Syst. Geogr. Pl. 79: 193-198 (2009) A historical record of Polysphaeria macrophylla (African Rubiaceae, Octotropideae) from Bioko. With a neotypification of the species Elmar Robbrecht National Botanic Garden, Domein van Bouchout, B - 1860 Meise, Belgium robbrecht@br.fgov.be Abstract. – The genus Polysphaeria was only in 2007 recorded for the Gulf of Guinea Islands. A specimen collected on Bioko (‘Fernando Po’) by Mildbraed in 1911 remained unnamed in the Rubiaceae until 1967 when it was erroneously filed under Coffea. It is in fact Polysphaeria macrophylla K.Schum., also known from the African mainland (Liberia to Cameroon). Polysphaeria macrophylla is easily distinguished from all other species of the genus by the presence of swollen internodes inhabited by ants, an apparent feature overlooked in the past. The species is neotypified and illustrated, and a key to the Guineo-Congolian species of the genus is provided. Key words: taxonomy, biogeography, Guineo-Congolian flora, myrmecophily. During routine determinations of unnamed Rubiaceae of HBG, I was able to identify a historical specimen as Polysphaeria macrophylla K.Schum. Johannes Mildbraed collected it when he explored the then-Spanish island “Fernando Poo” (at present Bioko), one of the Gulf of Guinea Islands, as a part of the great second German expedition to Central Africa (Mildbraed 1922). Polysphaeria, an Afromadagascan genus of about 25 species, is on the whole poorly documented; only a short conspectus of the genus (Verdcourt 1980) and some flora treatments are available. It is therefore judged to be worthwhile to publish here the information gathered during the identification of this historical record. Identification. In 1967, Wiemann, a staff member of HBG, identified Mildbraed 6796 as “cf. Coffea liberica Bull ex Hiern.” The specimen was seen by Stoffelen for his monograph of Central African Coffea (Stoffelen 1998) and annotated by him “non Coffea liberica”. The genus Coffea can indeed immediately be ruled out because of the absence of the characteristic ‘coffee beans’, seeds with a deep and curved adaxial invagination. Mildbraed’s specimen has mature fruits with two hemi-ovoidal, deeply ruminate seeds, very characteristic for Polysphaeria. In GuineoCongolian Rubiaceae, the combination ruminate seeds and an inflorescence (infructescence) position which is axillary and paired at the nodes, is only recorded from the genus Polysphaeria (Robbrecht & Puff 1986: table 8). The combination also exists in the Afromontane genus Galiniera occurring from Kivu (RD Congo) to East Africa and the Zambezian genus Kraussia, these two also members of the tribe Octotropideae but not recorded from the Guineo-Congolian rain forest block. Systematics and Geography of Plants is subject to copyright. All rights reserved.© 2009 National Botanic Garden of Belgium Permission for use must always be obtained from the National Botanic Garden of Belgium. ISSN 1374-7886 Figure 1. Polysphaeria macrophylla. Caption see next page. Syst. Geogr. Pl. 79 (2009) 䉳䉳 Figure 1. Polysphaeria macrophylla: A, fruiting twig; B & C, myrmecophilous part of an internode, cut longitudinally in C; D, pattern of the intersecondary venation; E, two paired inflorescences at the terminal node; F, flower bud subtended by bracts and bracteoles, the upper pairs fused into calyculi, the lower pair free; G, parts of F laid open; H, corolla & style in position; I, inside of half corolla with two attached stamens; J, style; K, calyx and ovary cut longitudinally; L, placenta with one pendulous ovule; M, seed, adaxial view & N, its median cross-section showing rumination and embryonal cavity with transversely cut cotyledons. Drawn after material from Cameroon (vegetative parts except stipules, fruits and seeds after Sonké 305, BR; stipules, inflorescence and flower after Goetghebeur 4860, BR). The specific identification too leaves no doubts. Polysphaeria macrophylla is the only species of the genus with myrmecophilous twigs. Some of the internodes are distinctly swollen, perforated in their uppermost part, and inhabited by ants. This diagnostic feature is represented at the base of the sole twig portion collected by Mildbraed. The myrmecodomatia were not mentioned nor used in his key to species by Verdcourt (1980), while they are very diagnostic. Verdcourt also considered the identity of P. macrophylla problematic because no original material survived. I therefore judge the present study a good occasion to designate a neotype for the destroyed holotype Preuss 1382, see below. In the protologue, Schumann (1903) stressed the sessile inflorescences and very large leaf-blades as diagnostic. These two elements are largely sufficient to establish the identity of Schumann’s P. macrophylla. Indeed, the species occurs outside the two diversity centres of Polysphaeria, viz. East Africa