. Sept.-Oct. Stony quartz outcrops on clay
slopes, SW (Swellendam to Bredasdorp)* (Klak 2003).
Erepsia simulans (L.Bolus) Klak Laxly branched succulent shrublet
to 16 cm. Leaves shortly fused toward base, narrowed near tips,
acute to obtuse. Flowers silvery white, tipped with pink. Capsules
7-locular. Oct. Limestone flats, AP (Agulhas Peninsula: Brand-
fontein)* (Klak 2003).
Esterhuysenia mucronata (L.Bolus) Klak is the correct name for
Lampranthus mucronatus L.Bolus (Klak 2003).
Lampranthus gydouwensis (L.Bolus) H.E.K.Hartmann is a synonym of
Phiambolia incumbens (L.Bolus) Klak (Klak 2003).
Machairophyllum acuminatum L.Bolus is a synonym of M. bijliae
(N.E.Br.) L.Bolus (Kurzweil & Chesselet 2003).
Machairophyllum baxteri L.Bolus is a synonym of M. bijliae (N.E.Br.)
L.Bolus (Kurzweil & Chesselet 2003).
Machairophyllum bijliae (N.E.Br.) L.Bolus Tufted perennial to 20
cm. Leaves pale green, trigonous, angles acute. Flowers solitary on
38
Bothalia 35,1 (2005)
pedicels to 75 mm long, golden yellow with red reverse or red,
50-60 mm diam. Mostly Oct. -Nov. Rocky slopes and rock
crevices, KM, SE (Swartberg Pass to Uniondale and George to
Humansdorp and E Cape). Revised description and range, no
longer endemic (Kurzweil & Chesselet 2003) (not M. bijlii).
Machairophyllum cookii (L. Bolus) Schwantes is a synonym of M.
albidum (L.) Schwantes (Kurzweil & Chesselet 2003).
Machairophyllum latifolium L. Bolus is a synonym of M. brevifolium
L.Bolus (Kurzweil & Chesselet 2003).
Machairophyllum stayneri L. Bolus does not occur in the CFR.
Oscularia comptonii (L. Bolus) H.E.K. Hartmann is a new name for O.
ebracteata (L.Bolus) H.E.K. Hartmann. Stony slopes, NW
(Namaqualand to Olifants River Valley). Corrected distribution, not
endemic (Klak 2003).
Phiambolia Klak is a new genus for some species previously of
Lampranthus and Ruschia (Klak 2003).
Phiambolia franciscii (L. Bolus) Klak Succulent shrub to 60 cm.
Leaves 25 — 40 x 3-6 mm, fused at base into a sheath, 4-5 mm long,
papillate-velvety. Flowers in cymes, pink, 30-45 mm diam. Fruits
funnel-shaped. Oct. -Nov. Mainly stony sandstone slopes, NW
(Cold Bokkeveld to Karoopoort).*
Phiambolia hallii (L. Bolus) Klak Prostrate succulent rooting at
nodes, branches to 30 cm. Leaves joined for 3-4 mm, free parts 20-
35 mm, 4—6 mm diam. Flowers to 35 mm diam., solitary on pedi-
cels, 15-25 mm long, pink. Fruits funnel-shaped. July. Sandstone and
shale outcrops, NW (Swartruggens: Katbakkies to Karoopoort).*
Phiambolia incumbens (L. Bolus) Klak is a new combination for
Ruschia incumbens L.Bolus.
Phiambolia mentiens Klak Shrublet to 35 cm, with smooth inter-
nodes, 25-40 mm long. Leaves basally fused, subterete, 13-25 x 4-6
mm, with recurved mucronate tips. Flowers in few-flowered cymes,
petals magenta, filamentous staminodes in a cone, white with
magenta tips. Fruits funnel-shaped. Sept. Sandstone rocks, NW
(Cederberg Mtns to Karoopoort).*
Phiambolia persistens (L.Bolus) Klak is a new combination for
Lampranthus persistens (L.Bolus) Klak Sandy and stony slopes,
NW, KM (Bokkeveld Mtns and Witteberg).* Revised range.
Phiambolia stayneri (L.Bolus ex Toelken & Jessop) Klak is a new
name for Lampranthus dissimilis (G.D. Rowley) H.E.K. Hartmann)
Stony slopes, NW (Ceres: Warm Bokkeveld).* Corrected range.
Phiambolia unca (L.Bolus) Klak is a new combination for Lam-
pranthus uncus (L.Bolus) H.E.K. Hartmann Stony slopes and flats,
NW (Bokkeveld Mtns to Ceres).* Revised range.
Ruschia knysnana (L.Bolus) L.Bolus Sandstone slopes, SE (Knysna
to Grahamstown). Revised distribution, not endemic.
Ruschia pulchella (Haw.) Schwantes is now regarded as an insuffi-
ciently known species and must be removed from the account of
Ruschia (Chesselet & Smith 2001).
Current total: genera 77; species 658 (previously 76 genera and 661
species).
ALLIACEAE
Tulbaghia capensis L. SW, AP, LB, SE (Namaqualand, Cape Penin-
sula to Long Kloof). Revised range, not endemic.
AMARANTHACEAE
Sarcocornia sp. 1. Woody jointed shrublet to 60 cm, branches 2 mm
diam. Leaves fleshy. Flowering time? Saline washes in renoster-
veld, NW (Eendekuil Flats).*
Current total: genera 1 2; species 23 (previously 1 2 genera and 22 species).
AMARYLLIDACEAE
Ammocharis longifolia (L.) M.Roem. is the correct name for Cybi-
stetes longifolia (L.) Milne-Redh. & Schweick. (Snijman & Archer
2003).
Brunsvigia elandsmontana Snijman Bulbous geophyte to 20 cm.
Leaves dry at flowering, 4—6, prostrate. Flowers 6-18 in a compact,
hemispherical head, actinomorphic, deep pink, tepals widely flared,
stamens about as long as tepals, central. Capsules 3-angled.
Mar.-May. Stony flats, SW (Hermon)* (Snijman 2001a).
Carpolyza Salisb. is now included in Strumaria (Meerow & Snijman
2001).
Clivia Lindl is added to the flora (Rourke 2002).
Clivia mirabilis Rourke Rhizomatous perennial to 80 cm. Leaves sev-
eral, linear, often with a whitish midline, 30-40 mm wide. Flowers
tubular, nodding, reddish with green to yellow tips, pedicels red-
dish, ± 25 mm long. Oct.-Nov. Wooded scree, NW (Bokkeveld
Mountains)*.
Cybistetes Milne-Redh. & Schweick. is now included in Ammocharis
(Snijman & Archer 2003).
Cyrtanthus debilis Snijman is a new species for Cyrtanthus sp. 2
(Snijman 2001b).
Gethyllis linearis L.Bolus Bulbous geophyte to 6 cm, forming com-
pact clumps. Leaves dry at flowering, spreading, tightly coiled,
glabrous and subsucculent. Flowers white tinged pink, anthers 6.
Oct.-Nov. Gravelly flats, NW (S Namaqualand to Bokkeveld Mtns)
(Manning et al. 2002).
Gethyllis oliverorum D.Miill.-Doblies Bulbous geophyte up to 40
mm. Leaves dry at flowering, linear, glabrous or subglabrous, curved
and spreading on ground or slightly ascending. Flowers white to
pale pink, anthers 6. Fruit almost dry. Nov.-Dee. Shallow soil on
rocks, NW (Heerenlogement, near Vanrhynsdorp, N Cederberg).*
Strumaria spiralis (L’Her.) Snijman is the correct name for Carpolyza
spiralis (L’Her.) Salisb.
Current total: genera 15; species 97 (previously 16 genera and 93
species).
ANTHERICACEAE
Chlorophytum comosum (Thunb.) Jacques (not Jacq.).
APIACEAE
Centella with 49 species is transferred here from Araliaceae.
Current total: genera 24; species 121 (previously 23 genera and 72
species).
APOCYNACEAE (P. Bruyns pers. comm.)
Pachypodium succulentum (Jacq.) Sweet (not (L.f.) A. DC.).
Stapelia obducta L.C. Leach Leafless succulent with erect, 4-angled
stems, 10-20 mm diam., forming dense clumps, 10-25 cm; sap
clear. Flowers 50 mm diam., button-like with strongly recurved
lobes, purple-brown, softly hairy on inner surface throughout.
Mar. -Oct. stony sandstone slopes, SE (Great Winterhoek Mtns).*
Not conspecific with S. hirsuta.
Tromotriche choanantha (Lavranos & H.Hall) Bruyns (not (Lavranos
& A.V.Hall) Bruyns).
Current total: genera 36; species 113 (previously 36 genera and 112
species).
ARALIACEAE
Centella with 49 species is transferred to Apiaceae.
Current total: genera 3; species 6 (previously 4 genera and 55 species).
ASPHODELACEAE
Bulbine cremnophila Van Jaarsv. Dwarf perennial to 30 cm, roots
fleshy, grey. Leaves rosulate, fleshy, narrowly lanceolate, glaucous.
Flowers in a lax raceme, yellow. Capsules ovoid, erect. Aug.-Feb.
Sandstone cliff faces, SE (Humansdorp)* (Van Jaarsveld & Van
Wyk 1999).
Bulbine melanovaginata G.Will. (Williamson 2003) is included in
Bulbine foleyi E.Phillips.
Bulbine meiringii Van Jaarsv. Dwarf geophyte to 30 cm, clustered,
rootstock a small tuber with fleshy roots. Leaves slender and fleshy,
semi-terete, surrounded at base by a short fibrous neck. Flowers in
a lax raceme, yellow. Capsules ovoid, erect. June-Sept. Rocky
sandstone ledges, KM (Swartberg Mtns)* (Van Jaarsveld 2003).
Bulbine navicularifolia G.Will. (Williamson 2003) is included in
Bulbine succulenta Compton.
Bulbine ramosa Van Jaarsv. Branching geophyte to 50 cm, forming
clusters, stems globose below with fleshy roots. Leaves linear-
lanceolate, bright green. Flowers in a lax raceme, yellow. Capsules
ovoid, erect. Nov. Sandstone cliff faces, KM (Calitzdorp: Bads-
poort)* (Van Jaarsveld 2003).
Gasteria polita Van Jaarsv. Like G. acinacifolia but smaller, up to 60
cm in flower, and usually solitary; raceme usually unbranched and
flowers smaller, 35 mm long. Oct.-Nov. Slopes and embankments
in forest, SE (Plettenberg Bay)* (Van Jaarsveld 2001).
Gasteria vlokii Van Jaarsv. KM (Swartberg Mtns to Willowmore).
Range correction.
Haworthia bruynsii M.B. Bayer does not occur in the CFR.
Haworthia pubescens M.B. Bayer NW (Worcester-Robertson Karoo)*.
Range correction.
Bothalia 35,1 (2005)
39
Current total: genera 8; species 161 (previously 8 genera and 157 species).
ASTERACEAE ( Othonna and Senecio corrections, P.V.Bruyns pers.
comm.)
Arctotheca marginata Beyers Prostrate perennial rooting at nodes,
softly hairy. Leaves linear-elliptic, margins crisped and rolled
under, softly hairy above and felted beneath. Flower heads radiate,
solitary on felted scapes, yellow with rays reddish outside, involu-
cral bracts felted. Achenes woolly, pappus wanting. Oct.-Nov.
Sandy edges of pans, NW (Bokkeveld Mtns)* (Beyers 2002).
Athanasia trifurcata (L.) L. NW, SW, AP, KM, LB, SE (Springbok to
Port Elizabeth). Revised range, not endemic.
Capelio B.Nord. is a new name for Alciope DC. (Nordenstam 2002a,
2003a).
Capelio caledonica B.Nord. is a new species for Alciope sp. 1. (Nor-
denstam 2002a, 2003a).
Capelio tabularis (Thunb.) B.Nord. is a new combination for Alciope
tabularis Thunb. (Nordenstam 2002a, 2003a).
Capelio tomentosa (Burm.f.) B.Nord. is a new name for Alciope lana-
ta (Thunb.) DC. (Nordenstam 2002a. 2003a).
Dimorphotheca nudicaulis (L.) DC. (not (L.) B.Nord.).
Felicia josephinae J.C.Manning & Goldblatt is a new species for
Felicia sp. 1. (Manning & Goldblatt 2002).
Gazania pectinata (Thunb.) Hartweg (not (Thunb.) Spreng.) (Mabber-
ley 1980).
Macledium Cass, is now recognized in the Cape flora for two of the
four species of Dicoma in the flora (Ortiz 2001).
Macledium relhanioides (Less.) S. Ortiz is a new combination for
Dicoma relhanioides Less. (Ortiz 2001).
Macledium spinosum (L.) S. Ortiz is a new combination for Dicoma
spinosum L. (Ortiz 2001).
Marasmodes oligocephala DC. (not M. oligocephalus DC.).
Masarmodes polycephala DC. (not M. polycephalus DC.).
Oedera epaleacea Beyers Twiggy shrublet to 60 cm. Leaves oblanceo-
late, spreading, gland-dotted. Flower heads radiate, few in slender
peduncles in lax umbels, yellow, disc florets female-sterile.
Achenes hairy. June-Sept. Sandstone outcrops in cracks, NW
(Swartruggens)* (Beyers 2001).
Osteospermum australe B.Nord. is a new species for Osteospermum
sp. 2 (Nordenstam 2004).
Osteospermum burttianum B.Nord. Densely leafy, glabrescent
rounded or spreading shrublet to 1 m. Leaves oblong-lanceolate,
leathery, ascending, margins minutely scabrid. Flower heads radi-
ate, solitary on short, roughly hairy peduncles, yellow. Achenes
obscurely ribbed, 5-7 mm long. Mainly Dec-May. Rocky sand-
stone slopes, LB (Langeberg near Heidelberg)* (Nordenstam 2004).
Osteospermum potbergensis A. R. Wood & B.Nord. Decumbent to
prostrate shrublet to 20 cm. Leaves leathery, petiolate, often sparse-
ly dentate, margins curved under. Flower heads radiate, single on
white-woolly peduncles, yellow. July-Dee. Stony lower slopes, SW
(Potberg)* (Wood & Nordenstam 2003).
Othonna alba Compton Like O. cylindrica but leaves long and slen-
der, almost filiform and rays usually white. Aug.-Oct. Sandy
plateaus, NW (Botterkloof, Cederberg, Anysberg).* Previously in-
cluded in O. cylindrica.
Othonna carnosa Less. Succulent shrublet with short, erect or sprawl-
ing branches, 10-30 cm. Leaves fleshy, ovoid to fusiform. Flower
heads radiate, few in lax, terminal cymes on slender peduncles, yel-
low. Mainly Apr. -Oct. Stony flats and slopes, NW, SW, AP, KM,
LB, SE (Namaqualand to Worcester to E Cape).
Othonna cylindrica (Lam.) DC. NW, SW (S Namibia to Langebaan).
Corrected range and taxonomy.
Othonna floribunda Schltr. does not occur in the flora, and is restricted
to Namaqualand.
Othonna spinescens DC. Spiny shrub, older branches slender and stiff.
Leaves tufted on short shoots, narrowly oblanceolate, coriaceous,
felted in axils. Flower heads radiate, solitary on short terminal
peduncles, yellow; pappus of marginal florets elongating in fruit.
Sept. Rocky sandstone slopes, NW, KM (Cederberg and Swart-
ruggens to Witteberg).*
Pteronia undulata DC. Like P. divaricata but leaves strongly undu-
late or crisped, glandular-papillate. Sept.-Oct. Rocky slopes, NW
(Namaqualand, Swartruggens).
Pteronia viscosa Thunb. Twiggy shrublet, 30-100 cm, branchlets
whitish, glabrous. Leaves oblong-lanceolate, keeled, leathery, setu-
lose-ciliate. Flower heads discoid, solitary at branch tips, yellow,
20-25 x ±15 mm; bracts rough, margins obscurely fringed.
Oct.-Dec. Rocky slopes, NW, KM (Namaqualand, W Karoo to E
Cape, Swartruggens, Little Karoo: Ladismith).
Roodebergia B.Nord. is a new monotypic genus and species for the
flora. It is probably close to Felicia (Nordenstam 2002b).
Roodebergia kitamurana B.Nord. Diffuse perennial rooting at nodes
to 20 cm. Leaves opposite, elliptic, roughly hairy. Flower heads dis-
coid, solitary, reddish purple. Jan. Rocky sandstone slopes, 1 850
m, NW (Hex River Mtns)* (Nordenstam 2002b).
Schistostephium umbellatum (L.f.) K. Bremer & Humphries (not S.
umbellata (L.f.) K. Bremer & Humphries).
Senecio addoensis Compton Like S. scaposus but leaves apically
toothed or lobed. Mar. Stony sandstone slopes, SE (Great Winter-
hoek Mtns and E Cape).
Senecio articulatus (L.) Sch.Bip. KM, SE (Montagu to Uitenhage and
Great Karoo). Revised range, not endemic.
Senecio corymbiferus DC. Gnarled or erect succulent shrub with
cane-like stems, 30-200 cm. Leaves fusiform, glaucous, striate.
Flower heads discoid, in sparse corymbs clustered apically, yellow.
Mar. -July. Rocky hills, often granite, NW (S Namibia to Cederberg
and Swartruggens).
Senecio ficoides (L.) Sch.Bip. KM (Swartberg Mtns to Suurberg).
Revised range, not endemic.
Senecio haworthii (Sweet) Sch.Bip. Thick-stemmed.white-felted shrub
to 70 cm. Leaves in terminal clusters, cylindric or fusiform, succu-
lent, white-felted. Flower heads discoid, yellow, large, mostly soli-
tary on thickly felted peduncles, involucres calycled. Nov.-Mar.
Rocky slopes, KM (Richtersveld and W Karoo to Witteberg Mtns).
Senecio littoreus Thunb. NW, SW (Namaqualand: Koekenaap to Cape
Peninsula and Napier). Revised range, not endemic.
Senecio ovoideus (Compton) H. Jacobsen (= Kleinia ovoidea Comp-
ton, Senecio sp. 5) Like S. crassulaefolius but leaves thicker,
ovoid and obtuse. Dec.-June. Dry stony slopes, KM (western Little
Karoo).* Previously included in S. crassulaefolius (DC.) Sch.Bip.
Stoebe nervigera (DC.) Sch.Bip. NW, SW, LB (Namaqualand to
Albertinia). Revised range, not endemic.
Syncarpha aurea B.Nord. is a new species for Syncarpha sp. 4
(Nordenstam 2003b).
Syncarpha chlorochrysum (DC.) B.Nord. is a new combination for
Syncarpha sp. 1 (Nordenstam 2003b).
Syncarpha mucronata (P.J.Bergius) B.Nord. is a new combination for
Syncarpha sp. 2 (Nordenstam 2003b).
Syncarpha staehelina (L.) B.Nord. is a new combination for the
species to which the name Syncarpha virgata (P.J.Berg.) B.Nord.
was misapplied (Nordenstam 2003b).
Syncarpha virgata (P.J.Bergius) B.Nord. is to be applied to Syncarpha
sp. 3 in Cape plants (Nordenstam 2003b).
Tarchonanthus littoralis P.P.J. Herman is a new species segregated
from T. camphoratus L. for the genus in the CFR. The revised range
is SW, AP. ?KM, LB, SE (Cape Peninsula to S KwaZulu-Natal)
(Herman 2002).
Current total: genera 123; species 1 047 (previously 121 genera and
1 035 species).
BORAGINACEAE
Trichodesma africanum (L.) Sm. (not (L.) Lehm.) (Mabberley 1980:
605).
BUDDLEJACEAE
Buddleja L. with three species has been removed to Scrophulariaceae.
Current total: genera 0; species 0 (previously 1 genus; 3 species).
CAMPANULACEAE (Merciera corrections, C.N. Cupido pers. comm.;
Grammatotheca correction, E. Knox pers. comm.).
Grammatotheca sp. 1 of Cape plants is Lobelia thermalis Thunb. (a
species included in Cape plants).
Merciera azurea Schltr. Rigid, closely leafy shrublet to 30 cm. Leaves
imbricate, stiffly linear, pungent, shortly hairy, margins slightly
revolute and roughly ciliate. Flowers subsessile in upper axils, blue
to purple, tube narrowly funnel-shaped, 10-25 mm long, petals
elliptic-lanceolate. Nov.-Feb. Sandstone slopes, SW (Sir Lowry’s
Pass to Bredasdorp).*
Merciera brevifolia A. DC. Like M. leptoloba but leaves shorter, less
than 8 mm long and petals ovate. Nov.-Feb. Shale or granite slopes,
SW (Houwhoek to Caledon Swartberg).* Revised description and
range.
Merciera eckloniana H.Buek. ex Eckl. & Zeyh. Like M. azurea but
plants slender and leaves scattered. Oct.-Feb. Rocky slopes, NW,
SW (Tulbagh to Groenland Mtns).* Previously a synonym of M.
brevifolia A. DC.
40
Bothalia 35,1 (2005)
Merciera leptoloba A. DC. Rigid, closely leafy shrublet to 30 cm.
Leaves imbricate, stiffly linear, pungent, shortly hairy, margins
lightly revolute and harshly ciliate, axillary leaves glabrous.
Flowers subsessile in upper axils, white, tube slender, 3-6 mm
long, petals linear-lanceolate. Nov. -Mar. Sandy flats and lower
slopes, SW, AP (Kogelberg to Bredasdorp).* Revised description
and range.
Merciera tenuifolia (L.f.) A. DC. Like M. azurea but axillary leaf
tufts present and corolla tube cylindrical. Dec.-Jan. Rocky slopes,
SW (Houwhoek to Kogelberg).* Revised description and range.
Merciera tetraloba C.N.Cupido Like M. leptoloba but floral parts in
fours, petals tipped mauve. Nov.-Feb. Clay and granite flats, SW
(Du Toitskloof to Gordon's Bay)* (Cupido 2002).
Wahlenbergia debilis H.Buek is the correct name for W. ramulosa
E.Mey. ex DC. (T. Lammers pers. comm.).
Current total: genera 13; species 186 (previously 13 genera and 184
species).
CELTIDACEAE
Chaetacme Planch, is added to the flora (Wilmot-Dear 1999).
Chaetacme aristata Planch. Monoecious scrambling shrub or small
tree with zig-zag branches and paired axillary spines. Leaves ellip-
tic, aristate, glossy. Flowers unisexual, in axillary cymes, greenish
or cream. Flowering time? Coastal and riverine forest, SW, SE
(Knysna to tropical Africa and Madagascar).
Current total: genera 2; species 2 (previously 1 genus and 1 species).
CERATOPHYLLACEAE
Ceratophyllum muricatum Cham. Monoecious, free-floating aquatic
herb to 3 m. Leaves whorled, aristate and mostly 3- or 4-branched,
margins sparsely cuspidate. Flowers unisexual, usually solitary at
nodes. Fruit warty with three slender spines. Flowering time?
Sluggish and stagnant fresh water, LB, SE (George to Old World
Tropics) (Wilmot-Dear 1997).
Current total: genera 1; species 2 (previously 1 genus and 1 species).
CONVOLVULACEAE
Calystegia sepium (L.) R.Br. Glabrous climber to 3 m. Leaves hastate-
sagittate. Flowers white or pink, 50-55 mm long, sepals broadly
lanceolate; bracts ovate-cordate, longer than calyx. Dec.-Jan. Bush,
SW (Cape Peninsula, northern hemisphere native now naturalized
along the Atlantic coasts of both hemispheres) ! (Meeuse & Welman
2000).
Calystegia soldanella (L.) R.Br. Glabrous creeping perennial to 50
cm, forming large mats. Leaves reniform and emarginate, subsuc-
culent. Flowers pink to pale purple, 25 — 40 mm long, sepals ovate;
bracts ovate-suborbicular, shorter than calyx. Nov.-Dee. Coastal
sands, AP (Stilbaai, nearly pantemperate).
Current total: genera 6; species 18 (previously 5 genera and 17 species).
CORNACEAE
Curtisia dentata (Burm.f.) C.A.Sm. has been removed to Curtisiaceae.
Current total: family no longer represented in the flora (previously 1
genus; 1 species).
CRASSULACEAE (P. Bruyns pers. comm.)
Adromischus bicolor Hutchison does not occur in the CFR.
Adromischus maculatus (Salm-Dyck) Lem. KM (Willowmore
District, Georgida, and E Cape). Corrected distribution.
Adromischus subdistichus Makin ex Bruyns Succulent perennial to
30 cm. Leaves suborbicular, brownish green without waxy bloom.
Flowers in a spicate cyme, greenish with purple stripes, petals
grooved, ovate, fused basally, anthers just exserted. Jan.-Feb. N-facing
sandstone and quartzite ridges, KM (Swartberg Mtns).*
Adromischus triflorus (L.f.) Berger KM, SE (Touws River to Great
Karoo and E Cape). Corrected distribution.
Cotyledon muirii Schonl. (inch C. eliseae Van Jaarsv.) Small shrublet
to 60 cm. Leaves obovate, green. Flowers several in a pedunculate
cyme, nodding, reddish, usually glandular, lobes twice as long as
tube. Mainly Oct.-Dec. Stony slopes, LB (Gouritz River Valley).*
Cotyledon papillaris L.f. Delicate sprawling shrublet with decumbent
branches to 25 cm long, rooting at nodes. Leaves linear-oblanceo-
late to fusiform, green. Flowers several in a pedunculate cyme, nod-
ding, reddish, usually glandular, lobes twice as long as tube. Mainly
Oct.-Dec. Gravelly slopes, KM (Little and Great Karoo).
Crassula badspoortense Van Jaarsv. Like C. perfoliata but leaves
broadly ovate and rounded inflorescence. Nov.-Feb. Sandstone
cliffs, KM (Calitzdorp: Badspoort).*
Crassula cremnophila Van Jaarsv. & A.E.van Wyk Like C. hemi-
spherica but leaves broadly obovate, flowers in a rounded thyrse,
petals 7 mm long and anthers black. Aug.-Feb. Sandstone cliff
faces, SE (Baviaanskloof and Kouga).*
Crassula deceptor Schonl. & Baker. Listed twice in Cape plants.
Crassula perfoliata L. Densely papillate, few-branched perennial to
1.5 m. Leaves opposite, lanceolate to triangular, green to grey
sometimes with purple blotches. Flowers in flat-topped, peduncu-
late clusters, tubular, white, pink or red, petals 3-6 mm long.
Oct.-Jan. Dry lower slopes, SE (Karoo and Uitenhage to Limpopo).
Tylecodon albiflorus Bruyns Succulent shrublet to 20 cm. Leaves dry
at flowering but not abscising, oblanceolate. Flowers in a narrow
cyme, funnel-shaped with spreading lobes, green but white with
reddish stripes in throat, tube 12-15 mm. Nov.-Feb. Renosterveld,
KM (Montagu to Barrydale).*
Tylecodon stenocaulis Bruyns Succulent shrublet to 30 cm. Leaves
dry at flowering but not abscising, oblanceolate. Flowers in a deli-
cate cyme, urn-shaped, yellowish green but purple in throat, tube
11-13 mm. Sept.-Mar. Dry slopes, NW (Swartruggens andTanqua
Karoo).
Current total: genera 5; species 129 (previously 5 genera and 123
species).
CUCURBITACEAE
Kedrostis psammophila Bruyns Monoecious tuberous perennial,
prostrate with stems to 1 m long, without tendrils. Leaves palmate.
Flowers borne at ground level on subterranean peduncles, male fas-
cicled, female solitary, greenish. Fruits berry-like, subterranean.
Apr. -June. Reddish sands, NW (Namaqualand to Redelinghuys)
(Bruyns 1993).
Current total: genera 5; species 8 (previously 5 genera; 7 species).
CURTISIACEAE
Curtisia dentata (Burm.f.) C.A.Sm. is transferred here from Comaceae.
Current total: genera 1; species 1 (family not previously included in the
flora).
CYPERACEAE
Isolepis incomtula Nees (not /. incomptula Nees).
DENNSTAEDTIACEAE
Hypolepis villoso-viscida (Thouars) Tardieu Rhizomatous perennial.
Fronds suberect to arching, to 1 m long, stipe hairy, lamina 3-pin-
nate-pinnatifid. Sori in 1 to 3 pairs on ultimate segments, pseudo-
indusium often strongly modified, receptacle haired. Perennial
streambanks and seeps, ± 50-760 m, SW (Peninsula to Genadendal
and E Cape, also S Atlantic Islands) (Roux 2001).
Current total: genera 4; species 6 (previously 4 genera; 5 species).
ERICACEAE
Erica amalophylla E.G.H.Oliv. & I.M.Oliv. Sprawling diffuse shrub-
let. Flowers small, cup-shaped, white, thinly hairy. Dec. Sandstone
shelters, NW (Twenty Four Rivers Mtns)* (Oliver & Oliver 2002b).
Erica anemodes E.G.H.Oliv. Compact shrublet to 50 cm. Flowers
small, campanulate, white. Nov. Sandstone slopes, NW (Hex River
Mtns and Keeromsberg)* (Oliver & Oliver 2001a).
Erica annalis E.G.H.Oliv. & I.M.Oliv. Erect shrublet to 1 m. Flowers
large, tubular, orange-red, shortly hairy. July-Oct. Quartzite rock
faces, KM (Kammanassie Mtns)* (Oliver & Oliver 2002b).
Erica blaerioides E.G.H.Oliv. Compact shrublet to 30 cm. Flowers
small, urceolate, white, shortly hairy. Dec.-Jan. Sandstone slopes at
high alt., KM (Swartberg Mtns)* (Oliver & Oliver 2001b).
Erica breviflora Duller is a synonym of Erica plukenetii L. (Oliver &
Oliver 2002a).
Erica casta Guthrie & Bolus is a synonym of Erica regia Bartl. (Oliver
& Oliver 2002a).
Erica cavartica E.G.H.Oliv. & I.M.Oliv. Diffuse procumbent shrub-
let. Flowers small, shortly tubular, hairy. Mar. Sandstone shelters,
NW (Cederberg Mtns)* (Oliver & Oliver 2002b).
Erica chionodes E.G.H.Oliv. Compact shrublet to 60 cm. Flowers
small, urceolate, white, finely hairy. Sept.-Nov. Sandstone seeps,
KM (Grootswartberg Mtns)* (Oliver & Oliver 2001b).
Bothalia 35,1 (2005)
41
Erica cymosa E.Mey. ex Benth. Sprawling diffuse shrublet. Flowers
small, campanulate, white to pale pink, sparsely hairy. Oct.-Mar.
Sandstone shelters, NW, SW (Hex River and Du Toitskloof Mtns to
Keeromsberg).* Revised description and range.
Erica comptonii T.M. Salter is a synonym of Erica banksii Andrews
(Oliver & Oliver 2002a).
Erica decora Andrews is a synonym of Erica viscaria L. (Oliver &
Oliver 2002a).
Erica dolfiana E.G.H.Oliv. Compact shrublet to 50 cm. Flowers
small, campanulate-urceolate, white or tinged pink. Oct.-Dec. Sand-
stone slopes at high alt., KM (Grootswartberg Mtns)* (Oliver &
Oliver 2001b).
Erica gallorum L. Bolus is a synonym of Erica viscaria L. (Oliver &
Oliver 2002a).
Erica gilva J.C.Wendl. is a synonym of Erica mammosa L. (Oliver &
Oliver 2002a).
Erica grandiflora L.f. is a synonym of Erica abietina L. (Oliver &
Oliver 2002a).
Erica humidicola E.G.H.Oliv. Bushy or lanky shrublet to 1 m. Flow-
ers small, broadly campanulate, shortly hairy, pink. Sept.-Oct.
Sandstone seeps. SW (Kogelberg)* (Oliver & Oliver 2000).
Erica intermedia Klotzsch ex Benth. Erect, rigid shrub to 1.2 m.
Flowers medium/large, tubular, white or green to yellowish, with
far-exserted anthers. Jan.-Dee. Sandstone slopes, LB, SE (Lange-
berg Mtns at Swellendam to Outeniqua Mtns at George)* (Oliver &
Oliver 2002a).
Erica jananthus E.G.H.Oliv. & I.M.Oliv. Compact or loose, single-
stemmed shrublet to 15 cm. Flowers small, urceolate, viscid, white.
Aug.-Nov. Rocky, S-facing sandstone slopes, KM (Swartberg Mtns:
Snyberg)* (Oliver & Oliver 2004).
Erica limnophila E.G.H.Oliv. Sprawling, tangled shrublet to 15 cm.
Flowers small/medium, urceolate, thinly hairy, white. Dec. Marshy
soils, SW (Du Toitskloof and Wemmershoek Mtns)* (Oliver &
Oliver 2001a).
Erica lineata Benth. is a synonym of Erica plukenetii L. (Oliver &
Oliver 2002a).
Erica lithophila E.G.H.Oliv. & I.M.Oliv. Compact brittle shrublet to 20
cm. Flowers medium, urceolate, pink. Aug.-Nov. N-facing sandstone
crevices, KM (Swartberg and Kammanassie Mtns)* (Oliver & Oliver
2002b).
Erica tnariae Guthrie & Bolus is a synonym of Erica regia Bartl.
(Oliver & Oliver 2002a).
Erica onosmiflora Salisb. is a synonym of Erica viscaria L. (Oliver &
Oliver 2002a).
Erica oreotragus E.G.H.Oliv. Compact shrublet to 40 cm. Flowers
small, urceolate with 4 basal bulges, densely hairy, pinkish. Dec.-Mar.
Sandstone slopes, KM (Swartberg Mtns)* (Oliver & Oliver 2001b).
Erica penduliflora E.G.H.Oliv. Erect shrublet to 1 m. Flowers large,
inflated tubular to um-shaped, white or green. Apr.-July. Sandy hills and
flats, SW, AP (Pearly Beach to Viljoenshof)* (Oliver & Oliver 2001c).
Erica petrusiana E.G.H.Oliv. & I.M.Oliv. Low woody shrublet. Flow-
ers medium, funnel-shaped, sparsely hairy, slightly sticky, dull yel-
low. Mar. Stony shale band and sandstone, SW (Kogelberg Mtns)*
(Oliver & Oliver 2002a).
Erica phylicifolia Salisb. is a synonym of Erica abietina L. (Oliver &
Oliver 2002a).
Erica pilaarkopensis H.A.Baker not E. pillarkopensis H.A.Baker (Oliver
2004).
Erica porteri Compton is a synonym of Erica thomae L. Bolus (Oliver
& Oliver 2002a).
Erica primulina (Bolus) E.G.H.Oliv. & I.M. Oliv. is a synonym of
Erica viridiflora Andrews (Oliver & Oliver 2002a).
Erica richardii E.G.H.Oliv. Prostrate to erect shrublet to 30 cm.
Flowers small, globose-urceolate, white, shortly hairy. May-July.
Crevices in N-facing quartzite outcrops, KM (Grootswartberg:
Witberg)* (Oliver & Oliver 2001a).
Erica rimarum E.G.H.Oliv. Compact gnarled shrublet to 10 cm.
Flowers small, campanulate, leathery, maroon. Oct.-Dec. S-facing
sandstone cliffs, NW, SW (Hex River and Du Toitskloof Mtns)*
(Oliver & Oliver 2000).
Erica rusticula E.G.H.Oliv. Compact shrublet to 30 cm. Flowers
small, widely funnel-shaped, pink. Apr. -May. Sandy flats, NW
(Cold Bokkeveld)* (Oliver & Oliver 2000).
Erica salicina E.G.H. Oliver (= E. viminalis E.G.H.Oliv.) Willowy
shrub to 1.5 m. Flowers small, ovoid, white. Jan. Moist, sheltered
sandstone cliffs, NW (Hex River Mtns: Milner Peak)* (Oliver &
Oliver 2001a; Oliver 2004).
Erica schelpeorum E.G.H.Oliv. & I.M.Oliv. Erect twiggy shrublet to
1.5 m. Flowers small, globose-urceolate, pink. Mainly May-Jul.
Dry renosterveld. KM (Swartberg and Kammanassie Mtns)* (Oliver
& Oliver 2002b).
Erica taylorii E.G.H.Oliv. Prostrate shrublet to 20 cm. Flowers me-
dium, ovoid-urceolate, finely hairy or smooth, pink. Oct.-Dec. Sand-
stone slopes at high alt., NW, KM (Cederberg, Swartberg Mtns)*
(Oliver & Oliver 2001b).
Erica tenax L. Bolus is a synonym of Erica thomae L. Bolus (Oliver &
Oliver 2002a).
Erica tragomontana R.C. Turner Erect, single-stemmed shrublet to
35 cm. Flowers broadly funnel-shaped, pink. Sep.-Nov. S-facing
quartzite slopes. NM (Cold Bokkeveld)* (Turner & Oliver 2004).
Erica umbratica E.G.H.Oliv. & I.M.Oliv. Delicate brittle shrublet to
50 cm with drooping branches. Flowers urceolate, white, sticky.
Jan.-Dee. Moist southern slopes, KM (Swartberg Mtns: Meirings-
poort)* (Oliver & Oliver 2002b).
Current total: genera 1; species 667 (previously 1 genus; 658 species).
EUPHORBIACEAE
Hyaenanche globosa (Gaertn.). Lamb. & Vahl has been removed to
Picrodendraceae.
Current total: genera 11; species 79 (previously 12 genera; 80 species).
FAB ACE AE
Aspalathus albens L. NW, SW (Namaqualand: near Hondeklipbaai to
Cape Peninsula). Revised range, not endemic.
Aspalathus hispida Thunb. NW, SW, AP. LB, SE (Namaqualand near
Springbok, Gifberg to Alexandria). Revised range.
Aspalathus spinescens Thunb. NW, SW (Namaqualand: near Honde-
klipbaai to Malmesbury). Revised range, not endemic.
Melolobium lampolobium (E.Mey.) A.Moteetee & B.-E.van Wyk
Rigid, thorny, scarcely glandular shrublet to 60 cm, with brown-
velvety stems. Leaves 3-foliolate, leaflets oblanceolate. Flowers
many along the thorns, yellow, fading reddish orange. Pods falcate,
shining. May-Jan. Karroid scrub, 900-1 530 m, SW, KM (Robertson
Karoo and Little Karoo mountains)* (Moteetee & Van Wyk 2001).
Podalyria myrtillifolia (Retz.) Willd. is the current name for P.
cuneifolia Vent. (Campbell & Van Wyk 2001).
Psoralea glaucescens Eckl. & Zeyh. NW (Richtersveld, Kamiesberg
to Bokkeveld Mtns). Revised range, not endemic.
Rafnia perfoliata (L.) Willd. is the current name for R. acuminata
(E.Mey.) G.J. Campbell & B.-E.van Wyk (Campbell & Van Wyk
2001).
Wiborgia obcordata (P.J.Bergius) Thunb. NW, SW, LB (Namaqua-
land: near Port Nolloth, and Bokkeveld Mtns to Mossel Bay).
Revised range, not endemic.
Current total: genera 37; species 761 (previously 37 genus; 760
species).
GENTIANACEAE
Sebaea amicorum I.M.Oliv. & Beyers Annual to 50 cm. Leaves nar-
rowly lanceolate. Flowers 4-lobed, yellow, calyx lobes slightly
keeled, corolla tube shorter than lobes, ± 4 mm long. Oct.-Dec.
Sheltered S-facing sandstone ledges, KM, SE (Klein Swartberg,
Great Winterhoek Mtns)* (Oliver & Beyers 2001).
Sebaea albens (L.f.) Sm. (not (L.f.) Roem. & Schult.) (Mabberley
1980: 605).
Sebaea aurea (L.f.) Sm. (not (L.f.) Roem. & Schult.) (Mabberley 1980:
605).
Current total: genera 3; species 32 (previously 3 genera; 31 species).
GERANIACEAE
Pelargonium senecioides L’Her. NW, SW, KM (Namaqualand to
Cape Peninsula and Witteberg). Revised range, not endemic.
HYACINTHACEAE
Albuca L. is now included in Ornithogalum; the 27 new combinations
in Ornithogalum resulting from this action are listed by Manning et
al. (2004).
Daubenya zeyheri (Kunth) J.C. Manning & Goldblatt Bulbous geo-
phyte to 10 cm. Leaves prostrate, shiny green, bracts small to 10
mm. Flowers clustered between leaves, tubular below, white, fila-
ments orange with purple base. May-June. Coastal limestone
flats, NW (Paternoster to Saldanha).* This species was incorrect-
ly identified as Daubenya angustifolia (L.f.) A.M.van der Merwe
& J.C.Manning.
42
Bothalia 35,1 (2005)
Dipcadi Medik. is now included in Omithogalum (Manning et al. 2004).
Drimia barkerae Oberm. ex J.C. Manning & Goldblatt is a new
species for Drimia sp. 1 (Manning & Goldblatt 2003).
Drimia ciliata (L.f.) J.C. Manning & Goldblatt is the correct name for
D. ciliata (L.f.) Baker (Manning & Goldblatt 2003).
Drimia fragrans (Jacq.) J.C. Manning & Goldblatt NW (Namaqua-
land: Hondeklipbaai, and Bokkeveld Mtns to Hex River Valley).
Revised range, not endemic.
Drimia hesperantha J.C. Manning & Goldblatt is the correct name for
Drimia revoluta (A.V.Duthie) J.C. Manning & Goldblatt (Manning
et al. 2004).
Lachenalia corymbosa (L.) J.C. Manning & Goldblatt is a new com-
bination for Polyxena corymbosa (L.) Jessop. (Manning et al. 2004).
Lachenalia ensifolia (Thunb.) J.C. Manning & Goldblatt is a new com-
bination for Polyxena ensifolia (Thunb.) Schonl. (Manning et al.
2004).
Lachenalia maughanii (W.F.Barker) J.C. Manning & Goldblatt is a
new combination for Polyxena maughanii W.F.Barker (Manning et
al. 2004).
Lachenalia paucifolia (W.F.Barker) J.C. Manning & Goldblatt is a new
combination for Polyxena paucifolia (W.F.Barker) A.M.van der
Merwe & J.C. Manning (Manning et al. 2004).
Ledebouria ensifolia (Eckl.) S. Venter & T.J. Edwards is a new name
for Ledebouria sp. 1 of Cape plants (Edwards & Venter 2003).
Massonia bifolia (Jacq.) J.C. Manning & Goldblatt is a new combina-
tion for Whiteheadia bifolia (Jacq.) Baker.
Massonia grandiflora Lindl. is now included in M. depressa Houtt.
(Manning et al. 2004).
Neopatersonia Schonland is now included in Ornithogalum (Manning
et al. 2004).
Ornithogalum cirrhulosum J.C. Manning & Goldblatt is a new name
for Dipcadi ciliare (Zeyh. ex Harv.) Baker. (Manning et al. 2004).
Ornithogalum cremnophilum (Van Jaarsv. & A.E.van Wyk)
J.C. Manning & Goldblatt (= Albuca cremnophila Van Jaarsv. &
A.E.van Wyk) Pendent, bulbous geophyte to 2 m, bulb usually epi-
geal, greyish green, scales firm, truncate above. Leaves lanceolate,
firm. Flowers erect on long pedicels, subsecund on an inclined pe-
duncle, white with pale greenish keels, inner tepals cowled, outer
anthers smaller. Dec. -Feb. Cliffs, SE (Baviaanskloof Mtns).* (Man-
ning et al. 2004).
Ornithogalum crispum (Baker) J.C. Manning & Goldblatt is a new
combination for Dipcadi crispum Baker (Manning et al. 2004).
Ornithogalum malodorum J.C. Manning & Goldblatt is a new name
for Dipcadi brevifolium (Thunb.) Fourc. (Manning et al. 2004).
Ornithogalum thermarum (Van Jaarsv. & A.E.van Wyk) J.C. Man-
ning & Goldblatt (= Albuca thermarum Van Jaarsv. & A.E.van
Wyk) Like O. cremnophilum but leaf bases persistent and fibrous.
Nov.-Dee. Sandstone cliffs, KM (Calitzdorp: Badspoort).* (Man-
ning et al. 2004).
Ornithogalum uitenhagense (Schonl.) J.C. Manning & Goldblatt is a
new combination for Neopatersonia uitenhagensis Schonland (Man-
ning et al. 2004).
Ornithogalum viride (L.) J.C. Manning & Goldblatt is a new name for
Dipcadi viride (L.) Moench. (Manning et al. 2004).
Polyxena Kunth is now included in Lachenalia (Manning et al. 2004).
Spetaea Wetschnig & Pfosser is a new monotypic genus for the species
identified as Scilla plumbea (Wetschnig & Pfosser 2003).
Spetaea lachenaliiflora Wetschnig & Pfosser is a new name for the
species identified as Scilla plumbea Lindl. in Cape Plants (Wetsch-
nig & Pfosser 2003).
Whiteheadia Harv. is now included in Massonia (Manning et al. 2004).
Current total: genera 10; species 193 (previously 14 genera; 192 species).
IRIDACEAE
Aristea capitata (L.) Ker Gawl. is the correct name for A. major Andrews
(Goldblatt et al. 2002).
Aristea bracteata Pers. is the correct name for A. monticola Goldblatt
(Goldblatt et al. 2002).
Aristea macrocarpa G.J. Lewis is now a synonym of A. bakeri Klatt
(Goldblatt et al. 2002).
Aristea nana Goldblatt & J.C. Manning Evergreen rhizomatous peren-
nial to 10 cm, with compressed, winged, unbranched flowering
stem with the terminal intemode elongated. Flower paired in termi-
nal clusters within green spathes, on long pedicels, anthers yellow.
Capsules ovoid. Aug.-Sept. Stony sandstone slopes, LB, SE (Robin-
son’s Pass to E Cape) (Goldblatt et al. 2005).
Babiana cuneata J.C. Manning & Goldblatt Acaulescent cormous
geophyte 0.8-1 .5 cm. Leaves abruptly truncate at widest point.
loosely pleated, smooth or sparsely hairy. Flowers several in dense
spikes at ground level, pale to deep blue, the lower lateral tepals
with white spear-shaped markings. Mainly Sept. Rocky sandstone
or dolerite slopes and flats, NW (Bokkeveld Mtns to Swartruggens
and Western Karoo) (Goldblatt & Manning 2004b).
Babiana fragrans (Jacq.) Goldblatt & J.C.Mannning is the correct
name for Babiana disticha Ker Gawl. (Goldblatt & Manning 2004b).
Babiana longiflora Goldblatt & J.C. Manning Cormous geophyte
1 .5-2 cm, with suberect stem. Leaves loosely pleated, softly hairy.
Flowers several in an inclined spike, zygomorphic, purple, tube
elongate, anthers erect, violet. Aug.-early Sept. Seasonally wet
stony flats, NW (Piketberg and Porterville)* (Goldblatt & Manning
2004b).
Babiana regia (G.J. Lewis) Goldblatt & J.C. Manning Cormous geo-
phyte 0.5-1 .2 cm, with stems arching outward. Leaves firm, erect,
tightly plicate. Flowers several in an inclined spike, actinomorphic,
violet with a deep red centre, anthers erect, pollen brown. Aug.-
Sept. Seasonally wet sandy flats, SW (Klipheuwel to Stellen-
bosch)* (Goldblatt & Manning 2004b).
Babiana spiralis Baker is the correct name for the plant called B. fim-
briata (Klatt) Baker in Cape plants (Goldblatt & Manning 2005).
Babiana truncata G.J. Lewis does not occur in the CFR.
Babiana sp. 1 (= Babiana inclinata Goldblatt & J.C. Manning ined.)
Cormous geophyte 1.5-3 cm, with stems arching outward. 1-eaves
firm, erect, tightly plicate. Flowers numerous in an inclined spike,
zygomorphic, violet with white and darker blue marks on the lower
tepals, dorsal tepal and stamens facing the spike apex. Sept.-Oct. Stony
clay flats and lower slopes in renosterveld SW (Piketberg to Paarl).*
Babiana sp. 2 (= Babiana melanops Goldblatt & J.C. Manning ined.)
Cormous geophyte 1-2 cm, with stems suberect. Leaves erect,
lanceolate, softly hairy. Flowers several in an erect spike, actino-
morphic, violet to purple with a dark centre, anthers arrow-shaped
with wide connective, blackish. Aug.-Sept. Clay slopes in renos-
terveld, NW, SW (Tulbagh valley to Mamre).*
Babiana sp. 3 (= Babiana noctiflora J.C. Manning & Goldblatt ined.)
Like B. odorata but flowers larger, with a tube 35-50 mm long,
narrow at base and wider in the upper 1 5-25 mm. Sept.-Oct. Rocky
outcrops in renosterveld, Paardeberg S of Malmesbury).*
Babiana sp. 4 (= Babiana papyracea Goldblatt & J.C. Manning ined.)
