New Species and Combinations of Catostemma and Pachira (Bombacaceae) from the Venezuelan Guayana

A. Steyermark had completed an extensive study of the Bombacaceae of southern Venezuela, but died before he was able to complete a treatment for the Flora of the Venezuelan Gua¬ yana.This paper lays the final groundwork for pub¬ lication of this treatment and includes two new spe¬ cies, Catostemma durifolius and Far/lira yapacanae, and nineteen new combinations in Pa¬ chira , into which all Venezuelan Guayanan species of Pochota, Bombacopsis, and Rhodognaphalopsis are placed. Catostemma and RelativesCatostemma Bentham is a small genus ol 10-14 species found in forests of northern Brazil, south¬ ern Venezuela, the Guianas, and Colombia (Stey¬ ermark, 1987;Paula, 1969).This genus and the closely related Aguiaria Ducke (1 species) and Scleronema Bentham (4-5 species) are placed in the tribe Catostemmatae of the Bombacaceae, based on their unifoliolate-compound leaves, few(l 4)-seeded fruits, and their largely free staminal filaments (Hutchinson, 1967).After reading a description ol the palmately compound juvenile leaves of Catos¬ temma commune Sandwith (Sandwith, 1931), Hutchinson surmised that the adult leaves of Ca¬ tostemma were unifoliolate-compound.A more re¬ cently discovered species, Catostemma digitatum J. I).Shepherd & W. S. Alverson, retains palmately compound (i.e., polyfoliolate) leaves into adulthood (Shepherd & Alverson, 1981).In contrast, no pal¬ mately compound juvenile or adult individuals of Aguiaria and Scleronema are known, but Hutch¬ inson concluded that both genera were unifoliolate because of the swollen distal pulvinus on each pet¬ iole, a morphological trait shared by all three genera.However, as fresh material of other genera ol Bom¬ bacaceae has become available in recent years, it is clear that most neotropical taxa share this trait.Thus, a swollen distal pulvinus cannot be used as a synapomorphy for the Catostemmatae.Nevertheless, similarities of flower and pollen structure (cf.Nilsson & Robyns, 1986) probably indicate a close relationship among all three genera.Affinities to other Bombacaceae and Malvales are much less clear.During preparation of the Bombacaceae treat¬ ment for the Flora of the Venezuelan Guayana, the following material could not be identified with keys in Paula (1969) or Steyermark (1987).