ABSTRACT
The Neotropics contain globally significant areas of seasonally dry forests and savannas. Seasonally dry tropical forests are found in scattered areas, with the most species-rich in Mexico, but similar levels of diversity may be found elsewhere (e.g. Peru, Bolivia). Levels of floristic similarity between areas are often low, and some areas (e.g. Brazilian caatingas, Peruvian inter-Andean valleys, Mexican Pacific coast) are rich in endemic species, whereas others (e.g. Bolivian chiquitano) are not. By far the largest savannas are the cerrados of central Brazil and the Llanos of Venezuela and Colombia. The cerrados have the highest species richness and endemism, and the woody flora of other areas is often just a subset of cerrado species. The fossil record and dated phylogenetic trees suggest that seasonally dry forests are at least as old as the Miocene, but that savannas dominated by grasses photosynthesizing using the C4 pathway may not have risen to dominance until the late Pliocene. Fossil pollen evidence and models of the distribution of species and vegetation indicate that savanna and seasonally dry tropical forest species changed their ranges during the climatic oscillations during the Pleistocene, but the extent of these shifts is uncertain. Clarification of whether disjunct distributions of species in both vegetations reflect migration through previously continuous formations or long-distance dispersal might be provided by population genetic studies of widespread species. Conservation strategy for seasonally dry forests has failed to take account of floristic patterns across the Neotropics, and some areas rich in endemic species (e.g. Peruvian inter-Andean valleys) lack protection entirely. However, a pragmatic approach, protecting any areas where social and political opportunities permit, is also necessary in these severely impacted ecosystems.
