A Synopsis of the Buxaceae in Africa South of the Sahara

Abstract

Summary. The African species of Buxus and Notobuxus are reviewed and their taxonomic relationship discussed. It is concluded that the 9 African species are best accommodated in 3 sections of Buxus, viz. 1 species in Sect. Tricera, 4 in Sect. Notobuxus and 4 in Sect. Buxella. The name Sect. Notobuxus is proposed in a new rank and combination; the new name Buxus acutata is proposed for Notobuxus acuminata when transferred to Buxus. The apparently rare B. nyasica, previously only known from the type gathering, has been documented by a new collection. This review, with some additional observations and conclusions, of African Buxaceae has been prompted by the writing of an account of the family for the Flora Zambesiaca, where the rarely collected species Buxus/Notobuxus nyasica occurs. The first review of African Buxaceae was published by Baillon (1859) who knew of one species, B. madagascarica Baill. from Madagascar. Baillon divided Buxus into two sections: Eubuxus in the Old World and Tricera in the New World; apart from the geographical distinction, the two sections could be separated on account of differences in the shape of the inflorescences. BailIon's arrangement was adopted by Mueller (1869), with the addition of more species, but none from Africa. Later, Baillon himself (Baillon 1873) added B. hildebrandtii from Somaliland, placing it with the other Old World species of Buxus. Oliver (1882) described a new species from Natal, placing it in a new, closely related genus, Notobuxus, on account of two extra stamens, making a total of 6, all with sessile anthers, and the absence of any rudiment of a gynoecium in the male flower. Later, Oliver (1886) described another species in Buxus, B. macowanii from southern Natal; this species had chiefly unisexual inflorescences with a single, terminal flower, and long exserted anthers. Van Tieghem (1897) reviewed the whole family, suggesting that in anatomical characters the African species were more closely related to one another than to the other species of Buxus. He further discussed the possibility of placing the African species in infrageneric taxa under Notobuxus, or to raise these taxa to generic rank. His conclusion was that all groups merited generic rank, and described therefore the genera Buxanthus and Buxella to accommodate the African species not placed in Notobuxus. Gilg (1899) described a new species in a new genus of Buxaceae, Macropodandra acuminata Gilg, from the northeast part of what is now Zaire, and described a new Buxus, B. benguellensis Gilg, from Angola, the latter being closely related to B. macowanii. Pax (1900) reduced van Tieghem's two new genera Buxanthus and Buxella to synonymy of Buxus sect. Eubuxus, while Hutchinson (1912) proposed