ABSTRACT
The controversies over explanations of endemism and the patchiness of South American seasonally dry tropical forests are potentially resolved with molecular phylogenetics. The genera Coursetia and Inga, for example, each show an ecological predilection to a different forest type. Sister species of Coursetia are often confined to the same narrow geographical region of dry forest, in contrast to Inga, in which closely related species often are widely distributed throughout neotropical rain forests. If these patterns are general, then the geographical phylogenetic structure of dry forest clades can be explained by low dispersal rates among isolated forest patches, a rate that does not replace the accumulation of endemic species by resident speciation. Recent critiques of vicariance and refugia hypotheses, the usual historical explanations for endemism and patchiness, are compelling. However, clades of sister species narrowly confined to dry forest patches commonly trace their ancestry back in time to at least several millions of years. Such minimum ages suggest that dry forest patches may be more persistent floristically and geographically than the critics of refugia hypotheses contend. If the phylogeny of Inga is representative of rain forest clades, then it is the rain forests rather than the dry forests that are more dynamic floristically and geographically.