Cormous geophyte to 1.5 cm, with stems reaching ground level.
Leaves firm, linear, erect, tightly plicate. Flowers several in short
erect spikes, actinomorphic, purple with a darker centre, anthers
erect, pollen cream, floral bracts dry, papery and attenuate. Sept.-
Oct. Clay flats, NW (Bokkeveld Plateau).*
Babiana sp. 5 (= Babiana radiata Goldblatt & J.C. Manning ined.)
Cormous geophyte to 1-1.5 cm, with stems reaching ground level.
Leaves firm, linear, erect, softly hairy. Flowers several in short
erect spikes, actinomorphic, purple with a red centre and margins
often white below, anthers erect, pollen cream. Aug.-Sept. Sandy
flats, KM (Little Karoo near De Rust).*
Dietes grandiflora N.E.Br. Evergreen rhizomatous perennial, 30-50
cm. Leaves sword-shaped. Flowers white with violet style arms,
outer tepals with yellow marking at limb base and a line of yellow
hairs on claws, lasting 2-3 days. Mainly Sept.-Dee. Margins of
evergreen thicket, SE (Hankey to KwaZulu-Natal).
Ferraria divaricata Sweet Cormous geophyte to 45 cm, stem reach-
ing well above ground, much-branched above. Leaves sword-shaped,
crowded basally. Flowers brown to maroon with lighter brown mar-
gins or golden brown with darker margins, claws broad, forming a
wide cup, nectaries pale green, large, anther lobes divergent, capsule
beaked. Mainly late Sept.-Nov. Deep sands, NW, SW (Hondeklipbaai to
Cape Flats). Revised species circumscription and range.
Ferraria uncinata Sweet Flowers blue to violet with brown margins.
Aug.-Sept. Mainly sandstone slopes and outcrops, NW, SW (Klawer
to Malmesbury)*. Revised description and range, now endemic
(Goldblatt & Manning 2005).
Ferraria variabilis Goldblatt & J.C. Manning Cormous geophyte, 6-
20 cm, branching mostly near base. Leaves sword-shaped, crowded
at base, sheathing the stem. Flowers dull yellow, yellow-green or
brown, with banded or speckled markings and darker margins, often
putrid-smelling, claws broad, forming a wide cup, nectaries basal, dark-
coloured, anther lobes diverging, capsule beaked. Aug.-Nov. Sandy
and shale flats and rock outcrops, NW, SW, KM, LB (S Namibia to
Clanwilliam, Caledon to Little Karoo) (Goldblatt & Manning 2005).
Freesia fucata J.C. Manning & Goldblatt Cormous geophyte, 15-30
cm. Leaves linear, glaucous. Flowers white flushed mauve on re-
verse, sweetly scented, bracts tricuspidate. July. Renosterveld, SW
(Bosjesveld between Villiersdorp and Breede River)* (Manning &
Goldblatt 2001a).
Bothalia 35,1 (2005)
43
Freesia sp. 1 (= Freesia marginata J.C. Manning & Goldblatt ined.)
Like F. caryophyllacea but leaves leathery with thickened submar-
ginal veins. May-June. Gravelly washes in succulent karoo, SW
(Robertson Karoo).*
Freesia sp. 2 (= Freesia praecox J.C. Manning & Goldblatt ined.) Like
Freesia alba but stems mostly unbranched and without axillary
cormels. June-July. Rocky sandstone slopes, SW (Riviersonderend
Mtns).*
Gladiolus wilsonii (Baker) Goldblatt & J.C. Manning Cormous geo-
phyte, 30-50 cm. tunics fibrous. Leaves linear, whip-like. Flowers
in slender spikes, short-tubed, bilabiate, white or flushed lilac.
Oct.-Nov. Open grassland, SE (Humansdorp to E Cape). Range
extension, new for the flora.
Hesperantha ciliolata Goldblatt Like H. pilosa but leaves terete or
oval in section with four to several deep grooves, scabrid-ciliate in
the rib edges, and flowers violet, musk-scented. Aug.-Sept. Stony
sandstone slopes, KM (Voetpadsberg (Touwsrivier) and Roggeveld
Escarpment) (Goldblatt 2003).
Hesperantha malvina Goldblatt Like H. pilosa but leaves sparsely
long-hairy, and flowers pale mauve, larger with tepals ± 14 x 4—5 mm.
Sandstone cliffs, KM (Little Karoo: Anysberg)* (Goldblatt 2003).
Hesperantha sufflava Goldblatt Like H. falcata but leaves always
3, crowded at base, flowers pale yellow and perianth tube 12-16
mm long, exceeding tepals. Late July-Aug. Sandy gravel slopes in
renosterveld, SW (Malmesbury)* (Goldblatt 2003).
ixia atrandra Goldblatt & J.C. Manning Cormous geophyte to 50 cm.
Leaves narrowly lanceolate, usually four, stem 1- or 2-branched.
Flowers crowded in a dense spike, pink or cream-coloured with a
large dark centre, tube filiform, stamens fully exserted. blackish,
anthers broad, arrow-shaped with exposed connective. Sept.-Oct.
Renosterveld, SW (Bosjesveld between Villiersdorp and the Breede
River)* (Manning & Goldblatt 2001a).
Ixia superba J.C. Manning & Goldblatt Cormous geophyte to 60 cm.
Leaves lanceolate, twisted. Flowers crowded in a dense 3-6-flow-
ered spike, salver-shaped, pale to deep pink, purple to blackish in
the centre, lightly scented, tube filiform below, stamens blackish,
anthers linear. Aug.-Sept. Loamy lower slopes, KM (Little Karoo,
Montagu)* (Goldblatt & Manning 2004a).
Moraea cantharophila Goldblatt & J.C. Manning Like M. lurida but
flowers always cream, tepals claws shorter, forming a shallow cup,
and anthers partly exserted from floral cup, not foul-scented.
Aug.-Sept. Loamy clay and shale bands, SW (Sir Lowry's Pass to
Sandy’s Glen)* (Goldblatt & Manning 2002b).
Moraea lilacina Goldblatt & J.C. Manning is a new species for Moraea
sp. 1 (Goldblatt & Manning 2002b).
Moraea minuta Goldblatt is the correct name for M. minutiflora Gold-
blatt.
Moraea monticola Goldblatt is the correct name for M. obscura Gold-
blatt.
Moraea neopavonia R.C. Foster is now a synonym of M. tulbaghensis
L. Bolus (Goldblatt & Manning 2002b).
Moraea regalis Goldblatt & J.C. Manning is the correct name for M.
derustensis Goldlbatt & J.C. Manning nom. nud.
Moraea simplex Goldblatt & J.C. Manning Like M. elsiae but foliage
leaf solitary, flowers pale yellow, and style branches undivided, fili-
form, extending between bases of anthers. Sept.-Oct. Gritty sandy
flats, NW (eastern foot of the Piketberg Mtns)* (Goldblatt & Man-
ning 2004a).
Romulea discifera J.C. Manning & Goldblatt Cormous geophyte,
10-20 cm, with symmetrical, depressed bell-shaped corm, lower
margins forming a spreading ridge, stem branching above ground.
Leaves 3-5. Flowers cup-shaped, bright yellow with a darker yel-
low cup. Mid July-early Aug. Sandy flats, NW (Bokkeveld
Plateau)* (Manning & Goldblatt 2001b).
Romulea lilacina J.C. Manning & Goldblatt Cormous geophyte, 1-3
cm, corms rounded at base. Basal leaf solitary, sticky with adhering
sand grains. Flowers solitary, lilac with darker stripes in a pale cup.
Ripe capsules recurved. May-June. Deep sands in washes, NW
(Cold Bokkeveld: Katbakkies)* (Manning & Goldblatt 2001b).
Romulea papyracea Wolley-Dod is now a synomym of R. schlechteri
Beg. (Manning & Goldblatt 2001b).
Thereianthus montanus J.C. Manning & Goldblatt is a new name for
Thereianthus sp. 1 of Cape plants (Manning & Goldblatt 2004).
Tritoniopsis bicolor J.C. Manning & Goldblatt Like T. parviflora but
leaves narrowly lanceolate, pseudopetiolate, and fdaments shorter,
6-7 mm long. Dec. Seasonally waterlogged sandstone plateau, SW
(Bredasdorp Mtns)* (Manning & Goldblatt 2001c).
Tritoniopsis flava J.C. Manning & Goldblatt Like T. parviflora but
more robust with lanceolate leaves, 2- or 3-veined and flowers yel-
low. Dec. Seasonal marshes, SW ( Kogelberg Reserve near Palmiet
River mouth)* (Manning & Goldblatt 2001c).
Tritoniopsis toximontana J.C. Manning & Goldblatt Cormous geo-
phyte, 30-65 cm. Leaves lanceolate, 3-veined, pseudopetiolate.
Flowers pink, tube elongate, to 20 mm long. March-May. Sand-
stone outcrops, NW (Gifberg and Matsikamma Mtns)* (Manning &
Goldblatt 2001c).
Current total: genera 28; species 684 (previously 28 genera; 663 species).
ISOETACEAE
Isoetes toximontana L.J.Musselman & J.P.Roux Tufted geophyte,
rootstock 3-sided, with homy tricuspidate scales. Sporophylls 3-
10, to 42 mm long. Sporangium lacking velum. Megaspores grey-
green. Seasonal pools and seepage areas, 300-560 m, NW (Gifberg
and Cederberg Mtns)* (Musselman & Roux 2002).
Current total: genera 1; species 3 (previously 1 genus; 2 species).
LOGANIACEAE
Nuxia Comm, ex Lam. with N. floribunda Benth. has been removed to
Stilbaceae.
Current total: genera 1; species 2 (previously 2 genera; 3 species).
MARSILEACEAE
Pilularia americana A. Braun Minute herb, rhizome creeping, branch-
ed. Leaves simple, terete, to 19 mm long. Sporocarps globose, to
2.5 mm diam., densely haired. Rim of ephemeral pools, NW (Bok-
keveld Plateau and N and S America) (Roux 2002).
Current total: genera 2; species 5 (previously 1 genus; 4 species).
MENYANTHACEAE
Villarsia manningiana Omduff Like V. capensis but usually smaller.
Leaf blades up to 50 mm long. Sept.-Dee. Peaty soils and stream
margins (SW, SE: Cape Peninsula to Hermanus and Outeniqua
Mtns at Knysna)* (Omduff 2001).
Current total: genera 1; species 3 (previously 1 genus; 2 species).
ONAGRACEAE
Ludwigia octovalvis (Jacq.) P.H. Raven Softly woody shrub to 4 m,
shortly hairy on branches. Leaves linear to lanceolate. Flowers soli-
tary in upper axils, yellow, sepals 4. Oct.-Jan. Wet places and river
banks, NW (pantropical but probably naturalized in the Olifants
River Valley)!
Current total: genera 1; species 2 (previously 1 genus; 2 species).
PICRODENDRACEAE
Hyaenanche globosa (Gaertn.) Lamb. & Vahl (not (Gaertn.) Lam.), trans-
ferred from Euphorbiaceae.
Current total: genera 1; species 1 (not previously recognized in the flora).
PLANTAGINACEAE
Ilysanthes dubia (L.) Bemh., transferred from Scrophulariaceae.
Limosella africana Gluck, transferred from Scrophulariaceae.
Limosella grandiflora Benth., transferred from Scrophulariaceae.
Current total: genera 3; species 6 (previously 1 genus; 3 species).
POACEAE
Agrostis polypogonoides Stapf (not A. polygonoides Stapf).
Cenchrus incertus M.A.Curtis (not C. incertis M.A.Curtis).
Chaetobromus involucratus (Schrad.) Nees (= C. dregeanus Nees).
Cymbopogon nardus (L.) Rendle (= C. validus (Stapf) Stapf ex Burtt
Davy).
Cymbopogon pospischilii (K.Schum.) C.E.Hubb. (= C. plurinodis (Stapf)
Stapf ex Burtt Davy).
Digitaria scalarum (Schweinf.) Chiov. (= D. abyssinica of authors not
D. abyssinica (A. Rich.) Stapf, misapplied name).
Eragrostis mexicana (Homem.) Link (= E. virescens J.Presl. & C.Presl.).
Hordeum geniculatum All. (= H. marinum Huds.).
Leptochloa fusca (L.) Kunth (= Diplachne fusca (L.) P.Beauv. ex
Roem. & Schult.).
Miscanthus ecklonii (Nees) Mabb. is the correct name for M. capen-
sis (Nees) Andersson (Mabberley 1984: 442).
44
Bothalia 35,1 (2005)
Pentaschistis heptamera (Nees) Stapf Perennial to 30 cm. Leaves
basal, linear. Spikelets 5-6 mm long, in a dense panicle, lemmas
5-9-awned. Nov.-Dee. Coastal sands, SE (Humansdorp to East
London).
Current total: genera 61; species 208 (previously 61 genera; 207 species).
POLYGALACEAE
Muraltia bondii Vlok Erect, single-stemmed, closely leafy shrublet to
50 cm. Leaves subsessile, linear-lanceolate and semi-terete,
mucronate. Flowers solitary in axils, white tipped purple.
Aug.-Nov. Rocky sandstone slopes, KM (Little Karoo, Anys-
berg).*
Current total: genera 3; species 142 (previously 3 genera; 141 species).
RESTIONACEAE
Ceratocaryum caespitosum H.P.Linder Dioecious, caespitose peren-
nial to 1 m, culms simple. Nuts to 10 mm long, tuberculate at apex.
Nov. Sandstone slopes, 100-200 m, SW (False Bay to Hermanus)*
(Linder 2001a).
Ceratocaryum persistens H.P.Linder Dioecious, spreading perennial
to 1.5 m, culms simple. Nuts to 10 mm long, smooth. Mar. -Apr.
Sandstone slopes, 300-500 m, SW (Hottentots Holland Mtns)*
(Linder 2001a).
Ischyrolepis gaudichaudiana (Kunth) H.P.Linder NW, SW, AP, KM,
LB (Namaqualand to Uniondale). Revised range, not endemic
(Linder 2001b).
Current total: genera 19; species 320 (previously 19 genera; 318 species).
RHAMNACEAE
Noltea africana (L.) Endl. (not (L.) Rchb.f.).
ROSACEAE
Cliffortia ruscifolia L. NW, SW, KM, LB, SE (Richtersveld to Humans-
dorp). Revised range, not endemic.
SAPINDACEAE
Dodonaea viscosa Jacq. (= Dodonaea angustifolia L.f.).
SCROPHULARIACEAE
Buddleja L. with B. glomerata H.L.Wendl., B. saligna Willd., and B.
salviifolia (L.) Lam. moved here from Buddlejaceae.
Freylinia helmei Van Jaarsv. Erect slender resprouting shrub to 2.5 m.
Leaves narrowly elliptic, 20-30 mm long. Flowers in short
racemes, subpendulous, tubular, white to mauve, 25-27 mm long,
stamens unequal, style over half as long, 15-17 mm long. Oct-
Nov. Steep shale slopes in renosterveld, SW (Botrivier)* (Van
Jaarsveld & Thomas 2003).
Gomphostigma Turcz. is added to the flora.
Gomphostigma virgatum (L.f.) Baill. Slender, willowy, closely leafy
shrublet to 1 m. Leaves opposite, linear, usually greyish. Flowers in
long narrow racemes, white, scented, tube subcampanulate, 2—4
mm long, tepals spreading. Nov.-Sept. Along watercourses in run-
ning water among boulders, NW, KM (Namaqualand, Bokkeveld
and Swartberg Mtns, Namibia and Zimbabwe).
Halleria, with H elliptica Thunb., H. lucida L., and H. ovata Benth.,
removed to Stilbaceae.
Ilysanthes, with /. dubia (L.) Bemh., removed to Plantaginaceae.
Ixianthes, with I. retzioides Benth., removed to Stilbaceae.
Limosella , with L. africana Gliick and L. grandiflora Benth., removed
to Plantaginaceae.
Nemesia cheiranthus E.Mey. ex Benth. NW (Namaqualand: Komag-
gas, and Bokkeveld Mtns to Piketberg). Revised range, not endemic.
Selago gloiodes Hilliard is the correct name for Selago gliodes Hilliard
(Hilliard 1999).
Selago subspinosa Hilliard KM (W Little Karoo and Roggeveld
Escarpment). Revised range, not endemic (Hilliard 1999).
Current total: genera 31 ; species 408 (previously 33 genera; 411 species).
SOLANACEAE
Lycium strand veldcnse AM. Venter Dioecious thorny shrub to 1.5 m.
Leaves densely fascicled on short branches, succulent, narrowly
ovate or obovate. Flowers solitary, from centre of leaf clusters,
actinomorphic, tubular, deep purple, male flowers with fertile sta-
mens and vestigial style lacking a stigma, female flowers with long
style, anthers lacking pollen. Sept.-Dee. Sandy flats and dunes,
NW (Namaqualand to Velddrif) (Venter & Venter 2003).
Current total: genera 2; species 19 (previously 2 genera; 18 species).
STILBACEAE
Halleria elliptica Thunb., transferred from Scrophulariaceae.
Halleria lucida L„ transferred from Scrophulariaceae.
Halleria ovata Benth., transferred from Scrophulariaceae.
Ixianthes retzioides Benth., transferred from Scrophulariaceae.
Nuxia floribunda Benth., transferred from Loganiaceae.
Current total: genera 9; species 19 (previously 6 genera; 14 species).
TAMARICACEAE
Tamarix usneoides E.Mey. ex Bunge Willowy tree with slender,
drooping branches, to 9 m. Leaves scale-like. Flowers in massed
panicles, minute, pink to grey. Mainly Mar. -June. Stream banks or
dry river courses, KM (Little Karoo to Great Karoo and E Cape).
Current total: genera 1; species 1 (not previously included in Cape
plants).
THYMELAEACEAE
Gnidia denudata Lindl. ?KM, LB, SE (Touws River and Langeberg
Mtns to Mozambique: Inhaca Island). Revised range, not endemic
(Edwards, Beaumont & Styles 2001).
Passerina comosa C.H.Wright NW, SW, KM, LB (Kamiesberg, W
Karoo, and Cold Bokkeveld to Klein Swartberg Mtns). Revised dis-
tribution.
Passerina esterhuyseniae Bredenk. & A.E.van Wyk Like P. comosa
but bracts helmet-shaped, brownish, and pale yellow flowers turn-
ing red to brown after pollen release. Flowering time? High rocky
slopes and peaks (NW: Pakhuis Mtns and Redelinghuys)*
(Bredenkamp & Van Wyk 2003).
Passerina falcifolia (Meisn.) C.H.Wright SW, KM, LB, SE (Caledon
District and Outeniqua Mtns to Alexandria), revised range, not
endemic.
Passerina filiformis L. NW, SW, LB (Clanwilliam to Peninsula, Hex
River Mtns to Langeberg at Attaquaskloof).* Revised distribution.
Passerina montevaga Bredenkamp & A.E.van Wyk Like P. filiformis
but leaves below flowers swollen at the base. Flowering time?.
Rocky slopes, LB, SE (Mossel Bay to N Tanzania) (Bredenkamp &
Van Wyk 2002c).
Passerina nivicola Bredenkamp & A.E.van Wyk NW (Cold Bokke-
veld and W Karoo) Revised range. (Bredenkamp & Van Wyk
2002a).
Passerina quadrifaria Bredenkamp & A.E.van Wyk Like P. comosa
but leaves less hairy. Flowering time? Rocky sandstone slopes at
high alt., LB, KM, SE (Langeberg and Little Karoo mountains to
Great Winterhoek Mtns)* (Bredenkamp & Van Wyk 2002b).
Passerina truncata (Meisn.) Bredenkamp & A.E.van Wyk is the cor-
rect name for Passerina glomerata Thunb. and the distribution is
corrected to NW, SW, KM, LB, SE (Namaqualand and Bokkeveld
Mtns to Baviaanskloof) (Bredenkamp & Van Wyk 2003).
Current total: genera 4; species 127 (previously 4 genera; 124 species).
ZYGOPHYLLACEAE
Roepera cordifolia (L.f.) Beier & Thulin (= Zygophyllum cordifolium
L.f.).
Roepera cuneifolia (Eckl. & Zeyh.) Beier & Thulin (= Zygophyllum
cuneifolium Eckl. & Zeyh.).
Roepera debilis (Cham. & Schltdl.) Beier & Thulin (= Zygophyllum
debile Cham. & Schltdl.).
Roepera flexuosa (Eckl. & Zeyh.) Beier & Thulin (= Zygophyllum
flexuosum Eckl. & Zeyh.).
Roepera foetida (Schrad. & J.C.Wendl.) Beier & Thulin (= Zygo-
phyllum foetidum Schrad. & J.C.Wendl.).
Roepera fulva (L.) Beier & Thulin (= Zygophyllum fulvum L.).
Roepera fuscata (Van Zyl) Beier & Thulin (= Zygophyllum fuscatum
Van Zyl).
Roepera lichtensteiniana (Cham. & Schltdl.) Beier & Thulin (= Zygo-
phyllum lichtensteinianum Cham. & Schltdl.).
Roepera maculata (Aiton) Beier & Thulin ( = Zygophyllum macula-
turn Aiton).
Roepera maritima (Eckl. & Zeyh.) Beier & Thulin (inch Zygophyl-
lum uitenhagense Sond.).
Bothalia 35.1 (2005)
45
Roepera morgsana (L.) Beier & Thulin (= Zygophyllum morgsana L.).
Roepera pygmaea (EckJ. & Zeyh.) Beier & Thulin (= Zygophyllum
pygmaeum Eckl. & Zeyh.).
Roepera rogersii (Compton) Beier & Thulin (= Zygophyllum rogersii
Compton).
Roepera sessilifolia (L.) Beier & Thulin (= Zygophyllum sessilifolium L.).
Roepera spinosa (L.) Beier & Thulin (= Zygophyllum spinosum L.;
inch Z. procumbens Adamson) NW, SW (Namaqualand: near
Kleinzee, and Lambert’s Bay to Cape Peninsula) Revised range, not
endemic. (Van Zyl 2000).
The undescribed species of Zygophyllum listed in Cape plants should
also be included in Roepera.
Tetraena retrofracta (Thunb.) Beier & Thulin (= Zygophyllum retro-
fractum Thunb.).
Current total: genera 5: species 22 (previously 4 genera; 23 species).
ACKNOWLEDGEMENTS
Fieldwork in southern Africa was supported by the
National Geographic Society (grants 5994-97, 6704-00
and 7103-01). We thank Clare Archer, R.K. Brummitt,
Peter Bruyns, C.N. Cupido, Cornelia Klak, Eric Knox,
Thomas Lammers, Bertil Nordenstam, Dee Snijman, and
J.J.A. Vlok for their help in compiling the corrections
presented here.
REFERENCES
ANGIOSPERM PHYLOGENY GROUP ( APG). 1998. An ordinal
classification for the families of flowering plants. Annals of the
Missouri Botanical Garden 85: 531-553.
ANGIOSPERM PHYLOGENY GROUP (APG). 2003. An update of
the Angiosperm Phylogeny Group classification for the orders
and families of flowering plants. Botanical Journal of the
Linnean Society 141: 399^436.
APPLEQUIST, W.L. & WALLACE, R.S. 2001. Phylogeny of the por-
tulacaceous cohort based on ndhF sequence data. Systematic
Botany 26: 406-419.
BEIER, B.-A., CHASE, M.W. & THULIN. M. 2003. Phylogenetic
relationships and taxonomy of subfamily Zygophylloideae
(Zygophyllaceae) based on molecular and morphological data.
Plant Systematics and Evolution 240: 11-39.
BEYERS, J.B.P. 2001. A new species of Oedera from Western Cape.
South Africa (Asteraceae). Bothalia 31: 41, 42.
BEYERS, J.B.P. 2002. A new species of Arctotheca from Northern
Cape, South Africa (Asteraceae). Bothalia 32: 185-187.
BREDENKAMP, C.L. & VAN WYK. A.E. 2002a. A new species of
Passerina from Western Cape, South Africa (Thymelaeaceae).
Bothalia 32: 76-79.
BREDENKAMP, C.L. & VAN WYK, A.E. 2002b. Passerina quadri-
faria (Thymelaeaceae): a new species from the southern Cape
and Little Karoo. South African Journal of Botany 68: 304—307.
BREDENKAMP, C.L. & VAN WYK. A.E. 2002c. Taxonomy of the Pas-
serina filiformis complex (Thymelaeaceae). Bothalia 32: 29-36.
BREDENKAMP, C.L. & VAN WYK. A.E. 2003. Taxonomy of the
genus Passerina (Thymelaeaceae). Bothalia 33: 59-98.
BRUYNS, P.V. 1993. A new species of Kedrostis from the Western
Cape (Cucurbitaceae). Bothalia 23: 233-235.
CAMPBELL. G.J. & VAN WYK, B.-E. 2001. A taxonomic revision of
Rafnia (Fabaceae, Crotalarieae). South African Journal of
Botany 67: 90-149.
CHASE, M.W.. SOLTIS, D.E., SOLTIS, P.S., RUDALL, P.J., FAY,
M.F., HAHN, W.H., SULLIVAN, S„ JOSEPH, J„ MOLVRAY,
M., KORES, P.J., GIVNISH, T.J., SYTSMA, K.J. & PIRES,
J.C. 2000. Higher-level systematics of the monocotyledons: an
assessment of current knowledge and a new classification. In K.
Wilson & P. Weston, Systematics and biology of the monocots:
3-16. Royal Botanic Gardens, Sydney.
CHESSELET, P. 2001. A new combination in Antimima N.E.Br.
(Mesembryanthemaceae). Aloe 38: 17.
CHESSELET, P, HAMMER, S. & OLIVER, I. 2003. Brianhuntleya , a
new genus endemic to the Worcester-Robertson Karoo. South
Africa (Mesembryanthemaceae). Bothalia 33: 160-164.
CHESSELET, P. & SMITH, G.F. 2001. Antimima aristulata comb. nov.
(Mesembryanthemaceae Fenzl); a heterophyllous succulent
from Cape Town, South Africa. Bradleya 19: 121-123.
CUPIDO, C.N. 2002. A new species of Merciera from Western Cape,
South Africa (Campanulaceae). Bothalia 32: 74—76.
EDWARDS, T.J., BEAUMONT, A.J. & STYLES, D. 2001. New
records and distributional disjunctions from South Africa,
Zimbabwe and Mozambique. Bothalia 31: 199-202.
EDWARDS, T.J. & VENTER, S. 2003. A revision of Ledebouria
(Hyacinthaceae) in South Africa. 3. The reinstatement of L.
ensifolia, L. galpinii and L. sandersonii. Bothalia 33: 49-51.
GOLDBLATT, P. 2003. A synoptic review of the African genus
Hesperantha (Iridaceae: Crocoideae). Annals of the Missouri
Botanical Garden 90: 390-443.
GOLDBLATT, P, DOLD, A.P & MANNING, J.C. 2005. Three new
species of Aristea from southern Africa (Iridaceae). Bothalia
35: 1-6.
GOLDBLATT, P. & MANNING, J.C. 2000. Cape plants. A conspectus
of the Cape flora of South Africa. Strelitzia 9. National Botanical
Institute, Cape Town & Missouri Botanical Garden, St Louis.
GOLDBLATT. P. & MANNING, J.C. 2002a. Plant diversity of the
Cape region of southern Africa. Annals of the Missouri
Botanical Garden 89: 281-302.
GOLDBLATT, P. & MANNING, J.C. 2002b. Notes and new species of
Moraea (Iridaceae: Iridoideae) from winter-rainfall southern
Africa. Novon 12: 352-359.
GOLDBLATT, P. & MANNING, J.C. 2004a. New species of Ixia
(Crocoideae) and Moraea (Iridoideae), and taxonomic notes on
some other African Iridaceae. Novon 14: 288-298.
GOLDBLATT, P. & MANNING, J.C. 2004b. Taxonomic notes and
new species of the southern African genus Babiana (Iridaceae:
Crocoideae). Bothalia 34: 87-96.
GOLDBLATT. P. & MANNING, J.C. 2005. Taxonomic notes on
Babiana and Ferraria in Namaqualand (Iridaceae). Bothalia
35: 71-74.
GOLDBLATT. P, MANNING, J.C. & GEREAU, R.E. 2002. Nomen-
clatural changes in Aristea section Racemosae (Iridaceae) in the
Cape Flora of South Africa. Novon 12: 190-195.
HARTMANN. H.E.K. 2001. Illustrated handbook of succulent plants:
Aizoaceae. A-E. Springer, Heidelberg.
HERMAN. P.P.J. 2002. Revision of the Tarchonanthus camphoratus
complex (Asteraceae-Tarchonantheae) in southern Africa.
Bothalia 32: 21-28.
HILLIARD. O.M. 1999. The tribe Selagineae (Scrophulariaceae).
Royal Botanic Garden. Edinburgh.
KLAK, C. 2003. New combinations, a new genus and five new species
in the Aizoaceae. Bradleya 21: 107-120.
KURZWEIL, H. & CHESSELET. P. 2003. Studies in the genus Ma-
chairophyllum (Mesembryanthemaceae), with notes on some
related genera. Bothalia 33: 19-39.
LINDER. H.P. 2001a. Two new species of Ceratocaryum (Restiona-
ceae). Kew Bulletin 56: 465^477.
LINDER, H.P. 2001b. The African Restionaceae. Contributions from
the Bolus Herbarium 20. CD-ROM.
LOWRY, P.P, PLUNKETT, G.M. & WEN. J. 2004. Generic relation-
ships in Araliaceae: looking into the crystal ball. South African
Journal of Botany 70: 382-392.
MABBERLEY, D.J. 1980. Generic names published in Salisbury’s
reviews of Robert Brown’s works. Taxon 29: 597-606.
MABBERLEY. D.J. 1984. Pallas’s buckthorn and two and a half cen-
turies of neglected binomials. Taxon 33: 433^144.
MANNING, J.C. & GOLDBLATT, P. 2001a. Two new renosterveld
species of Crocoideae from South Africa (Iridaceae). Bothalia
31: 189-192.
MANNING, J.C. & GOLDBLATT. P. 2001b. A synoptic review of
Romulea (Iridaceae: Crocoideae) in sub-Saharan Africa, the
Arabian Peninsula and Socotra, including new species, biolog-
ical notes, and a new infrageneric classification. Adansonia ser.
3, 23: 59-108.
MANNING, J.C. & GOLDBLATT, P. 2001c. Three new species of
Tritoniopsis (Iridaceae: Crocoideae) from the Cape Region of
South Africa. Bothalia 31: 175-181.
MANNING, J.C. & GOLDBLATT, P. 2002. A distinctive new species
of Felicia (Astereae) from Western Cape, South Africa (Aster-
aceae). Bothalia 32: 193-195.
MANNING, J.C. & GOLDBLATT, P. 2003. A new species and new
combinations in Drimia (Hyacinthaceae: Urgineoideae). Botha-
lia 33:109-111.
MANNING, J.C. & GOLDBLATT, P. 2004. A new species of Therei-
anthus (Crocoideae) from Western Cape. South Africa, no-
menclatural notes and a key to the genus. Bothalia 34: 1 03—
106.
46
Bothalia 35,1 (2005)
MANNING, J.C., GOLDBLATT, P. & FAY, M.F. 2004. A revised gene-
ric synopsis of Hyacinthaceae in sub-Saharan Africa based on
molecular evidence, including new combinations and the new tribe
Pseudoprospereae. Edinburgh Journal of Botany 60: 533-568.
MANNING, J.C., GOLDBLATT, P. & SNIJMAN, D. 2002. The color
encyclopedia of Cape bulbs. Timber Press, Portland, OR.
MANNING, J.C. & VAN DER MERWE, A.M. 2002. A new combina-
tion in Daubenya (Hyacinthaceae). Bothalia 32: 63-83.
MEEROW. A.W. & SNIJMAN, D.A. 2001. Phylogeny of Amaryl-
lidaceae tribe Amaryllideae based on nrDNA ITS sequences
and morphology. American Journal of Botany 88: 2321-2330.
MEEUSE, A.D.J. & WELMAN, W.G. 2000. Convolvulaceae. Flora of
southern Africa 28,1: 48-51.
MOTEETEE. A. & VAN WYK, B.-E. 2001 . The identity of Melolobium
lampolobium (Fabaceae: Papilionoideae). Bothalia 31: 209-212.
MUSSELMAN, L.J. & ROUX, J.P Isoetes toximontana (Isoetaceae), a
new quillwort with green megaspores from the northern Cape
of South Africa. Nown 12: 504-507.
NORDENSTAM, B. 2002a. Capelio B.Nord., a new name for a South
African genus of the Senecioneae, and the description of a new
species. Compositae Newsletter 38: 71-75.
NORDENSTAM, B. 2002b. A new monotypic genus of the Com-
positae-Astereae from the Cape Province, South Africa. Acta
Phytotaxonomica et Geobotanica 53: 101-105.
NORDENSTAM, B. 2003a. Alciope versus Capelio — a nomenclatural
ordeal. Compositae Newsletter 39: 48-51.
NORDENSTAM, B. 2003b. Further contributions to the genus Syn-
carpha (Compositae-Gnaphalieae). Compositae Newsletter 39:
52-57.
NORDENSTAM, B. 2004. Two new species of Osteospermum (Com-
positae-Calenduleae) from southwestern Cape Province, South
Africa. Edinburgh Journal of Botany 60: 259-265.
OLIVER, E.G.H. 2004. Nomenclatural changes in Erica (Ericaceae).
Bothalia 34: 38.
OLIVER, E.G.H. & OLIVER, I.M. 2000. Three new species of Erica
(Ericaceae) from Western Cape, South Africa. Bothalia 30:
147-153.
OLIVER. E.G.H. & OLIVER, I.M. 2001a. Four new species of Erica
(Ericaceae) from Western Cape, South Africa. Bothalia 3 1 : 1-8.
OLIVER, E.G.H. & OLIVER, I.M. 2001b. Five new species of Erica
(Ericaceae) from the Swartberg Range, Western Cape, South
Africa and a note on E. esterhuyseniae. Bothalia 31: 155-165.
OLIVER, E.G.H. & OLIVER, I.M. 2001c. Taxonomic problems in the
Erica filipendula complex. Yearbook of The Heather Society
2001:27-34.
OLIVER, E.G.H. & OLIVER, I.M. 2002a. The genus Erica (Ericaceae)
in southern Africa: taxonomic notes 1. Bothalia 32: 37-61.
OLIVER, E.G.H. & OLIVER, I.M. 2002b. Six new species and one
new subspecies of Erica (Ericaceae) from Western Cape, South
Africa. Bothalia 32: 167-180.
OLIVER, E.G.H. & OLIVER, I.M. 2004. Two new species of Erica
(Ericaceae); one from Western Cape and one from KwaZulu-
Natal, South Africa. Bothalia 34: 11-15.
OLIVER, I.M. & BEYERS, J.P.B. 2001. Anew specks of Sebaea from
the Swartberg Range, Western Cape, South Africa (Gentian-
aceae). Bothalia 31: 207-209.
OLMSTEAD, R.G., DEPAMPHILIS, C.W., WOLFE, A.D., NELSON,
N.D., ELISENS, W.J. & REEVES, PA. 2001. Disintegration of
the Scrophulariaceae. American Journal of Botany 88: 348-
361.
ORNDUFF, R. 2001. A(nother) new species of Villarsia (Menyan-
thaceae) from South Africa. Novon 1 1 : 437-439.
ORTIZ, S. 2001 . Reinstatement of the genus Macledium Cass. (Aster-
aceae, Mutisieae): morphological and phylogenetic arguments.
Taxon 50: 733-744.
ROURKE, J.P. 2002. Clivia mirabilis (Amaryllidaceae: Haemantheae)
a new species from Northern Cape, South Africa. Bothalia 32:
1-7.
ROUX, J.P. 2001. Hypolepis villoso-viscida new to the Flora of south-
ern Africa (Dennstaedtiaceae-Pteropsida). Bothalia 31: 195.
ROUX, J.P. 2002. First report of the genus Pilularia from continental
Africa (Marsileaceae-Pteropsida). Bothalia 32: 82, 83.
SAVOLAINEN, V., FAY, M.F., ALBACH, D.C., BACKLUND, A.,
VAN DER BANK, M„ CAMERON, K.M., JOHNSON, S.A.,
LLEDO, M.D., PINTAUD, J.-C., POWELL, M„ SHEAHAN,
M.C., SOLTIS, D.E, SOLTIS, PS., WESTON, P, WHITTEN,
W.M., WURDACK, K.J. & CHASE, M.W. 2000. Phylogeny of
the eudicots: a nearly complete familial analysis based on rbcL
gene sequences. Kew Bulletin 55: 257-309.
SNIJMAN, D.A. 2001a. A new species of Brunsvigia (Amaryllideae)
from Western Cape, South Africa (Amaryllidaceae). Bothalia
31:34-37.
SNIJMAN, D.A. 2001b. A new species of Cyrtanthus (Cyrtantheae)
from the southern Cape, South Africa (Amaryllidaceae).
Bothalia 31:31-34.
SNIJMAN, D.A. & ARCHER, R.H. 2003. Amaryllidaceae. In G.
Germishuizen & N.L. Meyer, Plants of southern Africa: an
annotated checklist. Strelitzia 14: 957-967 . National Botanical
Institute, Pretoria.
TURNER, R. & OLIVER, E.G.H. 2004. A new species of indehiscent-
fruited Erica from the central Kouebokkeveld, Western Cape,
South Africa (Ericaceae). Bothalia 34: 39 — 4 1 .
VAN JAARSVELD, E. 2001. Gasteria polita Van Jaarsv., a new
species from the Western Cape. Cactus and Succulent Journal
(U.S.) 73: 127-130.
VAN JAARSVELD, E. 2003. Four new cliff-dwelling Bulbine taxa
(Asphodelaceae) from the Eastern and Western Cape. Aloe 40:
4-10.
VAN JAARSVELD, E. & THOMAS, V. 2003. Freylinia helmei. Flow-
ering Plants of Africa 58: 112-1 16, t. 2197.
VAN JAARSVELD, E. & VAN WYK, A.E. 1999. Five new cremno-
philous taxa from semi-arid regions in South Africa. Aloe 36:
71-75.
VAN ZYL, L. 2000. Zygophyllum in the South African region. Ph.D.
thesis, University of Stellenbosch, Stellenbosch.
VENTER, A.M. & VENTER, H.T.J. 2003. Eycium strandveldense
(Solanaceae) a new species from the western coast of South
Africa. South African Journal of Botany 69: 476-479.
WETSCHNIG, W. & PFOSSER, M. 2003. The Scilla plumbea puz-
zle— present status of the genus Scilla sensu lato in southern
Africa and description of Spetaea lachetialiiflora , a new genus
and species of Massonieae (Hyacinthaceae). Taxon 52: 75-92.
WIKSTROM, N„ SAVOLAINEN, V. & CHASE, M.W. 2001. Evolu-
tion of the angiosperms: calibrating the family tree. Proceed-
ings of the Royal Society, London, B 268: 2211-2220.
WILLIAMSON, G. 2003. Three new species of Bulbine Wolf (Aspho-
delacaea) from the Western and Northern Cape Provinces of
South Africa. Aloe 40: 16-23.
WILMOT-DEAR, C.M. 1997. FSA contributions 8: Ceratophyllaceae.
Bothalia 27: 125-128.
WILMOT-DEAR, C.M. 1999. FSA contributions 13: Ulmaceae. Bothalia
29: 239-247.
WOOD, A.R. & NORDENSTAM, B. 2003. An interesting new species of
Osteospermum (Asteraceae-Calenduleae) from the Western Cape
Province, South Africa, providing a link to the genus Chry-
santhemoides. South African Journal of Botany 69: 572-578.
Bothalia 35,1:47-53 (2005)
Commiphora kaokoensis (Burseraceae), a new species from Namibia,
with notes on C. dinteri and C. namaensis
W. SWANEPOEL*
Keywords: Burseraceae, Commiphora Jacq., endemism, Kaokoveld, morphology, Namibia, new species, taxonomy
ABSTRACT
Commipora kaokoensis W.Swanepoel, here described as a new species, is known only from the Kaokoveld Centre of
Endemism, an arid region in northwestern Namibia. Illustrations of the plant and a distribution map are provided. Diagnostic
characters include the petiolate or subsessile, all-simple and relatively large leaves with the lamina obovate or elliptic. New
information is provided on the leaf morphology and geographical distribution of C. namaensis Schinz and C. dinteri Engl.,
species with which the new species shares some similarities. When without leaves or fruit, the three species can easily be con-
fused. A comprehensive table with diagnostic morphological features to distinguish between the three species is presented.
INTRODUCTION
A new species. Commiphora kaokoensis W.Swane-
poel, is described. Apart from one specimen ( Giess 9427
in WIND) collected in 1966, the new species remained
uncollected until recently. This can be attributed to its
limited range in remote parts of the Kaokoveld, north-
western Namibia. In addition, when without leaves or
fruit, the plants are virtually indistinguishable from C.
namaensis Schinz (Van der Walt 1973) and C. dinteri
Engl. (Van der Walt 1974), species for which it could
have been mistaken in the past.
New information is provided on the leaf characteris-
tics of C. namaensis and C. dinteri. Diagnostic morpho-
logical features to differentiate between these two
species and C. kaokoensis are presented. Previous mis-
identifications of C. dinteri are corrected and apparently
wrongly indicated localities, including quarter-degree
grids, of two herbarium specimens are rectified.
Apart from examining the herbarium collections of C.
kaokoensis , C. dinteri and C. namaensis in WIND and
PRE, numerous populations of the three species were
studied all over their respective ranges in Namibia dur-
ing two years of extensive field work. Morphological
characters were all determined from fresh material.
1. Commiphora kaokoensis W.Swanepoel , sp.
nov., C. namaensi Schinz habitu foliisque simplicibus
similis sed cortice in ramis caulibusque senioribus
plerumque longitudinaliter subcristato, foliis semper
simplicibus, plerumque maioribus, obovatis vel ellipti-
cis, margine plerumque crenato-serrato cum dentibus
pluribus usque ad duplo numero, raro subintegro, petio-
latis vel subsessilibus; petiolo elliptico vel lunato, in sec-
tione transversali maiore; floribus minoribus, solitariis
vel fasciculatis, parte distali loborum disci non ad hypan-
thium adnata, in floribus masculis lobis distincte bifidis,
filamentis staminum infra plerumque nec applanatis nec
incrassatis; fructu ovoideo, ellipsoideo vel obovoideo.
* RO. Box 21168, Windhoek, Namibia.
MS. received: 2004-04-02.
loculo fertili saepe ad loculum sterilem flexo, apiculato,
plerumque maiori, exocarpio valde glutinoso; angulo
inter loculos in apice putaminis plerumque valde minori,
putaminis sutura loculum fertilem versus convexa, locu-
lo sterili aspectu suturali variabili, subconcavo, paene
piano, triangulari, convexo vel e basi convexo ad con-
cavum apicem versus varianti; brachiis commissurae
pseudarilli plerumque longioribus, aliquando putaminis
apicem attingentes, differt.
TYPE.— Namibia, 1913 (Sesfontein): Kharokhaobvlak-
te, (-BC), 1 030 m, 10-05-2002, Swanepoel 1 (WIND,
holo.!; PRE, iso.!).
Illustrations: Steyn: 43, 44 & 87 (2003).
Dioecious, shrub-like tree, 0.3-3. 0 m tall, 0.3-6. 5 m
diam. Trunk branching repeatedly just above or below soil
level into many stems, rarely with a single trunk up to 200
mm high, up to 180 mm diam.; stems relatively thick, with
many thinner side branches (Figure 1). Bark pale grey to
reddish grey or yellowish cream to brown with small, dark
spots, with slightly raised, almost parallel longitudinal
ridges on stems and older branches, not peeling, trunk
rarely corky and irregularly cracked from soil level up to
150 mm. Branches and branchlets glabrous with scattered
small lenticels, not spine-tipped; branchlets relatively
short, stout, often scarred (Figure 2). Exudate milky, gluti-
nous, aromatic, producing a hard, colourless to pale green-
ish transparent resin, often cracked or disintegrated into
small glass-like pieces. Leaves simple, clustered on dwarf
lateral branchlets, spirally on shoots, glabrous, green; lam-
ina obovate to broadly obovate or elliptic to broadly ellip-
tic, (6-)15-34(-58) x (4 — ) 1 2— 20(— 32) mm, apex obtuse,
retuse or truncate, base cuneate or cuneate and abruptly
attenuate onto the petiole, rarely obtuse; margin finely cre-
nate-serrate or rarely serrate-dentate with ( 7— ) 1 4 — 20(— 25 )
teeth on each side, rarely subentire, entire near base;
midrib yellowish green, conspicuous abaxially towards
lamina base, prominently raised ab- and adaxially, espe-
cially towards lamina base (Figure 3); petiolate or subses-
sile, petiole from less than 1 mm up to 2 1 mm long, ellip-
tic to crescent-shaped in t/s with 5-12 vascular bundles,
dimensions in t/s (0.8—) 1.1—1 .6(— 1 .9) x (0.7—) 1 .0—1 -2(— 1 .4)
mm.
48
Bothalia 35,1 (2005)
FIGURE 1 . — C. kaokoensis in its natu-
ral habitat, ± 1 .2 m tall.
Inflorescence : flowers borne in clusters or solitary,
axillary. Flowers sessile or subsessile, unisexual, perigy-
nous, appearing before or with leaves or occasionally
flowering continuously until leaves have been shed.
Bracteoles ovate, apex acute, glandular, up to 1.2 mm
long. Calyx green to maroon or cherry, continuous with
hypanthium, usually sparsely glandular otherwise
glabrous; lobes ovate to triangular, apex acute. Petals
greenish yellow to yellow, glabrous, narrowly elliptic to
oblanceolate, recurved towards apex, but minute tip
indexed, inserted on hypanthium. Disk cylindrical with 4
fleshy lobes, adnate to hypanthium but distal part of lobes
free. Male flowers 2.8— 4.9 mm long; calyx 1. 6-3.4 mm
long, lobes 0.8-1. 4 mm long; petals 2.4-4. 0 mm long;
disk lobes distinctly bifid at apex; stamens 8, 4 long sta-
mens with filaments 1.6-2. 8 mm long, inserted on margin
of disk lobes, 4 short stamens with filaments 0.8-2. 1 mm
long, inserted on margin of disk between lobes; anthers
0.7- 1.0 mm long, equal in length on short and long sta-
mens; filaments subterete, rarely flattened and broadened
over lower part; gynoecium rudimentary (Figure 4A-C).
Female flowers 2. 0-2.7 mm long; calyx 1. 6-2.0 mm
long, lobes 0. 5-0.9 mm long; petals 1.7-2. 5 mm long;
disk lobes obscurely bifid; staminodes 8, 4 long and 4
short; ovary half inferior, sparsely glandular; style rela-
tively long, sparsely glandular, sutures deeply grooved;
stigma obscurely lobed; pistil 1.0-1. 6 mm long (Figure
4D-F). Fruit a drupe, ovoid, ellipsoid or obovoid, apicu-
late, flattened, asymmetrical (Figure 5), fertile locule
often bent over towards the sterile locule, appearing
FIGURE 2. — C. kaokoensis. Close up
of branches.
Bothalia 35,1 (2005)
49
slightly falcate in sutural view, 9-13 x 7-8 x 5-7 mm;
pericarp 2-valved; exocarp glabrous, glutinous, maroon
in ripe fruit; mesocarp not very fleshy; putamen flattened,
FIGURE 4. — C. kaokoensis. A-C, male flower: B, calyx and corolla
partly removed; C, disc as seen from above to depict position
of stamens (black) and rudimentary ovary (circle). D-F, female
flower: E, calyx and corolla partly removed; F, disc as seen
from above to depict position of stamens (black). G-I, putamen
with pseudo- aril: G, side with fertile locule; H, lateral view, fer-
tile locule (convex side) left, sterile locule (concave side) right;
I, side with sterile locule. A, B, Swanepoel 52; D, E, Swanepoel
73; G-I, Swanepoel 1. Scale bars: A, B, D, E. 1 mm; G-I, 5
mm. Artist: Anne Stadler.
asymmetrically ovoid, ellipsoid or rarely subglobose with
one fertile and one sterile locule, shghtly rugose, 5. 0-8. 7
x 4. 0-5. 8 x 3.0-4. 2 mm; fertile locule convex in sutural
and apical view; sterile locule dorsally ridged, variable in
sutural view: either shghtly concave, almost flat, triangu-
lar, convex or varying from convex at base to concave
towards apex, ± triangular in apical view; suture convex
towards fertile locule; angle between locules at apex
(420-)53°-730(-91°); pseudo-aril orange to red, fleshy,
cupular, covering ( 1 5— )20 — 25(— 34)% of fertile locule and
( 1 5— )25 — 40(^48)% of sterile locule, with 2 commissural
arms and 2(1) short facial lobes, extent of commissural
arms (relative to length of putamen with pseudo-aril
removed) (46-)60-80(-100)%, facial lobes convex or tri-
angular, 0. 5-1.1 mm on fertile locule, 0.3-2. 1 mm on
sterile locule, lobe on fertile locule often undeveloped
and completely absent; apical pits small, often absent
(Figure 4G-I). Flowering time : August to March, occa-
sionally throughout the year. Pollination : probably by
small ants, often observed on flowers.