in the wide sense (including Sudan, Somalia, Katanga and Zambia) and Madagascar, each with ten endemic species. Only three species are Guineo-Congolian (phytogeographical categories here and below follow White 1979). Polysphaeria subnudifaux Verdc. is a linking element of the Lower Guinean and Congolian subcentres of endemism. It has much smaller leaf-blades than the ones described for Preuss 1382. The two other species, P. arbuscula and P. macrophylla are linking elements of the Upper and Lower Guinean subcentres of endemism. It is unlikely to confuse them because they were simultaneously described by Schumann (op. cit.), who offered obvious diagnostic criterions, viz. the size and shape of the leaves and the sessile or pedunculate nature of the inflorescences. To these features the absence or presence of myrmecodomatia (they were obviously not collected in the parts collected by Preuss) should be added. The Guineo-Congolian species of Polysphaeria may be distinguished with the key herewith. Neotypification. Most original Preuss specimens were destroyed in Berlin, including number 1382, the holotype of Polysphaeria macrophylla. Preuss duplicates are conserved in BM, BR, EA, HBG, K and P, but according to Aluka (www.aluka.org; accessed 23 September 2009) this is not the case for Preuss 1382. The elements set out in the preceding section permit to designate a neotype, that I have chosen from the area in SW Cameroon were the holotype was collected. It also corresponds as closely as possible to Schumann’s description. The collector of the neotype mentions that the species is common in that locality; two more specimens were there made by him, Sonké 299 and Sonké 305, here used for preparing figure 1. The first record for Bioko. A verification of the literature learned that the genus remained long unrecorded from the island. Mildbraed himself published floristic lists of all the areas investigated during the expedition. He identified his specimen 6796 only at family level, and it is not included in his “Pflanzenliste von Fernando Poo” (1922: 175-195). The species’ presence in Bioko remained 196 E. Robbrecht, Polysphaeria macrophylla (African Rubiaceae, Octotropideae) Key to Guineo-Congolian species of Polysphaeria 1. Inflorescences or infructescences on a distinct, mostly reflexed peduncle 3-7 mm long. Leaf blades narrowly ovate (‘lanceolate’). Upper and Lower Guinea. . . . P. arbuscula K.Schum. 1’. Inflorescences or infructescences sessile. Leaf blades broader, ovate or elliptic. . . . . . . . . . 2 2. Upper part of some internodes swollen and perforated, inhabited by ants. Leaf blades very large, up to 22 ⫻ 8 cm. Upper and Lower Guinea. . . . . . . . . . . . . . . . . P. macrophylla K. Schum. 2’. No ant-inhabited structures. Leaf-blades much smaller, 5-12(15) ⫻ 2-5 cm. Lower Guinea and Congolia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P. subnudifaux Verdc. unrecorded in other floristic works on the island (Exell 1973) or covering it (Hutchinson & Dalziel 1931; Hepper 1963). The presence in Bioko was not mentioned in a short revision of Polysphaeria (Verdcourt 1980) and in the World Checklist of Rubiaceae (Govaerts & al. 2009) either. Recently, however, Davis & Figueiredo (2007) for the first time reported the presence of Polysphaeria macrophylla on Bioko. The distribution pattern of Polysphaeria macrophylla compares well many other angiosperms. The flora of Bioko has indeed many elements in common with the opposite African mainland, because the island was connected by a broad land bridge during the last ice age (Figueiredo 1994; Jones 1994). Present status on the island. Exell’s (op. cit.) floristic report on Bioko was based on his own visits to all the Gulf of Guinea islands (Frodin 2000: 489), and the species should normally not have escaped his attention. Davis & Figueiredo (2007) cited only one representative specimen, collected in 1990, near the River Tiburones along the road from Malabo to Luba. This seems to indicate that the species is nowadays rare on the island. The lowland forests up to 800 m were much disturbed in the 19th century, for the plantation of cocoa. In 1992 it was estimated that there were only small forest patches along the SE of the island, in addition to the forest along the South coast (Juste & Fa 1994). Both collections known of Polysphaeria macrophylla originate from the West of the island. Polysphaeria macrophylla K.Schum., Bot. Jahrb. Syst. 33: 349 (1903). Types: Cameroon, zwischen Isongo [= Izongo Point, 04 05 N; 09 01 E] und Bakingele [= Bakingili, 04 04 N; 09 02 E] im Urwald, Preuss 1382 (B† holo-). Cameroon: SW Province, Idenau, Onge [Mbonge River, 04 14 N; 08 59 E], 13 Dec. 1992, Sonké 304, BR S.P. 510 490; neotype designated here; isoneo-: BR S.P. 514273; K not seen; SCA not seen). Shrub 1-3 m tall [see note], with straight lateral