Dallas, T; Melbourne, BA; Hastings, A

Checkerboard distributions-mutually exclusive species co-occurrences-are a common observation in community ecology and biogeography. While the underlying causes of checkerboard distributions have remained elusive, a long-standing argument is that they are representative of strong competitive interactions and/or dispersal limitation. We explore this using a stochastic two-patch metacommunity model combined with an experimental two-patch system of competing Tribolium species, quantifying checkerboard distributions using the abundance-based index A(st). We find that maintenance of checkerboard distributions is possible in a limited parameter space consisting of low dispersal rates, low population growth rates and high interspecific competition. Checkerboards were not maintained in experimental metacommunities. Our model, parameterized using independent data, echoed this finding, providing a clear link between model and experiment, and suggested that only small regions of parameter space would allow for checkerboard distributions between patches with equally hospitable environments. These findings may provide insight into when interspecific competition and dispersal limitation would promote checkerboard distributions.