Tropical East Africa Research Articles

The Genus Crossandra (Acanthaceae) in the African Continent

The main purpose of this paper is to summarise information gained during work carried out on the genus Crossandra for Flora of Tropical East Africa and Flora Zambesiaca. It was felt that a proper monograph of the whole genus should not be attempted until the Madagascan and Indian species have been studied in full. It was therefore decided that a short synopsis would be an appropriate way to show the present knowledge of the taxonomy of the African species of Crossandra. The present paper deals with all taxa known to occur in Africa and Arabia. Apart from the taxa treated in this paper, there are a further 15-20 species on Madagascar and one species on the Indian subcontinent.

A New Species of Tricliceras (Turneraceae)

In the same year that the account of the Turneraceae for the Flora of Tropical East Africa was published (1954) I discovered, in the Tsavo National Park, Kenya, the striking new species of Tricliceras Thonn. ex DC. described below. At the time of the Flora account the genus was known as Wormskioldia Thonn. but the totally overlooked Tricliceras was resuscitated by Fernandes (1975: 15); it is a pity the well known name was not conserved.

Notes on African Euphorbiaceae: XXIV: Drypetes (ii)

Summary. A new species and two new varieties are described for the genus Drypetes from Tanzania. The Phyllanthoideae of East Africa were clearly the least-known subfamily of Euphorbiaceae at the time of publication of the 'Flora of Tropical East Africa' account, because, since its appearance, several changes have been found necessary. Two new genera have had to be described for the subfamily, and it is now apparent that the treatment of the genus Drypetes was far from complete. Collections of this genus made by Lovett, Lovett & Congdon, by Bidgood, Mwasumbi & Vollesen, by Gereau & Lovett, and by the author with Pope & Goyder in recent years have included the following novitates: Drypetes gerrardino7des A. Radclife-Smith sp. nov. inter species sectionis Stemonodisci (Pierre) Pax propter discum in centro floris maris stamen gerentem ponenda, et hic D. argutae (Muell. Arg.) Hutch. proxima, sed ab ea omnibus partibus glabris, foliis ovatis nec lanceolatis, stipulis non basi fimbriatis, staminibus paucioribus, fructibus fusiformibus nec subglobosis recedens, necnon D. parvifoliae (Muell. Arg.) Pax & K. Hoffm. arcte affinis, sed ab ea foliis minus valde asymmetricis, stipulis aliqua ex parte indurescentibus ac persistentibus nec facile cadentibus, pedicellis in ambobus sexibus plerumque longioribus, staminibus numerosioribus, ovario glabro nec tomentoso, fructibus fusiformibus nec subglobosis discedens. Typus: Tanzania, Bidgood, Mwasumbi & Vollesen 794 (holotypus K; isotypi B, BR, C,

A New Species of Hildegardia (Sterculiaceae) from Somalia

An interesting collection of Hildegardia made by Gillett & Hemming in Somalia came to our attention while we were preparing a treatment of the Sterculiaceae for the "Flora of Tropical East Africa." Initially determined as the West African H. barteri (Mast. in Oliv.) Kosterm., it is clearly not that species nor does it match the only species known previously from East Africa, H. migeodii (Exell) Kosterm.

The Identity of Clerodendrum kibwesense Moldenke (Verbenaceae)

Whilst preparing an account of the genus Clerodendrum for the Flora of Tropical East Africa it was necessary to ascertain the identity of the above plant. The holotype, stated to be in Brussels, could not be found there. Having read the quite detailed description several times, and recalling which Verbenaceae grew in the well-collected area of the type locality, I guessed that it was in fact a species of Premna common in the locality. Moldenke drew attention to the fact that it resembled a Premna. A search at Kew revealed

A New Species of Vitex (Verbenaceae) from Tanzania

During the preparation of an account of the family Verbenaceae for the Flora of Tropical East Africa several taxa were found to be undescribed. These will mostly be validated in the flora account but the name of the Vitex described below is required for another purpose. In the Mpwapwa district of central Tanzania a species of Vitex, long confused with the more widespread V. strickeri Vatke & Hildebrandt, appears to be not uncommon. Local collectors such as B. D. Burtt and Mr & Mrs

