Cienfuegosia Cav. Extended to Madagascar

Abstract

My earlier revision of the genus Cienfuegosia (Fryxell, 1969) did not recognize any representatives of that genus on the island of Madagascar, although reference was made (p. 198) to Ulbrich's statement that C. hildebrandtii Garcke occurred there. No basis was known then, nor is known now, for Ulbrich's assertion. However, an examination of the description and illustration of Hibiscus humbertianus Hochr. in the Flore de Madagascar (Hochreutiner, 1955) suggested the possibility that this species was a Cienfuegosia. Consultation of the original description revealed that Hochreutiner had acknowledged but rejected this possibility. In his later treatment he does not mention this similarity, but in placing it in Hibiscus states only (Hochreutiner, 1955: 68), Espece tres distincte, ne ressemblant en rien aux autres especes de la section [Spatula Hochr.] et meme du genre. The examination of authentic material confirms Hochreutiner's statement that this species is poorly accommodated in Hibiscus and also confirms his earlier suspicion that a placement in Cienfuegosia might be correct. Hochreutiner rejected a placement of Hibiscus humbertianus in Cienfuegosia for two reasons: (a) H. humbertianus has a 5-lobed stigma and a 5-celled fruit, whereas these structures are typically 3-4-merous in Cienfuegosia; and (b) H. humbertianus lacks the black punctiform glands (gossypol glands) in the calyx that are prominent in many species of Cienfuegosia. However, Cienfuegosia drummondii (A. Gray) Lewton has a (4-) 5-merous gynoecium, and several species have the gossypol glands restricted in distribution from certain plant parts, including absent from the calyx (or nearly so) in C. hildebrandtii Garcke and C. hearn ii Fryx. The transfer of H. humbertianus to Cienfuegosia, therefore, does not unduly stretch the generic boundary of Cienfuegosia. Characters that support such a transfer include the following: