Functional and phylogenetic similarity among communities

Abstract

Ecological studies often rely on coefficients of intercommunity (dis)similarity to decipher effects of ecological, evolutionary and human-driven mechanisms on the composition of communities. Yet, two main criticisms have been levelled at (dis)similarity coefficients. First, few developments include information on species' abundances, and either phylogeny or functional traits. Secondly, some (dis)similarity coefficients fail to always provide maximum dissimilarity between two completely distinct communities, that is, communities without common species and with zero similarities among their species. Here, we introduce a new family of similarity coefficients responding to these criticisms. Within this family, we concentrate on four coefficients and compare them to Rao's dissimilarity on macroinvertebrate communities, and simulated data. Our new coefficients correctly treat maximally dissimilar communities: similarities are always zero between two completely distinct communities. The originality of these new coefficients is even more profound as the existence of maximally dissimilar communities was not a requirement for the new coefficients to behave differently than Rao's dissimilarity coefficient. Our new family of similarity coefficients relies on the abundances or occurrences of species within communities and on phylogenetic, taxonomic or functional similarities among species. We demonstrate that this new family embeds many of the recent developments in both functional and phylogenetic diversity. It provides a unique framework for comparing traditional compositional turnover with functional or phylogenetic similarities among communities. © 2014 British Ecological Society.