New Species of Panicum (Gramineae) from Southern Tropical Africa

I have recently prepared an account of the genus Panicum for the Flora Zambesiaca; 82 species are recognized, which is substantially more than those recorded for the Flora of Tropical East Africa (68 species) and no doubt reflects the greater richness of the grass flora of the Zambesi region. Not surprisingly a few specimens could not be accommodated in the known taxa of Panicum in Africa and they are described for the first time in the following pages.

A Note on the Typification of Vallisneria spiralis (Hydrocharitaceae)

Vallisneria spiralis L. is an aquatic species which is common and widespread in the tropics and warm temperate regions. It is also well-known in cultivation and is frequently planted as an oxygenator in tropical aquaria. While preparing an account of Hydrocharitaceae for Flora of Tropical East Africa, I was unable to find any reference to a type for this species, and since it is such a well-known plant it seemed desirable to investigate this further. In LINN there are two specimens (Savage Catalogue nos. 1157-1 and 1157-2) which, without doubt, can be assigned to V. spiralis. However these appear to have been collected after 1753, and as there is no reference to them in any of Linnaeus' work they must be discounted as possible type material. Before 1758, Vallisneria was described in Linnaeus' Hortus Cliffortianus (1738), Flora Lapponica (1737) and Genera Plantarum (1738). However there are no specimens of Vallisneria in Linnaeus' Lapland Herbarium at the Institut de France, Paris, nor in other herbaria containing Linnaean material (C. Jarvis pers. comm.). There is also a mention of it in van Royen's Florae Leydensis prodromus (1740) but, again, no specimen can be found in van Royen's herbarium at L. In Species Plantarum three polynomials are cited in synonymy. The first and third of these are taken from Micheli's Nova plantarum genera (1729), where they are accompanied by detailed and accurate illustrations in Plate 10, Figs. 1 & 2 of the book. Although Micheli believed that these polynomials described separate taxa, they do in fact refer respectively to the female and male plants of V. spiralis and this was correctly acknowledged by Linnaeus in Species Plantarum. Unfortunately there are no specimens corresponding to V. spiralis in Micheli's herbarium at FI. The second polynomial is quoted as being taken from Boccone's "Museo di piante rare della Sicilia", p. 289 (1697). However this would appear to be incorrect as Boccone's work only has 196 pages and there is no mention of any such polynomial in the index. The same polynomial and reference are given by Micheli, and therefore it is probable that Linnaeus obtained them from this source, and as a result the error was perpetuated in his own work. The correct source of this polynomial cannot be traced.

The Identity of Clerodendrum eupatorioides Bak. from Cameroon

Summary. Clerodendrum eupatorioides Bak. is Eremomastax speciosa (Hochst.) Cufod. (Acanthaceae). Whilst examining Clerodendrum in connection with an account of the Verbenaceae for the Flora of Tropical East Africa I noted that Huber (1963:442) mentioned he considered the plant described by Baker as Clerodendrum eupatorioides doubtfully belonged to the genus. I examined the type of this and a glance is enough to see that the specimen is an 'Acanth'; the shrinking of the basal part of the petioles, the cystoliths on the leaves, leaf shape and calyx structure give it away at once. It is very surprising how many eminent botanists have examined this without realising what it was. The fact that no other specimen had ever been associated with it is now explained. A few minutes established its identity with an Acanthacea widespread in Africa and the synonymy is set out below in full; as far as I am aware the information has not previously been brought together before, nor is any Flora likely to present it in the future.

