Introduction

In considering how forests will react to global change, understanding the distinctions between vegetation types is important. If we are to pinpoint the species that might thrive in Amazonia if the rain forest there ‘dies back’ because of drying and more seasonal climates, then characterising the vegetation types growing currently in seasonally dry areas of the Neotropics is critical. This is one motivation for this paper. Another is to add impetus to the preservation of dry-adapted vegetation because it is highly threatened and relatively neglected by conservationists compared with rain forests.

Our approach is not to re-visit the labyrinthine debates of vegetation defined by subtleties of taxonomic composition and relative abundance of species (e.g. Mucina 1997; Poore 1955). Instead, we use an approach that asks whether major biome settings, as defined by physiognomies of their component plants and ecological factors (e.g. presence of fire), represent distinctive theatres of evolution for constituent woody floras. We build upon an approach developed to study the phylogenetic structure of local communities at small spatial scales (e.g. Webb et al. 2002) and apply it at near-continental scales. If certain kinds of dry-adapted vegetation represent evolutionary theatres, then detecting lineages repeatedly moving between biomes (e.g. ecological speciation) is expected to be less common than detecting clades of species that are all confined to one kind of dry-adapted vegetation (phylogenetic niche conservatism; Crisp et al. 2009; Donoghue 2008).