Landscape composition, not connectivity, determines metacommunity structure across multiple scales

Authors: W. A. Ryberg and L. A. Fitzgerald

Metacommunities occupy landscapes, yet few studies of metacommunity structure consider contributions of multiple landscape elements explicitly. Previous studies have focused on measures of focal habitat connectivity as proxies for dispersal to distinguish among alternative metacommunity models. However, it is also clear that dispersal of species and metacommunity structure are also shaped by landscape composition and configuration. Slow integration of explicit landscape heterogeneity into the metacommunity concept has both limited our understanding of how landscape–dispersal interactions shape metacommunity structure and constrained the practical relevance of metacommunity theory. In this study, we encourage integration by developing an approach that characterizes how multiple landscape elements simultaneously contribute to metacommunity structure. We demonstrate the utility of this approach by characterizing how landscape heterogeneity shapes metacommunity structure of lizards at multiple spatial scales in the Mescalero Sands. We found diversity and spatial configuration of habitats in the surrounding landscape matrix best explained the pattern of nested subsets with clumped species loss observed in this lizard metacommunity across scales. Although our analyses included connectivity metrics for focal habitats, they were not shown to be important at any scale in this study. Our results were consistent with mass effects models of metacommunity theory as interpreted through landscape contrast. By integrating explicit landscape heterogeneity into the metacommunity framework, our approach provided a spatially explicit description of how and where landscape–dispersal interactions shaped metacommunity structure across scales and is directly applicable to many systems, especially those with indistinct boundaries. Our approach complements other analytical methods designed to tease apart relative roles of environmental filtering and spatial structure in metacommunity data. Finally, our approach enhances the practical relevance of metacommunity theory for future research in ecosystems subject to increasing landscape contrast as environments continue to be fragmented, habitats become further degraded and homogenized, and matrix habitats become more inhospitable.

Suggested Citation