FIGURE 5. — C. kaokoensis. Dwarf lateral branchlet with leaves and
fruit. Scale bar: 10 mm.
50
Bothalia 35,1 (2005)
Diagnostic characters and affinities'. Commiphora kao-
koensis differs from C. dinteri and C. namaensis mainly
in leaf and fruit characters, apart from minor differences
in the flowers, as well as in geographical distribution. C.
kaokoensis can be readily distinguished from these and
all other southern African succulent-appearing Commi-
phora species, by its all-simple, relatively large leaves,
(6 — ) 1 5— 34(— 58) x (4—)12-20(-32) mm, that are petiolate
or subsessile, with the lamina obovate or elliptic. The
midrib is conspicuous abaxially towards the lamina base
and prominently raised ab- and adaxially, especially
towards lamina base. The petiole in t/s is relatively thick,
(0.8—) 1.1 — 1.6(— 1.9) x (0.7 — ) 1 .0—1 .2(— 1 .4) mm, crescent-
shaped or elliptic, with 5-12 vascular bundles.
C. namaensis also has simple leaves, but rarely develops
a few additional trifoliolate leaves. The leaves are
rotund, orbicular, ovate or cordate, usually much smaller
than in C. kaokoensis , (5— )7— 1 2(— 1 5) x (4 — )5— 1 1 (—14)
mm and always petiolate; the midrib is inconspicuous
and not, or only slightly raised ab- and adaxially towards
the lamina base; the petiole in t/s is smaller, 0.5-0. 7 x
0.5-0. 7 mm, always crescent-shaped and with 3-7 vas-
cular bundles only.
C. dinteri usually has predominantly trifoliolate
leaves with a few additional simple leaves. Only rarely,
in a few restricted localities, does it either have only
simple leaves, or predominantly simple leaves with a
few additional trifoliolate leaves, or simple and trifoli-
olate leaves together in equal numbers on the same
plant. Simple leaves in C. dinteri are variable in shape,
and on individual plants may vary between obovate,
elliptic, ovate, cordate, oblate or slightly oblong. Un-
like C. kaokoensis , the simple leaves of C. dinteri are
always petiolate, with the lamina usually much small-
er, about half the size, (3— )7— 1 8(— 27 ) x (3— )6— 1 5 (-25)
mm; the midrib is inconspicuous and not, or only
slightly raised, ab- and adaxially towards the lamina
base; the petiole in t/s differs from C. kaokoensis by
being triangular, pentagonal or reniform in shape, usu-
ally smaller, 0.7-0. 8 x 0.6-0. 7 mm, and with only 3-7
vascular bundles. For comparable lamina size, the peti-
oles of both C. namaensis and C. dinteri are slender
when compared to those of C. kaokoensis. Petioles of
C. kaokoensis with t/s dimensions similar to those of
C. namaensis and C. dinteri , are only found on excep-
tionally small leaves, occasionally present amongst the
usually larger leaves.
Additional differences between these three taxa are
provided by fruit morphology: the suture of the putamen
in C. kaokoensis is convex towards the fertile locule and
the angle between locules at the apex is 42°-91°. In C.
namaensis the suture is rectilinear and the apical angle
between locules is 80°-150°, whereas in C. dinteri the
suture is rectilinear but curved towards the sterile locule
at the apex and the apical angle is 51°-120°. Through
examination of herbarium specimens and plants in the
field, a comprehensive comparative table of diagnostic
characters to differentiate between the three taxa was
compiled (Table 1 ).
Etymology : the specific epithet refers to the Kaoko-
veld of northwestern Namibia. The distribution of C.
12 14 16 18 20 22 24 26
FIGURE 6. — Known distribution of C. kaokoensis. •; C. dinteri, ®;
and C. namaensis, A.
kaokoensis partly falls in the previous politically de-
marcated Kaokoland and Damaraland, now called the
Kunene Region. Both these regions are included in the
broader biogeographical concept of ‘Kaokoveld’ as a
centre of plant endemism (Van Wyk & Smith 2001 ).
Distribution: C. kaokoensis is known from several
isolated localities, all within the Kaokoveld Centre of
Endemism in northwestern Namibia (Figure 6). More
specific localities include the upper reaches of the
Obias River (1813CC; 1913AA); just south of the
Giribesvlakte (1913AB); the Kharakhaobvlakte south
of Sesfontein (1913BC; -CB); Grootberg area
(1913DD; 2014AA); Huab River Valley (2014AC;
-AD); Petrified Forest area (2014BC); catchment area
of the Goantagab River in the Doros Crater area
(2014CB; -CD); Goedgenoeg area (2014DA; -DC) and
along the lower Ugab River (2014CC). C. kaokoensis is
locally common to rare within these areas, growing in
loose colonies of a few plants each amongst other
species of Commiphora , such as C. oblanceolata , C.
saxicola, C. virgata and C. wildii. Unlike most of the
many other species of Commiphora occurring in the
Kaokoveld, C. kaokoensis is often absent from areas
with seemingly suitable habitat.
Habitat and ecology: habitat requirements of C.
kaokoensis are quite specific. It occurs in the Namib
Desert and pro-Namib, 45-130 km from the coast at alti-
tudes of 200-1 100 m, where the annual rainfall is
75-150 mm. It grows in rocky areas preferring terraces of
calcrete and outcrops consisting of metasedimentary and
metamorphic rocks of the Damara Supergroup, including
marble, marble-conglomerate, quartzite, gneiss and
inclined strata of schist (Mendelsohn et al. 2002). It is
restricted to calcrete formations, wherever it occurs with-
in the extensive areas of Etendeka Group basalt, e.g. at
Palm near Palmwag. In spite of extensive searching,
nowhere could it be found on formations from the Karoo
Supergroup and associated lavas of the Etendeka Group,
not even in areas where these occur within only a few
metres of C. kaokoensis specimens growing on rocks of
the Damara Supergroup.
Bothalia 35,1 (2005)
51
TABLE 1. — Salient morphological differences between Commiphora kaokoensis, C. namaensis, C. dinteri and C. dinteri (form: Uis area)
Giess & Muller 14284 (WIND).
52
Bothalia 35,1 (2005)
2. C. dinteri Engl.
Hitherto C. dinteri was regarded as having trifoliolate
leaves only, with no mention of simple leaves (Van der
Walt 1986). However, thorough investigation revealed that
many herbarium specimens of C. dinteri in PRE and
WIND have in addition to trifoliolate leaves, a small per-
centage of simple leaves, variable in shape, including
ovate, obovate, elliptic, cordate, oblate, slightly oblong, or
orbicular. During field work throughout the range of C.
dinteri , both simple and trifoliolate leaves were found to a
varying extent on virtually all individual plants examined.
The presence of the occasional simple leaf on plants was
probably overlooked (or at least not sampled) when sever-
al specimens were collected. For example. Van der Walt
201 , 207 in WIND and PRE respectively, have trifoliolate
leaves only, whereas the duplicates in PRE and WIND
respectively, each have a few additional simple leaves.
Moving north, between the Khan and Omaruru Rivers in
central-western Namibia, a gradual increase in the percent-
age of simple leaves on specimens of C. dinteri was
observed. Specimens from the south of this area (Stingbank
and Ebony areas, 2215AB), have a small percentage of
simple leaves, similar to those observed further afield in
south-central Namibia at Maltahohe (2416DD) and in the
Tsaris Mountains (2416CD). To the east of the Schwarze
Kuppen (2115CA; -CB), the percentage simple versus tri-
foliolate leaves is ± equal, whereas in the vicinity of the
Omaruru River (Nai-nais & Okombahe areas, 2115AC;
-AD), the majority of leaves are simple with only a few tri-
foliolate leaves, both on dwarf lateral branches and on long
shoots. This is likewise the case with plants from the Uis
area (2114BD), which were previously considered as an
isolated population of C. namaensis (Van der Walt 1986),
probably on account of the simple leaves, which resemble
those of C. namaensis to a certain extent. In addition to
both the simple and trifoliolate leaves resembling those of
C. dinteri , the floral structure and fruit of the plants in the
Uis area resemble those of C. dinteri as well (Table 1).
Therefore, it is concluded that all these specimens belong
to C. dinteri and that all specimens from the Uis area (pre-
dominantly simple-leaved) were previously misidentified
as being C. namaensis.
In the Purros area of the Kaokoveld (1812DA), pre-
dominantly simple-leaved plants (Swanepoel 26 & 27)
were found alongside plants with predominantly trifolio-
late leaves (Swanepoel 28, 29 & 30). The leaves were
relatively large, especially on the predominantly simple-
leaved plants and the petioles were long and slender,
similar to those in some C. dinteri specimens from the
Okombahe area (211 5 AD) in central-western Namibia.
In the Flora of southern Africa (Van der Walt 1986), C.
dinteri specimens with large leaves are also mentioned
for plants from the Orupembe area, 60 km to the north. A
specimen with predominantly trifoliolate leaves,
Jacobsen & Moss K154, displays the same characteris-
tics typical of C. dinteri from the Orupembe area. A speci-
men with simple leaves, Moss & Jacobsen K195 and
from the same locality as Jacobsen & Moss K154, was
previously identified as C. cf. namaensis, probably on
account of the simple leaves, which resemble those of C.
namaensis superficially. This specimen displays the
same characteristics as Swanepoel 26, 27 with leaves
large and petioles relatively slender. It lacks, however.
the presence of a few typical C. dinteri trifoliolate leaves
as found on Swanepoel 26, 27 (only discovered on these
plants upon thorough examination). The presence of the
occasional trifoliolate leaf on plants was probably over-
looked (or at least not sampled) when Moss & Jacobsen
K195 was collected. Therefore, it is concluded that both
Swanepoel 26, 27 and Moss & Jacobsen K195 belong to
C. dinteri, and are similar to specimens from the Uis
area. In all the examples mentioned above, simple and
trifoliolate leaves occur on both short lateral branchlets
and on long shoots. Lateral leaflets are smaller than ter-
minal leaflets to a variable extent. Hence C. dinteri is
considered a heterophyllous species as it usually devel-
ops both simple and trifoliolate leaves on the same plant.
3. C. namaensis Schinz
According to Van der Walt (1986), C. namaensis occurs
as an isolated outlier population in the Uis area of north-
western Namibia, in addition to being widespread in south-
ern Namibia and the adjacent parts of South Africa. As
explained above, all the plants from the Uis area should be
reclassified as C. dinteri. Trifoliolate leaves and transi-
tional forms of simple to trifoliolate leaves on plants under
cultivation are mentioned by Van der Walt ( 1 986) and were
found on many C. dinteri plants during the present study,
whereas only simple leaves were found on C. namaensis
during extensive field work in southern Namibia. Trifolio-
late and intermediate leaves, however, are present on a
single herbarium specimen of C. namaensis ( Giess &
Muller 14284) from southern Namibia (27 1 7AC). It is con-
cluded that C. namaensis is confined to southern Namibia
and adjacent areas in South Africa and that it only very
rarely develops a few trifoliolate leaves in addition to the
usually simple leaves.
SPECIMENS EXAMINED
The location given for the previously unidentified
specimen of C. kaokoensis, Giess 9427 (WIND), namely
Spaarwater 7 1 1 , is from a locality underlain by Etendeka
lava (basalt) associated with the Karoo Supergroup. This
is in stark contrast with the geological formations at all
the other known localities for this species which are sit-
uated on formations of the Damara Supergroup. As
expected, extensive searching for C. kaokoensis on the
Farm Spaarwater 711 produced no specimens, nor could
any suitable habitat be found. However, on the Farm
Palm 708, 20 km to the northwest of Spaarwater, a few
small specimens of C. kaokoensis were found growing
on an isolated formation of calcrete within the otherwise
homogeneous Etendeka basalt area. Twelve kilometres
to the south of Spaarwater, on the Farm Fonteine 717,
also in the basalt area, C. kaokoensis was found growing
on an isolated outcrop of dolomite. Further explorations
on Spaarwater for any calcrete formation outcrops would
in all probability produce some C. kaokoensis specimens
at that locality.
The combination of distance, bearing and quarter-
degree square given for Moss & Jacobsen K195 and
Jacobson & Moss K154 , both C. dinteri, is incorrect.
Investigation revealed that a point 25 km to the northwest
of Purros is situated within quarter-degree square 1812DB
Bothalia 35,1 (2005)
53
and not within 1812DD as given for the locality of these
specimens. In addition, none of the plants in question could
be found during field work in 1812DD, nor could any be
found in a wide area around a point situated 25 km to the
northwest of Purros. At a point, west-northwest, 25 km
from Purros on the Purros-Orupembe District road
(D3707), situated within 1812DA, many plants were found
in habitat similar to that given for the two specimens.
Therefore it seems appropriate to consider the locality for
both Moss & Jacobsen K195 and Jacobsen & Moss K154
to be as follows: 25 km from Purros on road D3707, in
quarter-degree square 1812DA.
Acocks 15665, 18168 (3) PRE.
Basson 204 (2a) PRE. Biggs 209 (2a) PRE. Botha 683 (3) PRE.
Buhrman BUH 1/32, BUH 1-104 (2a) WIND. Burgovne 3527 (2a)
PRE. Burke 9759 (3) WIND.
Craven 2254 (2f) WIND. Curtis 1/9 (2a) WIND; CUR 1/172 (2b)
WIND.
De Winter 3563 (3) PRE, WIND; 6035 (2b) PRE. De Winter &
Leistner 5737 (2a) PRE, WIND. Dinter958, 5148 (3) PRE. Dreyer 480
(2a) PRE.
Friedrich FR 12/112 (2a) WIND.
Gerstner 6343 (3) PRE. Giess 9427 ( 1 ) WIND; 13490A (2a) PRE;
14520 (3) PRE; 14525, 14561 (3) WIND. Giess & Leipert 7466 (2a)
WIND. Giess & Muller 11717 (2a) WIND; 11709 (2b) PRE, WIND;
11739 (2f) PRE, WIND; 12267 (3) PRE, WIND; 14284 (3) WIND;
14323 (3) PRE. WIND. Giess, Volk & Bleisner 5325, 6908, 6931 (3)
WIND.
Jacobsen & Moss K154 (2b) WIND. Jankowitz 90/813 (3) WIND.
Mannheimer CM2317 (3) WIND. Mannheimer & Mannheimer
CM1380 (2b) WIND. Merxmiiller & Giess 935 (2b) PRE. WIND; 1688
(2f) PRE. WIND; 2703 (3) PRE. Moss & Jacobsen K154 (2a) PRE;
K195 (2e) PRE. Muller 1339 (2b) PRE, WIND. Muller & Giess 361
(2g) PRE. WIND.
Oliver & Miiller 6401 (3) PRE. Ortendahl 204 (3) PRE.
Robinson & Knouwds 63 (2a) WIND.
Strey 2296, 2612 (2a) PRE. Strobach 3445 (3) WIND. Strobach &
Dauth 3752 (3) WIND. Strobach & Kubirske 3031 (3) WIND.
Swanepoel 1 (1) PRE, WIND; 13-25, 31, 32, 40. 42-52, 72-74, 726 (1)
WIND; 30, 33, 34 (2b) WIND; 28, 29, 35, 38, 41 (2c) WIND; 26. 27,
36, 37, 39 (2d) WIND.
Van der Walt 201 (2a) WIND; 207 (2a) PRE. WIND; 267 (2a) PRE;
201 (2b) PRE; 267 (2b) WIND; 263, 305 (3) PRE. WIND; 307 (3)
WIND. Volk 11493 (2a) WIND.
Walter 1514, 11821 (2a) WIND; 1719 (2b) WIND; 2173 (3) WIND.
Walter & Walter 2258 (3) WIND. Ward 10805, 10889 (2b) PRE,
WIND. Wendt 48 (3) WIND. Wiss 1852 (3) WIND. Werger 1507 (3)
PRE.
2a, trifoliolate leaves only.
2b, predominantly trifoliolate with few simple leaves.
2c, ± equal number of trifoliolate & simple leaves.
2d, predominantly simple with few trifoliolate leaves, excluding speci-
mens from the Uis area.
2e, simple leaves only.
2f, specimens from the Uis area with predominantly simple & few tri-
foliolate leaves.
2g, specimens from Uis area with simple leaves only.
ACKNOWLEDGEMENTS
I would like to thank Prof. A.E. van Wyk, University
of Pretoria, for advice and support. Dr H.F. Glen, South
African National Biodiversity Institute, Durban, for
translating the diagnosis into Latin, Anne Stadler for the
line drawings and the staff of the National Botanical Re-
search Institute, Windhoek, especially Silke Bartsch,
Marianna Uiras, Coleen Mannheimer and Salome Kruger,
for their co-operation during visits to the National
Herbarium of Namibia. The National Herbarium of
Namibia and the South African National Biodiversity
Institute are also thanked for the use of information from
their databases: SPMNDB, Flora DB. The curator.
National Herbarium, SANBI, Pretoria, is thanked for
access to their collections; the assistance of Dr C.L.
Bredenkamp during visits to the herbarium is acknowl-
edged with thanks. Many thanks to Peter and Marilyn
Bridgeford for the use of a microscope. I am especially
grateful to my wife Hannelie for assistance during field
trips.
REFERENCES
MENDELSOHN. J„ JARVIS, A., ROBERTS. C. & ROBERTSON, T.
2002. Atlas of Namibia. David Philip, Cape Town.
STEYN, M. 2003. Afield guide, southern Africa Commiphora/S; Veld-
gids, suider-Afrika Commiphora. Published by the author, Polo-
kwane.
VAN DER WALT, J.J.A. 1973. The South African species of Commiphora.
Bothalia 11: 92-95.
VAN DER WALT, J.J.A. 1974. A preliminary report on the genus
Commiphora in South West Africa. Madoqua 1,8: 9,10.
VAN DER WALT. J.J.A. 1986. Burseraceae. Flora of southern Africa
18,3: 5-34.
VAN WYK, A.E. & SMITH, G.F. 2001. Regions offloristic endemism
in southern Africa. A review with emphasis on succulents.
Umdaus Press. Hatfield, Pretoria.
'
.
| }' V
Bothalia 35,1: 55-61 (2005)
Two new species of Asteraceae from Northern and Western Cape,
South Africa and a new synonym
J.C. MANNING* and P. GOLDBLATT**
Keywords: Asteraceae, Cape Floral Region, Chrysosoma hantamensis J.C. Manning & Goldblatt, new species, Oncosiphon, Senecio speciosissimus
J.C. Manning & Goldblatt South Africa, sp. nov., systematics
ABSTRACT
We recognize two new species of Asteraceae from the winter rainfall belt of South Africa and reduce a third to synon-
omy. Senecio speciosissimus sp. nov. has been confused with S. coleophylhts Turcz. in the past but is distinguished by its
taller stature, larger and more finely serrated leaves, and congested synflorescences containing (6-) 15-40 flowerheads. The
two species are also geographically separated: S. speciosissimus occurs in the Hottentots Holland and Franschhoek
Mountains of the southwestern Cape, whereas 5. coleophyllus is endemic to the Riviersonderend Mountains. Chrysocoma
hantamensis sp. nov. is a distinctive new species endemic to the Bokkeveld and Roggeveld Plateaus. It is distinguished by
its resprouting habit, 3-5-fid leaves and large capitula, 12-15 mm in diameter, with lanceolate, 3-veined involucral bracts,
the largest 9-10 x 2 mm. Investigation of the variation in leaf morphology of the two radiate species of Oncosiphon, O.
africanum (P.J.Bergius) Kallersjo and O. glabratum (Thunb.) Kallersjo, reveals that only one species can be maintained,
and O. glabratum is accordingly reduced to synonomy in O. africanum.
INTRODUCTION
During the preparation of the account of Asteraceae
for Cape plants (Goldblatt & Manning 2000) it was evi-
dent that several undescribed species were represented
among the collections at the Compton Herbarium. The
more distinctive of these taxa were included in the treat-
ment of the relevant genera as numbered entries. Some
of these species have since been described (Manning &
Goldblatt 2002; Nordenstam 2003). One of the more dis-
tinctive of the unnamed species was Senecio sp. 3, which
we describe here as S. speciosissimus. In addition, an
unusual species of Chrysocoma with pinnatisect leaves
from the Bokkeveld and Roggeveld Escarpment in
Northern Cape is described as the new species C. hanta-
mensis. At the same time we take the opportunity of
reducing to synonomy the poorly understood species
Oncosiphon glabratum , which further study reveals to be
conspecific with O. africanum.
1 . Senecio L.
Senecio L., with some 1 250 species worldwide, is by
far the largest genus in the tribe Senecioneae (Bremer
1994). As the central genus of subtribe Senecioninae, it
is certainly paraphyletic as currently defined (Jeffrey
1986, 1992) and its circumscription is consequently
rather broad. The genus is best represented in South
America (± 500 species) and Africa (± 350 species)
(Bremer 1994). It is especially common in southern
Africa, where some 300 species have been recorded
(Herman et al. 2000). Around one third of the southern
African species have been recorded from the Cape Floral
Region, where 107 named species plus four unnamed
* Compton Herbarium, South African National Biodiversity Institute,
Private Bag X7, 7735 Claremont. Cape Town.
** B.A. Krukoff Curator of African Botany, Missouri Botanical
Garden, P.O. Box 299, St. Louis. Missouri 63166, USA.
MS. received: 2004-11-10.
species were recognized by Goldblatt & Manning
(2000). Among the unnamed species that they listed, was
an unusually striking and distinctive taxon that is de-
scribed and named here.
Senecio speciosissimus J. C. Manning & Goldblatt ,
sp. nov.
Herba perennis robusta Senecio coleophyllo Turcz.
affinis sed 0.9-1. 8 m alta, foliis grandioribus serra-
tioribus (30— )40— 90(— 1 00 ) x 1 5— 25(— 35 ) mm, capitulis
synflorescentia congestis (6-) 15-40.
TYPE. — Western Cape, 3418 (Simonstown): Kogel-
berg Forest Reserve, NE slopes of Voorberg, ± 550 m,
(-BD), 2 October 1971, C. Boucher 1650 (NBG, holo.;
PRE, iso.).
Robust, single-stemmed perennial, 0.9- 1.8 m tall,
branching above; stems and branches densely leafy in
upper parts but leafless below, thinly to moderately
densely cobwebbed at first, later glabrescent. Leaves
closely imbricate, alternate, erect or lower leaves
spreading to reflexed, ovate or elliptical to lanceolate,
decreasing in size acropetally, (30-)40-90(-100) x
1 5— 25(— 35) mm, sessile and obscurely decurrent on
stem, obtuse at base, apiculate, hard and leathery,
glabrous or thinly cobwebbed at first adaxially, later
glabrous or nearly so except along midrib, usually
more densely and persistently cobwebbed or felted
abaxially, margins revolute and sparsely 4-10-serru-
late, rarely entire, 3-5-veined from base. Capitula het-
erogamous, radiate, few to several in congested
corymbs aggregated into rounded or corymbose pani-
cles of (6-) 15-40 heads, terminal corymbophore ± 40
mm long, individual peduncles 10— 30(— 50) mm long,
cobwebbed, with 1-5 scattered, lanceolate bracts.
Involucre campanulate, calyculate, 7-12 mm diam.;
involucral bracts uniseriate, 13-18, 9-13 x 1. 8-3.0
56
Bothalia 35,1 (2005)
FIGURE 1. — Senecio speciosissimus, Boucher 1650, holo. (NBG).
mm, ± equalling disc, lanceolate, acute or attenuate,
mostly with scarious margins, ciliate-penicillate at
tips, veins resinous, especially in lower half.
Receptacle flat, glabrous. Ray florets female, 11-13;
tube compressed-cylindrical, ± 4 mm long, abaxial
(outer) side sparsely glandular-pubescent in upper half;
lamina spreading, elliptic-oblong, 4(-6)-veined, 18-20
x 5-8 mm, pink or mauve, rarely white. Style branch-
ing just below mouth of tube, branches narrowly
oblong, 1.5 mm long, lateral margins stigmatic, apices
obtuse, shortly papillate. Disc-florets bisexual, many, ±
6 mm long, glabrous, yellow; lower part of tube cylin-
drical, ± 4.5 mm long, limb narrowly campanulate, ± 2
mm long, 5-lobed; lobes triangular, 1.8 x 0.8 mm, with
submarginal veins and median resin duct. Anthers 2.5
mm long including ovate apical appendage; anther
base minutely sagittate, ecaudate. Ovary narrowly
ellipsoid, 8-10-ribbed, clavate-pubescent between
ribs; style terete with swollen base on distinct sty-
lopodium, branching just below mouth of tube, branch-
es ± 1 mm long, lateral margins stigmatic, apices trun-
cate with crown of papillae. Cypselas narrowly ellip-
soid, ± 6.0 x 1.8 mm, 8-10-ribbed, clavate-pubescent be-
tween ribs, reddish brown. Pappus present in all florets,
bristles numerous, uniseriate, white, barbellate, decid-
uous, 4-5 mm long. Flowering time: June to Novem-
ber. Figure 1.
Distribution and ecology : Western Cape, restricted to
the southwestern coastal mountains between 600 and
1 500 m, from Bainskloof in the north to Kogelberg in
the south, a distance of ± 40 km (Figure 2). Plants grow
■as 2000
16° 18° 20° 22° 24°
I -J L I I
FIGURE 2. — Distribution of Senecio speciosissimus, •; S. coleophyl-
lus, O; and Chrysocoma hantamensis, ▲.
in scattered communities, never very abundantly, in
moist fynbos vegetation, in seepage areas or along
streams. Rainfall in these mountains is relatively high, in
parts averaging more than 2 500 mm per annum, falling
predominantly during the winter months, although sum-
mer cloud condensing against the upper slopes provides
some moisture through the summer. The single-
stemmed, willowy growth form indicates that the species
is a reseeder, re-establishing after fire through the germi-
nation of dispersed seeds. Plants seem to be relatively
short-lived and the species is apparently a member of
early successional plant communities that do not persist
into more mature fynbos older than 10 years. Flowering
of the species begins in winter, in July, and continues into
late spring and early summer, in November or December,
with peak flowering between August and October.
Diagnosis and relationships: Senecio speciosis-
simus is a distinctive species. Plants are single-
stemmed, with erect, wand-like stems up to 1.8 m tall
that are branched near the tops and densely leafy to-
wards the tips. The ovate to lanceolate, leathery leaves
are usually cobwebbed beneath, and the flowerheads
are clustered in paniculate synflorescences, with pink
to mauve (rarely white) ray florets. In its habit, foliage
and pink ray florets, it approaches Senecio coleophyl-
lus Turcz. (Figure 3), a smaller species, 0. 5-1.0 m tall,
of similar moist montane habitats in the Rivier-
sonderend Mountains to the east, and the two are
apparently geographic vicariants. S. speciosissimus is
readily distinguished from S. coleophyllus by its taller
stature, 0.9- 1.8 m tall, and congested, paniculate syn-
florescences comprising several, relatively short inflo-
rescences in the upper leaf axils of the flowering
branches, producing an accumulated total of (6-) 15-40
flowerheads. S. coleophyllus , in contrast, produces a
solitary, slender peduncle ± 100 mm long at each
branch tip, bearing 1— 3(— 5) flowerheads. The leaves in
that species are also smaller, 20-30 mm long vs 30-90
mm long, and proportionately more deeply and coarse-
ly incised. S. glastifolius L.f., from moist mountain
Bothalia 35,1 (2005)
57
FIGURE 3. — Senecio coleophyllus , Oliver 10919 (NBG).
slopes between George and Humansdorp, is another tall,
purple-flowered species that bears a superficial resem-
blance to S. coleophyllus and S. speciosissimus. It is
readily distinguished from both by its thinner-textured
leaves with the margins flat or only slightly revolute, the
complete lack of indumentum on leaves and inflores-
cence, its diffuse synflorescence, and by the narrower,
linear involucral bracts, at most 1.5 mm wide.
History : this striking species appears to have been
first collected by the German botanist Rudolf Schlechter
in the mountains above Franschhoek in November 1896.
Several collections have been made since then, all of
which have been referred to S. coleophyllus, and the
plant has been illustrated under that name in various wild
flower guides (Anonymous 1980; Burman & Bean
1985). This persistent confusion has obscured the true
identity of the species. During the preparation of Cape
plants (Goldblatt & Manning 2000), when we had the
opportunity to examine recent collections of S. coleo-
phyllus from the Riviersonderend Mountains, it was
clear to us that the plants from the Hottentots Holland
Mountains represented a distinct species, and it was
accordingly included in the account as Senecio sp. 3. The
type of Senecio coleophyllus is from the Riviersonderend
Mountains, as are all subsequent collections that match
it. The recognition of the populations from the Hottentots
Holland and adjacent mountains as a distinct species, S.
speciosissimus, confirms that S. coleophyllus is endemic
to the Riviersonderend Mountains, where it is found
along the length of the range, from Jonaskop in the west
to Pilaarkop in the east (Figure 2).
Other material examined
Senecio speciosissimus
WESTERN CAPE. — 3319 (Worcester): lower NE slopes of Seven
Sisters Mountain, above Witte River Valley, (-CA), 29 November
1959, E. Esterhuysen 31995 (BOL. S); Tierkloof on lower slopes of
Wemmershoek Mountains, (-CC), 5 November 1950, E. Esterhuysen
17699 (BOL); Franschhoek Pass, (-CC), 2 December 1928, H. Herre
s.n. STE8989 (NBG); Franschhoek. (-DD), 19 November 1896, R.
Schlechter 9266 (BOL, PRE); Franschhoek Peak, (-DD), 6 October
1946, R.H. Compton 18550 (BOL). 3418 (Simonstown): Sneeukop,
(-BB), 7 November 1938, T.P. Stokoe PRE44832 (PRE); Nuweberg
Forest Reserve, north slope of Sneeukop, (-BB), 26 November 1969,
M.F. Thompson 993 (NBG, PRE); sheltered valley on E side of
Somerset Sneeuwkop, (-BB), December 1939, E. Esterhuysen 3533
(BOL); Hottentots Holland Mtns, (-BB), November 1923, T.P Stokoe
BOL17561 (BOL); Hottentots Holland, (-BB), 13 November 1930, T.P.
Stokoe PRE20583 (PRE); Sir Lowry's Pass (bought in Adderley
Street), (-BB), September 1917, BOL14010 (BOL); Steenbras Dam on
Farm Rockview, (-BB), 7 June 1982, C.M. van Wyk 988 (NBG); foot
of Kogelberg Peak, (-BB), 18 August 1970, F.J. Kruger KR1047
(NBG, PRE); Kogelberg State Forest, southern end of Five Beacon
Ridge, (-BB), 8 October 1980, C. Boucher 4976 (NBG); Kogelberg,
kloof running down to Steenbras Dam, moist area near stream, (-BD),
27 August 1971, E.G.H. Oliver 3461 (NBG, PRE); Palmiet River
Mountains, (-BD), August 1924, T.P. Stokoe 977 (PRE); mountains
near Palmiet River, (-BD). April 1936 (fr.), T.P. Stokoe s.n. (BOL);
Kogelberg, (-BD), August 1924, Stokoe 962A (BOL, PRE); between
Kogelberg and Cape Hangklip, (-BD), October 1920, T.P. Stokoe 623
(PRE); Hangklip, (-BD), 16 October 1923, T.P. Stokoe PRE44828
(PRE); Hangklip, (-BD), September 1917, Marloth 7745 (PRE); Pringle
East Peak, sheltered SW slopes and cliffs, (-DD), 16 September 1951,
E. Esterhuysen 18859A (BOL); Pringle East Peak, steep S slope below
summit, 2500 ft, (-DD), 21 September 1952, E. Esterhuysen 20409
(BOL, PRE); swamp on S slopes of mountains near Betty’s Bay,
(-BD), E. Esterhuysen 13711 (BOL). 3419 (Caledon): between Vil-
joen’s Pass and Somerset Sneeuwkop, (-AA), 3 October 1938, T.P.
Stokoe 7039 (BOL); Grabouw-Boland trail, upper Riviersonderend
River. 600 m. (-AA), 29 October 1983, C. Burman 1243 (BOL).
Senecio coleophyllus
WESTERN CAPE. — 3319 (Worcester): Riviersonderend Mtns,
Jonaskop, moist slopes among rocks, (-CD), 25 January 1982, J.P.
Rourke 1772 (MO, NBG, PRE); Onklaarberg, 20 miles S of Worcester,
(-DC), December 1924, T.P. Stokoe 1073 (PRE). 3419 (Caledon):
Riviersonderend Mtns. Schilpadkop, steep, marshy, S slope, (-BA), 30
November 1952, E. Esterhuysen 20791 (BOL); Riviersonderend Mtns,
Pilaarkop, moist rocky area, (-BB). 28 October 1997, E.G.H. Oliver
10919 (NBG); Riviersonderend Mtns, (-BB), November 1940, T.P.
Stokoe SAM57807 (SAM). October 1945, T.P. Stokoe SAM57546
(SAM); Riviersonderend, (-BB), K.H. Barnard 472 (SAM); mountains
near Riviersonderend, Appelskraal. (-BB), November 1830, Zevher
2953 (K, PRE!, S, SAM!, iso.).
2. Chrysocoma L.
Chrysocoma L., a genus of 20 species, is endemic to
southern Africa, mainly the drier western and southwest-
ern regions, with a single species extending into
Mozambique (Bayer 1981; Herman et al. 2000). In the
tribe Astereae the genus is distinguished by its shrubby
habit and linear or oblanceolate leaves that are usually
viscid, mostly entire and often ericoid, usually solitary
capitula borne on elongate, naked peduncles, biseriate
pappus with an outer series comprising a row of minute,
persistent scales and an inner series of caducous bristles,
and flattened cypselae with thickened margins contain-
ing apical resin sacs beneath the marginal ribs. All but
two species have discoid capitula and just a single
species, C. tridentata DC. is known to have some of the
leaves toothed or lobed (Bayer 1981). The species de-
scribed here as C. hantamensis is anomalous in
58
Bothalia 35,1 (2005)
Chrysocoma in its distinctly pinnatifid leaves and unusu-
ally large capitula but accords with the genus in other
respects, particularly the fruit characters.
Chrysocoma hantamensis J.C. Manning & Gold-
blatt, sp. nov.
Species insignis suffrutescens, ramis decumbentibus,
foliis 3(-5)-fidis, capitulis magnis 12-15 mm diam. solitari-
is, bracteis involucri lanceolatis 3-costatis, 9-10 x 2 mm.
TYPE. — Northern Cape, 3119 (Calvinia): 12 km E of
Nieuwoudtville, 2.5 km S of Calvinia road, (-AC), 12
September 2004, P Goldblatt & L.J. Porter 12418
(NBG, holo.; E, K, MO, PRE, S, iso.).
Suberect subshrub with strong taproot; stem partly sub-
terranean, compactly branched, producing annual flowering
shoots; branches decumbent, woody at base and closely
leafy, becoming pedunculoid and almost naked distally,
150-250 mm tall, sparsely villous with hairs 0.5-0.75 mm
long. Leaves patent or suberect, lowermost opposite and
decussate with bases connate but soon becoming alternate,
rarely subtending dwarf axillary shoots, mostly trifid or
subdigitately pinnatisect and then 5-lobed, outline spathu-
late, 10-14 x 4—6 mm, lobes linear-lanceolate, 3-6 x
0.8- 1.0 mm, obtuse, lower leaves and often uppermost
becoming progressively oblong-lanceolate, 1.0-1 .5 mm
wide, leathery, margins strigose with coarse hairs 0.75 mm
long and scattered glandular hairs. Capitula homogamous,
discoid, solitary, terminal, pedunculate; peduncle sparsely
villous but densely villous apically beneath capitula,
(40-)60-100 mm long, naked or with one or two linear-
lanceolate bracts 6-8 mm long. Involucre broadly hemi-
spherical, 7-8 x 12-15 mm; involucral bracts 4(5)-seriate,
lanceolate, thinly hairy or subglabrous, with narrow, scari-
ous, fimbriolate margins, 3-veined, outer bracts 5-6 x 1-2
mm, acute, sparsely or more closely ciliate with hairs ± 0.5
mm long, inner bracts 9-10 x 2 mm, acuminate-attenuate,
glabrous. Receptacle convex, epaleate, alveolate. Florets
bisexual, yellow or apparently reddish at tips, tube cylindri-
cal but widening slightly in upper 2.0-2.5 mm, middle third
sparsely pubescent with tapering, eglandular hairs, 4.5-5 .0
mm long, 5-lobed; lobes recurved, triangular with thickened
margins, ± 0.8 x 0.4 mm. Anthers ± 2 mm long including
ovate, somewhat keeled apical appendage; anther bases
obtuse, ecaudate. Ovary obovate, flattened with thickened
margins, adpressed-hairy; style terete, branches ± 1 mm
long, incurved, linear, margins stigmatic, apical appendage
triangular, papillate with sweeping hairs at base. Cypsela
obovate, ± 3.0 x 1.5 mm, flattened with thickened margins,
moderately densely adpressed-hairy, with two small apical
resin sacs beneath ridges. Pappus biseriate; outer series of
minute, obtuse scales united basally in ring; inner series of
± 20 bristles, 3^1 mm long, barbellate above but sub-
plumose basally, caducous. Flowering time'. August and
September. Figure 4.
Distribution and ecology. Northern Cape, known from
two populations on the Bokkeveld and northern Roggeveld
Escarpments (Figure 2). This region, known locally as the
Hantam (Manning & Goldblatt 1997), is a significant cen-
tre of endemism (Van Wyk & Smith 2001). C. hantamen-
sis appears to be restricted to doleritic clays, growing in
succulent karoo vegetation. The fine-grained, red dolerite
soils of the Hantam support a wealth of edaphic endemic
taxa adapted to their peculiar characteristics, including
other recently described species of Asteraceae (Manning &
Goldblatt 2001). C. hantamensis bears a remarkable super-
ficial resemblance to another distinctive, narrow endemic
of this region, Euryops minis B.Nord. The two species,
which grow together at the type locality east of Nieu-
woudtville, are extraordinarily similar in vegetative form,
sharing short, partially subterranean stems and branches
bearing pinnatifid leaves, and producing annual flowering
shoots with long, naked peduncles bearing solitary flower-
heads. It would appear to be a marked instance of ecologi-
cal convergence.
Diagnosis and relationships', the highly distinctive C.
hantamensis is readily separated from all other species of
Chrysocoma by its 3-5-fid leaves and large capitula,
12-15 mm diam. with lanceolate, 3-veined involucral
bracts, the largest 9-10 x 2 mm. In its distinctly lobed
leaves it approaches C. tridentata DC. (including C. pin-
natifida DC.) (Bayer 1981) but this species from the
Little Karoo is a divaricately branched shrublet with
rather fleshy, mostly subterete leaves, and with short
peduncles at most 30 mm long, bearing capitula 10-12
mm diam. In addition, the involucral bracts, like those of
all other species of Chrysocoma , are smaller, ± 5 x
1.0-1. 5 mm, linear-lanceolate and 1-veined. C. hanta-
mensis is probably most closely allied to C. oblongifolia
DC., which extends from Namaqualand through the
Hantam and into the Tanqua Karoo. Both species are
subshrubs with decumbent annual stems bearing rela-
tively broad leaves with strigose margins and large capi-
tula carried on long peduncles. The decussate lower
leaves of C. hantamensis and C. oblongifolia are also
evident in the few other species of Chrysocoma that are
subshrubs rather than true shrubs, although their opposite
arrangement is easily overlooked. C. oblongifolia differs
from C. hantamensis in its entire, oblanceolate leaves,
glandular-hairy rather than villous stems and peduncles,
slightly smaller capitula, 10-12 mm diam., and smaller,
1-veined involucral bracts.
History, this unusual species was brought to our atten-
tion by Simon Todd, who collected it as part of his vege-
tation studies in the Nieuwoudtville area. Subsequent
investigation brought to light an earlier collection from
the northern Roggeveld Escarpment, made by M.F.
Thompson in 1975, at which time it was filed among the
undetermined species of Chrysocoma. The specimen was
apparently overlooked by Bayer (1981) in her revision of
the genus.
Additional material examined
NORTHERN CAPE. — 3119 (Calvinia): 10 km towards Calvinia
from Nieuouwdtville along R27, ± 2.5 km S of road, (-AC), September
2003, S. Todd 302 (NBG); N end of Roggeveld Escarpment, De Hoop,
(-DD), 22 August 1975, M.F. Thompson 2534 (NBG, PRE).
3. Oncosiphon Kallersjo
The genus Oncosiphon Kallersjo (Anthemidae: Matri-
cariinae) (Kallersjo 1988) was established for a small
group of annual species from Western and Northern Cape
Bothalia35,l (2005)
59
FIGURE 4. — Chrysocoma hantam-
ensis, Todd 302 (NBG). A,
whole plant; B. variation in
leaf morphology; C, detail of
leaf margin; D, involucral
bracts (outermost on left,
innermost on right); E, floret;
F, single anther; G, style
branches; H. cypsela. Scale
bars: A, 10 mm; B, 5 mm; D,
2 mm; E, H, 1 mm; C, F, G,
0.5 mm. Artist: J. Manning.
that had been previously placed either in Matricaria
(Harvey 1865), or segregated between the genera Pentzia
(discoid species) and Matricaria (radiate species)
(Hutchinson 1917). As circumscribed by Kallersjo
(1988), Oncosiphon is defined by its annual habit, 4-
lobed corolla with a brittle, swollen tube, and 4-ribbed
cypselae that lack myxogenic hairs and bear a small,
unequal-sided pappus. Seven species are currently recog-
nized in the genus (Kallersjo 1988), just two of them with
radiate capitula. These two species, O. africanum
(P.J.Bergius) Kallersjo and O. glabratum (Thunb.)
Kallersjo, are endemic to the coastal lowlands of the
southwestern Cape (Goldblatt & Manning 2000), from
near Leipoldtville to Cape Town. They occur along the
margins of seasonal, often somewhat saline pans, and
flower in late spring as the pans dry out. The two species
were retained as distinct by Kallersjo (1988) on the basis
of the degree of dissection of the lower leaves. The leaves
in O. glabratum are described as simply pinnatisect,
whereas in O. africanum the lower leaves at least are bip-
innatisect. The degree of leaf dissection is notoriously
variable within species of Asteraceae (see Hilliard 1977).
Significantly, no well-preserved recent specimens have
been unequivocally associated with O. glabratum among
the collections of radiate Oncosiphon material that we
have examined. The collection Ecklon & Zeyher 213/323
(SAM) that was associated with this species by Harvey
(1865), actually has distinctly bipinnatisect leaves and
thus falls within the circumscription of O. africanum. In
contrast, the ample material determined as O. africanum
displays a range of leaf forms that suggests that the degree
of dissection is often associated with the luxuriance of the
60
Bothalia 35,1 (2005)
FIGURE 5. — Oncosiphori africanum , variation in lower leaves. A, B,
leaves from two different plants, Compton 9815 (NBG); C, leaf
from Compton 5086 (NBG); D-F, leaves from three different
stems on single plant, Compton 15129 (NBG). Scale bar: 10
mm. Artist: J. Manning.
plants (Figure 5). Well-grown, branched specimens typi-
cally have larger, more highly dissected leaves than
smaller, unbranched specimens. For instance, Compton
9815 (NBG) comprises a dozen depauperate plants, most
of which bear simple, pinnatisect leaves (Figure 5A) but
one of which has the lower leaves weakly bipinnatisect
(Figure 5B). While most of the leaves of Compton 5086
(NBG) are bipinnatisect (Figure 5C), the lower leaves of
different branches on the well-grown plants that comprise
Compton 15129 (NBG) and Hugo 667 (NBG) display a
range of degrees of dissection, from simply pinnatisect to
strongly bipinnatisect (Figure 5D-F). The leaves in all
other species of Oncosiphon are bipinnatisect (rarely trip-
innatisect) (Goldblatt & Manning 2000). Examination of
the type of O. glabratum, which comprises unbranched
plants, reveals that one of the leaves displays a secondary
lobe and is therefore by definition bipinnatisect. The dis-
tinction between the two taxa is clearly untenable and O.
glabratum is accordingly reduced to a synonym of O.
africanum.
Oncosiphon africanum (P.J.Bergius) Kallersjo in
Botanical Journal of the Linnean Society 96: 312 (1988).
Matricaria africana P.J.Bergius: 296 (1767). Type:
Western Cape, Milnerton, 2 March 1980, H.P. Linder
2208 (BOL, neo.!), designated by Kallersjo (1988).
Oncosiphon glabratum (Thunb.) Kallersjo in Botanical
Journal of the Linnean Society 96: 312 (1988). Matricaria
glabrata (Thunb.) DC.: 51 (1838). Chrysanthemum africanum
Thunb.: 161 (1800). Type: South Africa, Thunberg no. 20132 in
Herb. Thunb. (UPS, microfiche!), syn. nov.
Other material examined
WESTERN CAPE. — 3218 (Clanwilliam): outside Leipoldtville on
road to Sandberg, (-BC), 13 October 1976, L. Hugo 667 (NBG); Clan-
william, Bergvlei, (-BC), 23 September 1934, R.H. Compton 5086
(NBG); Piketberg, Papkuil Valley, (-CA), 22 September 1940, Compton
9523 (NBG); salt marsh opposite Berg River Marsh, (-CC), 14 October
1986, M. O' Callaghan 1159 (NBG); Berg River, (-CD), 21 September
1940, R.H. Compton 9469 (NBG); Berg River Station, (-CD), 1 October
1943, R.H. Compton 15129 (BOL, NBG). 3318 (Cape Town): halfway
between Yzerfontein and Langebaan turnoff, near Salt Pan, (-AA), 3
August 1985, Kallersjo 29 (BOL, S); north of Yzerfontein, (-AB), 21
November 1991, A. Craven 26 (NBG); Yzerfontein, (-AB), August 1931,
L. Bolus s.n. (BOL); Mamre Road, (-BD), 12 October 1940, R.H.
Compton 9815 (NBG); Cape Town, Rietvlei, (-CD), C.F. Ecklon & C.L.
Zeyher 213/323 (SAM); Observatory, Varschvlei, (-CD), 10 November
1891, C. Wolley Dod 3636 (BOL); Rugby near Milnerton, (-CD),
February 1939, M.R. Levyns 7016 (BOL); Paarden Island, (-CD), Decem-
ber 1920, Pole Evans BOL16870 (BOL).
ACKNOWLEDGEMENTS
Our thanks to the Curator of the Bolus Herbarium for
the loan of material; Michelle Engelbrecht for producing
the scans of the herbarium sheets (Figures 1, 3); Simon
Todd for alerting us to the new species of Chrysocoma\
and Barrie Low and Uschi Pond for their recent collec-
tions of Senecio speciosissimus. This study was support-
ed in part by grant 7316-02 from the US National
Geographic Society.
REFERENCES
ANONYMOUS. 1980. Wild flowers of South Africa. Struik, Cape Town.
BAYER, E. 1981. Revision der Gattung Chrysocoma L. (Asteraceae-
Astereae). Mitteilungen der Botanischen Staatssammlung
Munchen 17: 259-392.
BERGIUS, PJ. 1767. Descriptiones plantarum ex Capite Bonae Spei.
Salvius, Stockholm.
BREMER, K. 1994. Asteraceae: cladistics and classification. Timber
Press, Portland.
BURMAN, L. & BEAN, A. 1985. Hottentots Holland to Hermanns.
Wild Flower Guide 5. Botanical Society of South Africa, Cape
Town.
DE CANDOLLE, A.P. 1838. Prodromus systematis naturalis regni
vegetabilis 6. Treuttel & Wiirtz, Paris.
GOLDBLATT, P. & MANNING, J.C. 2000. Cape plants. A conspectus
of the Cape flora of South Africa. Strelitzia 9. National
Botanical Institute, Cape Town & Missouri Botanical Garden.
