Environment-driven variation in the supply of individuals to local assemblages can determine patterns of community structure. Alternatively, local environmental conditions can determine the type of species that can be successfully established in a given community. Most communities are probably found somewhere between these 2 extremes, but few studies have attempted to disentangle their effects in a community-wide context. Using multivariate ap-proaches in rocky shore communities, we showed that environmental variables (i.e. sea surface temperature and wind stress), the benthic abundance of 108 species of invertebrates and macroalgae, and recruitment rates of invertebrates all shared significant spatial and temporal patterns of variability across a 400 km shoreline marked by the presence of a prominent upwelling centre. Variance-partition analyses for the invertebrates with pelagic development showed that spatially structured environmental filtering alone explained only 7% of the variation in community structure. In contrast, the combination of environmental conditions and recruitment variation ex plained 45% of the variability in community structure and an additional 18% was jointly ex plained by recruitment and spatial relationships among sampling sites. Unexplained variation (41%) can be attributed to factors like local species interactions that are robust to environmental variability. Therefore, environment-driven variation in recruitment rates can have, in comparison to pure environmental filtering, stronger effects on the structure of this metacommunity. Our results can serve as a foundation for predictive models of the response of biodiversity to climate change and other human-induced disturbances, which are predicted to alter local environmental conditions and dispersal pathways.
- Environmental filtering