Disentangling the effects of local and regional processes on biodiversity patterns through taxon-contingent metacommunity network analysis

Metacommunity theory, which has gained a central position in ecology, accounts for the role of migration in patterns of diversity among communities at different scales. Community isolation has a main role in this theory, but is difficult to estimate empirically, partly due to the taxon-dependent nature of dispersal. Landscapes could be perceived as either fragmented or connected for organisms with contrasting dispersal abilities. Indeed, the dispersal ability of a taxon, and the spatial scale at which eco-evolutionary processes shape local diversity, determine a taxon-dependent metacommunity network. In this paper, we introduce a methodology using graph theory to define this taxon-dependent metacommunity network and then to estimate the isolation of local communities. We analyzed the relative importance of local conditions versus community isolation as determinants of community richness for 25 taxa inhabiting 18 temporary ponds. Although local factors have been the foci of most previous empirical and theoretical considerations, we demonstrate that the metacommunity network is an equally important contributor to local diversity. We also found that the relative effect of local conditions and the metacommunity network depend on body size and taxon abundance. Local diversity of larger species was more affected by patch isolation, while taxon abundances were associated with positive or negative effects of isolation. Our results provide empirical support for the proposed role of metacommunity networks as determinants of community diversity and show the taxon-dependent nature of these networks.