HARVEY, W.H. 1865. Compositae. In W.H. Harvey & O.W. Sonder,
Flora capensis 3: 44—530. Dublin.
HERMAN, P.P.J., RETIEF, E„ KOEKEMOER, M. & WELMAN, W.G.
2000. Asteraceae. In O.A. Leistner, Seed plants of southern
Africa: families and genera. Strelitzia 10: 101-170. National
Botanical Institute, Pretoria.
HILLIARD, O.M. 1977. Compositae in Natal. University of Natal Press,
Pietermaritzburg.
HUTCHINSON, J. 1917. Notes on African Compositae. Kew Bulletin
1916: 241-254.
JEFFREY, C. 1986. Notes on Compositae, IV: the Senecioneae in east
tropical Africa. Kew Bulletin 41 : 873-943.
JEFFREY, C. 1992. Notes on Compositae, VI: the tribe Senecioneae
(Compositae) in the Mascarene Islands with an annotated world
checklist of the genera of the tribe. Kew Bulletin 47: 49-109.
KALLERSJO, M. 1988. A generic re-classification of Pentzia Thunb.
(Compositae-Anthemideae) from southern Africa. Botanical
Journal of the Linnean Society 96: 299-322.
MANNING, J.C. & GOLDBLATT, P. 1997. Nieuwoudtville. Wild
Flower Guide 9. Botanical Society of South Africa, Cape Town.
MANNING, J.C. & GOLDBLATT, P. 2001. A new species of Emilia
(Senecioneae) from South Africa. Bothalia 31 : 46^18.
Bothalia 35,1 (2005)
61
MANNING, J.C. & GOLDBLATT, P. 2002. A distinctive new species
of Felicia (Astereae) from Western Cape, South Africa. Bothalia
32: 193-195.
NORDENSTAM, B. 2003. Further contributions to the genus Syncarpha
(Compositae-Gnaphalieae). Compositae Newsletter 39: 52-57.
THUNBERG, C.P. 1800. Prodromus plantarum capensium 2. Edman,
Uppsala.
VAN WYK, A.E. & SMITH, G.F. 2001. Regions offloristic endemism
in southern Africa: a review with emphasis on succulents.
Umdaus Press, Hatfield, Pretoria.
.
Bothalia 35,1:63-92 (2005)
Notes on African plants
VARIOUS AUTHORS
PROTEACEAE
A NEW LEUCADENDRON (PROTEEAE) FROM WESTERN CAPE, SOUTH AFRICA
The discovery of a large new species of Leucadendron
R.Br., strikingly different from its congeners in growth
habit, brings to 85 the number of Leucadendron species
currently known, thus making it the largest genus of
Proteaceae in southern Africa. Discovered by Dr Tony
Rebelo on the final annual field outing of the Protea Atlas
Project in March 2001, this remarkable species is here
described as L. immoderatum on account of its extraordi-
nary growth habit.
Leucadendron immoderatum Rourke, sp. nov.,
species distinctissma propter habitum dimorphum et
folia dimorpha: acicularia in ramis vegetativis, obovata
apicem versus in ramis reproductivis. Inflorescentiae
gemmis masculinis et femineis intra involucram conico-
acutam inclusis. Bracteae involucrales lanceolato-acutae
vel lineari-acuminatae, per anthesin cadentes.
TYPE. — Western Cape, 3319 (Worcester): Riviersonder-
end Mountains, Olifantsberg, north side between Doring
and Witte Rivers, on a saddle facing northwest, (-CD),
12-11-2001, J.P. Rourke 2224 (NBG, female specimen,
holo.!, PRE, K, MO, iso.!).
Robust, erect shrub up to 2 m tall with a single main
stem branching just above ground level; growth habit and
leaves dimorphic. Basal vegetative branches highly
divaricate with acicular terete leaves, forming a dense
mat up to 2 m diam; up to 0.7 m tall, from which emerge
5-14 stout, woody, reproductive branches, bare when
mature except for flat, obovate leaves at apex of each
shoot. Basal branches 3-5 mm diam., covered with
prominent, closely arranged leaf scars. Basal leaves acic-
ular-terete, 10-20 x 1.0- 1.5 mm, densely ascending to
slightly incurved, upper surface canaliculate, initially
sparsely sericeous, soon glabrous becoming slightly glau-
cous, apex mucronate. Reproductive branches stout, erect,
straight, 8-15 mm diam., elongating up to 0.7-1. 0 m, ini-
tially with broadly linear leaves, 15 x 5 mm, soon
caducous, leaving branches bare. Upper leaves flat,
oblanceolate to obovate, 25-55 x 10-20 mm, glabrous,
apices rounded, minutely uncinate. Male capitula in
groups of 3-6 on short shoots towards apex of reproduc-
tive branches. Bud stage enclosed in a conic-acute involu-
cre of ovate-acute bracts, 5-25 x 5-7 mm, glabrous, mar-
gins ciliate; caducous at anthesis. Inflorescence broadly
cylindric, 25 x 20 mm, minutely pedunculate. Floral
bracts broadly ovate, acute 1 x 1 mm, glabrous, margins
ciliate. Perianth glabrous, straight, 5-6 mm long, pale
yellow; perianth claws equally recurved at anthesis.
Anthers ± 3 m long. Style filiform, straight, 5-6 mm long,
glabrous. Pollen presenter clavate, ± 3 mm long. Hypo-
gynous scales linear-obtuse, 1 mm long.
Female capitula solitary or up to 3 on short branchlets
at apex of reproductive shoot surrounded by a loose
pseudowhorl of erect, greenish yellow/ivory leaves. Bud
stage enclosed in a conic-acute involucre of lanceolate-
acute to linear- acuminate bracts, 5-12 x 3^40 mm,
glabrous but margins densely ciliate-sericeous; caducous
at anthesis. Inflorescence cylindric, sessile, 40-50 x
10 mm. Floral bracts very broadly ovate-acute, 5 x
6 mm, horizontally patent, projecting, cartilaginous,
glabrous, but apex minutely crinite. Perianth 4-15 mm
long, zygomorphic, laterally compressed; lateral perianth
claws densely sericeous, adaxial and abaxial claws
sparsely sericeous to glabrous. Staminodes 3; anterior
staminode absent. Style straight ± 4 mm long, patent,
partially clasped by sterile anterior perianth limb. Pollen
presenter oblong-clavate, terminal, adaxial surface glan-
dular. Ovary ovoid-compressed, 1 mm long, sharply dif-
ferentiated from style base, glabrous. Hypogynous scales
lanceolate-obtuse, 1 mm long, camose. Mature cone
3CM-0 x 60-90 mm, prominently ridged. Fruit a round-
ed, flattened, black, samara, 8x10 mm, apically retuse.
Figures 1-3.
Diagnostic characters : this species is unique in the
genus on account of the conic-acute, 40-55 mm long
involucres that completely enclose both male and female
inflorescences in their bud stages giving the inflores-
cence buds the appearance of small species of Protea or
certain species of Pteronia (Asteraceae). These involucral
bracts are rapidly caducous at anthesis, linear-lanceolate
and up to 45 mm long in female inflorescences, ovate-
acute and up to 7 mm long in male inflorescences. (L. micro-
cephalum also has large involucral bracts but these only
partially enclose the inflorescences and persist into the
post-pollination phase).
Leucadendron immoderatum is further distinguished
by its dimorphic growth habit and dimorphic foliage with
a few stout, erect, reproductive branches bearing broadly
obovate leaves around the capitula, developing from a
dense, highly dichotomous basal growth, producing aci-
cular-terete leaves.
Affinities : morphological evidence, especially seed,
cone and leaf characters suggest that L. immoderatum is
best accommodated in Leucadendron sect. Alatosperma
subsect. Compressa. This subsection contains a number
of species with predominantly acicular-terete leaves that
have a tendency to become flattened and broader (very
narrowly oblanceolate) around the inflorescence in some
species (Williams 1972). Leucadendron comosum (Thunb.)
R.Br. subsp. comosum may be the most closely allied
species but the leaf dimorphism is very much more pro-
nounced in L. immoderatum. The extreme dimorphic
64
Bothalia 35,1 (2005)
FIGURE I. - Male inflorescences of Leucadendron immoderatum, Rourke 2224. A, fully open; B, previous year’s inflorescences and inflores-
cence bud enclosed in involucre; C, basal involucral bract; D, juvenile foliage; E, single juvenile acicular leaf; F, male flower and sub-
tending floral bract; G, single perianth segment and anther; H, style, pollen presenter and hypogynous scales. Artist: I.M. Oliver.
Bothalia 35.1 (2005)
65
FIGURE 2. — Female inflorescences of Leucadendron immoderation, Rourke 2224. A, complete branch system showing basal juvenile foliage,
reproductive branch with long growth increments and terminal obovate leaves; B, flowering female cone with remnants of involucral
bracts; C, obovate upper leaf; D, involucral bract; E, unopened female flower and subtending floral bract (cone scale); F, open female
flower; G, gynoecium with hypogynous scales; H, mature female cone, I, mature fruit. Artist: I.M. Oliver.
66
Bothalia 35, 1 (2005)
FIGURE 3. — Leucadendron immoderation Rourke. A, female and B, male plants in the type locality. The peak in the background, unnamed on
modern maps, is immediately adjacent to and west of Olifantsberg. C, male inflorescence; D. female inflorescence. Note the remnant
acuminate involucral bracts abscissing at the base of each inflorescence.
Bothalia 35.1 (2005)
67
growth habit of L. immoderatum, in which the adult
shrub is sharply differentiated into vegetative and repro-
ductive branch systems, represents a unique apomorphy
within this lineage.
Distribution : Leucadendron immoderatum is known
from a single population growing in mesic mountain fyn-
bos at about 1 300 m on the upper north slopes of
Olifantsberg in the Riviersonderend Mountains on a north-
west-facing saddle between the Doring and Witte Rivers
(Figure 4). When the type material was collected, only 16
mature adult plants were observed at this site. These flow-
ering specimens had escaped a bum that had occurred
approximately three years earlier. The original population
must have been considerably larger as some 30 seedlings
were observed in the area. This species is clearly an excep-
tionally narrow endemic in common with several other
narrow proteaceous endemics from the same area, like
Serruria williamsii and Sorocephalus alopecurus.
Biology. Leucadendron immoderatum is a serotinous,
non-lignotuberous seed regenerator. Seedling development
in the post-fire phase appears to be slow, eventually
resulting in the development of a dense, highly divaricate
mat-like shrub, ± 0.7 m tall and up to 2 m in diameter,
consisting of several main branches bearing numerous
slender axillary branchlets covered with slightly glau-
cous, acicular-terete leaves.
After persisting for several years in this juvenile non-
reproductive phase, a number of stout, erect, unbranched,
abruptly emergent shoots are produced from the main
basal branches, bearing a graduated series of acicular to
linear-oblong to oblanceolate leaves. These 1 m long
reproductive shoots consist of up to three annual growth
increments, some increments elongating by over 500 mm
in one year. This is in marked contrast to the very short
annual growth increments on the basal sterile branchlets.
Male or female cones surrounded by broadly obovate
leaves terminate these heavy, woody stems. Thereafter
annual growth slows dramatically but is continued by a
number of very short (80-100 mm long) axillary branch-
es arising below the terminal cones.
Flowering occurs in early summer, peaking between
the first and third weeks of November. Large numbers of
small, unidentified Diptera and Coleoptera as well as
large scarab beetles ( Trichostetha sp.) were observed on
both male and female inflorescences, apparently effect-
ing cross-pollination.
Conservation status: unless additional populations are
discovered, L. immoderatum must rank among the rarest
species in the genus. With under 20 adult plants and fewer
than 50 seedlings seen at the time of making the type col-
lection, it is clearly a naturally rare, very local endemic, but
currently not threatended by any obvious human activities
due to its isolated, somewhat inaccessible montane habitat.
As a slow maturing species, too frequent fires seem to be
the only immediate threat to its survival.
Other specimen examined
WESTERN CAPE. — 3319 (Worcester): northwest side of Olifants-
berg, ENE of saddle between Doring and Witte Rivers, (-CD), 22-3-2001 .
A.G. Rebelo Y1032106 (NBG, PRE).
ACKNOWLEDGEMENTS
I am most grateful to Tony Rebelo for showing me the
only known population of this species in the field and to
Colin Paterson-Jones who accompanied us and took the
photographs in Figure 4. Inge Oliver prepared the excel-
lent line drawings of male and female inflorescences.
REFERENCE
WILLIAMS, I.J.M. 1972. A revision of the genus Leucadendron
(Proteaceae). Contributions from the Bolus Herbarium No. 3.
I P. ROURKE*
* Formerly: Compton Herbarium, South African National Biodiversity
Institute, Private Bag X7, 7735 Claremont, Cape Town.
AMARYLLIDACEAE
A NEW VARIETY IN THE GENUS CLIVIA
The current classification for Clivia Lindl., a genus
endemic to southern Africa, recognizes five species of
which only one is infraspeciftcally divided into varieties
(Duncan 1999; Rourke 2002; Snijman & Archer 2003;
Swanevelder 2003). A still undescribed sixth species
from the Pondoland Centre of Endemism (Van Wyk &
Smith 2001) is currently under investigation (Swane-
velder 2003). Clivia miniata (Lindl.) Regel var. citrina
Watson was discovered in the latter part of the 19th cen-
tury and, on the basis of its yellow flowers, described by
W. Watson as a variety in 1899 (Watson 1899; Phillips
1931; Duncan 1985, 1992). Strictly speaking the rank of
68
Bothalia 35,1 (2005)
FIGURE 5. — Distribution of C. gardenii var. gardenii , •; and C.
gardenii var. citrina as well as the typical variety, ■.
forma should have been used for the sporadic occurrence
of the single colour mutation to which this variety name
applies (Stuessy 1990). Here we formally describe a dis-
tinct yellow-flowered form of Clivia gardenii Hook. Fol-
lowing the precedent set in Clivia miniata, we have decid-
ed to recognize the new infraspecific taxon at variety
level. For a complete description of C. gardenii, see
Hooker (1856).
Clivia gardenii Hook. var. citrina Z.H.Swanevel-
der, A.E.van Wyk & J.T.Trnter, var. nov., floribus pallide
luteis vel citrinis, non aurantiacis vel rubris ut in varie-
tate typico distinguitur.
TYPE. — KwaZulu-Natal, 2731 (Louwsburg): Ngome
Forest, (-CD), Swanevelder & Truter ZH10 (PRU, holo.).
Flowers pale yellow or lemon yellow, not orange or
red as in the typical variety.
The holotype of C. gardenii var. citrina was collected
in Ngome Forest (Ngotshe District, KwaZulu-Natal) on
22 June 2002. Visits to the forest confirmed previous
reports of a strong population of lemon- or pale yellow-
flowered C. gardenii, with the occasional pastel or orange-
flowered individual, at this locality. This stands in contrast
to the single yellow-flowered specimen of Clivia miniata
that was available when the variety citrina was described.
We therefore consider the establishment of a new yellow-
flowered variety in Clivia gardenii as fully justified.
In C. gardenii var. gardenii the flowers are various
shades of orange and red (Obermeyer 1972). We treat all
plants with flowers in shades of these two colours (at the
time of anthesis), as belonging to this variety. Clivia gar-
denii, as defined by Swanevelder (2003), is confined to
two disjunct areas in KwaZulu-Natal (Figure 5). Plants
from southern KwaZulu-Natal and adjacent parts of
Eastern Cape traditionally identified as C. gardenii are
now regarded as a new taxon and have been excluded
from the distribution of C. gardenii (Swanevelder 2003).
The main distribution range of C. gardenii extends from
Durban northwards to Empangeni and inland as far as
Kranskop, Greytown and Howick. A second outlier dis-
tribution area is located in the Ngome Forest between
Vryheid and Nongoma, and slightly further north. No
linking records were found between these two distribu-
tion areas.
Ngome Forest is part of the Ntendeka Wilderness Area
and the yellow-flowered plants are all confined to this
protected area. The population is healthy with a high per-
centage of plants producing flowers and seed. Hopefully
the relative inaccessibility of the plants would ensure the
survival of the population as it provides some protection
against illegal plant collecting. The conservation status of
the new variety on Table Mountain near Pietermaritzburg
is unknown as no additional herbarium specimens of this
taxon have been collected from there since 1949.
Additional material examined
KWAZULU-NATAL. — 2731 (Louwsburg): Ngome Forest, south-
facing gully about 100 m east of campsite, (-CD), June 1999, Rourke
2157 (NBG). 2930 (Pietermaritzburg): Table Mountain, Amatulu
Forest, (-DA), 28-05-1949, Killick 466 (PRE, NH).
ACKNOWLEDGEMENTS
Our thanks to Dr Hugh Glen for the Latin translation
of the diagnosis and to Ezemvelo KwaZulu-Natal Wildlife
for the necessary permit to collect the type specimen. The
South African National Biodiversity Institute is thanked
for the use of data from the National Herbarium, Pretoria
(PRE) Computerized Information System (PRECIS).
REFERENCES
DUNCAN, G. 1985. Notes of the genus Clivia Lindley with particular
reference to C. miniata Regel var. citrina Watson. Veld & Flora
71: 84, 85.
DUNCAN, G. 1992. Notes of the genus Clivia Lindley with particular refer-
ence to C. miniata Regel var. citrina Watson. Herbertia 48: 26-29.
DUNCAN, G. 1999. Grow clivias. In D. Snijman & C. Voget, Kirsten-
bosch Gardening Series. National Botanical Institute, Cape Town.
HOOKER, W.J. 1856. Clivia gardenii. Curtis's Botanical Magazine.
ser. 3, 12: t. 4895.
OBERMEYER, A. A. 1972. Clivia gardenii. The Flowering Plants of Africa
42: t. 1641.
PHILLIPS, E.P. 1931. Clivia miniata var. flava. The Flowering Plants
of South Africa 11: t. 41 1 .
ROURKE, J.P. 2002. Clivia mirabilis (Amaryllidaceae: Haemantheae)
a new species from Northern Cape, South Africa. Bothalia 32: 1-7.
SNIJMAN, D.A. & ARCHER, R.H. 2003. Clivia. In G. Germishuizen
& N.L. Meyer, Plants of southern Africa: an annotated checklist.
Strelitzia 14: 958, 959. National Botanical Institute, Pretoria.
STUESSY, T.F. 1990. Plant taxonomy. The systematic evaluation of
comparative data. Columbia University Press, New York.
SWANEVELDER, Z.H. 2003. Diversity and population structure of
Clivia miniata Lindl. (Amaryllidaceae): evidence from molecular
genetics and ecology. M.Sc. thesis. University of Pretoria, Pretoria.
VAN WYK, A.E. & SMITH, G.F. 2001 . Regions of floristic endemism in
southern Africa: a review with emphasis on succulents. Umdaus
Press, Pretoria.
WATSON. W. 1899. Clivia miniata var. citrina. The Garden 56: 388,
t. 1246.
Z.H. SWANEVELDER*, A.E. VAN WYK** and J.T. TRUTER***
* Department of Botany/Forestry and Agricultural Biotechnology Insti-
tute (FABI), University of Pretoria, 0002 Pretoria.
** H.G.W.J. Schweickerdt Herbarium, Department of Botany, University
of Pretoria, 0002 Pretoria.
*** P.O. Box 5085, 1502 Benoni South, South Africa.
MS. received: 2004-03-06.
Bothalia 35,1 (2005)
69
SCROPHULARIACEAE
NEMESIA ZIMBABWENSIS. A NEW RECORD FOR THE FSA REGION WITH NOTES ON ITS PHYTOGEOGRAPHICAL SIGNIFICANCE
Since its description by Rendle (1932), and subse-
quent treatment in the Flora zambesiaca (Philcox 1990),
Nemesia zimbabwensis Rendle (= N. montana Norl.) was
known only from the Masvingo District and Eastern
Highlands of Zimbabwe and the adjacent high-altitude
area of Gorongosa in Mozambique. The oldest known
herbarium collection of the species dates from 1929
when Rendle collected the type specimen at Great Zim-
babwe, Masvingo District.
Nemesia zimbabwensis Rendle in Journal of Botany
70: 95 (1932). Type: Zimbabwe, Great Zimbabwe, Rendle
329 n.v. (BM, holo.).
Nemesia montana Norl.: 100 (1951). Type: Zimbabwe, Mt In-
yangani. Fries, Norlindh & Weimarck 3586 (LD. holo.; K, PRE!, iso.).
Philcox (1990) considered Nemesia montana Norl.
(Norlindh & Weimark 1951) as conspecific with N. zim-
babwensis. Originally N. montana was characterized as a
perennial with sessile to shortly petiolate (up to 3 mm
long) leaves, and N. zimbabwensis as an annual with
petioles 3-7 mm long. Philcox (1990) pointed out that
both sessile and petiolate leaves may occur on the same
specimen, hence his recognition of only a single species,
described by him as either annual or biennial. Within its
range the species is easily distinguished from other mem-
bers of the genus by its broadly ovate leaves. Herbarium
specimens of plants from shady or marshy habitats have
weakly branched herbaceous stems with relatively long
intemodes and tend to have sessile leaves. This creates
the impression that they are annuals. Plants from more
exposed habitats, for example rock crevices or regularly
burned grassland along forest edges, are often more
branched and tend to have distinctly stalked leaves; they
clearly are short-lived perennials.
Recent collecting in the province of Limpopo, South
Africa, has confirmed the presence of N. zimbabwensis in
the Leolo Mountains, Sekhukhuneland, where it was re-
corded on 4 December 1999 (Van Wyk & Siebert 13454 in
PRU). A study of the Nemesia material in the National
Herbarium (PRE), Pretoria, has subsequently revealed three
earlier collections of the same species from Afromontane
vegetation along the Great Escarpment to the north of the
Leolo Mountains, near Haenertsburg, also in Limpopo. The
last-named three collections have sessile leaves and were
provisionally determined as showing affinity to N. montana
by Prof. O.M. Hilliard in 1985. These records confirm the
presence of N. zimbabwensis in the FSA region and show a
clear disjunction of ± 400 km from the nearest existing
record at Great Zimbabwe (Eigure 6).
On the Leolo Mountains, N. zimbabwensis is current-
ly only known from a single patch of relict Afromontane
Forest associated with rocky outcrops at an altitude of
1 800 m asl (Siebert et al. 2003). The mean annual rain-
fall is ± 900 mm and mist occurs frequently. Here it
occurs as a lithophyte, with plants sparsely distributed
along moist, rocky ledges in the forest where they grow
in pockets of sandy humus. At this locality the leaves are
FIGURE 6. — Known distribution of Nemesia zimbabwensis.
mainly distinctly petiolate. Plants tend to branch basally
and are clearly short-lived perennials. The flowers are
white, often tinged with lilac, and the nectar guides are
dark purple. In all the Limpopo collections of this taxon,
there is a complete lack of long glandular trichomes on the
calyx lobes (Figure 7). These trichomes are characteristi-
cally present in plants from Zimbabwe and Mozambique.
The Leolo Mountains fall within the core area of the
Sekhukhuneland Centre of Plant Endemism [SC]
(Siebert et al. 2003). Intriguing disjunct distributions
between the SC and parts of Zimbabwe have also been
recorded in other species, for example Melhania randii
(Verdoom 1981), Plectranthus venteri (Edwards et al.
2000), P. dolomiticus (Edwards et al. 2001) and Ra-
phionacme chimanimaniana (Victor 2002). In addition,
the SC also shows disjunct plant distributions with other
regions of South Africa, for instance with Eastern Cape
(forms of Asparagus intricatus and Schotia latifolia ),
Griqualand West in Northern Cape ( Gnidia polycephala
and Nuxia gracilis ), the Limpopo River Valley further
north in Limpopo ( Decorsea schlechteri and Sesa-
mothamnus lugardii) and North-West (Amphiglossa tri-
flora and Rhigozum obovatum). The SC is therefore not
only of special biogeographical significance for its high
levels of local endemism, but also for the observation
that it seems to have served as a refuge/repository for
relict plant taxa perhaps dating from several different
episodes of climatically-induced vegetational shifts over
a long period of time.
Owing to the restricted distributions of the Sekhu-
khuneland-Zimbabwe disjuncts, the threat status of these
taxa has been assessed, namely Melhania randii (Ster-
culiaceae), Plectranthus venteri (Lamiaceae) and Ra-
phionacme chimanimaniana (Apocynaceae) are all con-
sidered as Vulnerable (Siebert et al. 2002; Victor 2002).
Hilton-Taylor (1996) considered Plectranthus dolomiticus
(Lamiaceae) as Insufficiently Known. Siebert et al. (2002)
provisionally assessed N. zimbabwensis as Critically En-
dangered in South Africa owing to the ongoing destruction
70
Bothalia 35,1 (2005)
FIGURE 7. — A, Nemesia zimbabwen-
sis. Van Wyk & Siebert 13454
in PRU. Specimen collected
on the Leolo Mountains,
Sekhukuneland; B, note com-
plete lack of long glandular
trichomes on calyx lobes.
Scale bars: A, 20 mm; B, 5
mm.
by humans of the particular forest patches where it occurs.
For instance, a new and more accessible road has been
built to the summit of the Leolo Mountains and wood har-
vesters are now felling the last remaining large trees of
Primus africana and Kiggelaria africana. In due course
the forest microclimate will change and N. zimbcibwensis
will probably disappear. The Woodbush and Haenertsburg
collections date from 1913, 1935 and 1954 respectively.
Since then, agriculture and the large-scale establishment
of alien tree plantations have destroyed much of the natural
grassland and associated wetland vegetation in these two
areas. The lack of any recent collections from this area
suggests that the species is either very rare, or extirpated
due to habitat destruction.
There is an urgent need to assess the conservation sta-
tus, not only of rare plant species, but also that of the rare
plant communities that harbour disjunct satellite popula-
tions of these taxa. In addition to their considerable bio-
geographical significance, outlier populations may also
represent distinct ecotypes worthy not only of protection,
but perhaps even formal taxonomic recognition at the
infraspecific level. In this regard it is noteworthy that the
Limpopo collections of N. zimbcibwensis differ from
their northern counterparts in the complete lack of long
glandular trichomes (Figure 7B). Plants from Zimbabwe
and Mozambique are sparsely glandular-pubescent, the
long trichomes being most consistently present on the
calyx, the lobes of which are ciliate. Although the
Limpopo plants appear to be glabrous, minute, almost
sessile, glandular trichomes with globose heads are pre-
sent, especially on the lower surface of young leaves.
These structures are visible only under very high magni-
fication (40x) with a dissection microscope and although
Bothalia 35,1 (2005)
71
not previously reported, seem to be present in material
from the whole range of the species.
Specimens examined
LIMPOPO. — 2329 (Pietersburg): Woodbush, Mountain Home Farm,
(-DD). Mogg 14684 (PRE); 1 mile north of Haenertsburg, (-DD), Codd
8418 (PRE); Haenertsburg. (-DD). Pott 4813 (PRE). 2430 (Pilgrim's
Rest): summit of Leolo Mountains, (-CA), Van Wvk & Siebert 13454
(PRE, PRU).
ZIMBABWE. — 1832 (Mutare): Inyanga [Nyanga] Dist.. Gairesi
Ranch on P.E.A. border 6 miles north of Troutbeck, (-BB), Robinson
1979 (PRE, SRGH); Inyanga [Nyamga] Dist., Mount Inyangani
[Nyangani], (-BD), Goodier & Phipps 68 (PRE. SRGH), Fries,
Norlindh & Weimarck 3586 (PRE), Sturgeon s.n. GHS16956 (K.
SRGH), Watmough 740 (PRE); Inyanga [Nyanga], (-BD). Norlindh &
Weimarck4557 (PRE). 1932 (Melsetter): Vumba [Bvumba] Mountains,
(-BB), Ferrar 3952, 3953 (PRE), Obermeyer 2147 (PRE). 2030
(Masvingo); Masvingo Dist., Great Zimbabwe, (-BD), Balsinhas &
Kersberg 2179 (LMA, PRE), Wild 3036 (K, SRGH).
MOZAMBIQUE. — 1834 (Vila Paiva de Andrada): Gorongosa, Mount
Gogogo, (-AC), Schelpe 444 (BM).
ACKNOWLEDGEMENTS
We thank Ms Shirley Smithies, National Herbarium,
Pretoria, for valuable assistance and the Herbarium,
Royal Botanic Gardens, Kew, for confirming the identi-
fication of N, zimbabwensis. We appreciate the help of
Ms Sandra Turck and Ms Hester Steyn, National Bo-
tanical Institute, for scanning the specimen and produc-
ing the distribution map. The University of Pretoria,
Mellon Foundation and National Research Foundation
provided financial support.
REFERENCES
EDWARDS, T.J., BEAUMONT, A.J. & STYLES, D. 2001. New
records and distributional disjunctions from South Africa,
Zimbabwe and Mozambique. Bothalia 31: 199-202.
EDWARDS, T.J., PATON, A. & CROUCH. N.R. 2000. A new species
of Plectranthus (Lamiaceae) from Zimbabwe. Kew Bulletin 55:
459^(64.
HILTON-TAYLOR, C. 1996. Red Data List of southern African plants.
Strelitzia 4: 1-117. National Botanical Institute. Pretoria.
NORLINDH, T. & WEIMARK, H. 1951. Beitrage zur kenntnis der
Flora von Sud-Rhodesia IX. Botaniska Notiser 2: 97-102.
PHILCOX, D. 1990. 120. Scrophulariaceae. In E. Launert & G.V. Pope,
Flora zambesiaca 8,2: 9-14. Flora Zambesiaca Managing Com-
mittee, London.
RENDLE, A.B. 1932. African notes II. Journal of Botany 70: 89-96.
SIEBERT, S.J., VAN WYK, A.E., BREDENKAMP, G.J. & SIEBERT,
F. 2003. Vegetation of the rock habitats of the Sekhukhuneland
Centre of Plant Endemism, South Africa. Bothalia 33: 207-228.
SIEBERT, S.J., VICTOR, J.E., VAN WYK, A.E. & BREDENKAMP,
G. J. 2002. An assessment of threatened plants and conservation
in Sekhukhuneland. PlantLife 26: 7-18.
VERDOORN. I.C. 1981. Revision of Melhania in southern Africa.
Bothalia 13: 263-273.
VICTOR. J. 2002. South Africa. In J.S. Golding, Southern African
plant Red Data Lists. Southern African Botanical Diversity
Network Report No. 14: 93-120. SABONET, Pretoria.
S.J. SIEBERT* and A.E. VAN WYK**
* University of Zululand Herbarium, Department of Botany, University
of Zululand, 3886 KwaDlangezwa.
** H.G.W.J. Schweickerdt Herbarium, Department of Botany, Univer-
sity of Pretoria, 0002 Pretoria.
MS. received: 2002-10.
IRIDACEAE
TAXONOMIC NOTES ON BABIANA AND FERRAR1A IN ARID WESTERN SOUTHERN AFRICA
In preparation for an account of the Iridaceae for
Namaqualand and the winter rainfall Karoo, we have
found that several nomenclatural and taxonomic adjust-
ments are necessary. These concern the genera Babiana
Ker Gawl. and Fen-aria Burm. ex Mill. Both genera are
currently being actively studied but our conclusions will
not be published in the immediate future.
Taxonomic notes on Babiana
Babiana spiralis Baker, Handbook of the Irideae:
111 (1892). Type: South Africa, Northern Cape, Namaqua-
land, without precise locality or date, collector uncertain,
possibly J. Niven s.n. in Herb. Forsyth (K, syn.!: 2 sheets).
Babiana fimbriata was first collected by J.F. Drege in
1830 in Namaqualand between the Swartdoom and Groen
Rivers, between Bitterfontein and present day Garies. It
was assigned to the genus Antholyza by F.W. Klatt in
1867, 1868 and was transferred to Babiana by J.G. Baker
in 1877. Then, in her monograph of Babiana, G.J. Lewis
(1959) included as a synonym of B. fimbriata, a second
species, B. spiralis, which had been described by Baker in
1892. In Flora capensis. Baker (1896) recognized both B.
fimbriata and B. spiralis. Our field work in Namaqualand
over the past 10 years shows that Baker was correct in rec-
ognizing both species. They are readily distinguished from
each other by vegetative and floral features. Normally a
fairly tall plant, standing 250-550 mm high, B. spiralis has
a velvety hairy stem and smooth cataphylls and leaf
sheaths. The leaf blades are 1.5-2. 5 mm wide and loosely
twisted above with a distinct pseudopetiole up to 40 mm
long. The flowers are coloured bright pink or pale blue
with pale yellow nectar guides on the lower tepals, which
have auriculate lobes at the base of the limbs. The upper
lateral tepals curve outward shortly above the base and are
thereafter subpatent.
In contrast, Babiana fimbriata is 100-200 mm tall,
with a hairless or nearly hairless stem and cobwebby to
woolly cataphylls and leaf sheaths. The leaf blades are
3^4 mm wide. The flowers are dull purple and cream-
coloured and the lower lateral tepals also have prominent
auriculate lobes at the base of the limbs. The upper later-
al tepals are directed forward or slightly curved inward in
the proximal half, giving the flower a distinctly closed
appearance in comparison with that of B. spiralis.
Favouring deep sandy soils, Babiana spiralis is the
more widespread of the two species, extending from
Garies to Klawer in the south and along the coast
72
Bothalia 35,1 (2005)
between Groen River mouth and Kotzesrus in the west.
B. fimbriata , in contrast, is a fairly narrow endemic of
central Namaqualand, with a range limited to the area
between Garies and Nuwerus, and seems to prefer stony
ground rather than deep sands.
The only collection of Babiana fimbriata correctly
assigned to the species by Lewis (1959) is the type gath-
ering. Other specimens that she cited are B. spiralis.
Since the publication of her monograph, a handful of
additional collections have been made of plants corre-
sponding to the original Drege collection and they show
beyond question that B. fimbriata is distinct from B. spi-
ralis. The origin of the type collection of Babiana spi-
ralis has long been puzzling. The name Forsyth on the
type sheets at the Kew Herbarium is now thought to refer
to William Forsyth, whose herbarium collection was
bought in 1835 after his death by George Bentham of
Kew Gardens (C.E. Nelson pers. comm. 2003). We sus-
pect that the specimens were actually collected by James
Niven, who botanized in Namaqualand in 1799 and pos-
sibly later, gathering seeds and perhaps bulbs, for patrons
in Great Britain and France (Gunn & Codd 1981).
Material examined
Babiana fimbriata
NORTHERN CAPE.— 3017 (Hondeklipbaai): 550 m ESE of Waterklip
from Grootbrakfontein, near Garies, (-DB), August 1979, Van Berkel 129
(NBG).
WESTERN CAPE. — 3018 (Kamiesberg): between Bitterfontein and
Garies at Swartdoring River, (-CC), 2 September 2000, Goldblatt & Nanni
11452 (MO, NBG, PRE); 9 km N of Bitterfontein, (-CC), 20 August 2001,
Goldblatt & Porter 11711 (MO, NBG). 3118 (Vanrhynsdorp): Meer-
hofkasteel Farm, 15 km W of Nuwerus, (-AA), 28 August 1986, Hilton-
Taylor 1203 (NBG); 3 miles NW of Nuwerus, (-AB), 1 1 September 1971,
Hall 4141 (NBG); between Nuwerus and Lutzville, (-AB), 17 July 1964,
Lewis s.n. (NBG); rocky bank between Bitterfontein and Nuwerus, (-AB),
13 August 1997, Goldblatt & Manning 10666 (MO); Knersvlakte at
Grootgraafwater turnoff, stony east-facing slope, (-BC), 20 August 2001
(fr.), Goldblatt & Manning 11710 (MO).
Babiana spiralis
NORTHERN CAPE.— 3017 (Hondeklipbaai): lat. 30°32'67", long.
17o56'80\ (-DB), 26 April 1981, Van Berkel 378 (NBG); hills above
Garies, 4 August 1980, Greig & De Villiers 2 (PRE); 18 km from the
junction of the Garies-Groenrivier road toward Nuwefontein, (-DC), 3
September 1976, Boucher 3159 (K, NBG, PRE); Farm Waterval west
of Kotzesrus, (-DC), 28 August 2001, Goldblatt & Porter 11773 (MO,
NBG); sandy flats 1.8 km east of Kotzesrus, (-DC), Goldblatt & Porter
12080 (MO, NBG); top of Garies hill, ± 0.5 km along road to
Hondeklipbaai, (-DB), 2 September 2000, Goldblatt & Nanni 11453
(K, MO, NBG, PRE); 18 km towards Nuwefontein from junction with
Garies-Groenrivier road, (-DC), 3 September 1976, Boucher 3159 (K,
NBG, PRE); Farm Waterval, 14 km W of Kotzesrus, (-DC), 28 August
2001 (fr.), Goldblatt & Porter 11772 (MO, NBG).
WESTERN CAPE. — 3118 (Vanrhynsdorp): hills at Bitterfontein,
(-AA), 2 September 1897, Schlechter 11041 (K, MO, S); SW of
Bitterfontein, (-AB), 5 September 1955, Hall 1003 (NBG); Klawer,
(-DC), July 1848, Stokoe s.n. (SAM59866).
Locality uncertain: I km SW of Biesiesfontein, September 1974,
Nordenstam & Lundgren 1770 (MO, S).
Babiana planifolia (G.J. Lewis) Goldblatt & J.C. Man-
ning, comb, et stat. nov.
Babiana striata var. planifolia G.J. Lewis in Journal of South
African Botany, Suppl. 3: 130 (1959). Type: South Africa, Northern
Cape, 9 miles (14.4 km) west of Steinkopf, 30 June 1935, T.M. Salter
5554 (BOL, holo.!).
Although only two collections were known to Lewis
(1959) when she described Babiana striata var. planifo-
lia, several additional gatherings are now available and
make it clear that this taxon is best treated as a separate
species. Typical B. striata has a strongly inclined stem
and nearly horizontal spike, conspicuously undulate and
crisped leaves, and corm tunics of coarse fibres. What we
now recognize as B. planifolia has an erect stem and only
slightly inclined spike, almost plane leaves that are only
occasionally weakly undulate toward the base and loose-
ly twisted distally, and corm tunics of relatively fine
fibres. Plants of both taxa have cobwebby to woolly cat-
aphylls and lower leaf sheaths, and the stem and bracts
are usually hairless. The lower lateral tepals of B. plani-
folia also appear to lack the exaggerated auricular lobes at
the base of the limbs typical of most other species of sec-
tion Exohebeoides, including typical B. striata.
Plants from the Eksteenfontein area of the southern
Richtersveld associated with Babiana planifolia (as B.
striata var. planifolia) by Lewis have short, broad, weak-
ly undulate leaves 35-50 x 15-20 mm, twisted in the
upper half. These plants seem to differ consistently in
their short stature and additionally have long, soft,
almost silky hairs on the leaf margins and upper sheaths,
especially conspicuous in Goldblatt & Manning 9898
(MO, NBG). Although corms, cataphylls, and the lower
leaf sheaths of the Richtersveld plants are not known, we
provisionally include this form in B. planifolia. Addi-
tional material may show that it is a separate taxon.
Both Babiana striata and B. planifolia flower in May
and June, occasionally in July, and remain poorly repre-
sented in herbaria, although neither is rare in the wild.
Babiana planifolia grows on rocky slopes, mainly on
fine-grained soils, extending from the southern Richters-
veld to Garies and Soebatsfontein and locally in the
Knersvlakte.
Other material examined
NORTHERN CAPE.— 2817 (Vioolsdrif): 3 km N of Eksteenfon-
tein, (-CD), 3 August 1994, Goldblatt & Manning 9898 (MO, NBG);
near Eksteenfontein, (-CD), 27 March 1979, Van Berkel 100 (NBG).
2917 (Springbok): 23 miles W of Steinkopf, (-AB), June 1929,
Marloth 13254 (PRE); 26 km west of Springbok, Spektakel Pass, in
granitic ground, (-DA), 22 July 1976, Goldblatt 3658 (MO);
Ezelfontein, 14 miles west of Springbok, (-DA), 20 June 1965, Hall
3057 (NBG); near Paddagat, 22 miles NW of Springbok, (-DB), 28
May 1961, Leistner 2552 (NBG, PRE); 14 km S of Springbok, (-DD),
7 August 2000 (fr.), Goldblatt & Manning 11327 (MO, NBG). 3017
(Hondeklipbaai): 10 km N of Soebatsfontein, (-BA), 21 May 1986,
Duncan 114 (NBG); 14 miles NW of Kamieskroon, (-BB), 19 July
1957, Acocks 19323 (PRE); road to Hondeklipbaai near turnoff N of
Garies, (-DB), May-June. Loubser 2160 (NBG); 19 km S of Kotzesrus
on road to Landplaas, (-DD), 16 September 2001 (fr.), Goldblatt &
Porter 11898 (MO, NBG).
WESTERN CAPE. — 3118 (Vanrhynsdorp): Knersvlakte at Groot-
graafwater turnoff, stony east-facing slope, (-BC), 20 August 2001
(fr.), Goldblatt & Manning 11710A (MO).
Taxonomy of the Ferraria divaricata complex
When M.R de Vos revised the southern African genus
Ferraria in 1979, she treated F. divaricata as a single
Bothalia 35,1 (2005)
73
widespread species with four subspecies. The entire
range of the species extended from southwestern
Namibia to the southern Cape, and locally eastward into
the Karoo. Two subspecies, subsp. arenosa and subsp.
aurea were distinguished by a well-developed aerial
stem, usually extensively branched in the upper half and
flowers lasting a single day, whereas subsp. divaricata
and subsp. variabilis have short stems, branching to
some extent, but usually close to ground level and the
flowers last two, exceptionally three days. De Vos (1979)
also noted differences in the seeds between the two pairs
of subspecies, subsp. arenosa and subsp. aurea having
globose, matte brown seeds with reticulate sculpturing
and foveate epidermal cells, whereas subsp. divaricata
and subsp. variabilis have seeds typical of the rest of the
genus, being irregularly angled by pressure and pale
straw-brown in colour with a coat that is slightly wrin-
kled to ruminate.
Uniting the four subspecies is a similarity in floral mor-
phology, viz. the tepals have broad claws that together form
a wide floral cup, while the fringed tepal limbs extend hori-
zontally. The floral cup, 12-15 mm deep and 13-15 mm
wide at the rim, holds a pool of nectar secreted from small
nectaries a short distance above the tepal bases. Ferraria
divaricata belongs in section Macroscyphae of the genus,
defined by a beaked ovary and capsule and anthers with
divaricate lobes. Other species of the section, including the
Namaqualand F. macrochlamys (Baker) Goldblatt &
J.C. Manning (Goldblatt & Manning 2004) and F. katnies-
bergensis M.P.de Vos, and the Western Cape F. uncinata
Sweet, have relatively narrow tepal claws that form a nar-
row cup and smaller flowers.
After collecting plants of all four subspecies in the
field, we have become dissatisfied with this taxonomy.
The tall-stemmed subsp. arenosa and subsp. aurea , which
together have a distribution along the Cape west coast
from Hondeklipbaai in the north to the Cape Flats in the
south, have seeds unique in Ferraria , as well as a tall,
branched habit, and flowers lasting a single day and dif-
fering in coloration from those of subsp. divaricata and
subsp. australis , which last two or three days. A taxono-
my that better reflects the biological situation is to treat
subsp. arenososa and subsp. aurea as one species, and
subsp. australis and subsp. divaricata as another.
A second consideration relating to Ferraria divarica-
ta is the typification of the species. The type is an illus-
tration in The British flower garden (Sweet 1827), which
depicts a branch of a plant described as being 18 inches
(0.45 m) high, with chocolate-brown flowers with a pale,
darkly streaked floral cup. This matches exactly, popula-
tions of the plant currently called F. divaricata subsp.
arenosa from the Western Cape coast near Leipoldtville
and Sandberg (see Manning et al. 2002: 158, as F.
foliosa). It follows that subsp. arenosa and subsp. aurea ,
when raised to species rank, must be called F. divarica-
ta. This means that subsp. australis and subsp. divarica-
ta need a new name. There are no legitimate synonyms
available at species rank. The only synonym, F.
antherosa Ker Gawk, is a superfluous name for F. viridi-
flora Andr. which is itself a synonym of F. ferrariola
(Jacq.) Willd. We chose the name F. variabilis for the
species in view of the variable perianth coloration. This
ranges from red-brown tepal limbs with dull purple at the
base and pale claws (subsp. divaricata) to variously
greenish, dull yellow, or light brown with speckled mark-
ings on the limb bases (subsp. australis). Basic chromo-
some number is x = 10 in A divaricata and F. variabilis.
All counts are diploid, 2 n = 20, except in populations
referred by De Vos (1979) to F. divaricata subsp. aus-
tralis., which are tetraploid, 2 n - 40. We see no need to
distinguish subspecies within F. divaricata and F. vari-
abilis, based largely on tepal colour and patterning (and
chromosome number in the case of subsp. australis ), and
thus recognize just two species in place of F. divaricata
and De Vos’s four subspecies.
Ferraria divaricata Sweet , The British flower
garden 1: t. 192 (1827); M.P.de Vos : 354 (1979). Type:
illustration in Sweet, The British flower garden 1: t. 192
(1827), from South Africa, without precise locality, col-
lected by W. Synnot.
F. divaricata subsp. arenosa M.P.de Vos 45: 358 (1979). Type:
South Africa. [Western Cape], Clanwilliam, Nardouw Pass, Farm de
Lille, October 1973, Van Breda sub De Vos 2295 (STE, holo.!).
F. divaricata subsp. aurea M.P.de Vos 45: 359 (1979). Type: South
Africa, [Western Cape], 15 miles east of Lambert’s Bay, Langdam, 28
September 1973, De Vos 2297 (STE, holo.!).
Ferraria variabilis Goldblatt & J.C. Manning, sp.
nov.
F. antherosa Ker Gawl. : t. 751 (1804), nom. illeg. superfl. pro F.
viridiflora [as F. viridis] Andrews. Type: South Africa, without precise
locality or collector, illustration in Curtis’s Botanical Magazine 20: t.
751 (1804).
F. divaricata subsp. australis M.P.de Vos 45: 359 (1979). Type:
South Africa, [Western Cape], Clanwilliam. Langdam, 17 September
1945. R.H. Compton 17412 (NBG, holo.!).
Plantae 60-200 mm altae saepe caespitosae, prope
basem ramosae, foliis ensiformibus ad linearibus, rhipidiis
2-florum spatha interiore 40-80 mm longo exteriore usi-
tate pauce breviore, floribus usitate limbis pallide flavis
ad brunneis vel griseocaeruleis marginibus coloribus pal-
lidior vel atrobrunneis unguibus pallidis vel atrostriatis
cupulatis, tepalis exterioribus 27^f0(-50) x 10-15 mm,
interioribus 25— 40( — 45) x 8-10 mm, unguibus 10-15 mm
longis, filamentis connatis in columnam 8-13 mm longis
liberibus divergentibusque 3^1 mm, antheris 3. 5-5.0 mm
longis lobis late divergentibus, ovario rostrato, capsulis
ellipsoideis 30-50 mm longis.
TYPE. — Northern Cape, 3119 (Calvinia): Nieuwoudt-
ville, Klipkoppies, lower slopes in sand, (-AC), 15
September 1961, W.F. Barker 9537 (NBG, holo.!; MO,
iso.!).
Plants 60-200 mm tall, often branched just above
base, branches crowded and subequal in length, often
forming small tufts. Leaves sword-shaped to linear, usu-
ally about as long as stem, sometimes up to twice as long,
(2-)4— 10 mm wide, crowded basally, usually without vis-
ible midrib, often slightly striate, margins often weakly
thickened, rarely obscurely crisped, sheaths usually over-
lapping and concealing stem. Rhipidia 2-flowered; inner
spathes 40-80 mm long, outer slightly shorter to about as
74
Bothalia 35,1 (2005)
long, often arching outward in upper half. Flowers lasting
two or sometimes three days, predominantly pale to dull
yellowish to pale to middle brown, or dull grey-blue,
limbs solid dark brown to blackish purple at base, or with
scattered dark spots, margins darker or paler in colour,
claws uniformly pale or with dark longitudinal streaks or
with broad darker median streak, forming floral cup
12-15 mm deep, 13-15 mm wide at rim, usually slightly
putrid smelling, nectaries basal, pale or dark-coloured;
outer tepals 2740(-50) x 10-15 mm, claws 10-15 mm
long, inner tepals 25— 40(— 45 ) x 8-10 mm. Filaments
united in column 8-13 mm long, free and arching out-
ward in upper 24 mm; anthers 3. 5-5.0 mm long before
anthesis, shorter after dehiscence, lobes widely diverging.
