Studies in Ipomoea (Convolvulaceae) I. The Arborescens Group

Abstract

The taxon containing the arborescent species of Ipomoea is heterogeneous as it is presently recognized. A reclassification of the group has resulted in a taxon that can be characterized by several correlated features (involving habit, inflorescence type, sepal size, shape, texture, and pubescence, corolla color and pubescence, and seed pubescence) in contrast to its historical antecedent. The relationships of the ten constituent species are briefly discussed, and a detailed key and complete descriptions incorporate my conclusions concerning the limits of these species. Despite the long history and general acceptance of the taxon historically ancestral to it, the group comprising the arborescent species of Ipomoea and their close relatives has not been accurately circumscribed and the boundaries of its constituent species remain inadequately defined. As one result of a study of most of the species of Ipomoea in Mexico and Central America, I now recognize a taxon, the Arborescens group, which includes many of the erect, woody species traditionally considered closely related but excludes certain others. The composition suggested below permits the recognition of the group on a broader morphological basis than previously, and brings into association with the arborescent species certain woody vines. Choisy (1845) recognized the group Arborescentes and defined it as the subsection of the erect species (section Orthipomoea) of Ipomoea that are arborescent. In this subsection he placed I. murucoides, I. arborescens, I. batatilla, I. haenkeana, and I. coriacea. House (1908), in treating the North American species of the genus, added I. fistulosa (a synonym of I. carnea), I. glabriuscula, I. nicaraguensis, I. cuernavacensis, I. intrapilosa, I. calva, and I. wolcottiana. Matuda (1964), in dealing with the genus as it is represented in Mexico, added I. chilopsidis and I. pauciflora to the list of arborescent species, and he placed I. wolcottiana into synonymy with I. pauciflora. O'Donell (1950) added I. calodendron and I. vargasiana from Peru, and commented that the group is de sistematica bastante confusa. Among the names in the above roster, there are several that do not belong there. Ipomoea coriacea Choisy is a Brazilian species described as having herbaceous to membranous sepals, purple corollas, and glabrous seeds in contrast to the coriaceous sepals, white corollas, and pilose seeds of the Arborescens group. Therefore, although I have not seen the type, I have no hesitation in regarding this species as misplaced among those of the Arborescens group. Ipomoea glabriuscula House, although described by House as a tree, is in fact a vine closely related to and perhaps synonymous with I. microsticta Hallier f., as