Estimating metacommunity extent using data on species abundances, environmental variation, and phylogenetic relationships across geographic space

The metacommunity concept underlies many explanations for spatial variation in species diversity. However, to date, there are been only vague definitions of how to empirically construct descriptions of these important phenomena. We develop methods for constructing metacommunity descriptions based on the concept that each local community has a distinct metacommunity that contributes immigrants to that community. We call the community for which a metacommunity is constructed the focal community. We discuss three criteria for constructing metacommunity descriptions based on geographic distances, environmental distances, and phylogenetic distances. We develop objective measures for constructing subsets of local communities into the “best” metacommunity description for a given focal community using each of the three criteria. We illustrate these methods using data on abundance estimates on 1393 local surveys of North America breeding birds. For each survey route measures of environmental conditions were obtained and used to calculate environmental distances among the survey routes. In addition, using a phylogenetic hypothesis describing relationships among bird families, we obtained measures of phylogenetic distances among survey routes. Metacommunities constructed using geographic distances were closely correlated with those constructed using environmental distances, however, metacommunities constructed using family level phylogenetic distances were less representative of local communities than those constructed using the other distance measures. Our results imply that it is possible to obtain empirical descriptions of metacommunities related to communities of interest that can be used to develop rigorous empirical tests of mechanisms thought to be responsible for patterns in species diversity in space and time.

► Methods for estimation of the spatial extent of metacommunities are presented.
► The methods are illustrated using data on abundances of North American land birds.
► Our results are not consistent with neutral theories of biodiversity.