Trichodesma inaequale Edgeworth (Boraginaceae), a Species of Africa and India

Summary. The correct name is established for a relative of Trichodesma indicum (L.) Lehm. During preparation of an account of the family Boraginaceae for the Flora of Tropical East Africa, treatment of the genus Trichodesma R.Br. was made very easy by the previous work of R. K. Brummitt (1982, 1985). Only one of the species he did not deal with presented any problems. Since this occurs in both India and Africa a brief note was considered necessary, as the Flora account itself is not always consulted by workers outside the area with which it deals. In an excellent short paper Sedgwick* (1919) distinguishes between two species of Indian Trichodesma, frequently confused under the name T. indicum (L.) Lehm. ('R.Br.') and points out that although C. B. Clarke (1883) maintained two species he did not understand the fundamental differences between them. Sedgwick sets forth numerous characters based on his own careful field studies. The most striking character is that in T. indicum the anthers are 'combined into a long pointed cone remaining united when the corolla is split open and impossible to separate even by force, always far exserted beyond the patent limb' whereas in the other species which he refers to as T. amplexicaule 'DC. non Roth' they are 'merely connivent but not combined, separating at once when the corolla is split open, always included with the corolla throat'. He points out that Roth's description of his T. amplexicaule shows it to have the stamens of T. indicum so that the name is not available for the other species. T. amplexicaule is in fact often treated as a variety of T. indicum.

Two New Species of Parasitic Harveya (Scrophulariaceae) from East Africa

The genus Harveya Hook. (Scrophulariaceae) is composed of some 40 species of root parasites with a distribution from Madagascar to South Africa and northwards through East Africa to Rwanda, Sudan and Ethiopia, with the majority of the species in southern Africa. While a few of the species are widespread in savanna grassland, most of them have a confined distribution in forest where they parasitise tree roots. The vegetative parts are severely reduced, often white and drying black, with large coloured flowers at groundlevel. During research for the 'Flora of Tropical East Africa' two more species have been recognised and are here described.

The Typification of Nymphaea lotus L.

During the preparation of an account of the family Nymphaeaceae for the Flora of Tropical East Africa it has been necessary to typify Nymphaea lotus L., traditionally the name for the true Egyptian sacred lotus, a waterlily widespread in tropical and subtropical Africa, Madagascar and an outlying locality in NW. Roumania and also naturalised in the New World. Linnaeus (Sp. Pl.: 511 (1753)) cited four syntypes in the protologue, viz. 'Nymphaea foliis cordatis dentatis. Fl. Zeyl. 194*. Nymphaea indica, flore candido, folio in ambitu serrato Sloan. jam. 120. hist. 1 p. 252. Raj. suppl. 630. Lotus aegyptia. Alp. aegypt. 103. exot. 214. t. 213.216.218.220.222.226. Ambel. Rheed. mal. 11. p. 51., t. 26. Habitat in calidis Indiae, Africae, Americae.' There is also a Hasselquist specimen in the Linnaean Herbarium (673.5) which may constitute a fifth syntype. Choice of the Hermann specimen, the basis of Fl. Zeyl. 194, is clearly unacceptable for maintenance of traditional usage since the Indian subcontinent populations are now referred to as either N. lotus L. var. pubescens (Willd.) Trimen, N. pubescens Willd. (1799) or N. rubra Roxb. ex Salisb. (1805). Willdenow (Sp. P1., ed. 4, 2: 1153 (1799)), however, it must be noted, still cited Fl. Zeyl. 194 under N. lotus. The use of the name N. nouchali Burm.f. (1768) for the Indian plant, e.g. by Matthew (Fl. Tamilnadu Carnatic 3: 27 (1983)) is incorrect. Examination of the type kindly sent on loan from Geneva demonstrates that Burman's name is the earliest for a common Old World blue water-lily as pointed out by Van Royen (Sertulum Papuanum, 5, Nova Guinea Bot. 8: 110 (1962)). Although Conard (The Waterlilies: 198 (1905)) saw a Burman specimen in Herb. Delessert (G) and considered it the same as N. pubescens Willd., Burman's description (Fl. Indica: 120 (1768)) distinctly mentions blue flowers including blue anther-tips. The Sloane reference concerns a quite different plant-N. ampla DC.-as was pointed out by Caspary long ago (see Conard, The Waterlilies: 24 (1905)) and thus is also clearly unacceptable if traditional application is to be maintained. The Rheede name Ambel also refers to the Indian N. pubescens. The specimen in the Linnaean herbarium carries the name Lotus 3 and the sign for Hasselquist who collected in Egypt and Palestine and is clearly a possible choice, but to me the best is one of the plates used by P. Alpino [Alpini, Alpinus] (1553-1616) to illustrate his accurate account of the Egyptian lotus, presumably drawn up from life during his extensive visit to Egypt in 15801583, and from specimens sent to him later by M. Carbonus. These are historically the most obvious choice and I hereby choose Prospero Alpino, De plantis exoticis, illustration on p. 213 (1627) as lectotype (actually com-

