My work uses phylogenetic trees to address fundamental questions on the origin and distribution of biodiversity. In particular, I am interested in why some regions and some lineages contain more species than others. To date, approaches for studying these two patterns have been largely separate. Numerous studies have sought explanations for taxonomic imbalance, concentrating on the role of key biological traits, such as pollination syndrome or growth form, but in flowering plants (as in other groups) such traits apparently explain relatively little of the variation in species numbers. At the same time, ecological studies have explored the effects of environment on floristic richness within regions, but have not traditionally addressed evolutionary explanations. Phylogenetics offers a powerful means to combine both approaches. A better understanding of the processes shaping biodiversity patterns will also be critical if we wish to reduce current rates of biodiversity loss.
Evolution of species richness
I developed phylogenetic methods to explore the taxonomic and geographic distribution of species richness within the flowering plants (angiosperms). I used both molecular and supertree techniques to reconstruct the evolutionary relationships among higher angiosperm taxa, and mapped the location of significant shifts in diversification rates on the angiosperm tree of life. I demonstrated that rate shifts have been much more frequent than previously assumed, and that simple explanations based upon one or a few key traits are insufficient to explain the highly labile nature of diversification rate (Davies et al. 2004a PNAS 101:1904-1909). I explored environmental influences on rates of diversification by extending the comparative method to include geographic information system (GIS) data within a phylogenetic framework. I used sister group comparisons to demonstrate that net speciation rates were greater in high-energy environments, and that environmental energy might also drive rates of molecular evolution, but independently (Davies et al. 2004b Proc. R. Soc. 271:2195-2200).
Character divergence and species co-occurrence
Currently I am using phylogenetic methods to explore patterns of co-occurrence among species in relation to their morphological and ecological characteristics. In general, there are three possibilities. First, co-occurring organisms might tend to have similar ecomorphology, because they are adapted to the same physical environment. Second, co-occurring organisms might tend to have different ecomorphology and niche preferences, thereby reducing competitive interactions and permitting coexistence. The latter situation could arise either as a result of selection for character divergence or because only species with divergent niche preferences are able to colonise overlapping areas, i.e. species sorting. Finally, species differences and patterns of co-occurrence might be unrelated to one another, for example, if random patterns of dispersal determine which species coexist in an area and if measured species differences are not the target of strong selection by the environment or competition. Phylogenetic approaches allow assessment of ecological diversity within an evolutionary framework, and provide a simple null model for exploring the relationship between coexistence and evolutionary divergence.
Using extensive geographic and biological databases for carnivores, I show that morphological divergence in dental traits can predict patterns co-occurrence among closely related species (Davies et al. inpress). One of my future goals is to extent this approach to encompass communities comprised of many interacting species (Davies 2006 Curr. Biol. 16, R645-R647). Preliminary analyses indicate that there is much greater overlap than predicted from morphological divergence in regions where carnivore species richness is highest, for example East Africa. One possible explanation is that the high density of carnivores in this habitat inhibits geographical displacement to competition free space, and that niche space, via shifts in tooth size, is saturated.