Why phylogenies do not always predict ecological differences

The merger of phylogenies with ecology has given rise to the field of community phylogenetics, predicated on the assumption that ecological differences among species can be estimated from phylogenetic relationships (the phylogenetic distance/ecological difference, or PDED, hypothesis). A number of studies have failed to find strong support for this assumption, thus challenging the utility of phylogenetic approaches. This gap might highlight the fact that the PDED relationship is not useful for community assembly, but it is difficult to know because the lack of a relationship might also be due to a number of biological or methodological reasons, including inappropriate phylogenies, skewed distributions of phylogenetic distances, the lack of consideration of models of trait evolution, or the absence of sufficient niche space in experimental and observational venues. Each of these limitations, separately or combined, may confound recent experimental or observational results that examine relationships between phylogenetic distance and ecological differences. Notably, common evolutionary models can support alternative conclusions about the relationship between evolutionary distances and ecological differences than typically assumed and can change interpretations of community-based phylogenetic analyses. Here we review a number of issues that may lead to confounded effects in community phylogenetic analyses. In light of these potential pitfalls, we provide a number of guidelines for researchers to follow and stress that they need to address methodological shortcomings before concluding that ecological differences are unrelated to phylogenetic distances.