The Identity of Clerodendrum manettii Vis. (Verbenaceae)

Summary. Clerodendrum manettii Vis. is shown to be a synonym of C. tomentosum (Vent.) R. Br. The question of the identity of Clerodendrum manettii Vis. arose whilst I was preparing an account of the genus Clerodendrum for the Flora of Tropical East Africa and through the kindness of the Director of the 'Istituto di Botanica e Fisiologia Vegetale, Universita di Padova' I have been able to solve the problem satisfactorily. Giuseppe Manetti (fl. 1831-58) in a second supplement (1845) to his list of plants growing in the royal botanic garden at Monza near Milan records C. splendens G.Don and gives the locality Sierra Leone which is indeed where the type of Don's species was collected. Seeds of this or cuttings were sent to Visiani (1801-78) at Padua and when they flowered Visiani was aware that the plant could not be Don's species and he described and figured it as C. manettii and assumed it came from Sierra Leone. Manetti clearly obtained seed of doubtful provenance, identified the plants which resulted as C. splendens and gave the known locality of that species. It was probably not meant to be a clear indication of the origin of his material. This Manetti, who was a head gardener, should not be confused with the Florence physician Xarerio Manetti (1723-85) commemorated by the genus Manettia. Baker (1900) suggested that the West African C. buchholzii Gilg, described much later, might be identical with C. manettii 'a garden plant of uncertain origin'. When Berthold Thomas revised the genus for tropical Africa (1936) he also was convinced Visiani's name must apply to a West African species and accepted Baker's suggestion of affinity. Since C. buchholzii is recorded for East Africa, in fact occurs quite commonly, the problem was brought to my attention and clearly needed solution. Thomas stated in his revision 'eine Anfrage in Rom bzw. Florenz nach dem Typus blieb leider ergebnislos' and appeared to be unaware that Visiani was associated with Padua. My request there resulted in a loan of two sheets both numbered 7191 which I am

Studies in Cyperaceae in southern Africa: 13. The Bulbostylis schoenoides complex (Cyperaceae) in Africa, south of the Sahara

