Pollination Biology and Taxonomy of Dinemandra and Dinemagonum (Malpighiaceae)

Abstract

Dinemandra and Dinemagonum, the only two genera of Malpighiaceae in Chile, have calyx glands borne on stalks rather than appressed to the surfaces of the calyx lobes as is typical of malpighs. It is proposed that this positioning of the glandular secretory surfaces (elaiophores) on the ends of stalks provides an intergland distance and surface area needed for the robust Centris bees that serve as pollinators. The structure of the stalked calyx glands and similarities in pollen provide evidence for the close relationships of the two genera despite their traditional placement in different tribes. However, the sister group of the two genera is impossible to determine at the present time. Extensive collections of both genera in the desertic regions of northern Chile where they are endemic has shown that each is monotypic. The previous recognition of several species in each genus resulted from few available specimens and a tendency, particularly in the case of Dinemandra, for plants from small isolated populations to have rather distinctive morphologies. Of all of the countries in South America, Chile has the fewest taxa of Malpighiaceae, two gen- era, Dinemandra Adr. Juss. and Dinemagonum Adr. Juss., each with only one species. This low di- versity undoubtedly stems in part from the fact that Chile lacks moist tropical and subtropical habitats, in which most malpighs are found. Moreover, the subtropical portion of Chile is xeric and separated from similar areas of Ar- gentina (which could serve as sources of im- migration) by the Andes. The Malpighiaceae are generally poorly represented in xeric re- gions such as those in northern Mexico, the southwestern United States, and northwestern Argentina. Nevertheless, what the malpighs of Chile lack in terms of diversity, they make up in novelty. Morphologically, the two genera are unusual within the family in their mutual pos- session of stalked calyx glands. Most of the members of the Malpighiaceae have two oval glands appressed to the dorsal side of all (five), or most of, the calyx lobes. These glands, known as elaiophores, secrete oils that are collected primarily (as judged by documented visits to malpigh flowers) by female bees of the genus Centris (Anthophoridae). For many malpighs, oils are the only floral rewards produced, and Centris species are therefore the primary or ex- clusive pollinators. Centris presumably uses the oils collected from these glands similarly to those collected from other oil-secreting species (Neff and Simpson 1981; Vogel 1974), to mix with soil to form part of the nest lining and/or in com- bination with pollen and nectar as part of the larval provisions. Only four genera of the Mal pighiaceae are known to have stalked glands: Dinemagonum, Dinemandra, Heladena Adr. Juss., and Henleophytum Karsten. Dinemandra and Di- nemagonum are distinctive in having gland stalks 1.4-2.5 mm long. In addition, the stalked glands are basally connate, forming pairs between glands on adjacent sepals. As might be expected, this strange positioning of the glands plays a role in the pollination biology. Dinemandra and Dinemagonum also share the unique features of 8-colporate, reticulate pollen (Lowrie 1982). Nevertheless, Niedenzu (1928) placed the two genera in different tribes of his Pyramidotorae (=Gaudichaudioideae) because of their distinct fruit morphologies. Both have dry fruits that split into three winged segments, but the morphologies of the mericarps differ. Dinemandra has mericarps each of which has two large lateral wings on the dorsal surface. Dinemagonum has mericarps each with a single large median dorsal wing. These differences prompted Niedenzu (1928) to place Dinemandra in the Hiraeeae (=Hiptageae) subtribe Mascag- niinae and Dinemagonum in the Banisterieae subtribe Banisteriinae. Despite their unusual morphology, few botanists have discussed the relationships or biology of the two genera. Works on the vegetation of northern Chile, such as those of Reiche (1907) and Johnston (1929), merely commented on their presence. In this article I examine the unusual morx