A variety of processes are thought to be involved in the formation and dynamics of species assemblages. For example, various metacommunity theories are based on differences in the relative contributions of dispersal of species among local communities and interactions of species within local communities. Interestingly, metacommunity theories continue to be advanced without much empirical validation. Part of the problem is that statistical models used to analyze typical survey data either fail to specify ecological processes with sufficient complexity or they fail to account for errors in detection of species during sampling. In this paper, we describe a statistical modeling framework for the analysis of metacommunity dynamics that is based on the idea of adopting a unified approach, multispecies occupancy modeling, for computing inferences about individual species, local communities of species, or the entire metacommunity of species. This approach accounts for errors in detection of species during sampling and also allows different metacommunity paradigms to be specified in terms of species- and location-specific probabilities of occurrence, extinction, and colonization: all of which are estimable. In addition, this approach can be used to address inference problems that arise in conservation ecology, such as predicting temporal and spatial changes in biodiversity for use in making conservation decisions. To illustrate, we estimate changes in species composition associated with the species-specific phenologies of flight patterns of butterflies in Switzerland for the purpose of estimating regional differences in biodiversity.
Additional publication details
Models for inference in dynamic metacommunity systems
Ecological Society of America
Southeast Ecological Science Center, Patuxent Wildlife Research Center