The Bulbostylis schoenoides (Kunth) C.B. Clarke complex is distributed through sub-Saharan Africa from west and east tropical Africa to Angola and eastern South Africa and Madagascar, usually in damp situations at higher altitudes. Five species, B. schoenoides, B. cinnamomea (Boeck.) C.B. Clarke, B. megastachys (Ridley) C.B. Clarke, B. scleropus C.B. Clarke and B. erratica (Hooker f.) C.B. Clarke, are considered as falling within the limits of the complex. A brief account of each species is given. Relationships are considered, including opinions expressed by recent workers. Attention is drawn to the need for ecological assessment of habitat conditions, since the complex appears to comprise organisms which are linked clinally due to their apparent response to latitude and/or microhabitat. Alternative possibilities by which the relationships found to exist in nature may be incorporated into formal classification are proposed. Reduction to subspecies of the four associated species under the oldest described species, B. schoenoides, is favoured, but opinions are called for so that some consensus may be reached before changes are instituted, thus avoiding unnecessary nomenclatural proliferation. Further information on the complex is requested. Die Bulbostylis schoenoides (Kunth) C.B. Clarke-kompleks is versprei oor die hele sub-Sahara gebied van Afrika vanaf die wes- en oos-tropiese gebiede tot Angola en die oostelike gedeeltes van Suid-Afrika. Die plante kom gewoonlik onder nat hoogliggend toestande voor. Vyf spesies, B. schoenoides, B. cinnamomea (Boeck.) C.B. Clarke, B. megastachys (Ridley) C.B. Clarke, B. scleropus C.B. Clarke en B. erratica (Hooker f.) C.B. Clarke word as deel van die kompleks gereken. Die verwantskap van die groep word bespreek en die opinies van verskeie werkers aangehaal. Aandag word gevestig op die noodsaaklikheid van ekologiese inagneming van habitatstoestande. Veral daar die kompleks organismes insluit wat gekoppel word aan hulle skynbare reaksie op hoogte en/of mikroklimaat. Alternatiewe moontlikhede wat die verwantskappe wat natuurlik voorkom in gedagte hou ten opsigte van formele klassifikasie word voorgestel. Verlaging na die vlak van subspesies van die vier geassosieerde spesies onder die oudste beskrewe spesies B. schoenoides word voorgestaan. Voordat enige veranderinge aangebring word, en sodat ooreenstemming bereik kan word, word verdere inligting en bespreking oor die kompleks verwelkom. Die prosedure sal onnodige proliferasie van klassifikasie uitskakel.

A Revision of Ethulia (Compositae: Vernonieae)

Initial work on this genus was done by M.G.G. when employed in the East African Herbarium, Nairobi, prior to June, 1982. It was then put on one side when he was employed by the Ethiopian Flora Project to work at Kew on various other families. Later C.J. revised the East African species for the Flora of Tropical East Africa and as the taxonomic concepts of both authors agreed in all but a very few details, it was decided to pool results in the one paper. Authorities are indicated independently for each new taxon or combination. The Director of the Rijksherbarium, Leiden, is thanked for the loan of material to the East African Herbarium. The authors also thank the Directors of the following herbaria for the loan of critical material to Kew and/or for the provision of visiting facilities to one or other of the authors: BM, BR, BRLV, ETH, FT, MPU, P. All cited collections have been seen by at least one of the authors unless otherwise indicated.

Holmskioldia (Verbenaceae), a Genus New to the Flora of Tropical East Africa

Summary. The genus Holmskioldia Retz. (Verbenaceae) is reported from Tropical East Africa for the first time, and H. gigas Faden is described. encountered a tree which none of us had seen before. In its general appearance the tree bore a decided resemblance to Tectona grandis L. f., which is occasionally cultivated along the Kenya coast, but the distinctive fruiting calyx precluded that genus. Only one tree and a sapling were found, and the only materials that could be collected were fallen fruits, leaves and a small sample of bark. (Some of the leaves collected later proved to belong to a species of Ficus.) Mr Gillett subsequently sent fruits from this collection to the Royal Botanic

A New Genus Nogalia (Boraginaceae: Heliotropioideae) from Somaliland and Southern Arabia