branches; slightly anisophyllous (leaves on flowering branches somewhat larger and ± differing in shape from leaves on main branches); all parts drying blackish; vegetative parts glabrous; flowering branches at the base with one internode with swollen and perforated upper part inhabited by ants. Leaves with petioles 4-8 mm long; blades obovate, elliptic or obovate, 35-90 x 100-230 mm, cuneate (always so in leaves of main stem?) to rounded at base, long acuminate at apex; domatia absent. Interpetiolar stipules triangular, slightly fused at the base, more widely triangular when situated at the base of two inflorescences. Inflorescences ca. 8-flowered, congested and sessile, axillary and paired at the nodes, in bud already present on the terminal node of a flowering twig before the development of its leaves and then 197 Syst. Geogr. Pl. 79 (2009) looking terminal. Flowers sessile, 4-merous, the ovaries hidden in a series of triangular bracts and bracteoles embracing one another; upper bracts and bracteoles with a consistency much like the calyx and fused into a cupular structure; calyx cupular, its margin truncate or with minute teeth, slightly diaphanous, in the dried state with white spots corresponding to translucent crystal sand, outside glabrous, inside densely hairy; corolla pink, comprised of a gradually widening tube 5 mm long and four triangular lobes 3 mm long, glabrous outside, with hairy throat; anthers sessile inside corolla tube, 3 mm long, with triangular appendix at the top; style 2-armed, hairy in its upper 2/3, only the arms exserted; ovary glabrous, 2-celled, in each cell a placenta attached to top of dissepiment and with one pendulous ovule. Fruits sessile, globular, ca. 10 mm in diameter, crowned by remnants of the calyx, at maturity black, only a few per node; 2 hemispherical or rarely 1 ± spherical seed per fruit. Seeds ca. 8 mm high, with apical hilar scar; seed coat longitudinally grooved at the surface, the grooves corresponding with the rumination of the endosperm. – Fig. 1. Distribution. Reported from Liberia, Ghana, Nigeria, Cameroon (Verdcourt 1980) and Bioko (Equatorial Guinea) (Davis & Figueiredo 2007). Additional record from Bioko (Equatorial Guinea): Adolf Friedrich Herzog zu Mecklenburg. Zweite Expedition nach Central-Afrika 1910–1911. Fernando Poo: San Carlos (Westküste) Unterer Tropenwald bis 400 m ü. M. Ges. Oktrb 1911 von Dr. J. Mildbraed. No. 6796 (HBG). Note. Verdcourt (1980) already doubted the protologue where stating “Der Strauch wird 15 m hoch”, a size that is better applicable to a tree than to a shrub. This appears to be an error for 1.5 m. Acknowledgments. – I am grateful to Olivier Lachenaud for selecting unnamed Rubiaceae in HBG, and to the curator of this herbarium for sending these on loan. Marijke Meersman prepared an excellent illustration. Estrela Figueiredo made useful comments on the paper. References Davis A.P. & Figueiredo E. (2007) A checklist of the Rubiaceae (coffee family) of Bioko and Annobon (Equatorial Guinea, Gulf of Guinea). Systematics and Biodiversity 5(2): 159-186. Exell A.W. (1973) Angiosperms of the islands of the Gulf of Guinea. Bull. Brit. Mus. (Nat. Hist.), Bot. 4(8): 325-411. Figueiredo E. (1994) Diversity and endemism of angiosperms in the Gulf of Guinea islands. Biodiversity and Conservation 3: 785-793. Govaerts R., Ruhsam M., Andersson L., Robbrecht E., Bridson D., Davis A., Schanzer I. & Sonké B. (2009). World Checklist of Rubiaceae. The Board of Trustees of the Royal Botanic Gardens, Kew. Published on the Internet; http://www.kew.org/wcsp/rubiaceae/; accessed 17 July 2009. Hepper F.N. (1963), ed. Flora of West tropical Africa, by J. Hutchinson & J.M. Dalziel, second edition. Vol. II. London, Crown Agents for Oversea Governments and Administrations. Hutchinson J. & Dalziel J.M. (1931). Flora of West tropical Africa, vol. II, part I. London, Crown Agents for the Colonies. Jones P. J. (1994) Biodiversity in the Gulf of Guinea: an overview. Biodiversity and Conservation 3: 772-784. Juste B. J. & Fa J.E. (1994) Biodiversity conservation in the Gulf of Guinea islands: taking stock and preparing action. Biodiversity and Conservation 3: 759-771. Robbrecht E. & Puff C. (1986) A survey of the Gardenieae & related tribes (Rubiaceae). Bot. Jahrb. Syst. 108: 63-137. Schumann K. (1903) Rubiaceae africanae. Bot. Jahrb. Syst. 33: 333-374. Stoffelen P. (1998) Coffea and Psilanthus (Rubiaceae) in tropical Africa: a systematic and palynological study, including a revision of the West and Central African species. Thesis Univ. Leuven, Belgium. Verdcourt B. (1980) A conspectus of Polysphaeria (Rubiaceae). Kew Bull. 35: 97-130. White F. (1979) The Guineo-Congolian Region and its relationship to other phytochoria. Bull. Jard. Bot. Nat. Belg. 49: 11-55. Manuscript received July 2009; accepted in revised version October 2009. 198