Ovary with sterile beak; stigmas comprising small lobes
below tips of style arms, arching over anthers. Capsule
ellipsoid, 35-50 mm long, including beak; seeds rounded,
usually angled by pressure, coat dull and slightly wrin-
kled. Flowering time : August to November.
Distribution and ecology: sandy and shale flats and
granite outcrops, extending from southern Namibia to
Oudtshoom and into the Upper Karoo as far east as
Upington and Britstown, but absent from the western
coastal forelands of Western Cape.
ACKNOWLEGEMENTS
Support for this study by grants 6704-00, 7103-01 and
7316-02-02 from the National Geographic Society is
gratefully acknowledged. Collecting permits were pro-
vided by the Nature Conservation authorities of Northern
Cape and Western Cape, South Africa. We thank Ingrid
Nanni and Lendon Porter for their assistance and com-
panionship in the field and Roy Gereau for revising our
Latin descriptions.
REFERENCES
BAKER, J.G. 1 877. Systema Iridacearum. Journal of the Linnean Society
of London, Botany 16: 61-180.
BAKER, J.G. 1892. Handbook of the Irideae. Bell, London.
BAKER, J.G. 1896. Irideae. In W.T. Thiselton-Dyer, Flora capensis 6:
7-171. Reeve, London.
DE VOS, M.P. 1979. The African genus Ferraria. Journal of South
African Botany 45: 295-375.
GOLDBLATT, P. & MANNING, J.C. 2004. New species of Ixia
(Crocoideae) and Moraea (Iridoideae), and taxonomic notes on
some other African Iridaceae. Novon 14: 288-298.
GUNN, M. & CODD, L.E. 1981. Botanical exploration of southern
Africa. Balkema, Cape Town.
KER GAWLER, J. 1804. Ferraria antherosa. Curtis's Botanical Maga-
zine 20: t. 751.
KLATT, F.W. 1867-1868. Diagnoses Iridearum novarum. Linnaea 35:
380.
LEWIS, G.J. 1959. The genus Babiana. Journal of South African Botany,
Suppl. 3.
SWEET, R. 1827. Ferraria divaricata. Hie British flower garden 1: t. 192.
MANNING, J.C., GOLDBLATT, P. & SNIJMAN, D. 2002. The color
encyclopedia of Cape bulbs. Timber Press, Portland, Oregon.
P. GOLDBLATT* and J.C. MANNING**
* B.A. Krukoff Curator of African Botany, Missouri Botanical Garden,
P.O. Box 299, St. Louis, Missouri 63166, USA.
** Compton Herbarium, South African National Biodiversity Institute,
Private Bag X7, 7735 Cape Town.
MS. received: 2004-07-22.
LYCOPERDACEAE— GASTEROMYCETES
BOVISTA CAPENSIS, THE CORRECT NAME FOR BOVISTA PROMONTORU
In 1822 the botanical collector Carl Zeyher arrived at
the Cape of Good Hope with the purpose of collecting
mainly higher plants (Gunn & Codd 1981). Although he
initially had no interest in fungi, he amassed a fairly
extensive collection of macrofungi during his sojourn in
Uitenhage (near Port Elizabeth), Eastern Cape, during
the 1830s. Most of Zeyher’s fungal material was later
studied by the Rev. M.J. Berkeley in England, but some
specimens also found their way to E.M. Fries in Uppsala,
Sweden. It was from the latter material [Zeyher 106 sub
herb. E. Fries (UPS)] that Fries (1848) established
Lycoperdon capense, a new species accompanied by a
brief and rather ambiguous description. In a review of
South African Gasteromycetes, Bottomley (1948) was
unable to ascertain the identity of L. capense Fr. and
included it in her list of ‘Doubtful, unknown and insuffi-
ciently described species’.
In his classic monograph of the genus Lycoperdon
Pers.: Pers., Demoulin (1971) correctly excluded Lyco-
perdon capense from that genus and speculated that it
probably represented a species of Vascellum F.Smarda.
He overlooked the fact, however, that earlier, Kreisel
(1967: 106) had pointed out that Zeyher 106 at UPS is a
mixed collection, consisting of a Bovista Pers.: Pers. as
well as a Calvatia Fr. element. The two elements of
Zeyher 106 have subsequently been split and are now
filed in separate folders at UPS, with the Bovista and
Calvatia parts having been assigned the numbers Zeyher
106a and Zeyher 106b respectively. The numbers on
Zeyher specimens can be a source of confusion (Gunn &
Codd 1981). In some cases the number on the label may
be Zeyher’s collecting number but, most commonly, the
number refers to the collecting locality, with the second
part (if a double number) referring to the month of col-
lection. Therefore, specimens from the same locality
may belong to different gatherings, as seems to be the
case with Zeyher 106. Locality 106 is listed in Gunn &
Codd (1981) as ‘Uitenhage, Zuureberg, 2-3000’. [Note:
in addition to the specimens at UPS, Zeyher 106 consists
also of a third element, on which Berkeley (1843) based
the genus Scoleciocarpus (= Arachnion Schwein.) and
species Scoleciocarpus tener Berk. This third element,
originally from the herbarium of W.J. Hooker, has not
been examined by us but is in all probability still at K,
where it had been studied by Demoulin ( 1972).]
Despite the inadequacy of the original diagnosis of
Lycoperdon capense , reading it creates the distinct impres-
sion that it applies to the Bovista part of Zeyher 106 rather
Bothalia 35,1 (2005)
75
than the Calvatia part. Bottomley (1948) provided the fol-
lowing translation of Fries’s original diagnosis: ‘Peridium
has the habit of L. gemmatum but the structure rather of L.
pusillum. Exoperidium granular. Endoperidium membra-
nous, flaccid, opening by a small, obtuse mouth. Sterile
base wanting. Capillitium threads very lax. Spores brown,
not becoming olivaceous’. This description also corre-
sponds well with that provided by Kreisel (1967) for the
Bovista element (now Zeyher 106a): ‘Endoperidie papier-
stark, braunlich, glanzlos; Offnung unregelmassig eingeris-
sen, ca. 4 mm weit.’ and \ . . Sporenstaub umberbraun ohne
Olivton’. A recent examination of the two elements com-
prising Zeyher 106 at UPS convinced us that the impres-
sion created by Fries’s diagnosis is correct and that the
original description does indeed apply to the Bovista , and
not the Calvatia gathering. The Calvatia element of Zeyher
106 consists of three specimens, all of which dehisce irreg-
ularly in typical Calvatia style, not at all by means of
‘small mouths’ (= pore; ostiole) and with peridial charac-
ters irreconcilable with Fries’s description. Zeyher 106a ,
however, corresponds well with Fries’s original diagnosis,
both in terms of the roughness of the exoperidium and the
way the endoperidium dehisces. The name Lycoperdon
capense Fr. therefore clearly applies to the Bovista element
of Zeyher 106 and not to Zeyher 106b, and in view of that,
Zeyher 106a sub herb. E. Fries is here designated as lecto-
type of Lycoperdon capense Fr.
In 1967 the Zeyher 106a element at UPS served as the
basis for the description of a new species, Bovista promon-
torii Kreisel. In his publication, Kreisel (1967) does not
refer to the work of Fries (1848), which suggests that he
was not aware of Fries’s earlier name, L. capense. In the
light of the above discussion, the new combination Bovista
capensis is here proposed as the correct name for the
Bovista element of Zeyher 106 at UPS and Kreisel’s super-
fluous name is reduced to synonymy:
Bovista capensis (Fr.) J.C.Coetzee & A.E.van
Wyk, comb. nov.
Lycoperdon capense Fr. in J.A. Wahlbergii Fungi natalenses, adjec-
tis quibusdam capensibus: 30 (1848). Bovista promontorii Kreisel:
225-226 (1967). Type. South Africa. Pr. B. Sp. [= Promontorium
Bonae Spei; = Cape of Good Hope], Uitenhage, Zeyher 106a sub herb.
E. Fries (UPS!, lecto., here designated).
L. oblongisporum auct. non Berk. & Curt.: Lloyd: 235 (1905).
[Teste Kreisel (1967) & Ortega & Buendia (1985).]
L. polymorphum auct. non Vitt.: Bottomley: 557 (1948). [Teste
Kreisel (1967) & Ortega & Buendia (1985).]
Illustrations: Kreisel: figs 17c, 27b1 & 58 (1967); Ortega &
Buendia: figs 3, 4, 9, 10, 13, 14 (1985).
According to Kreisel (1967) and Ortega & Buendia
(1985) the ‘South African form’ of Lycoperdon polymor-
phum Vitt. [= Bovista polymorpha (Vitt.) Kreisel, the
correct name of which is Bovista aestivalis (Bonord.)
Demoulin] sensu Bottomley (1948) is conspecific with
Bovista capensis. Neither Kreisel nor Ortega & Buendia
have examined any of the specimens cited in Bottomley
(1948), however, and a re-examination of the South
African material is required to confirm their assertion. B.
capensis is separated from B. aestivalis only by the shape
of its spores (Kreisel 1967), and considering the opinions
of Calonge & Demoulin (1975) and Moyersoen &
Demoulin (1996), that in Europe, spore shape does not
allow the discrimination of taxa in the B. aestivalis com-
plex, the need for a re-assessment of B. capensis
becomes even more evident. The same applies to the
specimens referred to Lycoperdon oblongisporum Berk.
& Curt, by Lloyd (1905).
Kreisel (1994) cited Zeyher 106b under Calvatia
rugosa (Berk. & M.A.Curtis) D. A. Reid. However, the
complete absence of a subgleba and prominent slit-like
capillitial wall perforations clearly exclude it from that
taxon. We have little doubt that Zeyher 106b actually rep-
resents an undescribed, but not at all uncommon. South
African species belonging to Calvatia sect. Macrocalvatia
Kreisel sensu Coetzee & Van Wyk (2003). Followers of
Kreisel (1989) would, however, prefer to treat this fungus
in the segregate genus Handkea Kreisel.
ACKNOWLEDGEMENTS
We are indebted to the director of UPS for the loan
of the herbarium specimens in his care. The friendly
assistance of Dr S. Ryman of the same institution is
acknowledged with gratitude.
REFERENCES
BERKELEY, M.J. 1843. Enumeration of fungi, collected by Herr
Zeyher in Uitenhage. Hooker ’s Journal of Botany and Kew Garden
Miscellany 2: 507-524 (plus plate XXII).
BOTTOMLEY, A.M. 1948. Gasteromycetes of South Africa. Bothalia
4: 473-810.
CALONGE, F.D. & DEMOULIN, V. 1975. Les Gasteromycetes d'Espagne.
Bulletin trimestriel de la societe mycologique de France 9 1 :
247-292.
COETZEE, J.C. & VAN WYK, A.E. 2003. Calvatia sect. Macrocal-
vatia redefined and a new combination in the genus Calvatia.
Bothalia 33: 156-158.
DEMOULIN, V. 1971. Le genre Lycoperdon en Europe et en Amerique
du Nord. Etude taxonomique et phytogeographique. Doctoral
thesis, Universite' de Liege.
DEMOULIN, V. 1972. Observations sur le genre Arachnion Schw.
(Gasteromycetes). Nova Hedwigia 21 : 641-655.
FRIES, E.M. 1848. J.A. Wahlbergii Fungi natalenses, adjectis quibus-
dam capensibus. P.A. Norstedt et filii, Stockholm.
GUNN, M. & CODD, L.E. 1981. Botanical exploration of southern
Africa. Balkema, Cape Town.
KREISEL. H. 1967. Taxonomisch-Pflanzengeographische Monographic
der Gattung Bovista. Beihefte zur Nova Hedwigia 25: 1-244.
KREISEL, H. 1989. Studies in the Calvatia complex (Basidiomycetes).
Nova Hedwigia 48: 281-296.
KREISEL, H. 1994. Studies in the Calvatia complex (Basidiomycetes)
2. Feddes Repertorium 105: 369-376.
LLOYD, C.G. 1905. The Lycoperdons of the United States. Myco-
logical Writings 2: 221-238.
MOYERSOEN, B. & DEMOULIN, V. 1996. Les Gasteromycetes de
Corse: Taxonomie, Ecologie, Chorologie. Lejeunia n.s. 152:
1-128.
ORTEGA, A. & BUENDIA, A.G. 1985. Estudio de algunas especies
con esporas oblongas del genero Bovista Pers. Cryptogamie,
Mycologie 6: 281-288.
J.C. COETZEE* and A.E. VAN WYK**
* Department of Horticulture and Food Technology, Bell ville Campus,
Cape Peninsula University of Technology, P.O. Box 1906, 7535 Bellville.
**H.G.W.J. Schweickerdt Herbarium, Department of Botany, University
of Pretoria, 0001 Pretoria.
MS. received: 2004-06-07.
76
Bothalia 35,1 (2005)
LYCOPERDACEAE— GASTEROMYCETES
THE IDENTITY OF LYCOPERDON COMPLANATUM DESF. AND ITS NOMENCLATURAL IMPLICATION
INTRODUCTION
Lycoperdon complanatum Desf., nom. illeg. (non
Batsch 1786), was established by Desfontaines (1799) to
accommodate a fungus growing in arid soil from an
unspecified locality in Algeria, North Africa. The taxon
still remains known from the original collection only and
its taxonomic status has been uncertain ever since it was
first described. Desfontaines’s original diagnosis is rather
cryptic and ambiguous and of no use at all to establish
whether L. complanatum Desf. actually represented a good
species or not: ‘ Lycoperdon acaule, orbiculatum ; supeme
planum , leave; subtus lacunosum; margine acuto.
Orbiculatum, depression, sessile, planum, supeme leave,
subtus saepe lacunosum, irregulare; margine acuto, saepe
dentate-lacero. Diameter 1-2 decimeter’. Desfontaines
thus merely and, as would later be revealed, quite incor-
rectly, described it as being a stalkless, flat, round fungus,
100-200 mm diam., with an acute, lacerated margin, flat,
smooth upper surface and lacunose lower part. The illus-
tration accompanying the original description (tab. 261, not
161 as cited in the protologue; accessible also through the
Missouri Botanical Garden library’s rare books web site at
http://ridgwaydb.mobot.org/mobot/rarebooks/) depicts
dorsal (outer surface viewed from above) and lateral views
of the specimen, but those are also without any significant
diagnostic features.
In view of the illegitimacy of Lycoperdon complanatum
Desf., Rafinesque’s (1814) use of a new epithet was justi-
fied— whether intentionally or not — when he treated this
fungus as Omalycus erosus Raf. Durieu & Leveille, in
Durieu de Maisonneuve (1848), correctly concluded that L.
complanatum Desf. merely represented the sterile base of a
mature puffball of which the peridium had already disinte-
grated and the gleba was absent. However, as explained in
the note at the end of this paragraph, they erred in synony-
mising it in the protologue with their later (and, in that
sense, superfluous and also illegitimate) L. fontanesii
Durieu & Lev. According to Demoulin (1971), as well as
our own interpretation of the original material of L.
fontanesii at the cryptogamic herbarium of the Museum of
Natural History in Paris (PC), this latter fungus is the same
taxon as Calvatia utriformis (Bull.: Pers.) Jaap [or
Handkea utriformis (Bull.: Pers.) Kreisel, if one prefers to
accept the segregate genus Handkea Kreisel], In his classic
monograph of the genus Lycoperdon, Demoulin (1971) did
not provide any further clarity on the identity of L. com-
planatum Desf. and, since he was unaware of the existence
of the type material at P while revising Lycoperdon Pers.:
Pers. at PC (V. Demoulin pers. comm.), he only referred to
Desfontaines’s original ambiguous illustration. He never-
theless correctly concluded that Desfontaines’s fungus
could not have been a Lycoperdon , but reserved further
judgement regarding its true identity. [Explanatory note:
Lycoperdon fontanesii Durieu & Lev. is an illegitimate
(superfluous) name only because it was synonymised in the
protologue with the already existing L. complanatum Desf.
(the only legitimate name for which, at that stage — and
which should have been used in the protologue— was
Omalycus erosus ), not because of the synonymy with
Calvatia utriformis (such synonymy gives only the non pri-
ority of L. fontanesii).]
During a visit to the Botanisches Museum Berlin-
Dahlem (B) in 1998, the first author had the opportunity to
study Desfontaines’s herbarium (P-DESF at P) on micro-
fiche, from which the surprising discovery was made that
it included also two sheets of fungi, one of which repre-
sented the original material used by Desfontaines in draft-
ing the description of his Lycoperdon complanatum. This
specimen in P-DESV is a probable holotype (ICBN Art.
9.1, Note 1) but, as is the opinion also of V. Demoulin
(pers. comm.), it must rather be regarded as a lectotype
since no reasonable proof exists that it really was the only
material seen by Desfontaines. It is therefore here desig-
nated as such, conforming with ICBN Art. 9.9 (Greuter et
al. 2000). This material seems to have been overlooked by
all investigators since Durieu & Leveille (1848), although
enquiry confirmed that it still existed in good condition in
the phanerogamic herbarium (P) of the Museum of Natural
History in Paris [and not the cryptogamic herbarium (PC)
as might have been expected]. Since Desfontaines’s mate-
rial formed part of the ‘historical’ collection at P, it was
unfortunately not available on loan. A full-colour electron-
ic image of the material was obtained but, although it pro-
vided more information than Desfontaines’s original illus-
tration, it was still inadequate to allow identification. It did,
however, indicate the presence of small bits of glebal tissue
still adhering to the base of the specimen, a study of which
would certainly throw more light on the identity and status
of this fungus. To that puipose the first author undertook a
brief study visit to P in 2002, the outcome of which is
reported below.
EXAMINATION OF THE LECTOTYPE OF LYCOPERDON
COMPLANATUM DESF.
Methodology : the lectotype was examined macroscop-
ically and microscopically at P-DESF. Macroscopic obser-
vations were aided using a lOx magnifying hand lens.
Permission was obtained to remove a small tuft of glebal
tissue and a tiny piece of endoperidium from the lectotype
for microscopic study. The material was mounted in lac-
tophenol with aniline blue and briefly heated over an open
flame to determine the cyanophilic reaction as described
by Kreisel (1967). Initial microscopic observation at P-
DESF was made with a Nikon SE binocular light micro-
scope, but measurements were carried out in the first
author’s laboratory using a Reichert-Jung Polyvar
research microscope. Slides were sealed with clear nail
varnish and deposited in the slide collection of the
H.G.W.J. Schweickerdt Herbarium (PRU), Department of
Botany, University of Pretoria, Pretoria.
Macroscopic observations : the lectotype of L. com-
planatum Desf. consists of a single herbarium sheet on
which the two halves of a single, vertically sectioned and
pressed fungus are mounted, inside and outside surfaces
facing respectively. The specimen, that must have mea-
sured ± 90 mm diam. before sectioning, consists only of
Bothalia 35,1 (2005)
77
the flattened sterile base of a relatively large puffball of
which the gleba and surrounding upper section of the
peridium had almost completely disintegrated and disap-
peared. Small amounts of glebal tissue can, however, still
be observed in places adhering to the exposed upper sur-
face of the subgleba. Remnants of the basal part of the
endoperidium, just above and along the circumference of
the subgleba, are also still present. The outer surface of
the subgleba is reddish brown with a suede-like texture,
but a very thin glossy layer, pale brown in colour and
with a metallic sheen, is still present here and there in
surface folds. Remnants of what appears to have been
part of the exoperidium, now blackish brown, occur near
the very base of the specimen. The inner surface of the
subgleba is dull greyish brown.
Apart from the name ‘Lycoperdon complanatum' , the
herbarium label contains no additional information and
merely reads: ‘Herbier de la FLORE ATLANTIQUE
donne au Museum, par M. DESFONTAINES'. Included
also with the lectotype, however, is Desfontaines’s
(1799) original handwritten description as published in
Flora Atlantica.
Microscopic obsen’ations : capillitium septate, branched.
2.0-4.5 pm diam., occasionally slightly swollen at septa,
terminating in relatively blunt, rounded tips, ± 2 pm diam..
disarticulating at or rupturing between septa; capillitial
walls ± 0.25-0.75 pm thick, appearing smooth and imper-
forate at first glance but careful observation reveals seg-
ments densely pitted with small wall perforations <1 pm
wide, immediate cyanophilic reaction not intense but walls
staining bright blue over time. Spores globose, apedicel-
late, brownish, poorly cyanophilic, even over time, dis-
tinctly verrucose, omamentation'hp to 1 pm high, diameter
mostly 5-7 pm without and 6.5-9.0 pm with ornamenta-
tion. Endoperidium consisting of fragile, positively
cyanophilic, branched, septate, often bent and contorted
hyphae, breaking up into numerous short fragments when
pressure is applied; swollen, short, barrel-, spindle- or
irregularly shaped sphaerocyst-like elements present
between and continuous with unswollen peridial hyphae.
Taxonomic conclusion', after studying the material in
P-DESV, the current authors are quite convinced that L.
complanatum Desf. is conspecific with the common and
cosmopolitan puffball, Calvatia cyathiformis (Bose)
Morgan, and not with C. utrifonnis (= Lycoperdon
fontanesii) as has been suggested by Durieu & Leveille
(1848), De Toni (1888) and Mussat (1901). Macro-
scopically, Desfontaines’s material is reminiscent of both
C. utriformis and C. cyathiformis. Microscopically, how-
ever, C. utriformis is characterized by spores that are
smooth under the light microscope and by essentially
aseptate capillitium threads with slit-like wall perfora-
tions. The septate capillitium threads with numerous
small, not slit-like wall perforations, and the distinctly
verrucate spores of L. complanatum Desf. therefore
convincingly distinguish it from C. utriformis.
On the other hand, in terms of spore as well as capillitial
morphology, Desfontaines’s specimen closely matches C.
cyathiformis, a fungus that we are well acquainted with and
which, from the material/records at PC, also seems to be
quite common in Algeria. Although a cursory look at the
capillitium of Desfontaines’s fungus may create the impres-
sion that the walls are not perforated, careful observation
reveals many capillitial segments and fragments that are
densely pitted with small perforations, identical to and indis-
tinguishable from the capillitium of C. cyathiformis. The
spore ornamentation of L. complanatum Desf. appears to be
somewhat more pronounced than what we have become
accustomed to in C. cyathiformis, but is still within the range
as has been described for the latter fungus (Zeller & Smith
1964). A frequently overlooked diagnostic character of C.
cyathiformis is the occurrence of swollen, often irregularly
shaped sphaerocyst-like elements in its endoperidium. In the
course of our comparative studies on South African
Lycoperdaceae, and in concurrence with the opinion of V.
Demoulin (pers. comm.), we have found these elements,
previously described in Calonge & Demoulin (1975) and
Moyersoen & Demoulin (1996) and adequately illustrated
also in Dominguez de Toledo (1993) and Migliozzi &
Coccia (1999), to be a very constant, reliable and easily
observable diagnostic feature of C. cyathiformis. Hence the
presence of similar cells in the investigated peridium frag-
ment strengthens our conviction that Desfontaines’s L. com-
planatum and C. cyathiformis are conspecific.
NOMENCLATURAL IMPLICATION
When he established the genus Omalycus, Rafinesque
(1814) also included Lycoperdon complanatum Desf. in
his new taxon, renaming it Omalycus erosus Raf. No
original material of Omalycus viol acinus Raf., the type
species of the genus Omalycus , has survived, therefore,
it has never been possible to determine the taxonomic
status of Omalycus with certainty. Note, however, that
our acceptance of O. violacinus as type species is provi-
sional, and follows the interpretation of Farr et al.
(1979). There is some doubt as to whether the protologue
in Rafinesque (1814) provides enough evidence to justi-
fy the selection of O. violacinus as the type species.
Seeing that O. violacinus was not explicitly indicated as
type species by Rafinesque, and that the genus Omalycus
was not monotypic when established, then it may be
argued that O. erosus could, in the light of ICBN Art.
10. 1, Note 1, be regarded, by analogy, as a syntype of the
genus Omalycus. If the latter interpretation is followed,
then Farr et al. (1979) unintentionally lectotypified Oma-
lycus. In view of the existence of original material of L.
complanatum Desf., the appointment of O. erosus (= L.
complanatum Desf.) as lectotype might have been more
appropriate (ICBN Art. 9.10 and Art 10.2).
Although De Toni (1888), relegated Omalycus to synony-
my under Scleroderma Pers. (1801): Pers., more recent
authors listed it as a probable synonym of Calvatia
Hawks worth et al. 1995; Kirk et al. 2001 ). In the absence
of any substantial evidence, however, the question
remained: was Omalycus really a Calvatia or might it per-
haps have been a Scleroderma ? In the light of our conclu-
sion that Lycoperdon complanatum Desf., and therefore also
Omalycus erosus, is indeed a good Calvatia, and taking also
into account the opinion of V. Demoulin (pers. comm.) that
O. violacinus is the same species, it is our firm conviction
that Rafinesque’s Omalycus must be regarded as a synonym
of Calvatia, confirming earlier suggestions to that effect and
refuting its placement in the genus Scleroderma. The
nomenclatural implication of this, however, is far-reaching.
78
Bothalia 35,1 (2005)
Omalycus (1814) predates Calvatia Fr. (1849) by 35 years,
and its adoption to cover species of Calvatia would require
a considerable number of new combinations, something
which is highly undesirable. Since Calvatia is already a
nomen conservandum , it would be logical to add Omalycus
to the list of rejected names against it, which would not pre-
clude the use of Omalycus for a segregate including C.
cyathifonnis. A formal proposal to that effect has been sub-
mitted to the journal Taxon.
ACKNOWLEDGEMENTS
The first author wishes to express his gratitude to Dr
B. Hein (B); Dr M. Rignal and Prof. P. Morat (P); and
Dr B. Buyck and Prof. A. Coute (PC) for their hospitali-
ty and friendly assistance during visits to the respective
institutions. Dr Vincenzo Migliozzi is thanked for pro-
viding us with a copy of his paper on Calvatia cyathi-
fonnis. We are also indebted to Dr V. Demoulin,
Universite de Liege, for his valuable comments on an
earlier draft of this paper. Funding received from the
Peninsula Technikon research committee is acknow-
ledged with gratitude.
REFERENCES
BATSCH, A.J.G.C. 1786. Elenchus fungorum. Continuatio prima. Joh.
Jac. Gebauer. Halle.
CALONGE, F.D. & DEMOULIN, V. 1975. Les Gasteromycetes
d'Espagne. Bulletin trimestriel de la societe mycologique de
France 9 1 : 247-292.
DEMOULIN, V. 1971. Le genre Lycoperdon en Europe et en Amerique
du Nord. Etude taxonomique et phytogeographique. Doctoral
thesis, Universite de Liege.
DESFONTAINES, R.L. 1799. Flora Atlantica, siv e Historia plantarum
quae in Atlante, Agro Tunetano et Algeriensi Crescunt. Tomus
Secundus. Desgranges. Paris.
DE TONI, LB. 1888. Nidulariaceae, Lycoperdaceae et Hymenogastra-
ceae. In P.A. Saccardo, Sylloge fungorum omnium hucusque
cognitorum 7,1: 28-180. Sumptibus auctoris typis seminarii,
Patavii [Padua].
DOMINGUEZ DE TOLEDO, L. 1993. Gasteromycetes (Eumycota) del
centra y oeste de la Argentina. I. Analisis critico de los caracteres
taxonomicos, clave de los generos y orden Podaxales. Dar-
winiana 32: 195-235.
DURIEU DE MAISONNEUVE, M.C. 1848. Exploration scientifique
de VAlgerie 1,10. Imprimerie royale, Paris.
FARR, E.R., LEUSSINK, J.A. & STAFLEU, F.A. (eds). 1979. Index
nominum genericorum (Plantarum), vol. II. Regnum vegetabile
101. Bohn, Scheltema & Holkema, Utrecht.
GREUTER, W„ MCNEILL, J„ BARRIE, F.R., BURDET, H.M.,
DEMOULIN, V., FILGUEIRAS, T.S., NICOLSON, D.H.,
SILVA, PC., SKOG. J.E., TREHANE, R, TURLAND, N.J. &
HAWKSWORTH, D.L. 2000. International Code of Botanical
Nomenclature (Saint Louis Code) Regnum Vegetabile 138.
Koeltz Scientific Books, Konigstein.
HAWKSWORTH, D.L., KIRK, P.M., SUTTON, B.C. & PEGLER,
D.N. 1995. Ainsworth and Bisby's dictionary of the Fungi , edn
8. CAB International, Wallingford.
KIRK, P.M., CANNON, P.F., DAVID, J.C. & STALPERS, J.A. 2001.
Ainsworth and Bisby’s dictionary of the Fungi , edn 9. CABI
Publishing, Wallingford.
KREISEL, H. 1967. Taxonomisch-Pflanzengeographische Mono-
graphic der Gattung Bovista. Beihefte zur Nova Hedwigia 25:
1-244.
MIGLIOZZI. V. & COCCIA, M. 1999. Funghi del Lazio. XI. 48-50.
Descrizione di Xerocomus persicolor. Calvatia cyathiformis e
Myriostoma coliforme. Micologia Italiana 3: 46-55.
MOYERSOEN, B. & DEMOULIN, V. 1996. Les Gasteromycetes de
Corse: taxonomie, ecologie, chorologie. Lejeunia n.s. 152: 1-128.
MUSSAT, E. 1901. Synonymia generum, specierum subspecierumque
in vol. I-XIV descriptorum. In P.A. Saccardo, Sylloge fungorum
omnium hucusque cognitorum 15: 201. Octave Doin Edidit, Paris.
RAFINESQUE, C.S. 1814. Precis des decouvertes et travaux somio-
logiques de Mr. C.S. Rafinesque-Schmaltz entre 1800 et 1814 ;
ou choix raisonne de ses principales decouvertes en zoologie et
en botanique, pour servir d’ introduction a ses ouvrages futurs.
Royale Typographic Militaire, Palerme [Palermo],
ZELLER, S.M. & SMITH, A.H. 1964. The genus Calvatia in North
America. Lloydia 27: 148-186.
J.C. COETZEE* and A.E. VAN WYK**
* Department of Horticulture and Food Technology, Bellville Campus,
Cape Peninsula University of Technology, P.O. Box 1906, 7535 Bellville,
Cape Town.
** H.G.W.J. Schweickerdt Herbarium, Department of Botany, University
of Pretoria, 0002 Pretoria.
MS. received: 2004-05-20.
BORAGINACEAE
CODONOIDEAE, A NEW SUBFAMILY BASED ON CODON
The genus Codon was formally established by Carl
Linnaeus (1767) in the second volume of the 12th edition
of his Systema naturae. He placed the genus in his Class
X: Decandria, Monogynia. The generic name is derived
from the Greek word kodon , a bell (although the flowers
do not hang down), and alludes to the shape of the flow-
ers of C. royenii L., which are deeply cup-shaped. Codon
comprises two described species, C. royenii and C.
schenckii Schinz, both endemic to Namibia and South
Africa. A possible undescribed third species is found in
the southern part of Namibia and is currently under
investigation.
It was in France that a move towards more ‘natural’
groupings of plants was first made. It is clear from his
writings that Linnaeus recognized natural affinities, but
that ease of classification and identification were his
main objectives (Gunn & Codd 1981 ). Michel Adanson’s
Families des pi antes (1763-64) can be regarded as the
first ‘logically and philosophically sound basis for a clas-
sification of plants’ (Stafleu & Cowan 1976). In 1789
Antoine-Laurent de Jussieu followed with his Genera
plantarum. He published the description of ‘Borragineae’
as one of 100 orders (i.e. families). Many of his families
are still maintained in modern classifications. De Jussieu
based ‘Borragineae’ on the genus Borago L. He divided
28 genera into three different groups using fruit mor-
phology as a distinguishing character: 1, berry-like fruits;
2, one- or two-locular capsules; and 3, four separate nut-
lets. He regarded Codon as a genus of uncertain position.
Of the five genera of Hydrophyllaceae known to him,
De Jussieu (1789) assigned Hydrophyllum L., Phacelia
Juss. and Ellisia L. to ‘Borragineae’ and Nama L. and
Hydrolea L. to ‘Convolvuli’. R. Brown separated the for-
mer trio of genera as the natural order Hydrophylleae in
1810, and the latter two as the natural order ‘Hydroleae’
in 1818. Choisy ( 1 833) treated the Hydroleae in a mono-
Bothalia 35,1 (2005)
79
graph, recognizing the genera Hydrolea , Nama , Wigan-
dia Kunth and Romanzoffia Cham.; to these he added
Eriodictyon Benth. in 1846, and at the same time vigor-
ously defended the distinctness of the Hydroleae. De
Candolle (1846: 589) was the first to place Codon in the
family Hydrophyllaceae. Gray (1875) united all the gen-
era mentioned in the family Hydrophyllaceae, which he
divided into four tribes. Baillon (1890) merged Hydro-
phyllaceae under Boraginaceae, but his view was not fol-
lowed at the time. Hydrophyllaceae was restored by Brand
(1913) in a monograph of the family.
Until recently, most authors accepted the Hydrophylla-
ceae as a separate family. A comparison between Hydro-
phyllaceae and Boraginaceae in southern Africa based on
pollen and macromorphological characters, however,
shows a strong overlap of features. The surface structure of
pyrenes of Ehretia P.Browne shows similarity with the
seeds of Naina (compare Retief & Van Wyk 2001: 15 and
Chance & Bacon 1984: 832), although this may not be
meaningful, as the outer surfaces in these two structures are
obviously not homologous. Of more significance is the
likeness between pollen grains of Wellstedia Balf.f. and
those of Eriodictyon , Nama and Phacelia (compare
Constance & Chuang 1982 and Retief & Van Wyk 2005);
tapetal orbicules or Ubisch bodies — sporopollenin particles
usually lining the inner tangential tapetal cell walls of secre-
tory tapetums — of Wellstedia and Codon show similarity in
morphology (Retief et al. 2001 ). The broad family concept
of Baillon (1890) is followed here, and we agree with the
Angiosperm Phylogenetic Group (APG) (i998, 2003) and
Langstrom & Chase (2002) who regard Hydrophyllaceae
and Lennoaceae as synonyms of Boraginaceae s.l.
Modem views on the delimitation of Boraginaceae
differ, for example, 1, segregating a separate family,
Heliotropiaceae (Diane et al. 2002) from Boraginaceae
s.l. ; or 2, recognizing several segregate families: Bora-
ginaceae s. str., Cordiaceae, Ehretiaceae, Heliotropiaceae,
Hydrophyllaceae, Lennoaceae and Wellstediaceae (Lebrun
& Stork 1997; Gottschling et al. 2001; Gottschling 2003).
However, in neither of these two approaches has the
position of Codon been considered.
In Lerguson’s (1999) phylogenetic analysis of evolu-
tionary relationships within the Hydrophyllaceae, it is
concluded that the family is nested within a paraphyletic
Boraginaceae s.l., excluding Codon and Hydrolea.
Hydrolea is placed in a family of its own (APG II 2003).
Codon is included in Boraginaceae s.l. in a treatment of
this family for the Cape flora (Retief & Buys 2000: 374,
706). The genus Codon has traditionally been assigned to
the Hydrophyllaceae, where it seems to be unusual geo-
graphically, as the family is otherwise largely restricted
to the New World. Similarities between Codon and other
members of Boraginaceae indicate that Codon should be
placed in a subfamily of its own within Boraginaceae s.l.
A new subfamily, Codonoideae, is established here to
accommodate this southern African genus within
Boraginaceae s.l., the other local subfamilies being Well-
stedioideae, Ehretioideae, Cordioideae, Heliotropioideae
and Boraginoideae (Retief 2000: 179; Retief & Van Wyk
2001). The precise classification of the other, mainly
New World genera (see Brand 1913) of Hydrophyllaceae
s.str. within Boraginaceae s.l. has not yet been addressed
but they would most probably also require placement in
one or more additional subfamilies.
Phytogeographically the restriction of Codon to the arid
southwestern comer of Africa is of special interest. Its nearest
relatives appear to be those members of Boraginaceae s.l. pre-
viously placed in Hydrophyllaceae s. str., found mainly in
North America, but with members of both Nama and
Phacelia also occurring in South America (Deginani 1999).
Among plants, African-New World distributions are rather
unusual, but for southern Africa, involve as many as seven
families and many more genera (Goldblatt 1978). It is intrigu-
ing that several other plant groups with a distribution pattern
comparable to that of Codon, namely with a disjunct presence
in southwestern and in northeastern Africa, show links with
taxa in the New World. Among these are the boraginaceous
genus Wellstedia as well as members of Calliandra , Cae-
salpinia, Haematoxylon , Hoffmannseggia, Parkinsonia, Xero-
cladia (all Labaceae), Nicotiana (Solanaceae), Thamnosma
(Rutaceae) and Tumera (Tumeraceae) (Van Wyk & Smith
2001). These African-New World disjuncts may have been
established in different ways, but one possible explanation is
based on the proximity of Africa and South America during
Gondwana times and for quite some time after the break-up
of the supercontinent.
Members of the new subfamily Codonoideae, estab-
lished here, show the strongest affinity with the subfamilies
Wellstedioideae and Ehretioideae in Boraginaceae s.l. — in
similar pollen, inflorescence and trichome morphology.
However, the Codonoideae differ from Wellstedioideae in
the 10 or 12 (not 4) corolla lobes and in the seed which is
subglobose and glabrous (not truncate and pubescent). They
differ from Ehretioideae in habit in that they are annual or
short-lived perennial herbs (not shrubs or trees) and in the
fruit which is a capsule (not a drupe).
Codonoideae Retief & A.E.van W)’k, subfam. nov.
Type: Codon royenii L.
Herbae annuae vel breviter perennes, patentes ad erectae,
interdum basi lignescentes. Partes vegetativae aculeato-
pubescentes trichomatibus spiniformibus, et setis tri-
chomatibusque multicellularibus non ramosis. Folia
petiolata. Flores cymis scorpioideis, ebracteati. Calyx
profunde lobatus, lobi lineares. Corolla 10- vel 12-loba-
ta, tubus cylindricus vel campanulatus, fauce glabra.
Stylus terminalis, linearis, paene ad medium fissus, per-
sistens; stigma capitatum. Fructus capsula multisemi-
nalis. Semina globosa vel irregulariter angulosa, omata.
Spreading to erect, annual or short-lived perennial herbs,
sometimes woody at base. All vegetative parts prickly
pubescent with spine-like trichomes, also with setae and
unbranched, multicellular trichomes. Leaves petiolate. In-
florescence scorpioid. Calyx deeply lobed, lobes linear.
Corolla 10- or 12-lobed, tube cylindric or campanulate,
throat naked. Style terminal, linear, cleft to almost halfway,
persistent; stigma capitate. Fruit a capsule, many-seeded.
Seeds globose or irregularly angled, ornamented.
Genus: Codon L.
ACKNOWLEDGEMENT
We wish to thank Dr O.A. Leistner for translating the
subfamily description into Latin.
80
Bothalia 35,1 (2005)
REFERENCES
ANGIOSPERM PHYLOGENETIC GROUP. 1998. An ordinal classi-
fication for the families of flowering plants. Annals of the
Missouri Botanical Garden 85: 531-553.
ANGIOSPERM PHYLOGENETIC GROUP II. 2003. An update of the
Angiosperm Phylogenetic Group classification for the orders
and families of flowering plants: APG 11. Botanical Journal of
the Linnean Society 141: 399-436.
BAILLON, H.E. 1890. Hydrophylleae. Histoire desplantes 10. Hachette,
Paris.
BRAND, A. 1913. Hydrophyllaceae. In A. Engler, Das Pflanzenreich
59 (4, 251): 1-210.
BROWN, R. 1810. Prodromus florae novae Hollandieae. Johnson,
London.
BROWN, R. 1818. Observations, systematical and geographical, on
Professor Christian Smith’s collection of plants in the vicinity
of the river Congo. In J.K. Tuckey, Narrative of an expedition
to explore the river Zaire. John Murray, London.
CHANCE, G.D. & BACON, J.D. 1984. Systematic implications of
seed coat morphology in Nama (Hydrophyllaceae). American
Journal of Botany 71: 829-842.
CHOISY, J.D. 1833. Description des Hydroleacees. Memoirs de la
societe de physique et d’ histoire naturelle de Geneve 6: 95-122.
CHOISY. J.D. 1846. Hydroleaceae. In A. de Candolle. Prodromus 10:
179-185.
CONSTANCE, L. & CHUANG, T.I. 1982. SEM survey of pollen mor-
phology and classification in Hydrophyllaceae (Waterleaf fam-
ily). American Journal of Botany 69: 40-53.
DE CANDOLLE, A.P. 1846. Hydroleaceae? Prodromus 10: 588. 589.
Masson, Paris.
DEGINANI. N.B. 1999. Hydrophyllaceae. In O. Zuloaga & O. Morrone,
Catalogo de las Plantas Vasculares de la Repiiblica Argentina
II. Missouri Botanical Garden Press, Missouri.
DE JUSSIEU. A.L. 1789. Genera plantarum. Herissant & Barrois,
Paris.
DIANE, N„ FORTHER, H. & HILGER, H.H. 2002. A systematic
analysis of Heliotropium, Tournefortia, and allied taxa of the
Heliotropiaceae (Boraginales) based on ITS1 sequences and
morphological data. American Journal of Botany 89: 287-295.
FERGUSON, D.M. 1999. Phylogenetic analysis and relationships in
Hydrophyllaceae based on ndhF sequence data. Systematic
Botany 23: 253-268.
GOLDBLATT, P. 1978. An analysis of the flora of southern Africa: its
characteristics, relationships and origins. Annals of the Missouri
Botanical Garden 65: 369—436.
GOTTSCHLING, M. 2003. Phylogenetic analysis of selected Bora-
ginales. Unpublished Ph.D. thesis. Freie Universitat. Berlin.
GOTTSCHLING, M„ HILGER, H.H., WOLF, M. & DIANE, N. 2001.
Secondary structure of the ITS 1 transcript and its application in
a reconstruction of the phytogeny of Boraginales. Plant Biology
3: 629-636.
GRAY, A. 1875. A conspectus of the North American Hydrophyllaceae.
Proceedings of the American Academy of Arts and Sciences 1 0:
312-332.
GUNN, M. & CODD, L.E. 1981. Botanical exploration of southern
Africa. Balkema, Cape Town.
lAnGSTROM, E. & CHASE, M.W. 2002. Tribes of Boraginoideae
(Boraginaceae) and placement of Antiphytum, Echiochilon ,
Ogastemma and Sericostoma: a phylogenetic analysis based on
atpB plastid DNA sequence data. Plant Systematics and
Evolution 234: 137-153.
LEBRUN, J. & STORK, A.L. 1997. Enumeration des plantes a fleurs
d'Afrique tropicale 4: 374-386.
LINNAEUS, C. 1767. Svstema naturae per regna tria naturae , edn 12:
292.
RETIEF, E. 2000. Boraginaceae. In O.A. Leistner, Seed plants of south-
ern Africa: families and genera. Strelitzia 10: 178-183.
RETIEF, E. & BUYS, M.H. 2000. Boraginaceae. In P. Goldblatt & J.
Manning, Cape plants. A conspectus of the Cape flora of South
Africa. Strelitzia 9: 374—377. National Botanical Institute, Pretoria,
and Missouri Botanical Garden, Missouri.
RETIEF, E. & VAN WYK, A.E. 2001. The genus Ehretia (Boragi-
naceae: Ehretioideae) in southern Africa. Bothalia 31: 9-23.
RETIEF, E. & VAN WYK, A.E. 2005. The genus Wellstedia
(Boraginaceae: Wellstedioideae) in southern Africa. Bothalia
35. In prep.
RETIEF, E„ VAN WYK, A.E. & VAN DER MERWE. C.F. 2001.
Tapetal orbicules in southern African Boraginaceae. Proceed-
ings of the Microscopy Society of Southern Africa 31: 56.
STAFLEU. F.A. & COWAN, R.S. 1976. Taxonomic literature A-G:
49-52. Bohn, Scheltema & Holkema, Utrecht/Antwerpen.
VAN WYK, A.E. & SMITH, G.F. 2001. Regions offloristic endemism
in southern Africa: a review with emphasis on succulents.
Umdaus Press, Pretoria.
E. RETIEF*tand A.E. VAN WYK**
* National Herbarium, South African National Biodiversity Institute,
Private Bag X 1 0 1 , 0001 Pretoria.
t Student affiliation: Department of Botany, University of Pretoria,
0002 Pretoria.
** H.G.W.J. Schweickerdt Herbarium, Department of Botany, Univer-
sity of Pretoria, 0002 Pretoria.
MS. received: 2004-06-30.
POACEAE
NOTES ON ERAGROSTIS
A variant of Eragrostis gummiflua Nees?
Eragrostis gummiflua occurs in Botswana, Lesotho,
Mozambique, Namibia, Swaziland, South Africa and
Zimbabwe, usually on sand. Up to now, it has been one of
the easier Eragrostis species to identify in these regions as
it is a perennial with large, sticky, glandular patches below
the collar on the leaf sheaths and often at the nodes as well.
Sand grains or other pieces of material usually stick to these
areas, making them easy to see. The nodes and area below
the collar are often flushed purple, though sometimes the
glandular patch below the collar is yellow or brown. The
spikelet is purple to straw-coloured, with distinct, thick
nerves on the lemmas. At maturity, the palea and lemma
curve away from each other, leaving only their bases and
apices touching and then resembling the pincers of a crab.
In December 1985 the author collected specimens in
northeastern KwaZulu-Natal (2732 BA) that bore a close
resemblance to E. gummiflua but without any indication
of sticky glandular patches. A further search in the PRE
(National Herbarium, Pretoria) collection yielded two
more specimens without these glandular patches, one
from the Manzibomvu area (2732 DA) and a another
from southern Mozambique between Bela Vista and
Umbeluzi (2632 ?). The main differences between these
specimens and E. gummiflua are provided in Table 1 .
To date, the non-sticky specimens seen by the author are
from the biogeographical region known as Maputaland, an
area that has been identified as an important centre of
endemism and biodiversity in southern Africa (Siebert et ai
2004). It is bound by the Inkomati-Limpopo River in the
north, the Indian Ocean in the east, the Lebombo Mountains
in the west and by the St Lucia estuary in the south. Much
of the area is a Hat, low-level coastal plain with infertile
soils consisting of geologically recent fine-grained aeolian
sands. Climatically it lies within a transitional zone between
Bothalia 35,1 (2005)
81
TABLE 1 . — The main differences between E. gummiflua sensu stricto and the possible variant without the sticky glandular areas
* The Myre 1099 specimen had obviously been burnt and although sticky areas are present, the inflorescence is less dense tending towards that
of the variant.
** Ellis 5240 was originally thought to be the variant as it did not appear to be sticky, but closer examination showed differences in cell structure
on the sheath just below the collar where the sticky glandular patches are usually found.
TABLE 2. — The main differences between Eragrostis mexicana subsp. virescens and E. barrelieri
Character
E. mexicana subsp. virescens
E. barrelieri
Inflorescence
Inflorescence branches and/or pedicels
Spikelet
Lemma
copiously branched moderately branched, with branches stiff
eglandular crateriform glands
generally 1 .0—1 .2 mm wide but sometimes wider generally wider than 1.5-1. 8 mm
1.3-1 .7 mm long 1.7-2. 3 mm long
tropical and subtropical coastal conditions where many
tropical plants and animals reach the southern-most limit of
their distribution range (Siebert et al. 2004).
Future study and more specimens are needed to decide
whether the non-sticky form is a variant of E. gummiflua
or a new taxon. The author would be happy to receive any
duplicates of both forms from Maputaland and further
north in Mozambique either on loan or as exchange mate-
rial for PRE.
Specimens examined
E. gummiflua variant?
KWAZULU-NATAL.— 2732 (Ubombo): 2 km S of Phelendaba, (-BA).
along track. December 1985, Smook 5716 (PRE); 2 km S of Phelendaba.
December 1985, common around offices, deep sand, Smook 5727\
Manzibomvu area, east of Mbazwane stream, (-DA ), December 1971, open
grassland, sandy soil, common. Ward 7461 (PRE) (Siebert et al. 2004).
MOZAMBIQUE. — 2632 (Bela Vista): between Bela Vista and
Umbelusi, (-?AB), April 1949. Myre 523 (PRE).
E. gummiflua s. str.
KWAZULU-NATAL.— 2832 (Mtubatuba): northeast of Fanie’s
Island, (-AB), January 1960, in sandy soil, Feely & Ward 25 (PRE).