Whilst preparing an account of the Boraginaceae for the Flora of Tropical East Africa I attempted to name as much unidentified Heliotropium material collected in northeast Africa as possible. The genus is very complex in this region, often with affinities in the Middle East as well as with other areas of tropical Africa. A revision of the described and the many undescribed species would be a lengthy task. One outstanding Somali plant has always puzzled me ever since I examined it a few years after it was first received at the East African Herbarium for naming in 1945. Study of three more gatherings, including one in fruit, demonstrates that the plant cannot be accommodated in Heliotropium nor in the related Tournefortia and Argusia. A new genus has therefore been erected. A little hunting demonstrated that there was a very close relative in southern Arabia described long ago by Baker as Heliotropium drepanophyllum; the relationship is, indeed, so close there is difficulty deciding whether the Somali plant is a distinct species or only a subspecies. It is surprising this relationship has escaped the notice of routine determiners for over 40 years but worthy of note that the first person (the hand seems to be that of J. H. Hemsley), to examine the Kew duplicate of Glover and Gilliland 1110 annotated it '? Boraginaceae gen. indet.'. No fruits of the Arabian plant were available to Baker and his assignment of H. drepanophyllum to section Radula Bunge was a guess based on superficial resemblance. Despite similarity in floral characters the fruit precludes relationship with Heliotropium as the classification of the subfamily stands at present. Boissier (1875: 125) in his generic description of Heliotropium states 'vel rarius in caryopsem monospermam concretae'. This, however, refers to H. supinum L. in sect. Piptochlaena (Don) DC. (now subg. Piptochlaena (Don) Riedl) where the single nutlet is due to reduction; in the tropics this species often has 1-4 nutlets. There is lack of agreement as to how the rest of the subfamily Heliotropioideae should be classified, either as a single genus Tournefortia L., as advocated by Edmondson in Davis (1978: 246), or as two--Tournefortia L. and Argusia Boehmer (= Messerschmidia Hebenstreit)-as proposed byJohnston (1935: 161; 1951: 118), Dandy (1967, in Heywood 1972: 256), Valentine in Tutin (1972: 84), Riedl (1967: 8-9) and Heine (1976: 108). Argusia is a small genus of saline areas with a specialised fruit comprising A. argentea (L.f.) Heine, a widespread littoral plant of the Old World Tropics, A. sibirica (L.) Dandy (= Messerschmidia sibirica (L.) L.) from Roumania across Central Asia to Japan and A. gnaphalodes (L.) Heine (Mallotonia gnaphalodes (L.) Britton, Messerschmidia gnaphalodes (L.)

Reconstructing 19th-Century Vegetative Patterns in Kenya

Reconstructing 19th-Century Vegetative Patterns in Kenya

Notes on Peucedanum (Umbelliferae) and Allied Genera in Central and East Africa

Summary. The plant hitherto known as Peucedanum heracleoides Bak. is transferred to Physotrichia as P. muriculata (Welw. ex Hiern) Droop & Townsend. Erythroselinum and Lefebvrea are maintained as a genera distinct from Peucedanum and L. droopii sp. nov. is described from Tanzania. Peucedanum aberdarense Wolff is reduced to var. bipinnatum of P. friesiorum Wolff and P. canaliculatum Verdcourt to subsp. australe of P. harmsianum Wolff. The identity of P. winkleri Engl. is discussed. Five new tropical African Peucedanum species are described. The following are intended chiefly as precursory notes for the account of the Umbelliferae for the 'Flora of Tropical East Africa'. During the course of this I have profited greatly from discussion with Mr Stephen Droop, formerly of the Conservation Unit of the Royal Botanic Gardens, Kew. He had commenced a Ph.D. thesis for the University of Reading on Peucedanum and its allies at the same time as my work on the Umbelliferae for F.T.E.A. began, although I was at that time unaware of the fact. We have not totally agreed in every respect, so that the decisions taken in the current paper must be taken as my responsibility; but Mr Droop has discussed the work freely with me and I am grateful for his valuable help. His own taking up of full-time employment has halted work on his thesis, while my own impending retirement has necessitated acceleration towards completion of my F.T.E.A. account; hence our cooperation has been less full than it might otherwise have been.

New Species and Taxonomic Changes amongst the Succulent Trees and Shrubs of Euphorbia subgen. Euphorbia from East Africa

These descriptions and systematic rearrangements became necessary during work on the genus Euphorbia for the Flora of Tropical East Africa. The geographical subdivisions are those used for the Flora. Thanks are due to Miss C. M. Wilmot-Dear for checking the Latin descriptions, to Mrs Christine GreyWilson for the illustrations and especially to Dr H. Balslev of Aarhus University, Denmark, for the use of his excellent, very comprehensive field-notes on this difficult group of species, resulting from a study he made in 1974-5 when a student at the University. Specimens are in Kew unless otherwise stated.