Across biological communities ranging from microbes to plants and animals, the spatial distribution of phylogenies reflects the processes of diversification, dispersal, and community assembly. Phylogenetic beta-diversity, which quantifies the phylogenetic turnover of communities across the landscape, provides a powerful tool to study the ecology, biogeography and conservation of communities. Quantitative predictions for phylogenetic beta-diversity are necessary to better understand the forces underlying current patterns of diversity, to estimate ecologically relevant parameters such diversification and dispersal rates, and to inform conservation practices. Using sampling theory and a mediterranean flora dataset collected across four continents, we characterize the decay of phylogenetic similarity between communities with the geographic distance separating them. We provide an analytical relationship between the spatial turnover of species, phylogenetic tree shape, and phylogenetic beta-diversity under a null model of community assembly.

Results/Conclusions

We find that the decay in phylogenetic similarity with geographic distance is slower than the decay in compositional similarity, suggesting that a larger geographic distance between protected sites is necessary to protect phylogenetic diversity versus species richness. The relationship between the rates of decay in phylogenetic and compositional similarity is given by phylogenetic tree shape: high phylogenetic redundancy leads to a shallow decay in phylogenetic similarity with geographic distance, while high phylogenetic distinctness leads to a rate of decay in phylogenetic similarity approaching the rate of decay in compositional similarity. We show that a higher phylogenetic redundancy is expected for clades in which diversity is saturated in comparison to clades where diversity is expanding. We present a novel approach, based on the coalescent theory of population genetics, which allows us to test whether diversity is saturated or expanding over time. Using a wide array of phylogenies including birds, plants, lizards and mollusks, we show that most empirical phylogenies are consistent with the hypothesis that diversity is expanding. We conclude by discussing potential extensions of the coalescent-based approach for modeling phylogenies in a spatial context.