MPUMALANGA. — 2431 (Acomhoek): Arethusa Farm, Sabi Sands
Game Reserve, (-CB), April 1987, damp sandy soil in seepage area, Ellis
5240 (PRE), specimen without sticky erudite but indications of glandular
area on leaf sheaths.
MOZAMBIQUE. — Gaza Dist., Banhina National Park (area covers
part of 2232, 2233, 2322, 2323), October 1973, Tinley 2979\ (possibly
2433CD), between Guija and Macia near side track to S. Paulo de
Messano, June 1951, Myre 1099.
Eragostis mexicana subsp. virescens, a new record
for Botswana
The specimen Hansen 3363 at PRE is here identified
as Eragrostis mexicana (Homem.) Link subsp. virescens
(J.Presl.) S.D.Kock & Sanchez Vega (= Eragrostis vires-
cens J.Presl.). Previously the specimen at PRE was
wrongly identified as E. pilosa , whereas in Cope (1999:
138) the duplicates of Hansen 3363 housed at Kew and
the National Herbarium of Zimbabwe are cited under E.
barrelieri. In E. pilosa the palea falls at ± the same time
as the lemma, whereas in E. mexicana subsp. virescens
and E. barrelieri the palea persists long after the lemma
has fallen. The two last-named taxa differ as given in
Table 2 (Gibbs Russell et al. 1990; Cope 1999).
It is a new record of Eragrostis mexicana subsp.
virescens for Botswana, therefore for the Flora of south-
ern Africa ( FSA ) region, and it is the 115th Eragrostis
species recorded for the Flora zambesiaca region.
Specimen examined
BOTSWANA. — 2425 (Gabarone): Sebele Agriculture Research
Station. (-DB), Hansen 3363 (PRE).
ACKNOWLEDGEMENTS
I wish to thank Tom Cope, Herbarium, Royal Botanic
Gardens, Kew for always being willing to help and share
his vast knowledge with me. Thanks to Emsie du Plessis
and Gerrit Germishuizen for the editing and helping the
text to flow smoothly and be understandable.
REFERENCES
COPE, T. 1999. Flora zambesiaca 10,2: 138, 139.
GIBBS RUSSELL, G.E., WATSON, L., KOEKEMOER, M„ SMOOK,
L.. BARKER, N.P., ANDERSON, H.M. & DALLWITZ, M.J.
1 990. Grasses of southern Africa. Memoirs of the Botanical
Survey of South Africa No. 58.
SIEBERT, S.J., FISH. L.. UIRIS, M.M., & IZIDINE, S.A. 2004. Grass
assemblages and diversity of conservation areas on the coastal
plain south of Maputo Bay, Mozambique. Bothalia 34: 61-71.
L. FISH*
* South African National Biodiversity Institute, Private Bag X 1 0 1 . 000 1
Pretoria.
MS. received: 2004-10-17.
82
Bothalia 35,1 (2005)
POACEAE
NAME USED IN THE FSA REGION FOR THE CYMBOPOGON EXCAVATUS-CAESIUS-GIGANTEUS COMPLEX
Cymbopogon Spreng. is notorious for being easily
divided into complexes, but within the complexes the
individual species are both variable and intergrading and
are often based on rather ill-defined characters. Cym-
bopogon excavatus Stapf, C. caesius Stapf and C. gigan-
teus Chiov. are the only species in the Flora of southern
Africa (FSA) region belonging to a complex easily rec-
ognizable by the lowermost internode and pedicel of the
raceme that are swollen and fused. An additional charac-
ter often used, but nofalways easy to see, is the V-shaped
median groove on the lower glume of the sessile spikelet.
Cymbopogon caesius and C. excavatus have been sep-
arated from each other by characters such as width of the
wings of the lower glume of the sessile spikelet, shape of
the leaf base and the stronger tussock growth form of C.
excavatus (Soenarko 1977; Clayton & Renvoize 1982). C.
giganteus was separated from the two species mentioned
above by its broader leaves (Clayton & Renvoize 1982).
Clayton & Renvoize (1982) placed C. excavatus in syn-
onymy under C. caesius , keeping C. giganteus a separate
species. Sales (2002) kept C. excavatus as a synonym of
C. caesius but reduced C. giganteus to a subspecies of C.
caesius. In the FSA region the name C. excavatus was
used for both the narrow- and much broader-leaved speci-
mens (Chippindall 1956; Gibbs Russell et al. 1990).
Sales (2002) records both subsp. caesius and subsp.
giganteus for Botswana. It therefore seemed appropriate
to re-examine all specimens from the FSA region at PRE.
Working through the collection at PRE and using all
the available literature, a number of problems become
apparent. Accepting C. excavatus as a synonym of C.
caesius , the main characters used in the past to separate
C. caesius and C. giganteus are;
Leaf blades up to 8 (rarely 10) mm wide, ligule l^t(-5) mm
long; culms wiry or slender C. caesius
Leaf blades over 8 mm wide (e.g. Clayton & Renvoize 1982) or
(9-) 10-22 mm wide, ligule rarely longer than I mm;
culms robust C. giganteus
There is no definition of precisely what is meant by
‘wiry’, ‘slender’ and ‘robust’ nor any agreement on the
cut-off points of the leaf width. The width of the leaf
blade on herbarium specimens may not be representative
of the plant, as Sales (2002) notes ‘Lamina width can
vary greatly in an individual and measurements should
be taken of the widest leaves in a specimen’. In Sales
(2002) the specimen Skarpe 287 from Botswana housed
at Kew (K) is cited as subsp. giganteus , which according
to the key has ‘leaves (9-) 10-22 mm wide’ but in the
duplicate of Skarpe 287 at PRE the leaves are no wider
than 7 mm.
The ligule length of the specimens at PRE also does
not correlate with the above-mentioned characters. There
are, for example, specimens with leaf blades 9-10 mm
wide, ligules 3-5 mm long and culms varying from
robust through moderately robust to slender.
Distribution also does not separate the two taxa:
although specimens with, say, the broadest leaves may be
more common in one area than in another, both forms
occur right across the FSA region.
Therefore, until further studies are done in the field or
better characters are found to separate C. caesius from C.
giganteus, at PRE I shall treat all specimens of which the
lowermost intemode and pedicel of the raceme are swol-
len and connate, as C. caesius.
ACKNOWLEDGEMENTS
I wish to thank Tom Cope, Herbarium, Royal Botanic
Gardens, Kew for always being willing to help and share
his vast knowledge with me. Thanks to Emsie du Plessis
and Gerrit Germishuizen for the editing and helping the
text to flow smoothly.
REFERENCES
CHIPPINDALL, L.K.A. i955. A guide lo the identification of grasses
in South Africa. In D. Meredith. The grasses and pastures of South
Africa. Trustees of the Grasses and Pastures of South Africa Book
Fund, Johannesburg.
CLAYTON. W.D. & RENVOIZE, S.A. 1982. Flora of tropica l East
Africa Gramineae (Part 3): 759-761.
GIBBS RUSSELL, G.E., WATSON, L„ KOEKEMOER, M„ SMOOK, L„
BARKER, N.P., ANDERSON, H.M. & DALLWITZ. M.J. 1990.
Grasses of southern Africa. Memoirs of the Botanical Survey of
South Africa No. 58.
SALES, F. 2002. Flora zambesiaca 10,4: 76-80.
SOENARKO, S. 1977. The genus Cymbopogon Spengel (Gramineae).
Reinwardtia 9: 225-375.
L. FISH*
* South African National Biodiversity Institute, Private Bag X101, 0001
Pretoria.
MS. received: 2004-10-17.
HYACINTHACEAE
ORNITHOGALUM JUNCIFOLIUM VAR. EMS1I, A NEW CLIFF-DWELLING ORNITHOGALUM FROM EASTERN CAPE, SOUTH AFRICA
introduction particularly to the Cape Floristic Region (Obermeyer
1973; Manning & Goldblatt 2003). Recent field studies
The bulbous genus Ornithogalum L. consists of 200 on cliffs associated with dry river valleys revealed several
species of which ± 108 are confined to South Africa and plant taxa new to science (Van Jaarsveld & Van Wyk
Bothalia 35,1 (2005)
FIGURE 8. — Omithogalum juncifolium var. emsii , x 1, show-
ing basal vegetative bulbils. Artist: Lisa Strachan.
83
1999, 2003; Van Jaarsveld 2003). Among these, a new
variety of Omithogalum juncifolium was collected and is
described in the present paper.
Whereas many bulbous plants grow opportunistically
on cliffs, O. juncifolium var. emsii is encountered only in
this type of habitat and appears to be an obligate cremno-
phyte ( cremno is derived from the Greek for cliff and
phyton means plant). Owing to the vertical orientation of
cliffs, water run-off is extreme, resulting in a mostly very
dry habitat with succulents often a conspicuous feature
on precipices in South Africa.
Omithogalum juncifolium Jacq. var. emsii Van
Jaarsv. & A.E.van Wyk., var. nov., a var. typica tepalis
10-12 mm longis, bulbillis e basi dense proliferantibus et
florescentia medio aestate (mense Decembri) differt.
TYPE. — Eastern Cape, 3326 (Grahamstown): cliffs over-
looking the Koonap River, near the Koonap Reserve,
(-BA), 08-10-2002, Van Jaarsveld & Ems 16808 (NBG,
holo.).
Plants bulbous, epigeous and forming round clusters
up to 100 mm diam. and consisting of many bulbs and
bulbils. Bulbs globose, 15-20 mm diam. and high;
tunics grey, papery and exposing green live tissue;
basal part of bulb continuously proliferating, forming
many ovate to rounded bulbils, up to 5 mm in diam..
Leaves 2 or 3, synantherous, linear, half-terete, 95-150
x 1.5 mm, apex acute, dark green; adaxial surface shal-
lowly canaliculate; abaxial surface rounded; base
sheathing, tubular with short membranous neck, 5-8 x
2-3 mm; margin minutely ciliolate. Raceme 100-200
mm long, 8-12-flowered; scape terete, erect; bracts del-
toid-cuspidate, auriculate, up to 6 x 2 mm; pedicel up to
4-5 mm long, lengthening up to 6-7 mm in fruit.
Perianth stellate, white, up to 20-24 mm diam. Tepals
linear-lanceolate, 3 inner 10-12 x 3. 0-3. 5 mm, white
with green median stripe. Stamens 5 mm long; outer fil-
aments flattened, linear-acuminate, 1 mm diam. at base;
inner filaments shorter, ovate-triangular, up to 1 .5 mm
long; anthers 0.8 mm long, yellow. Ovary ovate, 3x2
mm, green, shortly stipitate; style erect, 4 mm long;
stigma capitate. Capsule ovoid, 5-7 x 3-4 mm. Seeds
triangular-ovate, 1.5 x 0.8 mm, black, denticulate, 24
per capsule. Figure 8.
Phenology : flowering mainly from early December to
January (summer). Seeds are dispersed by wind in sum-
mer and early autumn (October onwards).
Diagnostic features and affinities : Omithogalum jun-
cifolium var. emsii is at once distinguished from the typi-
cal variety by its proliferous production of bulbils at the
base of the bulb and its flowering time which is from
December to January. Var. juncifolium flowers from
spring to early summer and has slightly smaller flowers
(tepals 7-10 mm long) and does not have the dense basal
proliferation of small bulbils.
Distribution, habitat and cultivation: Omithogalum
juncifolium var. emsii is known only from vertical south-
facing shale cliffs of the Adelaide Subgroup, Beaufort
Group, Karoo Supergroup, overlooking the Great Fish
84
Bothalia 35,1 (2005)
FIGURE 9. — Known distribution of O. juncifolium var. emsii , •.
River (Figure 9). It grows sympatrically with O. bolusia-
num and is also an Albany Centre endemic (Van Wyk &
Smith 2001). On the other hand, var . juncifolius is wide-
ly distributed, ranging from the Western Cape north-
wards on or below the Great Escarpment to Gauteng and
North-West.
Plants of var. emsii are locally common, consisting of
small epigeous, globose clusters of green (photosynthet-
ically active, though covered with dry, grey tunics) and
two or three spreading, pendent, filiform leaves, an adap-
tation to shady cliffs. The prolific formation of basal bul-
bils results in a continuous release of vegetative propag-
ules, enabling the plant to establish in adjacent rock
crevices. It is often mat-forming and plants may com-
pletely fill a single crevice. This vegetative reproductive
strategy has also been observed in other cliff-dwelling
bulbous species such as Cyrtanthus montanus, C. labia-
tus, C. inaequalis and Omithogalum longibracteatum
(Van Jaarsveld & Van Wyk 2003).
The very thin, wiry leaves of O. juncifolium var. emsii
are evergreen and bundled together at the base, allowing
for maximum light penetration while also contributing to
photosynthesis. The spreading leaves sometimes become
pendent with age. The bulbs are also enveloped in grey
tunics which probably serve to block excessive light and
to reduce transpiration. Plants are long-lived with the dry
leaves persistent, withering from the base. The black,
angled, light seeds are shaken from the capsules and are
wind-dispersed in summer.
Plants are easily grown by bulbils and thrive in culti-
vation.
Etymology, the specific epithet emsii is named after
Paul Ems, horticulturist and botany student who first
spotted the plants, in recognition of his assistance in the
field.
ACKNOWLEDGEMENTS
Gerrit Germishuizen and Emsie du Plessis are thank-
ed for editing the text. Dr Hugh Glen for compiling the
Latin diagnosis and the artist Lisa Strachan for preparing
the drawing. Paul Ems is thanked for his assistance dur-
ing plant expeditions to explore cliff faces.
REFERENCES
MANNING, J.C. & GOLDBLATT. P. 2003. Omithogalum. In G.
Germishuizen, & N.L. Meyer, Plants of southern Africa: an
annotated check®. Strelitzia 14: 1065-1068.
OBERMEYER, A. A. 1978. Omithogalum : a revision of the southern
African species. Bothalia 12: 323-376.
VAN JAARSVELD, E.J. 2003. Southern African bulbs on the edge: a
survey of those taxa associated with cliffs. The Indigenous Bulb
Association of South Africa Bulletin 52: 115-126.
VAN JAARSVELD, E. & VAN WYK, A.E. 1999. Five new cremno-
philous taxa from semi-arid regions in South Africa. Aloe 36:
71-74.
VAN JAARSVELD, E.J. & VAN WYK, A.E. 2003. Hyacinthaceae and
Crassulaceae. Two new cremnophilous taxa from semi-arid
regions in South Africa. Bothalia 33: 115-118.
VAN WYK, A.E. & SMITH, G.F. 200 1 . Regions offloristic endemism
in southern Africa: a review with emphasis on succulents.
Umdaus Press, Hatfield, Pretoria.
E.J. VAN JAARSVELD*f and A.E. VAN WYK**
* South African National Biodiversity Institute, Kirstenbosch, Private
Bag X7, 7735 Claremont.
t Student affiliation: Department of Botany, University of Pretoria, 0002
Pretoria.
** H.G.W.J. Schweickerdt Herbarium, Department of Botany, University
of Pretoria, 0002 Pretoria.
MS. received: 2004-02-18.
RUBIACEAE
INFRASPECIFIC TAXA IN A SOUTHERN AFRICAN PAVETTA SPECIES
The genus Pavelta L. comprises ± 400 species occur-
ring in the Old World tropics (Bridson 2003) with 21
species in the summer rainfall areas of southern Africa
(Retief & Leistner 2000). In 1929 Bremekamp revised
the South African species of the genus Pavetta, followed
by a monograph of the group in 1934. Several authors,
e.g. Bridson ( 1 978) and Kok & Grobbelaar ( 1 984), main-
tained that Bremekamp not only recognized too many
species but that some of the species delimitations were
unsatisfactory. Subsequently several of his species were
placed in synonymy by Bridson (1978) and Kok &
Grobbelaar (1984). Although these later treatments did
help to make identifications easier, some taxa seem to be
'waste bins’ in which too many names were dumped, e.g.
Pavetta zeyheri Sond. Kok & Grobbelaar (1984) listed
10 names as synonyms under P. zeyheri including P.
microlancea K.Schum., P. middelburgensis Bremek. and
P. lasiopeplus K.Schum. Recently Bridson (2003) resur-
rected P. lasiopeplus and hinted that P. microlancea and
P. middelburgensis may deserve recognition as distinct
taxa. It seems that Coates Palgrave (2002) also upheld P.
lasiopeplus as separate from P. zeyheri.
Bothalia 35,1 (2005)
85
TABLE 1. — Comparison between the three subspecies of Pavetta zeyheri
B, Botswana; LIM, Limpopo; NW, North-West; G, Gauteng; M, Mpumalanga; FS. Free State.
* Calyx tube = 'hypanthium' + limb-tube.
Bremekamp (1934) separated P. zeyheri from P. mid-
delburgensis and P. microlancea on the grounds of the leaf
shape and ratio of leaf length to width. He distinguished
between P. middelburgensis and P. microlancea on the
grounds of corolla and leaf length. The leaves of P. micro-
lancea and P. middelburgensis are usually very small, but a
study of material housed in PRE and PRU showed no clear
discontinuity in size among the three taxa discussed in the
present contribution (Table 1 ) and these differences are not
reliable for distinguishing the three taxa. In all three species
the corolla tube is hairy inside.
As the differences in leaf shape and size, and the mor-
phology of the flowers show some integration, it is diffi-
cult to justify three separate species. However, because of
the differences in growth form and the disjunct geographi-
cal distribution, P. middelburgensis and P. microlancea are
formally reinstated as subspecies of P. zeyheri in the pre-
sent contribution. P. lasiopeplus K.Schum. is upheld, fol-
lowing Bridson (2003). Bridson (2003) distinguishes
between P zeyheri and P. lasiopeplus inter alia on the
grounds of the differences in the length of the calyx lobes:
0.5-1. 5(-2.0) mm in P. zeyheri and 2-3 mm in P. lasiope-
plus and inflorescence branches apparent in P zeyheri , but
suppressed in P. lasiopeplus. These were the only reliable
characters found to distinguish between these two species
in material studied at PRE. Silky hairs on the inflorescence
bracts were observed in almost all the P. zeyheri speci-
mens examined at PRE, yet Bridson (2003) maintained
they are without silky hairs. The shape of the stipules
varies too much to be used as a distinguishing character
between these two species.
P. dissimilis Bremek., P. pseudozeyheri Bremek., P.
zeyheri Sond. var. brevituba Bremek., P. zeyheri Sond.
var. pubescens Bremek. and P. zeyheri Sond. var. sonderi
Bremek. are all regarded as part of P. zeyheri subsp.
zeyheri , as the differences given by Bremekamp (1929,
1934) are only part of the variation range of this taxon.
Pavetta zeyheri Sond., Flora capensis 3: 21
(1865); Bremek.: 211 (1929); Bremek.: 183 (1934);
Launert: 23 (1966); Kok & Grobbelaar: 187 (1984);
Retief & P.P.J. Herman: 588 (1997); Coates Palgrave:
1123, 1 124 (2002); Bridson: 582, 583 (2003); Retief: 837
(2003). Type: Transvaal [Gauteng], Vaal River, Burke
s.n. (K, PRE, -photo.!; S!, TCD, -photo.!).
subsp. zeyheri sens. str.
Ixora zeyheri (Sond.) Kuntze: 287 (1891).
P. dissimilis Bremek.: 211 (1929); Bremek.; 183 (1934). Type:
Middleberg, Gray 4188 (PRE!).
P. pseudozeyheri Bremek.: 21 1 (1929); Bremek.: 183 (1934). Type:
Transvaal [North-West], Rustenburg, Bremekamp 103 (PRE!).
P. zeyheri Sond. var. brevituba Bremek.: 212 (1929); Bremek.: 184
(1934). Type: Transvaal [Gauteng], Pretoria District, Nelson s.n.
Transv. Mus. 11710 (PRE!).
P. zeyheri Sond. var. pubescens Bremek.: 212 (1929); Bremek.: 184
(1934). Type: Transvaal [Gauteng], Pretoria, Premier Mine, Poliak s.n.
PRE41394 (PRE!).
P. zeyheri Sond. var. sonderi Bremek.: 184 (1934). Type: Transvaal
[Gauteng], Pretoria, Magalisberg [Magaliesberg], Zeyher 768 (S!, PRE,
photo.!).
P. zeyheri subsp. zeyheri is a shrub or small tree, usu-
ally single-stemmed, 1^1 m tall. Leaves narrowly obo-
vate, oblanceolate to narrowly elliptic, (20.5)28.0-57.0
(68.0) x (4.5)5.0-12.0(15.5) mm. Inflorescence branch-
es apparent; bracts with silky hairs on inside; calyx lobes
0. 5-2.0 mm long. Grows mainly on rocky outcrops and
86
Bothalia 35, 1 (2005)
i i i i i I I I I I 1 I I L
12 14 16 18 20 22 24 26 28 30 32
FIGURE 10. — Distribution of P. zeyheri subsp. zeyheri, •; P. zeyheri
subsp. middelburgensis , A; and P. zeyheri subsp. microlancea ,
■. in southern Africa.
hillsides, but also on flats, often on sandy soil in Botswana,
Limpopo, North-West, Gauteng, Mpumalanga and the Free
State (Figure 10).
BOTSWANA. — 2425 (Gaborone): Mmokolodi, 1 2 km S of Gaborone,
(-DB), Cole 642 (PRE); Mokolodi Village. Mokolodi Reserve, 15 km
S of Gaborone, (-DD), Cole 1154 (PRE).
(-CD), Relief & Herman 171 (PRE, PRU).
MPUMALANGA. — 2529 (Witbank): Loskopdam Nature Reserve,
(-AD), Theron 1094 (PRE, PRU). 2630 (Carolina): Carolina, (-AA),
Nicholson PRE4312 (PRE).
FREE STATE. — 2727 (Kroonstad): Vredefort Dist., (-AB), Oliver
277 ( PRE).
subsp. middelburgensis (. Bremek .) P.PJ. Herman,
comb, et stat. nov.
P. middelburgensis Bremek. in Annals of the Transvaal Museum
13: 212 (1929); Bremek.: 184 ( 1934). Type: Mpumalanga, Middelburg.
without exact locality, Jenkins s.n. Transv. Mus. 9828 (PRE, holo.!).
Pavetta zeyheri subsp. middelburgensis is a dwarf,
compact shrub, almost bonsai-like, up to 0.5 m tall,
growing wedged amongst outcrops of rocks and boulders
or rocky sheets, confined to Middelburg area, Mpuma-
langa (Figures 10, 11). Leaves small, narrowly obovate
or oblanceolate, 17-36 x 2. 5-6.0 mm. Inflorescence
branches apparent; bracts with silky hairs on inside;
calyx lobes 0.5-1. 2 mm long.
MPUMALANGA.— 2529 (Witbank): Doornkop 273 J.S., ‘Ghost
Rocks’ W of Eerstekamp, (-CB), Du Plessis 984 (PRE, PRU);
Botsabelo, (-CD), Codd & Dyer 2869, Fouche s.n. Transv. Mus.
26971, Herman 1531, 1532 (PRE), Schlechter 3781 (PRE, PRU); kop-
pie W of Middelburg, (-CD), Brusse 1932 (PRE); Witbank Nature
Reserve, (-CD), Smit 1843 (PRU); Middelburg, without exact locality,
Thode A 1610 (PRE).
LIMPOPO. — 2427 (Thabazimbi): Waterberg Dist., Vaalwater,
Farm Malmanies River, (-BA), Hardy, Relief & Herman 5301 (PRE,
PRU). 2428 (Nylstroom): Waterberg Dist., 3 km W of Nylstroom-
Vaalwater road, on road to Loubad, (-CB), Reid 460 (PRE).
NORTH-WEST. — 2525 (Mafikeng): Botsalano Game Reserve,
(-DA), Phalatse 170 (PRE). 2527 (Rustenburg): 3 miles (4.8 km) S of
Rustenburg on road to Rustenburg Kloof, (-BA), De Winter 7800
(PRE). 2625 (Delareyville): E of Madibogo, Madibogo Hills, (-AC),
Gubb KMG10737 (PRE). 2627 (Potchefstroom): Dassiesrant, (-CA),
Botha & Ubbink 1819 (PRE). 2725 (Bloemhof): Leeuwfontein, 10 km
W of Wolmaranstad, (-BB), A.E. van Wyk 758 (PRE).
GAUTENG. — 2527 (Rustenburg): Hartebeestpoort Dam, Dr
Brassy’s Farm, (-DD), Mogg 34052 (PRE). 2528 (Pretoria): Fountains
Valley, (-CA), Repton 98 (PRE); Lukasrand, below Post Office Tower,
subsp. microlancea ( K.Schum .) P.PJ. Hetman, comb,
et stat. nov.
P. microlancea K.Schum. in Botanische lahrbiicher 28: 80 (1899).
Bremek.: 212 (1929); Bremek.: 184 (1934). Type: Mpumalanga,
Komatipoort, without exact locality, Schlechter 11760 (B, holo-
PRE!).
Pavetta zeyheri subsp. microlancea is a small shrub,
0.3-0.8(-l .0) m tall, growing on rocky slopes or loamy
flats in the Sabie, Komatipoort and Crocodile Bridge
areas of Mpumalanga (Figure 10). Leaves vary from nar-
rowly obovate or oblanceolate to narrowly elliptic, (17.0)
FIGURE 11.— Habit and habitat of
Pavetta zeyheri subsp. middel-
burgensis.
Bothalia 35.1 (2005)
87
23.5 — 3 1 .0( — 40.0) x 3.0-5.5(-7.0) mm. Inflorescence
branches suppressed; bracts with silky hairs on inside;
inflorescence few-flowered (mostly 4-6); calyx lobes
1-2 mm long.
Coates Palgrave (2002) referred to a large-leafed form
of P zeyheri which grows up to 3 m tall in the Sekhu-
khuneland area as probably P. microlancea, but it cannot be
the same taxon as described above because of the differ-
ences in stature of the plants. The so-called Sekhukhune-
land form of P. zeyheri is characterized by blue-green
leaves and may well represent a separate species — it is listed
as Pavetta sp. nov. (tree number 722.1) in Von Breitenbach
et al. (2001). Bridson (2003) mentioned that P. micro-
lancea differed from P. zeyheri sens, str., apart from size,
in having fewer-flowered, umbellate inflorescences sub-
tended by bracts with silky hairs.
MPUMALANGA. — 2531 (Komatipoort): Kruger National Park.
10 miles S of Lower Sabie camp, (-BB), Codd 6133 (PRE. KNP);
Kruger National Park, Lower Sabie/Crocodile River Bridge, near
Nthandanyathi waterhole. Van Rooyen 4133 (PRU); between Komati-
poort and Letaba River, on the Selati railway, (-BD), Rogers 12887
(PRE); 3 miles (4.8 km) N of Crocodile River Bridge, Kruger National
Park, (-BD), P. van Wyk 4735 (PRE. KNP).
CONSERVATION STATUS
Given the very restricted geographical distribution of
both P. zeyheri subsp. middelburgensis and subsp.
microlancea , these two taxa should probably be declared
Vulnerable, the population restricted to an area of occu-
pancy < 100 km2 (VU D2) according to the 1994 IUCN
Red List categories (Golding 2002).
ACKNOWLEDGEMENTS
The Officer in Charge of the Botsabelo Nature Reserve
(Mission Station) is thanked for permission to collect on
the property. The curators of the Skukuza Herbarium and
Herbarium Stockholm are thanked for the loan of speci-
mens and of the Schweickerdt Herbarium (PRU) for study-
ing their material on site. I wish to thank Ms T. Rampho
for the scanning of type material in Kew, Mrs S. Brink
for the scanning of the photograph, Ms H. Steyn for the
distribution map and Mr and Mrs C. de Jager for accom-
modation during weekend visits. The company of my
family on the weekend trips is greatly appreciated.
REFERENCES
BREMEKAMP, C.E.B. 1929. A revision of the South African species
of Pavetta. Annals of the Transvaal Museum 13: 182-213.
BREMEKAMP. C.E.B. 1934. A monograph of the genus Pavetta L.
Repertorium specierum novarum regni vegetabilis 37: 1-208.
BRIDSON, D.M. 1978. Studies in Pavetta (Rubiaceae subfamily
Cinchonoideae) for part two of Flora of tropical East Africa:
Rubiaceae. Kew Bulletin 32: 609-652.
BRIDSON, D.M. 2003. Pavetta L. Flora zambesiaca 5,3: 544-598.
COATES PALGRAVE, M. 2002. Keith Coates Palgrave Trees of south-
ern Africa, edn 3. Struik, Cape Town.
GOLDING, J.S. (ed.). 2002. Southern African plant Red Data Lists.
SABONET Report No. 14.
KOK, P.D.F. & GROBBELAAR, N. 1984. Studies on Pavetta (Rubi-
aceae) II. Enumeration of species and synonymy. South African
Journal of Botany 3: 185-187.
KUNTZE, C.E.O. 1891. Rubiaceae. Revisio generum plantarum 1: 287.
Charles Klincksieck, Paris.
LAUNERT, E. 1966. Rubiaceae. Prodromus einer Flora von Stidwest-
afrika 115: 1-27.
RETIEF, E. 2003. Rubiaceae. In G. Germishuizen & N.L. Meyer, Plants
of southern Africa: an annotated checklist. Strelitzia 14: 836,
837. National Botanical Institute, Pretoria.
RETIEF, E. & HERMAN, P.P.J. 1997. Plants of the northern provinces of
South Africa: keys and diagnostic characters. Strelitzia 6: 576-593.
RETIEF. E. & LEISTNER. O.A. 2000. Rubiaceae. In O.A. Leistner,
Seed plants of southern Africa: families and genera. Strelitzia
10: 476-495. National Botanical Institute, Pretoria.
SCHUMANN, K. 1899. Rubiaceae africanae. Botanische Jahrbticher
28: 80, 81.
SONDER. O.W. 1865. Rubiaceae. Flora capensis 3: 1-39. Hodges,
Smith, Dublin.
VON BREITENBACH, J„ DE WINTER, B„ POYNTON, R.. VAN
DEN BERG, E„ VAN WYK, B. & VAN WYK, E. 2001. Pocket
list of southern African indigenous trees: including selected
shrubs and woody climbers, edn 4, 1st abridged impression.
Briza Publications & Dendrological Foundation, Pretoria.
P.P.J. HERMAN*
* South African National Biodiversity Institute, Private Bag X101, 0001
Pretoria.
MS. received: 2004—06-28.
AGAPANTHACEAE
SYNONYMY IN AGAPANTHUS
INTRODUCTION
The sole member of the family Agapanthaceae,
Agcipanthus L'Her. is endemic to four southern African
countries, Lesotho, Mozambique, South Africa and
Swaziland (Duncan 2003). Represented by A. africanus
in the southwestern Cape, the genus is distributed from
the Cape Peninsula eastwards to the southern, southeast-
ern, eastern and northern parts of the subcontinent, ter-
minating in the Soutpansberg Mountains south of the
Limpopo River in Limpopo (Leighton 1965). Aga-
panthus has always presented a special problem in clas-
sification due to the extreme morphological variability
within the species and the paucity of reliable characters
on which to base them. The genus was last revised by
Leighton (1965) whose morphological and geographical
study resulted in the recognition of ten species, in which
she made the following statement: ‘I have explored every
avenue for characters on which to base the species of
Agcipanthus and have found very few which are reason-
ably constant’.
The modem multidisciplinary approach to classifica-
tion has resulted in more avenues becoming available for
delimitating species boundaries. Recent research using
somatic nuclear DNA content, as measured in picograms
(pg) using flow cytometry with propidium iodide, as well
as pollen colour and vitality, introduced novel criteria in
effectively delimitating the species of Agapanthus
(Zonneveld & Duncan 2003). Based on these criteria, as
Bothalia 35,1 (2005)
well as morphological and distribution data, it was pro-
posed that the number of species be reduced to six,
namely A. africanus (L.) Hoffmanns., A. campanulatus
F.M.Leight., A. caulescens Sprenger, A. coddii F.M.Leight.,
A. inapertus P.Beauv. and A. praecox Willd., with the
recognition of a high degree of variation within each
species. The DNA content confirmed several relation-
ships previously speculated by Leighton (1965), and
combined with pollen colour, the species were effective-
ly divided into two distinct groups: a group with lilac
pollen and a DNA content of 22.3-24.0 pg containing the
species A. campanulatus , A. caulescens and A. coddii ,
and a group with yellowish brown pollen and a DNA
content of 25.2—3 1 ..6 pg containing the species A.
africanus, A. inapertus and A. praecox. The narrow en-
demic A. walshii L. Bolus from the southwestern Cape
was most appropriately treated as a subspecies of the
more widespread A. africanus as A. africanus subsp.
walshii (L. Bolus) Zonn. & G.D. Duncan, A. comptonii
F.M.Leight. was found to be synonymous with A. prae-
cox subsp. minimus (Ker Gawl.) F-M.Leight., A. nutans
F.M.Leight. was found to be synonymous with A.
caulescens subsp. gracilis F.M.Leight., and A. dyeri
F.M.Leight. was found to be synonymous with A. inaper-
tus subsp. intermedins F.M.Leight. The new combination
A. africanus subsp. walshii has already been formalized
(Zonneveld & Duncan 2003), and discussed in more
detail (Duncan 2004). The opportunity is taken here to
formalize the synonomy of A. comptonii, A. dyeri and A.
nutans, as follows:
NEW SYNONYMY
Agapanthus praecox Willd. subsp. minimus ( Ker
Gawl.) F.M.Leight. in Journal of South African Botany,
Suppl. vol. 4: 22 (1965). Type: without locality and col-
lector [Botanical Register 9: t. 699 (1823), lecto., here
designated],
A. comptonii F.M.Leight. subsp. comptonii ; 24, 27 (1965), syn. nov.
Type: Eastern Cape. Bathurst: Kaffir Drift, Compton s.n. NBG397/45
(BOL, holo.!; NBG, iso.!).
A. comptonii F.M.Leight. subsp. longitubus F.M.Leight: 27, 29
(1965), syn. nov. Type: Eastern Cape, East London: Chalumna
Causeway, Smith s.n. NBG 1 35/45 (BOL, holo.!; NBG, iso.!).
Agapanthus praecox is an evergreen, extremely vari-
able species with regard to perianth size and shape, and
overall plant size and habit, and has a fairly wide distri-
bution extending from Knysna in the southern Cape to
Port Shepstone on the KwaZulu-Natal south coast.
Leighton (1965) recognized three subspecies: subsp.
praecox, subsp. minimus and subsp. orientalis. The sub-
species minimus is distinguished from the other sub-
species by the relatively small overall plant size, the
plants do not form dense clumps, they have fewer-flow-
ered inflorescences and a slender peduncle usually less
than 600 mm high. The dwarf species A. comptonii
occurs in Eastern Cape from Bathurst to East London,
and comprises subsp. comptonii and subsp. longitubus.
The subsp. comptonii is distinguished from subsp. longi-
tubus by its perianth tube being one third or less of the
perianth length, versus more than one third or nearly half
the perianth length in subsp. longitubus. There is no con-
vincing way of distinguishing subsp. comptonii and
subsp. longitubus morphologically from A. praecox
subsp. minimus, which has similar evergreen leaves and
overall plant size, and shows a gradation in perianth tube
length that includes those of subsp. comptonii and subsp.
longitubus. The distribution ranges of subsp. comptonii
and subsp. longitubus fall within that of A. praecox
subsp. minimus, and both subsp. comptonii and A. prae-
cox subsp. minimus occur in the Bathurst area (Leighton
1965). The nuclear DNA content of subsp. comptonii and
subsp. longitubus ranges from 25.40-25.61 pg which is
within the range for A praecox subsp. minimus, and they
are therefore considered synonymous with A. praecox
subsp. minimus (Zonneveld & Duncan 2003). A. praecox
subsp. orientalis remains valid and is recognized by its
dense clumps producing relatively broad, arching
foliage, its sturdy peduncles well over 600 mm, and its
large, dense inflorescences.
Agapanthus caulescens Sprenger subsp. gracilis
F.M.Leight. in Journal of South African Botany, Suppl.
vol. 4: 36 (1965). Type: KwaZulu-Natal, Ubombo-Ingwa-
vuma range, Gerstner 3189 (BOL, holo.!; NBG, iso.!).
A. nutans F.M.Leight.: 38 (1965), syn. nov. Type: KwaZulu-Natal.
Estcourt: Mooi River, Cheape s.n. NBG824/53 (BOL, holo.!; NBG,
iso.!).
Agapanthus caulescens is a deciduous, summer-grow-
ing species found in KwaZulu-Natal, Mpumalanga and
Swaziland. Leighton (1965) recognized three subspecies:
subsp. caulescens, subsp. cmgustifolius and subsp. gracilis.
Subspecies gracilis is distinguished by its more slender,
laxer growth form and its smaller flowers with the perianth
segments recurving markedly towards the apices. A. nutans
and A. caulescens subsp. gracilis both have similar dis-
tinctly caulescent shoots, and linear, deciduous leaves with
hyaline margins. The degree of recurving of perianth seg-
ments does not seem an important difference, as those of A.
nutans also show some degree of recurving. Leighton
(1965) also regarded A. nutans as distinct due to its nod-
ding flowers, but both nodding and spreading flowers may
occur within the same inflorescence and does not seem an
important difference. A. nutans only occurs as single
plants, not in populations; Leighton (1965) said of A.
nutans: ‘I am aware that the sporadic occurrence of all
those plants I have placed in A. nutans may very well mean
that they are all mutants which arise from time to time, and
1 regard this as a grouping of like forms rather than a close-
knit species’. A. nutans occurs within the distribution range
of A. caulescens and Leighton (1965) stated that A. nutans
has links with A. caulescens subsp. gracilis. Both A. nutans
and A. caulescens subsp. gracilis are recorded from
Ingwavuma in the Ubombo Mountains of KwaZulu-Natal.
The nuclear DNA content of A. nutans was found to be
23.38 pg which is within the range for A. caulescens, and it
is therefore considered synonymous with A. caulescens
(Zonneveld & Duncan 2003), most appropriately placed
under A. caulescens subsp. gracilis.
Agapanthus inapertus P.Beauv. subsp. inter-
medius F.M.Leight. in Journal of South African Botany,
Suppl. vol. 4: 45 ( 1 965). Type: Mpumalanga, Standerton:
near Palmford, Reynolds 7214 (BOL, holo.!; NBG,
iso.!).
Bothalia35,l (2005)
89
A. dyeri F.M.Leight.: 46 (1965), syn. nov. Type: Limpopo, Polo-
kwane [Pietersburg]: Blaauberg, Mohlakeng Plateau, Dyer & Codd
8765 (PRE. holo.!).
Agapanthus inapertus is a deciduous, summer-growing
species, widely distributed in Mpumalanga and Limpopo,
and also occurring in Gauteng and Swaziland. Leighton
(1965) recognized five subspecies: subsp. inapertus ,
subsp. hollandii , subsp. intermedius, subsp. parviflorus
and subsp. pendulus. Subspecies intermedius has flowers
25-45 mm long and the perianth segments range from
those which spread very little to forms in which they
spread appreciably (Leighton 1965). Its distribution covers
the whole territory of the species, occurring mostly on the
fringes (Leighton 1965). A. dyeri is recorded from two
widely separated localities, Mohlakeng on the Blouberg
Plateau near Polokwane (Pietersburg) in Limpopo, and at
Namaachas in southwestern Mozambique (Leighton
1965; Duncan 1993). Leighton (1965) said of A. dyeri:
‘This species is very close to A. inapertus subsp. inter-
medius, but the very short tube precludes it from being
placed in that species’. Tube length in A. dyeri does not
seem an important difference as it shows a gradation of
sizes to those of A. inapertus subsp. intermedius, and geo-
graphically, A. dyeri occurs within the same area of dis-
tribution as A. inapertus subsp. intermedius , such as on
the Blouberg Plateau (Leighton 1965). Dr R.A. Dyer,
after whom A. dyeri was named in 1965, himself ques-
tioned the validity of this species, as follows: ‘I am not so
modest that I should wish to see A. dyeri Leighton rele-
gated to synonymy, but I forsee the day when some work-
er will wish to know more precisely what its relationship
is to A. inapertus subsp. intermedius, which is also
recorded from the Blouberg Plateau, the type locality for
A. dyeri' (Dyer 1966). The nuclear DNA content of A.
dyeri was found to range from 24.99-25. 17 pg (including
the geographically isolated plants from Namaachas in
southwestern Mozambique), which is within the range of
A. inapertus subsp. intermedius, and it is therefore con-
sidered synonymous with A. inapertus subsp. intermedius
(Zonneveld & Duncan 2003).
REFERENCES
DUNCAN, G.D. 1993. Agapanthus dyeri. The Flowering Plants of Africa
52: t. 2062.
DUNCAN. G.D. 2003. Some notes on Agapanthus. The Plantsman,
new ser. 2: 174—178.
DUNCAN, G.D. 2004. Agapanthus africanus subsp. walshii. Curtis's
Botanical Magazine 21: 205-214.
DYER. R.A. 1966. Agapanthus inapertus subsp. pendulus. The Flowering
Plants of Africa 37: t. 1480.
KER GAWLER, J. 1823. Agapanthus umbellatus var. minimus. Edwards's
Botanical Register 9: t. 699.
LEIGHTON, F.M. 1965. The genus Agapanthus L'Heritier. Journal of
South African Botany, Suppl. vol. 4. National Botanic Gardens
of South Africa, Cape Town.
ZONNEVELD. B.J.M. & DUNCAN, G.D. 2003. Taxonomic implica-
tions of genome size and pollen colour and vitality for species
of Agapanthus L'Heritier (Agapanthaceae). Plant Systematics
and Evolution 241: 115-123.
G.D. DUNCAN*
* South African National Biodiversity Institute, Kirstenbosch National
Botanical Garden, Private Bag XI, 7735 Claremont, Cape Town.
MS. received : 2004-10-01.
OROBANCHACEAE
A NEW SPECIES OF HARVEYA FROM WESTERN CAPE. SOUTH AFRICA
The hemi- or holoparasitic genera of Scrophulari-
aceae sensu lato, in which the upper lip of the corolla is
fully or partially enclosed by the lower in bud, are now
recognized to comprise the family Orobanchaceae
(Angiosperm Phylogeny Group 1998). Among the four
holoparasitic genera of Orobanchaceae that are regarded
as native in southern Africa, the genus Harveya Hook, is
distinguished by its large, brightly coloured and often
attractive flowers with a 5-lobed, ± bilabiate corolla. The
well-developed corolla tube is straight or slightly curved
and the anthers, with one exception, are 2-thecous with
one theca reduced in size and sterile. Currently 40
species of Harveya are recognized from Africa and the
Mascarenes (Smithies 2000), with ± 25 species from
southern Africa, but the genus is urgently in need of revi-
sion. The South African species were last treated by
Hiem (1904), who recognized 21 species from the sub-
continent, 14 of them from the Cape Floral Region. The
description of three additional species from KwaZulu-
Natal by Hilliard & Burtt (1986) increased the number to
the current total. It was evident to Goldblatt & Manning
(2000), however, that several of the species from the
Cape Floral Region were of dubious status. In their
account of the genus they admitted just nine species from
the region. This treatment has recently been supported by
Randle (2004).
Among the undetermined collections at the Compton
Herbarium was a specimen from the Little Karoo, deter-
mined as H. hyobanchoides Schltr. ex Hiem, but the
accompanying colour notes indicated the anomalous
colour combination of white flowers among red floral
bracts. A recent colour photograph from a second popu-
lation of plants sent to the herbarium alerted us to the
likelihood that this taxon represented an undescribed
species and an expedition was mounted to the original
locality to collect fresh material. This confirmed that the
plant was distinct and we name it Harveya roseoalba.
Harveya roseoalba J.C. Manning & Goldblatt, sp.
nov.
TYPE. — Western Cape, 3321 (Ladismith): Gamkaberg
Nature Reserve, (-BD), stony slopes at 910 m, 24 Sep-
tember 2004, P. Goldblatt & J.H.J. Vlok 12547 (NBG,
holo.; E, K, MO, PRE, iso.).
Plantae holoparasiticae 100-180 mm altae, partibus
omnibus, praeter corollam, roseis ad carmineis, dense
viscideo-villosis vel lanatis, inflorescentia spica conges-
ta vel racemus spiciformis, bracteis obovato-rhom-
boideis 15-25 x 10-15 mm, adaxialiter glabris, calyce
90
Bothalia 35,1 (2005)
FIGURE 12. — Harveya roseoalba , Goldblatt & Vlok 12547. A, whole plant. B-D, flower: B, front view; C, side view; D, 1/s. E, calyces from two
different flowers laid out; F, corolla laid out; G, anther; H, stigma. Scale bars: A-F, 10 mm; G, H, I mm. Artist: John Manning.
campanulato, 25-30 x ± 10 mm, glabrescenti ad basem,
adaxialiter glabrato, ± aequilobato usque ad mediium,
sepalis deltoideis ad linearo-lanceolatis vel bilabiatis,
sepalo posteriore longissimo; tubo corollae pallide viri-
de, anguste et oblique campanulato, 38^10 mm longo,
circa 10 mm supra basin constricto, interne glabrato
praeter annulum setarum ad partem constrictum, petalis
5(6-7) albis sub anthesi, postea erubescentibus, subae-
qualibus, patentibus, suborbicularibus, 5-6 x 5-6 mm;
staminibus subdidynamis, in quarta parte inferiore tubi
corollae ad annulum setarum insertis, antheris glabratis
theca fertili una ±.2.5 mm longa, subulata, ad basem
uncinata et theca sterili ad lobum perparvum deminuta;
stigmate valde decurvato, oblongo, 5 x 1 mm.
Plants holoparasitic, 100-1 80 mm high, all parts deep
pink to carmine red except corolla, densely viscid-villous
or -lanate. Stem fleshy, simple, 15-25 mm diam.; caudal
scales imbricate, reduced below, grading into floral
bracts above, concave, lower scales transversely obo-
vate, upper scales obovate, 5-15 x 5-9 mm. Inflores-
cence a dense spike or spike-like raceme of 25-50 flow-
ers, all subsessile or lowermost shortly pedicellate with
pedicels up to 5 mm long; bracts sessile, adpressed, obo-
vate-rhomboidal, concave, 15-25 x 10-15 mm, fleshy,
densely villous abaxially, glabrous adaxially; bracteoles
arising from base of calyx or on pedicel if present, oppo-
site, linear-oblanceolate, 20-25 x 2-4 mm, canaliculate,
densely villous abaxially, glabrous adaxially. Calyx cam-
panulate, 25-30 x ± 10 mm, densely villous adaxially
above but glabrescent towards base, glabrous abaxially,
variously lobed, varying from ± equally lobed for up to
halfway with all lobes deltoid or lanceolate to linear-
lanceolate, 7-10 x 2-4 mm, or bilabiate with anterior lip
comprising four lobes divided for quarter to halfway and
a single posterior lobe divided from anterior lip for three
Bothalia 35,1 (2005)
91
FIGURE 13. — Distribution of Harveya roseoalba.
quarters. Corolla : tube pale greenish, narrowly and oblique-
ly funnel-shaped, ± straight or slightly curved at mouth,
38-40 mm long dorsally and 25-30 mm long ventrally,
5-6 mm diam. at base, constricted ± 10 mm from base
and then gradually expanded to apex, mouth oblique,
8-10 mm diam., pubescent in upper half but glabrescent
below, glabrous internally with exception of dense ring
of stiff hairs encircling tube at point of constriction; limb
white in bud and at anthesis but later turning pink, 5(6 or
7)-lobed, usually bilabiate with 2-lobed upper lip and 3-
lobed lower lip; lobes subequal, patent, suborbicular, 5-6
x 5-6 mm, glandular-hairy on both sides with shorter hairs
on adaxial surface. Stamens arising in lower quarter of
corolla tube at constriction among ring of hairs, subdidy-
namous, shorter anterior pair ± 20 mm long, posterior pair
exceeding anterior by ± 1 mm, exserted for up to 4 mm;
filaments linear, glabrous; anthers glabrous, 1-thecous
with sterile theca reduced to minute deltoid protrusion,
fertile theca subulate-fusiform, hooked at apex and dehisc-
ing from apical suture, ± 2.5 mm long. Ovary subglobose,
± 4 mm diam., glabrous; style exserted, ± 25 mm long,
strongly decurved apically, glabrous; stigma oblong to
slightly clavate, 5 x 1 mm. Fruit and seeds unknown.
Flowering time : September to October. Figure 12.
