A Synopsis of the African Species of Pandanus

Abstract

The taxonomy of the African species of Pandanus is presently in a very unsatisfactory state. Since Warburg's monograph in 1900, a number of species have been described which have never been critically compared. The treatment by Warburg was itself very sketchy, as several species were founded on the poorest of specimens and unfortunately some have also been incorrectly, or inexactly, described. The earliest known African species was published by Palisot de Beauvois in 1804 (Pandanus candelabrum). Next described was Heterostigma heudelotianus Gaudich., 1843, now called Panclanus heudelotianus (Gaud.) Balf. f. In 1887, Henriques described P. thomensis from Sdio-Thome' Island in the Gulf of Guinea; this was furnished with a particularly good description and figure. Rendle in 1894 in a paper entitled Tropical African Screw Pines attempted a complete review of the African species; he recognized eight species, five from West Africa and three from East Africa; of these, five were newly proposed species (P. barterianus, P. welwitschii, P. rabaiensis, P. kirkii, and P. livingstonianus). Up to this time, not much had been effected toward a subgeneric classification of the genus. Warburg paid special attention to the African species of Pandanus. His monograph of 1900 includes 16 species credited to Africa, of which eight were species which he had described. Equally important was his attempt to establish a sound subgeneric, or rather sectional classification of the genus. The African species were arranged under Section Keura, Vinsonia, and Sussea. Sect. Keura is now correctly called Pandanus, since it includes the type species of the genus. Both Vinsonia and Sussea are names which Warburg took from the work of Gaudichaud, but these being invalid names (according to our present nomenclatural rules) effectively date only from Warburg's usage in 1900. Since 1900, nine more species have been described. There are therefore at present 25 species to account for in any review of the African species. This does not include some nomina nuda and illegitimate names. Martelli (1933) showed that one of the species described by Warburg and which he thought was from Africa, Pandanus platycarpus Warb., was wrongly localized; it proved to be from Java. This reduces the total to 24. These 24 species are listed below according to the section of the genus to which they belong. After each is added a commentary on the present state of our knowledge of the species. Several names are reduced to the status of new synonyms.