Distribution and ecology: Harveya roseoalba appears to
be a local endemic of the Gamkaberg and adjacent Atta-
kwasberg in the central Little Karoo, south of Oudtshoom
(Figure 13). Recorded hosts are P hylic a lanigera (Rham-
naceae) and Hermannia velutina (Malvaceae).
Diagnosis and relationships: this distinctive species is
readily recognized by its bicoloured, spiciform inflores-
cence in which the white or cream-coloured corollas of
the fresh flowers contrast strongly with the deep pink to
carmine bracts and calyces. With age the flowers turn
pink and then blend in with the rest of the inflorescence.
In general appearance the dense, reddish spike of H.
roseoalba recalls the genus Hyobanche L., with which it
also shares a 1-lobed anterior calyx lip and 1-thecous
anthers. In Hyobanche, however, the corolla is ± 3-lobed,
and there is no doubt that H. roseoalba is appropriately
placed in Harveya.
Harveya roseoalba is most similar to H. hyoban-
choides , with which it shares a dense, spiciform inflores-
cence bearing narrowly funnel-shaped flowers with rela-
tively small corolla lobes, stamens that arise in the lower
quarter of the tube from a ring of stiff hairs, and an
oblong stigma. In addition, the 1-lobed anterior calyx lip
of H. hyobanchoides is sometimes evident in H. roseoal-
ba, although the latter species is rather variable in the
division of the calyx. Harveya roseoalba is also variable
in the number of corolla lobes, which may number 6 or 7
in the upper flowers of some spikes. The two species are
well-separated geographically, with H. hyobanchoides
restricted to the southeastern seaboard, from Humans-
dorp to East London (Goldblatt & Manning 2000).
Harveya roseoalba is readily distinguished from H.
hyobanchoides by its white, rather than yellow to yellow-
green corolla, completely glabrous adaxial surfaces of the
bracts, and by the truly 1-thecous anthers in which the
second theca is completely absent and represented mere-
ly by a minute bump on the connective. The only other
species in the genus with 1-thecous anthers, H. speciosa
Bemh., is a very different plant from the Drakensberg of
KwaZulu-Natal and Eastern Cape, in which the sterile
theca has apparently been independently lost.
Other material examined
WESTERN CAPE. — 3321 (Ladismith): Gamka Nature Reserve,
(-DA), eastern slopes in clay soil, 14 September 1981, J.H.J. Vlok 312
(NBG); mountain west of Fouriesberg near Cloete’s Pass, (-DD),
September 2003, A. Mohl s.n. NBG195623 (photo.).
ACKNOWLEGEMENTS
Support for this study was provided by grant 7316-02
from the US National Geographic Society. Collecting per-
mits were provided by the Nature Conservation authorities
of Western Cape, South Africa. We thank Lendon Porter
and Jan Vlok for their assistance in the field.
REFERENCES
ANGIOSPERM PHYLOGENY GROUP. 1998. An ordinal classifica-
tion for the families of flowering plants. Annals of the Missouri
Botanical Garden 85: 531-553.
GOLDBLATT, P. & MANNING. J.C.. 2000. Cape plants. A conspec-
tus of the Cape flora. Strelitzia 9. National Botanical Institute,
Cape Town & Missouri Botanical Garden.
HIERN, W.P 1904. Scrophulariaceae. In W.T. Thiselton-Dyer, Flora
capensis 4,2: 1 2 1 — 420. Reeve, Covent Garden.
HILLIARD, O.M. & BURTT, B.L. 1986. Notes on some plants of
southern Africa chiefly from Natal: XII. Notes from the Royal
Botanic Garden Edinburgh 43: 345—405.
RANDLE, C.P. 2004. The evolution and expression of rbch in holo-
parasitic sister genera Harveya Hook, and Hyobanche L. (Oro-
banchaceae) and systematics and taxonomic revision of south-
ern African species of Harveya. Ph.D. dissertation. The Ohio
State University, Ohio.
SMITHIES, S.J. 2000. Scrophulariaceae. In O.A. Leistner, Seed plants
of southern Africa: families and genera. Strelitzia 10: 508-537.
National Botanical Institute, Pretoria.
J.C. MANNING* and P. GOLDBLATT**
* Compton Herbarium. South African National Biodiversity Institute,
Private Bag X7, 7735 Claremont, Cape Town.
** B .A. Krukoff Curator of African Botany, Missouri Botanical Garden,
PO. Box 299, St. Louis. Missouri 63166, USA.
MS. received: 2004-10-04.
92
Bothalia 35,1 (2005)
SALICACEAE
SALIX : THE CORRECT APPLICATION OF THE NAME SALIX MUCRONATA , AND A NEW COMBINATION
Because of its heterophylly, Salix in southern Africa
has been subject to considerable name changes and dis-
similar species concepts. Thunberg (1794, 1807) collect-
ed and described four species of Salix from the Cape, S.
aegyptiaca, S. mucronata , S. capensis and S. hirsuta.
Burtt Davy (1922) sank S. aegyptiaca into S. mucronata
but recognized 10 species and varieties in South Africa.
Thunberg provided no locality data for the type of S.
mucronata but the type of S. aegyptiaca was collected
from Roodesand in rthe Tulbagh Division. Burtt Davy
concluded that the type of S. mucronata also came from
the vicinity of Tulbagh.
Immelman (1987) recognized only one species, S.
mucronata with five subspecies. Jordaan (2002a, b) fur-
ther reduced this number to four subspecies, sinking
subsp. wilmsii into subsp. woodii. From Jordaan (2002a,
b), we concluded that there is no meaningful difference
between S. mucronata subsp. mucronata and subsp. capen-
sis. Since S. mucronata is the oldest name for the taxon
occurring in the southern most parts of southern Africa,
this name has priority.
Salix subserrata is the present accepted name for the
widespread tropical taxon occurring from Arabia to pos-
sibly the Cape (Wilmot-Dear 1991). Burtt Davy recog-
nized this as a tropical African species under the later
name of S. safsaf Willd. Both Immelman and Jordaan
reduced S. subserrata to synonymy under subsp. mucro-
nata. Furthermore, Jordaan treated S. mucronata subsp.
mucronata as a widespread tropical taxon, which, in the
Flora of southern Africa region only occurs in the north-
ern parts of Namibia and Botswana, but neither of these
two authors discuss their reasons for sinking S. subser-
rata. However, Wilmot-Dear (1991) stated that the two
species, S. subserrata and S. mucronata are very similar
and may yet prove to be conspecific. Burtt Davy (1922)
and Wilmot-Dear (1991) stated that S. woodii intergrades
with typical S. subserrata in Zimbabwe.
Thus numerous authors in various ways imply that S.
mucronata is related to or cannot be kept specifically dis-
tinct from S. subserrata. It would therefore be prudent to
maintain S. subserrata as a subspecies under S. mucrona-
ta until a detailed study of the variation of the species in
the whole of southern and tropical Africa can be under-
taken. For a full description of the taxon S. mucronata,
see Jordaan (2005).
The following new combination and synonymy is
effected here.
Salix mucronata Thunb. subsp. mucronata.
Immelman: 173 (1987) p.p. (excluding tropical ele-
ments). Type: Thunberg s.n., sheet annotated 'Salix
mucronata 9’ ( UPS23065 , lecto.; IDC microfiche
1063/968!).
Salix mucronata subsp. capensis (Thunb.) Immelman: 173 (1987);
Jordaan: 255 (2002a); 122 (2002b), syn. nov. Type: Cape Province,
near rivers in mountains near Hantam, Thunberg s.n. ( UPS22958 , lecto.
designated by Jordaan (2005); IDC microfiche 1063/963!).
Salix mucronata Thunb. subsp. subserrata
(Willd.) R.H. Archer & Jordaan, stat. et comb. nov.
Salix subserrata Willd. in Species plantarum 4: 671 (1806).
Wilmot-Dear: 1 (1985); 121 (1991). Type: Egypt, near Cairo, Bulak,
Herb. Willd. 18137 (B-WILLD, holo.; IDC microfiche 7440-30/1313!).
REFERENCES
BURTT DAVY, J. 1922. The distribution and origin of Salix in South
Africa. Journal of Ecology 10: 62-86.
IMMELMAN, K.L. 1987. Synopsis of the genus Salix (Salicaceae) in
southern Africa. Bothalia 17: 171-177.
JORDAAN, M. 2002a. Abstract: The infraspecific classification of
Salix mucronata Thunb. (Salicaceae) in southern Africa. South
African Journal of Botany 68: 255.
JORDAAN, M. 2002b. Salix. In M. Coates Palgrave, Keith Coates
Palgrave Trees of southern Africa, edn 3: 121, 122. Struik,
Cape Town.
JORDAAN, M. 2005. FSA contributions 18: Salicaceae s. str. Bothalia
35: 7-20.
THUNBERG, C.P. 1794. Prodromus plantarum capensium, Uppsala.
THUNBERG, C.P. 1807. Flora capensis. Cotta, Stuttgardt.
WILLDENOW, C.L. VON. 1806. Species plantarum, vol. 4. Nauk.
Berlin.
WILMOT-DEAR, C.M. 1985. Salicaceae. Flora of tropical East Africa
Salicaceae: 1-7.
WILMOT-DEAR, C.M. 1991. Salicaceae. Flora zambesiaca 9, 6: 120-124.
R.H. ARCHER* & M. JORDAAN*
* South African National Biodiversity Institute, Private Bag X101,
0001 Pretoria.
MS. received: 2004-12-08.
Bothalia 35,1:93-99 (2005)
The systematic value of the leaf indumentum in Lobostemon (Bora-
ginaceae)
M.H. BUYS*
Keywords: Boraginaceae, indumentum, leaves, Lobostemon Lehm., trichomes
ABSTRACT
The character states pertinent to the indumentum of Lobostemon Lehm. leaves are tabulated and discussed. Lobostemon
possesses similar trichome and indumentum types as described for Echium L., with the exception of the ‘Alpine’ indumen-
tum type. Due to the environmentally induced variation, the leaf indumentum characters of Lobostemon are viewed to be
of limited phylogenetic value, although they do diagnose a number of taxa.
INTRODUCTION
In terms of Lobostemon Lehm. systematics, the most
recent revision by Levyns is noteworthy because she
delimits five sections based on floral characters and pre-
sents a branching diagram (Levyns 1934: 412) to eluci-
date relationships within the genus. The aforementioned
constitute falsifiable hypotheses.
One of the most striking features of the herbaceous
forms of the Boraginaceae is the covering of thick-
walled, harsh, unicellular trichomes (Metcalfe & Chalk
1950: 945), so much so that Klotz (1959) was of the
opinion that one cannot ignore these characters when
studying the Boraginaceae. Numerous studies of the leaf
indumentum in the Boraginaceae have been undertaken
e.g. Revedin (1902), Jonova (1926), Bider (1935), Klotz
(1959), Lems & Holzapfel (1968), Bramwell (1972) and
Selvi & Bigazzi (2001). Yet, opinions differ with regard
to the taxonomic value of these characters. Because the
largely European Echium L. is considered to be a sister
taxon of the South African endemic Lobostemon (Bohle
etal. 2001), the focus henceforth will for the moment fall
on Echium. Klotz (1959), during his revision of the
genus, used indumentum and trichome characters to key
out species, series and sections. Lems & Holzapfel’s
(1968) in depth study of the genus on the Canary Islands,
leads to an optimistic view about their use, claiming that
there are at least 14 criteria which can be applied in com-
paring the Canary Island Echium species with one anoth-
er, including not only the types of trichomes present, but
their relative abundance, their total coverage of the leaf
surface, the size of the pustules, and the orientation of the
trichomes. Gibbs (1971: 31, 32), confining himself to the
Spanish echiums, formed three groups based on indu-
mentum characters and used a number of characters to
key out species. Bramwell (1972), working in Macaro-
nesia, chose to differ and suggested that the indumen-
tum: ‘is more or less useless and certainly of secondary
importance ... in the consideration of the evolutionary
and ... the phenetic relationships’.
* A.R Goossens Herbarium, School of Environmental Sciences & De-
velopment: Botany Div., North-West University, 2520 Potchefstroom,
South Africa.
MS. received: 2003-07-24.
Levyns (1934), for the first time in Lobostemon,
mentioned that indumentum and trichome characters
are particularly influenced by environmental and tem-
poral factors. Failure to realise this earlier led especial-
ly De Candolle (1846) and Wright (1904) to describe a
myriad of names, most of which were placed in syn-
onymy by Levyns (1934: 403). Levyns (1934), in cer-
tain instances, still made use of vegetative characters in
her key, e.g. in the section Trichotomi Levyns. Dif-
ficulty in identifying species using this key invariably
occurs when using vegetative characters like indumen-
tum or trichome type.
While it is true that like all taxonomic criteria, epider-
mal characters must be interpreted with great circumspec-
tion, Barthlott (1981) voiced the opinion that the major
problem in their systematic application is that we do not yet
have enough data [compare also Cole & Behnke (1975)].
Despite the vast amounts of SEM micrographs published,
many of the data are not comparable because of a lack of
standardized terminology and often no structural interpre-
tation of the characters illustrated.
In the light of the aforementioned, this paper aims to
contribute to epidermal related data in the Boraginaceae.
Existing terminology is followed as far as possible, and
for this reason is expanded in Material and methods
below. Two issues are investigated, namely: 1, can SEM
analyses of indumentum characters reveal states that can
diagnose Lobostemon species, and 2, does the grouping
of taxa based on these characters correlate with existing
hypotheses of infrageneric relationships within the genus?
MATERIAL AND METHODS
Voucher specimens were collected for all taxa exam-
ined and are listed in Table 1 .
Material collected from plants in their natural habitat
in the field was fixed in FAA. The indumenta of the fixed
leaves were studied with a JOEL scanning electron micro-
scope (SEM), using secondary-electron detection and an
acceleration voltage of 4—5 kV. All fresh material used in
the SEM study was collected during the peak flowering
season between August and October. Additional observa-
tions were made with a light microscope on herbarium
94
Bothalia 35,1 (2005)
specimens as well as on potted plants. In the latter case,
both young to mature leaves were analysed.
Terminology
Lems & Holzapfel (1968) and Bramwell ( 1972) iden-
tified three basic trichome types in Macaronesian species
of Echium :
Glandular trichomes (Gl) occur mainly on the abaxi-
al surface of the midrib or more rarely over wide areas of
the abaxial surface of the leaf in most of the Macaro-
nesian species of Echium. Selvi & Bigazzi (2001 ) distin-
guished two types of glandular trichomes in the genera of
the tribe Boragineae: Type 6 trichomes consisting of a
single stalk cell, and Type 7 trichomes consisting of three
or more stalk cells.
Simple trichomes (Si) correspond to the Type 3 tri-
chomes of Selvi & Bigazzi (2001). They are mostly short,
stiff, hollow trichomes without a large pustular base, and
are present in the cotyledons of all the Echium species.
However, they do not persist into the adult stage in all
species. Simple trichomes occurring in mature leaves are
usually short, often curved and closely appressed to the leaf
surface [ the appressed trichomes correspond to the Type 2
trichomes of Selvi & Bigazzi (2001)]. Though the base of
the trichome is usually swollen, there is generally little or
no cell differentiation of the surrounding epidermal cells.
Lems & Holzapfel (1968) record simple trichomes longer
than 400 mm always possessing one or more rows of dif-
ferentiated cells around the base. It is well established that
the number of rows of epidermal cells involved in the pus-
tule formation in the Boraginaceae is a product of environ-
mental and temporal variables.
Pustular trichomes (Pu), a term coined by Lems &
Holzapfel (1968), correspond to the Hiigelborsten of Bider
(1935), nodular bristles of Metcalfe & Chalk (1950),
Hockerborsten of Klotz (1959) and Type 1 trichomes of
Selvi & Bigazzi (2001). One or more concentric rings of
strongly differentiated epidermal cells which contain
cystolith-like structures and whose walls are strongly im-
pregnated with calcareous material surround the bases of
the hollow, pustular trichomes. In very large trichomes,
some of the cells of the upper palisade layer of the meso-
phyll are also involved in the pustule structure and may
also be calcified. Uphof (1962) referred to these subepi-
dermal areas as pedestals.
Lems & Holzapfel (1968) furthermore discerned four
main indumentum types in the Boraginaceae:
Spinous indumentum (Sp) consists of stiff spines and
is found on the leaf surface or is often confined to the
margins and the midrib of the leaf;
Appressed to ascending silky indumentum (Ap). Here
the leaf surface is covered with a dense layer of
appressed trichomes that are either simple or with small
basal cells;
Umbonate indumentum (Um), spinous, consisting of
relatively sparsely distributed pustular trichomes with a
large, round, basal region on the otherwise glabrous leaf
surface. I designate the term umbonate to describe this
type of indumentum. Klotz (1959) referred to these as
TABLE If. — Distribution of indumentum types and trichome characteristics in Lobostemon (with voucher specimens housed in NBG)
MHB, M.H. Buys; Sp, spinous; Ap, appressed; Um, umbonate; Gl, glandular; Si, simple; Pu, pustular; 1, homomorphic or with a continuous
variation; 2, dimorphic; Ad, adaxial; Ab, abaxial; No, without papillae; Sm, smooth papillae; Un, undulate papillae.
Bothalia 35,1 (2005)
95
FIGURE 1. — Lobostemon regulareflorus : t/s through leaf showing
involvement of underlying parenchyma in pustule formation.
Scale bar: 50 pm.
diskusculig bases. Selvi & Bigazzi (2001) referred to the
trichomes forming this indumentum type as Type 4 tri-
chomes;
Alpine indumentum (Al), dense, ascending to erect,
with long trichomes with small bases, found only in sub-
alpine zone species.
Trichome and indumentum types were identified
according to those recognized by Lems & Holzapfel
(1968). All observations of indumentum presence, abun-
dance and type were confined to the surface of the lamina
(excluding margins). Leaves were considered to be
glabrous when no sign of any trichomes could be seen
(including on the midrib) with a light microscope or SEM.
Transverse sections of paraffin wax-embedded lami-
nae were cut with a rotary microtome and stained with a
mixture of Safranine O and Alcian green (Joel 1983).
Sections were taken through the middle of the laminae.
The cluster analysis was done using the Statistica 6. 1
package with the following settings: tree clustering;
Ward’s (1963) method of minimum-variance clustering
under the amalgamation rule and percentage disagree-
ment as a measure of distance. Characters were coded as
qualitative presence/absence data. Data for the spinous
indumentum type and glandular trichome type were ex-
cluded from the analysis due to the presence of unknown/
uncertain states in one or more taxa.
RESULTS
Table 1 presents a summary of the various leaf indumen-
tum and trichome characters codified for Lobostemon.
Trichomes
In Lobostemon, the tendency for epidermal cells at the
base of the trichomes to become organized into pustules
extends sometimes to several concentric rows, the num-
ber of which seem to vary with the climatic conditions
under which the plant grew and the stage of ontogenetic
development. Trichomes typically longer than 400 mm
tend to develop pustular bases. In trichomes with a large
base, the underlying parenchyma also becomes involved
in pustule formation (Figure 1). The pustular trichome
type occurs in all Lobostemon species (Figure 2). Most
of the species in Lobostemon possess glandular tri-
FIGURE 2. — Indumentum types in Lobostemon leaves. A, spinous in L. belliformis', B, appressed in L. montanus ; C, appressed in L. trichotomies', D,
umbonate in L. paniculatus. Scale bars: 100 pm.
96
Bothalia 35,1 (2005)
Micropapillae
Only a limited number of species seem to have tri-
chomes without micropapillae (Figure 5A). In the major-
ity of taxa, trichomes examined in young leaves as well
as flower buds display micropapillae. Micropapillae may
either be smooth (Figure 5B) or undulated (Figure 5C).
In terms of shape, both round and elongate micropapillae
have been observed on single trichomes (Figure 5D).
The round micropapillae in the aforementioned figure
are confined to the base of the trichome, whereas the
elongate micropapillae occupy the distal parts. L. decorus
Levyns and L. marlothii Levyns have been observed to
possess trichomes with or without micropapillae.
FIGURE 3. — Collapsed glandular trichomes in Lobostemon fruticosus
indicated by arrow. Scale bar: 100 pm.
chomes on their leaf surfaces. These trichomes have a
tendency to fall flat from an early age (Figure 3).
Length
With regards to the leaves, only four species of Lobo-
stemon, namely L. argenteus (P.J.Bergius) H.Buek, L. fru-
ticosus (L.) H.Buek, L. paniculatus (Thunb.) H.Buek and
L. stachydeus A. DC. clearly possess a dimorphic indumen-
tum in the adult stage (Figure 4A). The shorter trichome is
usually of the simple trichome type. Some forms of L.
echioides Lehm., L. gracilis Levyns, L. trichotomus (Thunb.)
A. DC. and L. paniculiformis A.DC. possess a heteromor-
phous indumentum. Here the variation in trichome length
is discerned to be continuous (Figure 4B).
Distribution
Species with trichomes on both the abaxial and adaxial
leaf surfaces are the most common. L. regulareflorus (Ker
Gawl.) M.H.Buys is unique in only its adaxial leaf surface
being hairy. L. collinus C.H.Wright is a good example of
how climate or age can influence trichome distribution.
Plants collected in spring generally have both sides of the
leaf hairy. Those collected in late summer (January-April)
have glabrous adaxial surfaces. In L. capitatus (L.)
H.Buek, however, the opposite holds true. Young leaves
collected in spring appear to be hairy only on the abaxial
surface. Older leaves from the previous year’s growth are
hairy on both the adaxial and abaxial surfaces.
Indumentum
Three main indumentum types can be recognized in the
genus. The spinous indumentum type (Figure 2A) is the
most prevalent. The appressed indumentum type (Figure
2B) is commonly found on those taxa exhibiting silvery
leaves. Forms of L. argenteus, L. curvifolius H.Buek, L.
echioides, L. fruticosus and L. trigonus (Thunb.) H.Buek
that grow in less arid and more sheltered conditions tend to
have appressed trichomes. Generally appressed indumenta
also become more spinous as the season progresses, i.e. as
it becomes drier and wanner. L. trichotomus and L. gracilis
appear to be the only species with an appressed indumen-
tum without an accompanying complement of spinous
indumentum (Figure 2C). The umbonate indumentum type
is the most difficult to identify due to the general absence
of calcified cells around the trichome bases (Figure 2D).
This indumentum type appears to be absent from sections
Argentei Levyns and Fruticosi Levyns, but most prevalent
in section Trichotomi Levyns. Only L. regulareflorus and
L. sanguineus Schltr. appear to possess an umbonate indu-
mentum type to the exclusion of a spinous indumentum on
the leaf surfaces. The remainder of species generally have
a spinous indumentum on the margins in addition to the
umbonate indumentum of both leaf surfaces. Although the
leaves of L. glaucophyllus (Jacq.) H.Buek appear to be
glabrous to the naked eye, this study has shown the preva-
lence of minute umbonate indumentum on especially the
abaxial surface (Figure 6).
A cluster analysis of the data in Table 1 (excluding char-
acters with unknown/uncertain states) created a number of
clusters with taxa possessing identical character states viz.:
FIGURE 4. Trichome lengths in Lobostemon. A, dimorphous in L. argenteus', B, continuous in L. echioides. Scale bars: 100 pm.
Bothalia 35.1 (2005)
97
FIGURE 5. — Trichome micropapillae in Lobostemon. A, no micropapillae in L. muiriv, B, round and smooth in L. paniculatus ; C, round and undu-
late in L. belliformis; D, round and elongate smooth in L curshfolius. Scale bars: 100 pm.
L. argenteus and L. fniticosus cluster; L. capitatus, L. colli-
nus, L. lucidus (Lehm.) H.Buek and L. strigosus (Lehm.)
H.Buek cluster; L. curvifolius, L. montanus H.Buek and L.
trigonus cluster; L. echioides, L. gracilis and L. trichoto-
mus cluster; L. glaucophyllus , L. hottentoticus Levyns, L.
laevigatus (L.) H.Buek, L. paniculiformis A.DC. and L.
pearsonii Levyns cluster; L. muirii Levyns and L. oederi-
aefolius A.DC. cluster (Figure 7).
DISCUSSION
I alluded above to Barthlott’s (1981) reasons as to
why indumentum and trichome characters have not been
applied to systematics with great success. Following this
study, and those by Levyns (1934) and Bramwell (1972),
FIGURE 6. — Minute umbonate indumentum on abaxial leaf surface of
Lobostemon glaucophyllus. Scale bar: 100 pm.
it seems clear that in the absence of data to the contrary,
a lack of knowledge concerning the influence of habitat
and ontogenetic development probably more than any-
thing else places a damper on the use of indumentum and
trichome data for systematic studies in the Boraginaceae.
However, even though there is uncertainty about their
systematic use, and grouping of plants based on epider-
mal features does not agree with those based on repro-
ductive features, the need remains to provisionally
describe and organize into a system, all data whereby
relationships between the various patterns can be mean-
ingfully approached in the future (Klucking 1995).
This study has revealed that trichomes in Lobostemon
should not be defined as purely epidermal but rather as
emergences. In Lobostemon, young plants possess
straight, simple, unicellular trichomes (Table 1: Si)
whose swollen bases are part of the epidermal layer.
From this simple type, present on juvenile leaves, differ-
ent developments may occur both in the ontogeny and in
the transition from juvenile to adult foliage.
Most of the species in Lobostemon possess glandular
trichomes on their leaf surfaces. By contrast, Klotz
(1959) in his revision of the genus Echium , found the
chief occurrence of glandular trichomes to be on the
stems of E. humile Desf. and E. trygorrhizum Pomel,
although E. gaditanum Boiss. was observed with glandu-
lar trichomes on their leaf surfaces.
Levyns (1934) correctly indicated that Lobostemon
tends to undergo a second growing period later in the
98
Bothalia 35,1 (2005)
L. argenteus
L. fruticosus
L. echioides
L. gracilis
L. trichotomus
L. daltonii
L. stachydeus
L. paniculatus
L. glaucophyllus
L. hottentoticus
L. laevigatus
L. paniculiformis
L. pearsonii
L. belliformis
L. capitatus
L. collinus
L. lucidus
L. strigosus
L. glaber
L. curvifolius
L. montanus
L. trigonus
L. decorus
L. marlothii
L. muirii
L. oederiaefolius
L. regnlarefloras
L. sanguineus
0.0 0.5 1.0 1.5 2.0
Linkage Distance
2.5
FIGURE 7. — A phenogram of Lobostemon based on indumentum characters employing Ward’s method of minimum-variance clustering and per-
centage disagreement as a measure of distance.
season that results in the indumentum becoming sparser
as the leaf size increases. Moreover, this study has shown
that trichomes tend to fall or break off in some species of
Lobostemon , leaving behind their hardened bases and
manifesting a sparser coverage.
Some taxa can be diagnosed on a single or a combi-
nation of indumentum characters. The possession of a
dimorphic indumentum in Lobostemon is confined to L.
argenteus , L. stachydeus , L. paniculatus and L. frutico-
sus. L. curvifolius and L. fruticosus have often been mis-
taken for each other: the two taxa are distinguishable —
the dimorphic indumentum being absent in L. curvi-
folius. L. regulareflorus and L. belliformis M.H.Buys,
although morphologically similar, differ in that L. belli-
formis possesses trichomes on both leaf surfaces where-
as L. regulareflorus possesses trichomes confined to the
adaxial surface.
Keeping the shortcomings of phenetics in mind, the
aforementioned cluster analysis created groups that
could not be correlated to current sectional divisions
sensu Levyns (1934: 412). Some clusters, however, can
be found as subgroups within Levyns’ sections e.g. the L.
decorus , L. marlothii, L. muirii and L. oederiaefolius
cluster and to an extent the L. glaucophyllus, L. hotten-
toticus Levyns, L. laevigatus and L. paniculiformis clus-
ter. The L. echioides , L. gracilis and L. trichotomus clus-
ter, although not all members of the same section, repre-
sent the basal taxa in Levyns’ branching diagram.
REFERENCES
BARTHLOTT, W. 1981 . Epidermal and seed surface characters in plants:
systematic applicability and some evolutionary aspects. Nordic
Journal of Botany 1: 345-355.
BIDER, J. 1935. Beitrage zur Pharmakognosie der Boraginaceen und
Verbenaceen. Vergleichende Anatomie des Laubblattes. Bider, Basel.
BOHLE, U.-R„ HILGER, H.H. & MARTIN, W.F. 2001 . Island coloniza-
tion and evolution of the insular woody habit in Echium L. (Bora-
ginaceae). Proceedings of the National Academy of Science 93:
11740-11745.
BRAMWELL, D. 1972. A revision of the genus Echium in Macaronesia.
Lagascalia 2: 37-115.
COLE, G.T. & BENKE, H.-D. 1975. Electron microscopy and plant
systematics. Taxon 24: 3-15.
DE CANDOLLE, A.R 1846. Boragineae. Prodromus systematis natu-
ralis regni vegetablis 10: 1-23. Treuttel & Wiirtz, Paris.
GIBBS, P.E. 1971. Taxonomic studies on the genus Echium. 1. An out-
line revision of the Spanish species. Lagascalia 1: 27-82.
JOEL, D.M. 1983. AGS (Alcian Green Safranin), a simple differential
staining of plant material for the light microscope. Proceedings
of the Royal Microscopical Society 18: 149-151.
JONOVA, M. 1926. Anatomie a morfologie trichomu u Borraginacei s
ohledem na systematiku teto celedi. Vestnik Kralovske ceske
spolecnosti nduk.Trtda mathematicko-pn'rodovedeckd 2: 1-66.
KLOTZ, G. 1959. Die systematische Gliederung der Gattung Echium
L., ein Beitrag zum Problem der Gliederung von Pflanzen-
sippen. Habilitation, Martin Luther University.
KLUCKING, E.P. 1995. Leaf venation patterns. The classification of
leaf venation patterns, vol. 7. Cramer, Berlin.
LEMS, K. & HOLZAPFEL, C.M. 1968. Evolution in the Canary Islands.
I. Phylogenetic relations in the genus Echium (Boraginaceae) as
shown by trichome development. Botanical Gazette 129:
95-107.
LEVYNS, M.R. 1934. A revision of Lobostemon. Journal of the Lin-
nean Society, Botany 49: 393-45 1 .
Bothalia 35,1 (2005)
99
METCALFE, C.R. & CHALK, L. 1950. Anatomy of the dicotyledons.
Clarendon Press, Oxford.
REVEDIN, P. 1902. Studio Sopra I Peli delle Borraginacee. Nuovo
Giomale Botanico Italiano 9: 154, 155.
SELVI, F. & BIGAZZI, M. 2001. Leaf surface and anatomy in Bora-
ginaceae tribe Boragineae with respect to ecology and taxono-
my. Flora 196: 269-285.
UPHOF, J.C.T. 1962. Plant hairs. Gebriider Bomtraeger, Berlin-Nikolassee.
WARD, J.H. 1963. Hierarchical grouping to optimize an objective func-
tion. Journal of the American Statistical Association 58: 236.
WRIGHT, C.H. 1904. Boragineae. In W.T. Thiselton-Dyer, Flora capen-
sis 4, 2: 2-45. Reeve, London.
Bothalia 35,1: 101-108(2005)
Ovule-to-seed development in Dovyalis caffra (Salicaceae: Flacour-
tieae) with notes on the taxonomic significance of the extranucellar
embryo sac
E.M.A. STEYN*, A.E. VAN WYK**f and G.F. SMITH*+
Keywords: Do\yalis caffra (Hook.f. & Harv.) Hook.f., embryology, Flacourtiaceae, Flacourtieae, Salicaceae, Salix L., seed coat, taxonomy
ABSTRACT
Dovyalis caffra (Hook.f, & Harv.) Hook.f. is a widespread and horticulturally important southern African endemic. Here
described, ovule-to-seed development represents the first embryological information on this genus of the tribe Flacourtieae,
Salicaceae sensu lato. Results are discussed in the light of data available on the embryology of the order Malpighiales in
general and on the tribe Flacourtieae in particular. It is clear that Dovyalis E.Mey. ex Am. shares many characters with other
members of the Flacourtieae. Ovules of D. caffra are anatropous, bitegmic and crassinucellate with both bisporic Allium-
and monosporic Polygonum- type embryo sacs. One of the most characteristic embryological features of the tribe is the pres-
ence of an unusual, extranucellar embryo sac that stretches halfway up into the nricropyle. Since this extraordinary-shaped
embryo sac also defines the embryo sacs of Salix L. and Populus L., it is suggested that the character provides strong sup-
port for the proposed close phylogenetic link between tribes Flacourtieae and Saliceae of Salicaceae sensu lato.
INTRODUCTION
Dovyalis E.Mey. ex Am. is a relatively small, mostly
African genus comprising about 15 species of shrubs or
relatively small trees, often thorny, with simple, alternate
and exstipulate leaves (Sleumer 1972). All species are
dioecious or rarely polygamous (Langenegger 1976)
with a tendency towards andromonoecy, i.e. the male
plants occasionally bear a few bisexual flowers and fruits
(Sleumer 1972). The apetalous flowers are unobtrusive,
greenish, nectariferous and borne singly or in few-flow-
ered fascicles. Fruits are indehiscent, fleshy and edible
berries in various species. Dovyalis is widespread in
tropical and subtropical Africa (excluding Madagascar),
with one species reported from Sri Lanka (Sleumer
1972). Six species occur in the Flora of southern Africa
region (Bredenkamp 2003) where they form a common
constituent of mixed scrub, riverine bush, open wood-
land and forest.
When the genus was first validly described, Amott
(1841) regarded the type species, Dovyalis zizyphoides
E.Mey. ex Am., as closest to Euphorbiaceae with some
resemblance in habit to Flacourtia L’Her. Although
Endlicher (1842) consequently listed Dovyalis as a dubi-
ous genus under his natural order Euphorbiaceae, the
association with Flacourtia and a placement of Dovyalis
in Flacourtiaceae DC. were subsequently endorsed by
most taxonomists during the 19th and 20th centuries (see
Warburg 1893 (as Doryalisf Gilg 1925; Hutchinson
1967; Lemke 1988). However, the beginning of the 21st
century has seen a radical reshuffling of genera tradi-
tionally placed in Flacourtiaceae. Molecular phylogenet-
* South African National Biodiversity Institute (SANBI), Private Bag X101,
0001 Pretoria.
** H.G.W.J. Schweickerdt Herbarium, Department of Botany,
University of Pretoria, 0002 Pretoria.
+ Affiliation: Acocks Chair, Department of Botany, University of Pretoria,
0002 Pretoria.
t Corresponding author, e-mail: braam.vanwyk@up.ac.za
MS. received: 2004-08-31.
ic studies, supported by data from several other botanical
fields, have indicated that most flacourtiaceous genera
fall into two clades, each more closely related to other
families within Malpighiales than to one another (Chase
et al. 2002). Flacourtiaceae were accordingly split into
two groups and the flacourtiaceous genera (not Aphloia
DC.) became part of two recircumscribed families, name-
ly Achariaceae sensu lato and Salicaceae sensu lato.
Consequently, all the southern African flacourtiaceous
genera were placed in Salicaceae sensu lato with the
exception of Kiggelaria L., Rawsonia Harv. & Sond. and
Xylotheca Hochst.; these three genera were included in
Achariaceae sensu lato (Chase et al. 2002).
Recent studies on ovule-to-seed development and
structure in the three southern African genera of the
Achariaceae sensu stricto (Steyn et al. 2001, 2002a, b,
2003) and in Kiggelaria L. (Steyn et al. 2003), offer
unequivocal embryological support for the proposed
close phylogenetic relationship between the herbaceous,
highly modified and monotypic genera of Achariaceae
sensu stricto and the African genus Kiggelaria of the
woody tribe Pangieae (Flacourtiaceae sensu stricto).
These taxa share a number of uncommon embryological
characters, e.g. sessile ovules with zigzag micropyles,
deep-lying embryo sacs covered by an epistase in the
ovule and seed, endotestal-exotegmic protective layers in
the seed coat, suspensorless embryos and sarcotestal
seeds with stomata in the epidermis (Steyn et al. 2003).
On the other hand, available embryological and other
structural support for emending the circumscription of
Salicaceae sensu stricto has been meagre and mainly
rests on stamen and pollen characters (Keating 1973;
Meeuse 1975; Kaul 1995; Bernhard & Endress 1999). Al-
though Meeuse (1975) listed some ovule characters shared
between Salicaceae sensu stricto and Flacourtiaceae sensu
stricto , these characters are not restricted to the two taxa;
they are commonly found among related violalean fami-
lies (Meeuse 1975). Nevertheless, the strong similarity in
embryological characters between Achariaceae sensu
102
Bothalia 35,1 (2005)
stricto and Kiggelaria (Steyn et al. 2003) suggests that
detailed embryological studies of flacourtiaceous species
representing the tribes assigned together with Saliceae
Rchb. to Salicaceae sensu lato might also reveal charac-
ters that are diagnostic for Flacourtiaceae sensu stricto
and Salicaceae sensu stricto. Such embryological char-
acters would be useful for testing hypotheses by molec-
ular phylogeneticists for a close evolutionary relation-
ship between these two families.
Here we present data on ovule-to-seed development
and structure in Dovyalis caffra (Hook.f. & Harv.) Hook.f.,
a representative of the tribe Flacourtieae DC. and a
species that is, like all the other members of Dovyalis ,
embryologically completely unknown. Results are com-
pared with findings recently obtained during a study of
seed structure in the polymorphic African willow, Salix
mucronata Thunb. (Steyn et al. 2004) and with informa-
tion found in classic embryological literature on Salica-
ceae and Flacourtiaceae. Representatives of Flacourtieae
are, according to data collected in multidisciplinary stud-
ies, including molecular phylogeny, possibly the closest
relatives of Salix L. and Populus L., both of the tribe
Saliceae (Chase et al. 2002).
Dovyalis caffra is the only southern African species of
the genus that belongs to the section Aberia Hochst., the
other five species represent section Dovyalis E.Mey. ex
Am. (Sleumer 1972). D. caffra occurs in the easternmost
parts of Western Cape and is widespread in the eastern,
subtropical regions of southern Africa, extending from
Eastern Cape northwards through KwaZulu-Natal, Swazi-
land, Mpumalanga and Limpopo into Zimbabwe. Com-
monly known as the Kei apple, D. caffra is also often cul-
tivated for its edible fruit and, because of its formidable
thorns, frequently used as a security hedge.
MATERIAL AND METHODS
Floral buds, open flowers and developing fruits were
collected from cultivated female trees of Dovyalis caffra
growing in the Pretoria National Botanical Garden and in a
private garden in Murrayfield, Pretoria. All flowering stages
were dissected to facilitate penetration of chemicals into
ovules and seeds and fixed in a 0. 1 M cacodylate buffered
solution containing 4% formaldehyde and 2.5% glutaralde-
hyde. Following the methods of Feder & O’Brien (1968),
material was dehydrated in an alcohol series and impreg-
nated with glycol methacrylate (GMA). All material was
imbedded in GMA, sectioned transversely or longitudinally
at 2-3 pm and subsequently stained with the periodic
acid/Schiff reaction and toluidine blue by using the proto-
cols of O’Brien & McCully (1981 ).
Seed coat terminology follows Corner (1976) as rec-
ommended by Schmid (1986).
RESULTS
Placentation, orientation and development of
megagametophyte
The gynoecium in the apetalous female flowers (Figure
I A) of Dovyalis caffra is usually hexa- or heptacarpellate
and very rarely pentacarpellate. The syncarpous, uniloc-
ular and superior ovary is globose and glabrous (Figure
IB) with six or seven styles, each ending in a bifid stig-
ma. Twelve to fourteen ovules, two per carpel, are borne
on parietal placentae developing approximately at the
same level, ± halfway down the locule (Figure 1A, B).
The ovules occur in pairs on the flanks of the fused
carpel margins. These thick structures extend deeply into
the locule (Figure 1A), limiting the space available to the
developing ovules so that they show some spatial adjust-
ment. Nevertheless, longitudinal sections showed that all
ovule stages are epitropous with a ventral raphe sensu
McClean & Ivimey-Cook (1956: 1392) and anatropous or
about to become anatropous (Figure 1C).
The youngest buds examined contained bitegmic,
crassinucellate young ovules in the process of curving
towards the anatropous position (Figure 1C). The nucel-
lus tissue holds a single megaspore mother cell covered
by one or two parietal cell layers. At this early stage the
ovule seems to be slightly stalked, but when the curva-
ture is completed at about the time meiosis takes place,
the ovule has no funicle (Figure ID).
Meiosis does not seem to result in a linear tetrad of
megaspores. After the first meiotic division, two dyad
cells are formed of which the chalazal cell undergoes the
second meiotic division, but mostly without the forma-
tion of a cell wall between the two megaspore nuclei
(Figure 2A); in only one ovule a partitioning wall was
seen resulting in two chalazal megaspores (Figure 2C).
The micropylar dyad cell degenerates without further
division (Figure 2B, C). The starting point of the mega-
gametophyte is therefore usually a bisporic cell with the
nuclei situated at opposite poles of the cell (Figure 2B).
After two mitotic divisions an eight-nucleate, bisporic
embryo sac is formed that conforms to the Allium- type of
megagametophyte development (Willemse & Van Went
1984). However, the presence of three megaspores in the
nucellus (Figure 2C) suggests that monosporic Poly-
gonum-type embryo sacs also occur in Dovyalis caffra.
Structure of the mature ovule
Ovules are pendant, sessile, ovoid and anatropous
structures reaching a size of about 90 x 60 pm at anthesis
(Figure 2D). The two integuments are of equal length so
that the micropylar canal is formed by the inner integu-
ment only, but in post-fertilization stages the outer integu-
ment usually lengthens to form an exostome (compare
Figure 2D, E) that is slightly out of line with the endos-
tome. The lower part of both integuments consists of about
five parenchymatic cell layers, but in the micropylar
region the integuments become thicker (Figure ID). The
outer epidermis of the inner integument is the most promi-
nent of the integumental layers. The lower part of this epi-
dermal layer is uniseriate and consists of meristematic
cells with large nuclei and dense protoplasm, whereas the
micropylar part is multiseriate and forms most of the cells
in the thick micropylar region of this integument (Figure
2D). In the centre of the ovule, the nucellus is an ovoid
structure with a slightly attenuate apex consisting of a
nucellus cap. The latter is formed by derivatives of the
nucellus epidermis and the parietal cell (Figure 2B, C). The
nucellus cells are parenchymatic, except at the base of the
Bothalia 35,1 (2005)
103
FIGURE 1. — Flacentation and orientation of ovules in Dovyalis caffra. A, apetalous female flower in t/s showing parietal placentation of devel-
oping ovules in unilocular ovary; B, female flower in superficial (tangential) 1/s showing position and orientation of young ovules; C, sagit-
tal section of bitegmic, dorsal epitropous ovule primordium with megaspore mother cell; D, anatropous, crassinucellate, sessile ovule in
sagittal section during formation of bisporic embryo sac. a, 6-carpellate ovary; b, locule; c, calyx lobe; f, megaspore mother cell; i, inner
integument; j, initial stage of bisporic embryo sac; n, nucellus; o, outer integument; p, placenta of locule; v, young ovule. Scale bars: A, B,
50 pm; C, D, 10 pm.
104
Bothalia 35,1 (2005)
FIGURE 2. — Formation of embryo sac and structure of ovule in Dovyalis caffra. A, dyad cells in nucellus with chalazal cell containing two megas-
pore nuclei; B. nucellus containing bisporic embryo sac in two-nucleate stage and disintegrated micropylar dyad cell; C, nucellus with two
chalazal megaspores and disintegrated micropylar dyad cell; D, structure of ovule at anthesis. Note micropyle (curved arrows) formed by
inner integument and embryo sac protruding into lower part of endostome; E, embryo sac at fertilization. Note dark-staining filiform appa-
ratus of synergid, extranucellar embryo sac and slightly skewed micropyle (curved arrows) formed by both integuments, e, embryo sac; i,
inner integument; j, initial stage of bisporic embryo sac; k, disintegrated micropylar dyad cell; m, chalazal megaspores; n, nucellus; ne,
nucellus epidermis; o, outer integument; oi, outer epidermis of inner integument; s, synergid. Scale bars: A, B, C, 5 pm; D, E. 10 pm.
Bothalia 35, 1 (2005)
105
embryo sac where the cell walls are noticeably thicker,
darkly stained and possibly lignified (Figure 2D, E). The
cells possibly represent a postament as described by
Shamrov (1998: 379), i.e. a column-like tissue located
below the sporogenous or gametophytic structures, con-
sisting of thin-walled or sometimes lignified cells.
The mature embryo sac is a narrow-elliptical structure
that occupies an unusual position in the ovule — it lies
partly inside and partly outside the nucellar tissue
(Figure 2D). In its growth towards maturity, the base of
the embryo sac does not succeed in penetrating the cha-
lazal nucellus, the thick-walled cells of the postament
possibly acting as a barrier tissue to stop the encroach-
ment of the embryo sac into the chalaza (McLean &
Ivimey-Cook 1956; Bouman 1984). However, in the
micropylar region the apical part of the embryo sac has
broken through the nucellus epidermis and, flanked by
the inner integument, stretches up halfway into the
endostomium (Figure 2D, E). The detailed structure of
the embryo sac elements could not be determined. At fer-
tilization the antipodals have disappeared, the central cell
nucleus lies near the egg apparatus in the extranucellar
apical region and the synergids show the unmistakable
presence of a filiform apparatus (Figure 2E).
Seed development
Embryo : fertilized ovules take about three months to
develop into mature, woolly seeds of ± 10x5 mm, embed-
ded in the fleshy pulp of the fruit, which is a berry. Embryo
development was not studied in detail and the embryo
could not be typified. The youngest embryo found was ±
eight weeks old. Against the wall of the embryo sac the first
layers of endosperm cells have been formed. At eight
weeks, the embryo is in an early globular stage with a long,
uniseriate suspensor stretching up into the micropyle (Figure
3A). The cells of the suspensor contain starch grains and
extend past the nucellus epidermis into the region original-
ly occupied by the extranucellar part of the embryo sac. It
seems possible that, by elongating, the suspensor has
pushed the embryo proper into the developing endosperm.
At seed maturity the embryo is without chlorophyll, erect
and spathulate with thin, expanded cotyledons and lies
imbedded in oily endosperm.
Seed coat : in Dovyalis caffra the seed coat is not mul-
tiplicative, i.e. periclinal divisions do not occur during
seed development to add extra layers to the seed coat.
Since the only mechanical layer in the mature seed coat
develops from the outer epidermis of the inner integu-
ment, the seed is exotegmic.
In the early stages of seed coat development (Figure
3B, D), the testa comprises about four to five cell layers,
except in the raphal region where the mesophyll is mul-
tilayered (Figure 3B). The mesophyll and inner epider-
mis of the testa consist of thin-walled parenchyma. The
outer epidermis of the testa possibly plays an important
part in the protection of the seed before the mechanical
layer of the tegmen matures. After fertilization, numer-
ous thin-walled, unicellular epidermal hairs are formed
which cover the young seed (Figure 3B, D) at the early
globular stage of the embryo (Figure 3A). The hairs grow
perpendicular to the seed surface (instead of becoming
depressed) and eventually permeate the space between
the seeds and the pericarp. The hairs possibly guard
against desiccation of the inner seed tissues by prevent-
ing the loss of water vapour through the numerous stom-
ata in the epidermis of the testa (Figure 3D). Apart from
the hairs, the outer epidermis also contains large num-
bers of tanniniferous cells with walls impregnated and
lumina filled with phenolic compounds. These polyphe-
nols increase the rigidity of the thin epidermal cell walls
and contribute towards the hardness of the seed coat
(Werker 1997: 100). Phenolic-containing cells offer re-
sistance to biodegradation and help to protect against
attacks by herbivores, insects and micro-organisms
(Boesewinkel & Bouman 1984). In the ripe fruit the
outer epidermis with its hairs persists, but the phenolic
substances have disappeared and the cells of the meso-
phyll and inner epidermis collapse (Figure 3C).
At the onset of seed coat development, the tegmen con-
sists of about five layers (Figure 2E). The outer epidermis
of the tegmen (exotegmen) show frequent anticlinal divi-
sions to form a single layer of thin-walled, radially flat-
tened cells (Figure 3D) that rapidly stretch in a direction
parallel to the longitudinal axis of the seed (Figure 3B).
The inner epidermis of the tegmen (endotegmen) also
divides anticlinally and forms a layer of small, tightly
packed cells containing large amounts of dark-staining
phenolic substances (Figure 3B, D). In the ripe fruit the
exotegmen forms the mechanical layer of the seed and
comprises thick-walled, radially flattened cells (Figure 3C)
with cell walls impregnated with phenolic compounds. The
mesotegmen has disintegrated and the endotegmen
remains as small, strangely formed flask-shaped cells adja-
cent to the persistent nucellus epidermis that separates the
endotegmen from the endosperm (Figure 3C).
DISCUSSION
A detailed comparison of ovule and seed characters in
the 36 families placed by Savolainen et al. (2000) in
Malpighiales — the clade previously termed Violales
(Chase et al. 2002) — is hampered by a lack of compara-
ble data for many of the families, as given in compendia
dealing with comparative embryology. Nevertheless,
according to such works (Davis 1966; Comer 1976; Johri
et al. 1992; Nandi et al. 1998) similarities include
bitegmy, anatropy and crassinucelli with the nucellus
epidermis participating in the formation of the nucellus
cap (tenuinucellate in Clusiaceae, Dichapetalaceae,
Linaceae, Ochnaceae, Trigoniaceae); both integuments
usually form the zigzag micropyle canal (inner integu-
ment only in Dichapetalaceae, Erythroxylaceae, Rhizo-
phoraceae, outer integument only in Salicaceae sensu
stricto)', a Polygonum- type embryo sac usually develops,
but Allium- and Adoxa- types also occur; endosperm for-
mation is nuclear, it later becomes cellular and is usually
copious (exalbuminous seeds in Caryocaraceae, Clusia-
ceae, Ochnaceae, Salicaceae sensu stricto)', the embryo-
geny varies considerably, but the mature embryo is typi-
cally straight, medium-sized to large and protected by a
fibrous exotegmen in the seed coat (mesotestal seed in
Clusiaceae, exotestal in Dichapetalaceae, Salicaceae).
As far as embryology is concerned, Dovyalis seems to
fit comfortably into the framework of Malpighiales (com-
106
Bothalia 35,1 (2005)
FIGURE 3. — Seed and seed coat formation in Dovyalis cajfra. A, micropylar region of young seed with developing embryo, note long suspensor
in micropyle; B, 1/s developing seed coal; C, t/s seed coat of seed in ripe fruit; D, t/s developing seed coat, i, inner integument (legmen);
li, inner epidermis of tegmen; mi, mesophyll of tegmen; mo, mesophyll of testa; n, nucellus; ne, nucellus epidermis; o, outer integument
(testa); oi, outer epidermis of tegmen; oo, epidermis of testa; o, testa; t, epidermal hair; u, suspensor; w, endosperm; x, guard cell of stoma;
y, embryo. All scale bars 10 pm.
Bothalia 35,1 (2005)
107
TABLE 1. — Comparison of selected ovule and seed characters in Dovyalis and Salix
pare Table 1 : Nos 2, 3, 4, 6, 7, 8, 9, 10 & 12). Reports in
classical literature (Davis 1966; Comer 1976; Johri et al.
1992) indicate that the genus also shares many of the
above-mentioned characters with other representatives
of the Flacourtieae, e.g. Are chav aletaia uruguayensis
Speg., Idesia polycarpa Maxim., Flacourtia indica
(Burm.f.) Merr. However, features that seem to stand
out as particular to Flacourtieae include the following:
an embryo sac that breaks through the nucellus and pro-
trudes into the endostome, a variation in the functional
behaviour of the megaspores and a fibrous exotegmen in
the seed (Johri et al. 1992). These are also characters
that proved to be definite for Dovyalis in the present
study.
Embryological studies on Salix and Populus L.
(Chamberlain 1897; Maheshwari & Roy 1951; Nagaraj
1952; Steyn et al. 2004) partly support the placement of
Salicaceae sensu stricto in Malpighiales (see Table 1 : Nos
1 , 2 (partly), 3, 6. 7, 8 (partly). 9 (partly), 1 0 (partly), 1 1 ).
It is noteworthy that two other embryological characters
of diagnostic value for Flacourtieae, namely the unusual
extranucellar embryo sac and the inconsistent behaviour
of the megaspores, are also characteristic for both Salix
and Populus (Chamberlain 1897; Nagaraj 1952). Fur-
thermore, the hairy seed coat in Dovyalis may be taxo-
nomically significant. Although the seed itself is glabrous
in Salix, the genus is characterized by the presence of
long, unicellular, intra-ovarian hairs as well as seed with
a hairy aril (Steyn et al. 2004). This propensity to produce
epidermal hairs in the ovary and its associated structures
such as ovules, seed and arils may well reflect a close
phylogenetic link between Flacourtieae and Saliceae.
Despite all the similarities between Salix and Dovyalis,
significant deviations of the basic flacourtiaceous pattern
also occur. In Salix the endosperm is ephemeral
(Hakansson 1954), the seed exalbuminous and exotestal
and the inner integument absent. Unitegmy results in the
lack of an endostome and the absence of a tegmen with its
characteristic fibrous exotegmen. Steyn et al. (2004) sug-
gested that the marked differences in seed coats between
Salix and animal-dispersed flacourtiaceous taxa like
Dovyalis may possibly be linked to seed adaptations for
different dispersal strategies — Salix seeds are mostly dis-
persed by wind and water, mechanical layers would only
unnecessarily increase seed weight, whereas Dovyalis seed
occur inside edible berries and need the protection of
exotegmic fibres against the onslaught of animal feeders.
ACKNOWLEDGEMENTS
The South African National Biodiversity Institute
(SANBI), Pretoria is acknowledged for providing the
infrastructure to execute this investigation. Thanks are
also due to Dr Robert Archer who provided much of the
material for the study.
REFERENCES
ARNOTT. G.A.W. 1841. On some South African plants. 10. Dovyalis.
Hooker's Journal of Botany 3: 251, 252.
BERNHARD, A. & ENDRESS. PK. 1999. Androecial development
and systematics in Flacourtiaceae. Plant Systematics and
Evolution 215: 141-155.
BREDENKAMP, C.L. 2003. Flacourtiaceae. In G. Germishuizen &
N.L. Meyer, Plants of southern Africa: an annotated checklist.
Strelitzia 14: 559-561.
108
Bothalia 35,1 (2005)
BOESEWINKEL, F.D. & BOUMAN, F. 1984. The seed: structure. In
B.M. Johri, Embiyology of angiosperms: 567-610. Springer,
Berlin.
BOUMAN, F. 1984. The ovule. In B.M Johri, Embryology of angio-
sperms : 123-158. Springer, Berlin.
CHAMBERLAIN, C.J. 1897. Contribution to the life history of Salix.
Botanical Gazette 23: 147-179.
CHASE, M.W., ZMARZTY, S„ LLEDO, M.D., WURDACK, K.J.,
SWENSEN. S.M. & FAY, M.F. 2002. When in doubt, put it in
the Flacourtiaceae: a molecular phylogenetic analysis based on
plastid rbcL DNA sequences. Kew Bulletin 57: 141-181.
CORNER, E.J.H. 1976. The seeds of dicotyledons, vol. 1. Cambridge
University Press, Cambridge.
DAVIS. G.L. 1966. Systematic embryology of the angiosperms. Wiley,
London.
ENDLICHER, A. 1 842. Euphorbiaceae. Genera dubia. Dovyalis. Genenun
plantarum. Suppl. 2: 91.
FEDER, N. & O'BRIEN, T.P. 1968. Plant microtechnique: some prin-
ciples and new methods. American Journal of Botany 55:
123-142.
GILG, E. 1925. Flacourtiaceae. In A.Engler & K. Prantl, Die natiirlichen
Pflanzenfamilien, edn 2, 21: 377—457. Engelmann, Leipzig.
HAKANSSON, A. 1954. Endosperm formation in Salix. Botaniska
Notiser 1954: 326-332.
HUTCHINSON, J. 1967. The genera of flowering plants, vol. 2. Clarendon
Press, Oxford.
JOHRI, B.M., AMBEGAOKAR, K.B. & SRIVASTAVA, PS. 1992. Com-
parative embryology of angiosperms, vol. 1. Springer, Berlin.
JUDD, W.S., CAMPBELL, C.S., KELLOGG, E.A., STEVENS, P.F. &
DONOGHUE, M.J. 2002. Plant systematics: a phylogenetic
approach, edn 2. Sinauer Associates, Sunderland.
KAUL, R.B. 1995. Reproductive structure and organogenesis in a cot-
tonwood, Populus deltoides (Salicaceae). International Journal
of Plant Sciences 156: 172-180.
KEATING, R.C. 1937. Pollen morphology and relationships of the Fla-
courtiaceae. Annals of the Missouri Botanical Garden 60: 273-
305.
LANGENEGGER, J.E. 1976. Dovyalis. In J.H. Ross, Flora of south-
ern Africa 22: 84—90. Botanical Research Institute, Pretoria.
LEMKE, D.E. 1988. A synopsis of Flacourtiaceae. Aliso 12: 29^43.
MAHESHWARI, P. & ROY, S.K. 1951. The embryo sac of Salix. Pliyto-
morphology 1 : 70-72.
MARTIN. A.C. 1946. The comparative internal morphology of seeds.
The American Midland Naturalist 36: 513-660.
McLEAN, R.C. & IVIMEY-COOK, W.R. 1956. Textbook of theoreti-
cal botany , vol. 2. Longmans, London.
MEEUSE. A.D.J. 1975. Taxonomic relationships between Salicaceae
and Flacourtiaceae: their bearing on interpretation of floral
morphology and dilleniid phylogeny. Acta Botanica Neerlandica
24: 437-457.
NAGARAJ, M. 1952. Floral morphology of Populus deltoides and P.
tremuloides. Botanical Gazette 114: 222-243.
NANDI, O.I., CHASE, M. & ENDRESS, P.K. 1998. A combined
cladistic analysis of angiosperms using rbcL and non-molecular
data sets. Annals of the Missouri Botanical Garden 85:
137-212.
O'BRIEN, T. P. & McCULLY, M.E. 1981. The study of plant structure :
principles and selected methods. Termacarpi, Melbourne.
RIDLEY, H.N. 1930. The dispersal of plants throughout the world.
Reeve, Ashford, Kent.
SAVOLAINEN, V., FAY, M.F., ALBACH, D.C., BACKLUND, A.,
VAN DER BANK, M„ CAMERON, K.M., JOHNSON, S.A.,
LLEDO, M.D., PINTAUD, J,-C„ POWELL, M„ SHEAHAN,
M.C., SOLTIS, D.E., SOLTIS, P.S., WESTON, P, WHITTEN,
W.M., WURDACK, K.J. & CHASE, M.W. 2000. Phylogeny of
the eudicots: a nearly complete familial analysis based on rbcL
gene sequences. Kew Bulletin 55: 257-309.
SCHMID, R. 1986. On Comerian and other terminology of angiosper-
mous and gymnospermous seed coats: historical perspective
and terminological recommendations. Taxon 35: 476-491.
SHAMROV, I.I. 1998. Ovule classification in flowering plants — new
approaches and concepts. Botanische Jahrbiicher 120: 377—407.
SLEUMER, H. 1972. A taxonomic revision of the genus Dovyalis
E.Mey. ex Am. (Flacourtiaceae). Botanische Jahrbiicher 92:
64-89.
STEYN, E M. A., VAN WYK, A.E. & SMITH, G.F. 2001. A study of
the ovule and young seed of Guthriea capensis (Achariaceae).
South African Journal of Botany 67: 206-213.
STEYN, E.M.A., VAN WYK, A.E. & SMITH, G.F 2002a. A study of
ovule-to-seed development in Ceratiosicyos (Achariaceae) and
the systematic position of the genus. Bothalia 32: 201-210.
STEYN, E.M.A., VAN WYK. A.E. & SMITH, G.F. 2002b. Ovule, seed
and seedling characters in Acharia (Achariaceae) with evidence
of myrmecochory in the family. South African Journal of
Botany 68: 143-156.
STEYN, E.M.A., VAN WYK, A.E. & SMITH, G.F 2003. Embryology
and systematic relationships of Kiggelaria (Flacourtiaceae).
Bothalia 33: 199-206.
STEYN, E.M.A., SMITH, G.F. & VAN WYK, A.E. 2004. Functional
and taxonomic significance of seed structure in Salix mucrona-
ta (Salicaceae). Bothalia 34: 53-59.
WARBURG, O. 1893. Flacourtiaceae. In A. Engler & K. Prantl, Die
natiirlichen Pflanzenfamilien, edn 1. Teil III, Abt. 6a: 1-56,
Engelmann, Leipzig.
WERKER, E. 1997. Seed anatomy. Bomtraeger, Berlin.
WILLEMSE, M.T.M. & VAN WENT, J.L. 1984. The female gameto-
phyte. In B.M Johri, Embryology of angiosperms: 159-196.
Springer, Berlin.
Bothalia 35,1: 109-110(2005)
Miscellaneous notes
AGAPANTHACEAE
CHROMOSOME COUNTS IN THE GENUS AGAPANTHUS
INTRODUCTION
The genus Agapanthus L’Her. consists of ten species
(Archer 2003), widely distributed in southern Eastern
Cape, southern Western Cape, southern KwaZulu-Natal,
Swaziland, Mpumalanga, Free State, Lesotho, Gauteng,
Limpopo and Mozambique (Duncan 1998).
Very little morphological variation exists in the genus
and the delimitation of species is mainly based on flower
type and whether the taxon is deciduous or not (Leighton
1965). The species also hybridize quite freely when grown
next to each other (Duncan 1998). Human intervention
(selection) and especially the production of numerous cul-
tivars (often resulting from interspecific hybrids) blur
species lines. The extent of natural hybrids is not known.
The objective of this study was to determine somatic
chromosome numbers in various subspecies and cultivat-
ed varieties of Agapanthus.
MATERIALS AND METHODS
Seeds of species of Agapanthus collected in the wild
or cultivars obtained from nurseries, were germinated in
a greenhouse at the University of the Free State, South
Africa. Germinated plants were watered heavily a day
before collecting the root tips. Root tips were treated with
cold water at 4°C for 48 hours (Jong 1995). Then the root
tips were fixed in Camoy’s fixative (Camoy 1886) for 72
hours. The Camoy’s fixative was replaced by 70%
ethanol. Root tips were hydrolysed with IN hydrochloric
acid for seven minutes and stained with Feulgen reagent
for two hours in darkness (Darlington & La Cour 1976).
The root tips were stored in 30% alcohol until squashing.
Cover slips were treated with Mayr’s albumen and
squashes were made in aceto-orcein according to
Darlington & La Cour (1976). Contrast between chromo-
somes and the cytoplasm was intensified by adding 45%
acetic acid, saturated with iron acetate (Thomas 1940).
Slides were made permanent by floating the cover slip off
in acetic acid, dehydrating in alcohol and mounting in
Euparol (Darlington & La Cour 1976).
Observation of the slides was done with an Olympus
CH2 light microscope. Cell positions were located with
an England Finder. At least ten cells per specimen were
studied. Chromosomes in the cells were photographed
with a Cool Pix digital camera, mounted on a Nikon
Microphot FXA microscope.
RESULTS AND DISCUSSIONS
Results were obtained from five species, nine sub-
species and 1 1 cultivars (Table 1). Chromosome numbers
observed for Agapanthus campanulatus subsp. campan-
ulatus, A. praecox subsp. praecox and A. praecox subsp.
orientalis support previous findings.
Somatic chromosome numbers of 2n =28 + 0-2B, 30
and 30 + 0-2B were observed (Table 1). The 2n = 28 + 0-
2B was observed in A. inapertus subsp. intennedius.
This species was also the only one with a chromosome
count less than 2n = 30. The species is morphologically
different from other Agapanthus species since it is the
only species with drooping flowers featuring the colours
Aconite violet 937/3 and Victoria violet 738 (Leighton
1965).
Chromosome counts of 2n = 30 and 30 + 0-2B were
the most frequent for the studied taxa and agree with pre-
vious observations (Guignard 1884; Belling 1928;
Darlington 1933; Geitler 1933; Stenar 1933; Matsuura &
Suto 1935; Mookeijea 1955; Lima-de-Faria & Sarvella
1958; Sharma & Sharma 1961; Riley & Mukerjee 1962;
Sharma & Mukhopadhyay 1963; Vijavalli & Mathew
1990). B-chromosomes were present in all species of the
genus. Chromosomes were considered to be B-chromo-
somes if the number of chromosomes varied between
different unbroken cells of the same individual. How-
ever, not all subspecies of a species or all cultivated
forms had B-chromosomes. In some cases these B-chro-
mosomes occurred in taxa where they have not been
described previously (Table 1). This study did not focus
on the occurrence of B-chromosomes, therefore it was
difficult to determine if B-chromosomes occur in all
Agapanthus specimens and whether they are restricted to
any part of the soma. It was also observed that the karyo-
type of Agapanthus comprised of chromosome pairs of
different sizes.
The chromosome counts indicated that the basic chromo-
some number for Agapanthus is x = 15, with a reduction to
x = 14 in at least A. inapertus subsp. intennedius. This is a
high basic chromosome number and suggests a palaeoploid
origin for Agapanthus. This study therefore added new
information for Agapanthus. since Darlington (1933) only
reported on the basic chromosome number of x = 15.
Further studies are needed to test the relationships in
Agapanthus and especially the function (if any) and origin
of the B-chromosomes. The other four species of Aga-
panthus should also be studied to determine their chromo-
some numbers and to see whether other basic chromo-
some numbers may be present.
ACKNOWLEDGEMENTS
The University of the Free State is thanked for finan-
cial support during this study.
110
Bothalia 35.1 (2005)
TABLE 1. — Specimens of Agapanthus taxa studied with their somatic (2n) chromosome numbers and voucher numbers or source
* first chromosome no. report.
REFERENCES
ARCHER, C. 2003. Agapanthaceae. In G. Germishuizen & N.L. Meyer,
Plants of southern Africa: an annotated checklist. Strelitzia 14: 954.
BELLING, J. 1928. Contraction of chromosomes during maturation
divisions in Lilium and other plants. University of California
Publications in Botany 14: 335-343.
CARNOY, J.B. 1886. La cytodierese de l'oeuf. Cellule 3: 1-92.
DARLINGTON, C.D. 1933. Meiosis in Agapanthus and Kniphofia.
Cytologia 4: 229-240.
DARLINGTON, C.D. & LA COUR, L.F. 1976. The handling of chro-
mosomes. Allen & Unwin, London.
DUNCAN, G. 1998. Grow agapanthus. Kirstenbosch Gardening Series.
Trident Press, Cape Town.
GEITLER, L. 1933. Das Verhalten der Chromozentren von Agapanthus
wahrend der Meiose. Osterreichische botanische Zeitschrift 82:
277-282.
GUIGNARD, L. 1884. Recherches sur la structure et la division du
noyau cellulaire chez les vegetaux. Annates des sciences natu-
re lies, botanique, set: 17: 5-59.
JONG, K. 1995. Laboratory manual of cytological techniques. Royal
Botanic Gardens, Edinburgh.
LEIGHTON, F.M. 1965. The genus Agapanthus L’Heritier. Journal of
South African Botany, Supplementary volume IV.
LIMA-DE-FARIA, A. & SARVELLA, P. 1958. The organization of
telomeres in species of Solanum, Salvia, Scilla, Secale,
Agapanthus and Ornithogalum. Hereditas 44: 337-346.
MATSUURA, H. & SUTO, T. 1935. Contributions to the idiogram
study in phanerogamous plants. I. Journal of the Faculty of
Science, Hokkaido Imperial University , ser. 5: 33-75.
MOOKERJEA, A. 1955. Cytology of amaryllids as an aid to the under-
standing of evolution. Caryologia 7: 1-71.
RILEY, H P. & MUKERJEE, D. 1962. Chromosomes of some species
of Agapanthus. Cytologia 27: 325-332.
SHARMA, A.K. & MUKHOPADHYAY, S. 1963. Chromosome study
in Agapanthus and the phylogeny of its species. Caryologia 16:
127-137.
SHARMA, A.K. & SHARMA, A. 1961. An investigation of the cytology
of some species of Liliaceae. Genetica Iberica 13: 25—42.
STENAR, H. 1933. Zur Embryologie der Agapanthus- Gruppe.
Botaniska Notiser 1933: 520-530.
THOMAS, PT. 1940. The Aceto-carmine Method for fruit material.
Stain Technology 15: 167-172.
VIJAVALLI, B. & MATHEW, P.M. 1990. Cytotaxonomy of the Lilia-
ceae and allied families. Continental Publishers, Kerala, India.
M. MUZILA* and J.J. SPIES**
* Department of Biological Sciences, University of Botswana, Private
Bag UB 00704, Gabarone, Botswana.
** Department of Plant Sciences: Genetics (62), University of the Free
State, P.O. Box 339, Bloemfontein 9300, South Africa.
MS. received: 2004-05-17.
Bothalia 35,1: 111-113 (2005)
Book Reviews
MEDICINAL PLANTS TRADED ON SOUTH AFRICA’S EASTERN
SEABOARD, by D. VON AHLEFELDT, N. CROUCH, G. NICHOLS,
R. SYMMONDS, S. MCKEAN, H. SIBIYA, & M.P. CELE (edited by
Prof. K.D. Gordon-Gray). 2003. Ethekwini Parks Department, P.O.
Box 3740, Durban 4000, and University of KwaZulu-Natal, Private
Bag X54001, Durban, 4001. Pp. 267. Soft cover: ISBN 0 620 31569 5,
price R 150.00.
Medicinal plants traded on South Africa’s eastern seaboard is a
comprehensive and scientifically accurate photographic guide to 330 of
the most commonly traded medicinal plants in Durban's informal muthi
markets. The book aims to assist in the identification of fresh, dried and
often fragmentary plant parts displayed in these markets and in many
other urban markets in the region. The correct identification of these
plants is essential for further scientific investigation and, most impor-
tantly, conservation planning and action.
The authors represent a wide range of fields of expertise including
systematic botany, ethnobotany, horticulture, conservation and tradi-
tional health care and are highly qualified for such an undertaking. The
book is the result of 15 years of collaborative research under the
umbrella of the Indigenous Plant Trade Research Associates of Durban
(IPTRAD) working group, a group with a common interest in medici-
nal plant conservation. Other than traditional healers and medicinal
plant traders themselves, institutional collaborators include eTheke-
weni Municipality (Silverglen Medicinal Plant Nursery), Ezemvelo
KwaZulu-Natal Wildlife Nature Conservation Services, the Ethno-
botany Unit of the National Botanical Institute [now South African
National Biodiversity Institute], and the University of Natal, Durban
[now University of KwaZulu-Natal],
Traditional health care requirements in South Africa create a huge
demand for hundreds of medicinal plant species sold in informal street
markets and ‘African chemists’ in many towns and cities throughout
the country. The trade in medicinal plants is probably greater now than
at any time in the past and is one of the most complex resource man-
agement issues facing conservation agencies, health care professionals
and resource users in South Africa today. It is acknowledged that the
harvesting of traditional medicinal plant material is a serious threat to
biodiversity in the region; on the other hand, the trade not only provides
healthcare for millions of consumers, it is also critical for the welfare
of all the people engaged in the industry.
A visit to any of these muthi markets reveals a mind-boggling array
of plant (and animal) parts displayed for sale, usually in the open air on
street pavements. The majority of the plants are impossible to identify
without seeing the plant in habitat or at least fresh parts thereof.
Vernacular names can be useful for identification but can also cause
great confusion as they are often applied in a general sense. For exam-
ple. a single name, ubulawu, refers to a suite of unrelated plant species
used for ritual washing. This book will go a long way towards solving
such perplexities but will however also invite new inquiries and facili-
tate further research. It is hoped that future research results will be trans-
lated into improved management strategies that go towards ensuring the
long-term survival of these species, as well as ensuring that they remain
accessible to those who rely on their trade for health care and liveli-
hoods.
A book of this nature is long overdue and the authors are to be con-
gratulated on producing this useful, pocket-sized, well-bound guide.
This type of information is often only available in large format, elabo-
rate scientific publications and scientific journals. It is useful to have a
book that can be of interest and value to diverse interest groups such as
traditional healers, herbalists, ethnobotanists, taxonomists, conserva-
tionists, anthropologists, students, ecotourists and the general public.
Although the book is based on research done in the Durban area, many
of the species are found throughout the southeastern coastline and it
will be equally useful in neighbouring regions such as Eastern Cape.
An important aspect of the book is its accessibility to non-scientists,
which will help towards demystifying, understanding and improving
the often derogatory perception the general public have of the medici-
nal plant trade and the traders themselves.
A detailed introduction, explaining the purpose of the book and how it
is laid out, includes explanatory paragraphs on the geographic scope of the
book, maps, identification, names and a description of the format of indi-
vidual plant entries. The book is divided into six colour-coded chapters
based on the part of the plant used for medicinal purposes. These include:
whole plants, fruits and seeds, climbers and creepers, succulents, bark,
stems and leaves, and underground parts. Within each chapter three species
are presented per page, arranged alphabetically by family, and then alpha-
betically by genus. Some species are repeated as different parts are used and
therefore appear in the relevant chapters. Each species entry includes com-
mon and vernacular names, growth form, distinguishing characteristics,
habitat. Red Data status, legal status in KwaZulu-Natal, voucher specimen
numbers and a distribution map for KZN and its neighbouring areas. Each
species is well illustrated with up to five photographs for some species,
showing the plant in habitat as well as plant parts, as they are displayed on
the market. Scale bars indicate actual size of the plants. The photographs are
clear but small, with some wasted space between each. The full page could
rather have been used to increase the size of the illustrations.
The uses of the plants are unfortunately not provided but the read-
er is referred to several books (included in the reference list) docu-
menting medicinal plants in the region. I believe the book would have
been strengthened greatly by including this information, if only broad-
use categories such as used by Arnold et al. (2002). Notes on propaga-
tion and cultivation of the species would also have been very useful but
would however, have increased the size and cost of the volume.
The main entries are followed by a comprehensive list of refer-
ences, a glossary and a single index, which includes subjects, species
names and vernacular names.
REFERENCE
ARNOLD, T.H., PRENTICE, C.A., HAWKER, L.C., SNYMAN, E.E.,
TOMALIN, M„ CROUCH, N.R. & POTTAS-BIRCHER, C.
2002. Medicinal and magical plants of southern Africa. Stelitzia
13. National Botanical Institute, Pretoria.
TONY DOLD*
* Selmar Schonland Herbarium, P.O. Box 101, 6140 Grahamstown,
South Africa.
SEED CONSERVATION: TURNING SCIENCE INTO PRACTICE
edited by ROGER D. SMITH, JOHN B. DICKIE, SIMON H. LINING-
TON, HUGH W. PRITCHARD & ROBIN J. PROBERT. 2003. Royal
Botanic Gardens, Kew, Richmond, Surrey TW9 3AB, UK. Pp. 1023.
Soft cover: ISBN 1-84246-052-8, price £59.95.
This is an impressive, comprehensive publication which is packed
with information, guidelines, research results, suggested research
options and discussion which challenges existing hypotheses and
dogma in the field of seed conservation. It is not possible to do full jus-
tice to this volume in this review, however, it is trusted that the
overview presented will whet the appetites of those who have an inter-
est in seed conservation. Plant genetic resource conservation acts as a
link between genetic diversity of a plant and its utilization or exploita-
tion by humans and it is clear that there is still much to be done to
enable seed conservation to meet the challenges of the 21st Century.
Furthermore, the adoption of the Global Strategy for Plant Conserva-
tion in 2003, the adoption of the International Treaty on Plant Genetic
Resources in 2001 and the launch of the Global Conservation Trust in
2002 are three relatively recent, large international initiatives which
have contributed to raising the profile of seedbanks.
The book is a synthesis of the proceedings of an International
Workshop held under the auspices of the Millennium Seedbank Project
112
Bothalia 35,1 (2005)
(MSBP), managed by the Royal Botanic Gardens, Kew, at the Millen-
nium Seedbank in 2001. The meeting was attended by 88 delegates,
including 29 MSBP collaborators from 27 countries. The aims of the
workshop were to: undertake a review of what is understood about the
collection and conservation of seeds and fruits of non-domesticated
plant species; identify similarities and differences in the process of seed
conservation for wild and cultivated species; and share progress in the
seedbanking of non-domesticated plants.
Exacerbation of the loss of plant diversity, as a result of a number
of factors which we are all too familiar with, such as the effects of cli-
mate change and land conversion, and building, agriculture and min-
ing, has enhanced the role of a seedbank in the conservation of dis-
placed diversity and feeding into plant breeding and conservation pro-
grammes. At present, seedbanking can only be effectively used for
species which are dessication tolerant.
Chapters 1-14 in the book, deal with planning and collecting seed
material. The Convention on Biological Diversity (CBD), as a legal
and ethical framework for seed conservation activities, is discussed.
Each partnership which has been set up by the MSB with organizations
in different countries is unique, to maximize the potential benefits from
the partnerships. However, there are many common elements which are
presented as model legal clauses, compliant with the Bonn Guidelines,
to assist with the drawing up of Access and Benefit Sharing Agree-
ments (ABSA) reflecting the characteristics and needs of the partners,
while taking into account CBD implementation at national level. A
helpful outline and explanation of key legal terms and conditions that
may require consideration, is also provided.
organizations which carry out bona fide non-commercial research
under a signed, legally binding Material Supply Agreement.
The MSBP is supported by the Seed Bank database (SBD), the
specimen database, which holds information about every seed sample
in the MSB, and the Seed Database (SID), containing taxon-based and
a synthesis of information contained in the SBD together with other
seed research data gleaned from project and literature searches.
Taxonomically based summaries are available online for conservation-
ists and scientists.
Principles of seed-drying methods accompanied by useful data are
presented to demonstrate the potential of charcoal which, although it is
not as* effective as silica gel, provides an alternative as a low-cost, low-
technology seed desiccant. Non-destructive measurement of seed mois-
ture by measuring the RH of air in equilibrium with seed samples is
regarded as a reliable alternative to gravimetric moisture content deter-
mination. Recommendations based on experiments are provided and
alternative lbw-cost approaches are briefly reviewed. Statistical models
to determine water content may also be used, and it is suggested that
techniques such as functional genomics and quantitative genetics be
further explored and assessed to determine whether they can be used to
identify global markers as prognostic tools. In a discussion on aspects
of desiccation tolerance and sensitivity and the handling of desiccation-
sensitive seeds, the authors of this chapter emphasize the gap that exists
between the current state of urgency to preserve germplasm and the
lack of understanding of the phenomena leading to desiccation tole-
rance or sensitivity. Seed recalcitrance for example, appears to be wide-
spread across families with little taxonomic relationship.
Factors to be considered when planning a programme of plant
genetic conservation are reviewed. There is a detailed and useful chap-
ter which includes website addresses which can be accessed to provide
guidance and suggest criteria to be used in the identification of target
species for plant conservation projects. Measurable criteria such as
genetic distinctiveness, probability of species extinction, threat of gene-
tic erosion and potential economic value of a taxon are suggested. Taxa
can also be selected by assembling information on geographic distrib-
ution, habitat preference, phenology, genetics and taxonomy.
GIS can assist in the planning process, although the most frequent
obstacle to its use is limited data availability. The use of genetic data
derived from techniques such as genetic fingerprinting, which only
requires small quantities of DNA and plant tissue, offers the opportuni-
ty to study some of the rare and endangered species for which this type
of information is critical. Understanding the effect of seed maturity on
stability, is dealt with in chapters on specific examples such as Millettia
leucantha and Az.adirachta indiea. Ideally seeds should be collected
when they are at the point of natural dispersal. Although various crude
indicators can be used to identify seed maturity, it is suggested that
measuring the seed equilibrium relative to humidity and comparing this
with the ambient RH, may be a useful and precise tool to identify this
point. It is important that the seed collection represents, as far as possi-
ble, the genetic diversity of target populations, taking into account the
ecology and distribution of the species, the geography of the collection
region, the likely breeding system and pollinators, natural seed disper-
sal mechanisms and seed quality indicators.
Practical guidance is also provided on the collection of seed from
non-dessicated species, whereas Chapters 10-15 deal with specific
case studies such as the collection of seed of temperate forages, orga-
nization of targeting and collection for conservation in Namibia and
Mexico, the development of the MSB conservation project in South
Africa and forest seed collection in Burkino Faso.
Chapters 15-32 which deal with processing and testing, begin with
a detailed review of seed and fruit structure, and cover the general prin-
ciples of seed processing for storage, explaining how this work is car-
ried out in diverse MSB collections. X-ray analysis can be an invalu-
able tool in determining the status of seed samples before, during or
after seed cleaning. The authors, however, admit that as there is a pos-
sibility for seed to incur. some genetic damage during the process, x-
rayed samples are not returned to the batches. The requirement for
security duplication of collections as a backup, is emphasized. In many
cases, up to half of each collection has already been deposited in the
country of origin. However, if suitable storage facilities do not exist in
these countries, collections are held at the MSB until the country of origin
requests repatriation of half of the seeds. Subsamples of seed material are
not supplied to private individuals, but may be supplied to users in
The theoretical and practical aspects of measuring seed germination
and viability and the relative advantages and disadvantages of each
type of test were assessed. There is a false impression that ‘viable’
seeds are synonomous with ‘germinable’ seeds but viable seeds are not
necessarily capable of germination into normal ‘seedlings’. The vital
stain fluorescein diacetate (FDA) is a fast, dependable and accurate
measure of viability but it can underestimate the viability of certain
species such as the Orchidaceae, due to problems with stain perme-
ability. Other approaches to test viability including the use of in vitro
techniques for the recovery of stored embryos, are outlined. The use of
molecular techniques may assist to indicate whether the problem is in
the embryo or endosperm, what kind of dormancy is involved and the
best method of breaking it.
Predictive classification of seed dormancy in relation to biogeogra-
phy and phylogeny has highlighted that much more research is needed
on within-class taxonomy of seed dormancy of many more species in
all terrestrial biomes. Lastly, methods to effectively remove physiologi-
cal dormancy in certain Western Australian Acacia species, the patterns
of seed germination in response to smoke in plants from the Cape flo-
ral kingdom and the effects of temperature on Mexican cactus seed in
the Sonoran Desert are also dealt with.
The final section of the book (Chapters 33-54) covers storage and
utilization issues including an update on the principles of seedbank
design to provide effective drying facilities for seed collections and,
once dried and packaged, their maintenance at cool temperatures.
Designs for these banks range from very simple ones, to cater for those
with limited budgets, to those involving greater technological and
financial input. As an example, a quick-testing protocol which utilizes
the colour change of self-indicating silica gel was used to select suit-
able seed containers to be used in the MSBP.
Viability equations such as the improved seed viability equation
developed by Ellis & Roberts (1980) which include constants that
explain the empirical effects of moisture content and temperature on
longevity, have formed the basis for ex situ conservation of plant genetic
resources in seed banks. While the temperature constants appear to be
universal, the moisture content is species specific. Seed-viability con-
stants are available for at least 66 species from 26 families. Qualifications
to the seed viability equations have contributed to the recent debate about
the validity of using extrapolated predictions to determine longevities in
the region of millennia for dry cold-stored seeds.
Optimized gene banking procedures are often interpreted as storing
seeds under environmental conditions that give them maximum
longevity. But this narrow view of optimized gene banking is difficult
to define or predict because the interactions that contribute to the life-
span of a seed are not yet completely understood. The authors suggest
Bothalia 35,1 (2005)
113
that the concept of optimal seed banking procedures should be broad-
ened to include the intended use of the germplasm as well as other ex
situ conservation steps, especially collection and regeneration schemes.
Longevity of seeds has been attributed to the presence of an intracellu-
lar glassy state based on the assumption that the high viscosity of glass
decreases molecular mobility and slows diffusion in cytoplasm, in turn
slowing the possible harmful chemical reactions and changes in struc-
ture and chemical composition during ageing. Specific cases such as
that of the Neem (Azadirachta indica ) are discussed in more detail.
A number of gaps have been identified in the knowledge of seed
biology of many Australian species and problematic dormancy is often
encountered. The authors of this particular chapter have identified a
number of areas for further research such as the impact of long-term
storage. Mature seeds of Matnmillaria supertexta are tolerant of desic-
cation to very low moisture contents and of storage at conventional
seedbank temperatures, suggesting they are orthodox seeds and thus
supporting the theory that ex situ conservation of arid and semi-arid
plants is possible by means of seed storage. The impact on viability of
the hydration conditions required to cryo-preserve in liquid nitrogen
was investigated in non-orthodox oily seeds of nine species of coffee.
The hydration window is highly variable across species but appears to
correspond to seed unfreezable water content, suggesting that seed sur-
vival depended strictly on intracellular ice formation.
Ten case studies outlining aspects of seed conservation in a number
of countries including Ethiopia, Spain. Greece, USA (Oregon), the Nether-
lands, Nordic countries, Jordan, India, Morocco and USA (Hawaii) are
detailed in further chapters. The seed of plants from Hawaii displayed
very low incidences of recalcitrance and it is suggested that the require-
ment for long distance dispersal of the original colonizers selected
against recalcitrant seeds. Oceanic islands are likely to show similar
patterns suggesting that they can be stored using conventional tech-
niques for orthodox seeds. It is notable in all these studies that the
effectiveness of the work being executed is greatly enhanced through
partnerships with a variety of individuals and institutions.
The concluding chapters of the book contain editorial perspectives
on the future of seedbanking and a broad spectrum of ideas, concepts
and approaches to seed conservation, some of which have been high-
lighted in other chapters, while others have been derived from other
sources, with a view to encourage active debate amongst seed conser-
vationists.
In the final analysis, the book is structured in such a way that it will
contribute to the updating of current practitioners such as geneticists,
plant breeders, seed biologists and taxonomists, be very useful to orga-
nizations in the process of establishing seedbanks, be valuable to stu-
dents and policy makers, and also update handbooks published by the
international Board for Plant Genetic Resources (IPGRI). Furthermore,
as the gaps in present knowledge are emphasized throughout, potential
avenues for future research are highlighted.
REFERENCE
ELLIS, R.H. & ROBERTS. E.H. 1980. Improved equations for the pre-
diction of seed longevity. Annals of Botany 45: 13-30.
M.M. WOLFSON*
* South African National Biodiversity Institute, Private Bag X101,
0001 Pretoria.
■
'
South African National Biodiversity Institute
FLORA OF SOUTHERN AFRICA
Vol. 5 Part 1, Fascicle 2 Asphodelaceae (First part): Kniphofia
L.E. Codd (2005)
One recently described species was added to the 47 original species treatments by the
late Dr L.E. Codd, former director of the Botanical Research Institute (1963-1973). A
detailed key to species is followed by a description and distribution of each species, with
an accompanying line drawing by Gillian Condy or a black and white copy of paintings by
Cythna Letty; 25 distribution maps; a list of references; and an index to species,
synonyms and exluded names. 240 x 180 mm. pp. 106.
Soft cover: ISBN 1-919976-03-5.
Price SADC countries R80. 00/other countries $20.00.
national
biodiversity
Seed plants of southern tropical Africa: families and genera
Southern African Botanical Diversity Network Report No. 26 (SABONET)
O.A. Leistner (2005)
This book documents the flora of Angola, Malawi, Mozambique, Zambia, and Zimbabwe.
The flora of the five countries, as reflected here, comprises 228 families, 2 032 genera,
and 1 1 637 species. The book is intended as a companion volume to Seed plants of
southern Africa: families and genera, which covered Namibia, Botswana, South Africa,
Swaziland and Lesotho, and is aimed at the serious student of our botanical diversity. It
provides identification keys to all families and genera of seed plants indigenous to and
naturalised in the region. Families and genera are critically described, together with
notes on their distribution and size, both local and worldwide, and with lists of the most
important literature. The latest views on relationships of families are reflected by means
of dendrograms, and the classification of genera within the larger families is given. For
easy reference, genera are arranged alphabetically within their family, and families are
presented in alphabetical sequence within the three major groups: gymnosperms,
dicotyledons and monocotyledons. A comprehensive glossary and an index to family and
genus names conclude the work. Published by the Southern African Botanical Diversity
Network (SABONET), c/o the South African National Biodiversity Institute (SANBI),
Pretoria. 297 x 210 mm. pp. 498.
Soft cover: ISBN 1-919976-07-8.
Price SADC countries R1 50.00/other countries $38.00.
■fkrwering Pfonts of
AFRICA'
FLOWERING PLANTS OF AFRICA
Vol. 59. Plates 2201-2220 (due June 2005)
Twenty full-colour plates and descriptions of plants appear in this biennial series, which
has become a collector's item of the South African flora. This issue includes a new
combination and new status of Hibiscus, a cliff-dwelling species of Aloe from Namibia,
an Ixia with glorious sprays of bright sea-green flowers, four members of the orchid
family and one of various species of Hoodia said to have been used for centuries by the
San and other people in southern Africa to curb the effects of hunger and thirst. The
botanical art is mainly the work of the resident artist, Gillian Condy; other artists
contributing to this issue are Sandie Burrows, Andrew Kamiti, Fay Anderson, Auriol
Batten, Elbe Joubert and Vicki Thomas. There is a guide for authors and artists and an
index to species. 250 x 190 mm. pp. 146.
Soft cover: ISBN 1-919976-15-9.
Price SADC countries R1 90.00/other countries $48.00.
Catalogue on
web site or
available from
SANBI Bookshop
Tel: +27 12 843 5001
Fax:+27 12 804 3211
bookshop situated at the SA National Biodiversity Institute in Pretoria
All postage and clearance fees subject to currency fluctuations
ORDER FORM
April 2005
ORDER FROM:
SANBI Bookshop
Private Bag XI 01
Pretoria
0001 SOUTH AFRICA
Tel.: +27 12 843 5001
Fax: +27 12 804-3211
E-mail: bookshop@sanbi.org
Web site: www.sanbi.org
PERSONAL DETAILS:
TITLE: INITIALS: SURNAME:
BUSINESS/ORGANISATION:
ACCOUNT Number STUDENT Number BOTSOC card number
POSTAL ADDRESS:
POSTAL CODE:
COUNTRY:
TEL. NO. (W): (H):
CELL: FAX:
SUBSCRIPTION: I would like to subscibe to BOTHALIA SADC customer
□ 1 year: SADC R200,00/other countries $50,00 Other
2 years: 2004/5 SADC R360.00/other countries US$ 72.00
2005/6 SADC R380. 00/other countries US$ 95.00
2006/7 SADC R400. 00/other countries US$ 100 00
PAYMENT:
A. Enclosed is my cheque/money order no for the amount of R
or US$ payable to the South African National Biodiversity Institute
B Please charge my credit card:
Visa MasterCard American Express □
Expiry date:
SIGNATURE: DATE:
"SADC countries: Angola, Botswana, Lesotho, Malawi, Mauritius, Mozambique, Namibia, South Africa, Swaziland, Tanzania,
Zambia, Zimbabwe
ANNUAL SUBSCRIPTION (2005)
SADC R200,00 Other countries US$50.00
TWO-YEAR SUBSCRIPTION
2004/5: SADC R360,00 Other countries US$72.00
2005/6: ” R380,00 ” ” US$95.00
2006/7: ” R400,00 ” ” US$100.00
BOTHALIA SPECIALS
Colour plates in Bothalia
Vol. 9, 3 & 4: 27 plates of Kniphofia spp. by Cythna Letty and others
Vol. 16,1: Kniphofia splendida by Cythna Letty
Vol. 27,2: Nivenia parviflora by Fay Anderson
Vol. 28,2: Cyrtanthus erubescens by M.E. Connell
Vol. 29,2: Gladiolus rhodanthus by Auriol Batten
Vol. 30,1: Clivia miniata by Barbara Jeppe
Vol. 32,1: Clivia mirabilis by Auriol Batten
Vol. 33,2: Cyrtanthus macmasteri by Auriol Batten
Erica xwillmorei by R. Mills
Jamesbrittenia bergae by Gillian Condy
Price per volume
SADC: R30.00 / Other: US$6.00
SADC: R10. 00 / Other: US$2.00
SADC: R85.00 / Other: US$17.00
SADC: R95.00 / Other: US$19.00
SADC: R95.00 / Other: US$19.00
All prices include VAT. Prices are subject to change from time to time. Postage is excluded. Please consult the latest catalogue.
Available from: The Bookshop, South African National Biodiversity Institute, Private Bag X101, Pretoria 0001, RSA
Tel. (012) 843-5000 • Fax. (012) 804-3211 • email: bookshop@nbi.ac.za
BOTHALIA
Volume 35,1
May 2005
CONTENTS
1. Three cryptic new species of Aristea (Iridaceae) from southern Africa. P. GOLDBLATT, A.P. DOLD and
J.C. MANNING 1
2. FSA contributions 18: Salicaceae s. str. M. JORDAAN 7
3. Three new species and a new synonym in Strumciria (Amaryllidaceae: Amaryllideae) from southern
Africa. D.A. SNIJMAN 21
4. Lamarck’s new species of Mesembryanthemaceae and the types of their names. P. CHESSELET and M.
PIGNAL 29
5. Cape plants: corrections and additions to the flora. 1. P. GOLDBLATT, J.C. MANNING and D. SNIJMAN . . 35
6. Commiphora kaokoensis (Burseraceae), a new species from Namibia, with notes on C. dinteri and C.
namaensis. W. SWANEPOEL 47
7. Two new species of Asteraceae from Northern and Western Cape, South Africa and a new synonym.
J.C. MANNING and P. GOLDBLATT 55
8. Notes on African plants:
Agapanthaceae. Synonymy in Agapanthus. G.D. DUNCAN 87
Amaryllidaceae. A new variety in the genus Clivia. Z.H. SWANEVELDER, A.E. VAN WYK and
J.T. TRUTER 67
Boraginaceae. Codonoideae, a new subfamily based on Codon. E. RETIEF and A.E. VAN WYK . . 78
Iridaceae. Taxonomic notes on Babiana and Ferraria in arid western southern Africa. P. GOLD-
BLATT and J. C. MANNING 71
Lycoperdaceae-Gasteromycetes. Bovista capensis , the correct name for Bovista promontorii.
J.C. COETZEE and A.E. VAN WYK 74
Lycoperdaceae-Gasteromycetes. The identity of Lycoperdon complanatum Desf. and its nomen-
clature implication. J.C. COETZEE and A.E. VAN WYK 76
Poaceae. Name used in the FSA region for the Cymbopogon excavatus-caesius-giganteus Com-
plex. L. FISH 82
Poaceae. Notes on Eragrostis. L. FISH 80
Hyacinthaceae. Ornithogalum juncifolium var. emsii, a new cliff-dwelling Ornithogalum from
Eastern Cape, South Africa. E.J. VAN JAARSVELD and A.E. VAN WYK 82
Orobanchaceae. A new species of Harveya from Western Cape, South Africa. J.C. MANNING
and P. GOLDBLATT ’ 89
Proteaceae. A new Leucadendron (Proteeae) from Western Cape, South Africa. J.P. ROURKE . . 63
Rubiaceae. Infraspecific taxa in a southern African Pavetta species. P.P.J. HERMAN 84
Salicaceae. Salix: the correct application of the name Salix mucronata, and a new combination.
R.H. ARCHER & M. JORDAAN 92
Scrophulariaceae. Nemesia zimbabwensis, a new record for the FSA region with notes on its phyto-
geographical significance. S.J. SIEBERT and A.E. VAN WYK 69
9. The systematic value of the leaf indumentum in Lobostemon (Boraginaceae). M.H. BUYS 93
10. Ovule-to-seed development in Dovyalis caffra (Salicaceae: Flacourtieae) with notes on the taxonomic
significance of the extranucellar embryo sac. E.M.A. STEYN, A.E. VAN WYK and G.F. SMITH . . 101
1 1 . Miscellaneous notes:
Agapanthaceae. Chromosome counts in the genus Agapanthus. M. MUZILA and J.J. SPIES ... 109
12. Book reviews Ill
Abstracted, indexed or listed in • AETFAT Index • AGRICOLA • AGRIS • BIOSIS: Biological Abstracts/RRM • CABS • CABACCESS • CAB
ABSTRACTS • ISI: Current Contents. Scisearch, Research Alert • Kew Record of Taxonomic Literature • Taxon: reviews and notices.
ISSN 006 8241
© Published by and obtainable from: South African National Biodiversity Institute, Private Bag X101, Pretoria 0001, South Africa. Tel. (012) 843-5000.
Fax (012) 804-3211. email: bookshop@nbi.ac.za website: www.nbi.ac.za/pubs. Typesetting and page layout: S.S. Brink (NBI). Printing: Afriscot
Printers. PO. Box 75353, Lynnwood Ridge, 0040 Pretoria. Tel (012) 349-2800/1 . Fax (012) 349-2802.
I