Nested species assemblages of amphibians and reptiles on islands in the Laurentian Great Lakes

Aim The amphibians and reptiles of the Great Lakes are diverse in their basic biology and natural history. These species are also widely distributed and overlap highly in their spatial extent. The Great Lakes basin includes 35,000 islands that differ in aspects of their geography, geology and climate, but that share a common post‐glacial history. This system provides a unique opportunity to assess the relative importance of biotic vs. abiotic factors in the development of nestedness. Our goal was to document and explain the patterns of nestedness across a variety of spatial and taxonomic scales. Location We studied amphibian and reptile assemblages (forty‐five species) occurring on 107 islands among four archipelagoes (Lake Erie, St Lawrence, Georgian Bay, Apostle) in the Laurentian Great Lakes of North America. The entire basin was recolonized in the post‐Pleistocene by amphibians and reptiles following the final retreat of the Wisconsin ice sheet. The present configuration of lakes and islands has existed over the past 5000 years. Most islands are landbridges; however, lake level fluctuation in Lake Huron inundated some islands of low relief. Presumably most islands would contain primarily extinction – as opposed to colonization – driven faunas. Methods We used data from four recent and thorough surveys to construct presence–absence matrices. To quantify nestedness of assemblages we used the nested temperature (T) method. To assess the association of nestedness with area and isolation we used the departures method. To quantify nestedness at the species level we used the Wilcoxon two‐sample method. To analyse species richness and patterns of nestedness we used multiple regression, and two‐wayANOVA. Results Islands supported diverse assemblages of both amphibians and reptiles that were in harmony with their mainland species pool(s). We found highly significant nestedness across the entire basin, all archipelagoes and all taxonomic levels. However, the degree of nestedness did not differ significantly among archipelagoes. The degree of nestedness did differ significantly (most to least) among classes (reptiles, amphibians) and orders (snakes, turtles, frogs, salamanders). Patterns of nestedness were more complex at the species level where some species were nested in all archipelagoes, while other species were in some locations but not others. Species richness and nestedness were strongly associated with area but not isolation. Main conclusions The similarity of the insular and mainland faunas indicate that the entire Great Lakes basin shares virtually the same species pool. Although some abiotic differences exist among the archipelagoes, their common history and similar contemporary habitats can explain the lack of differences among archipelagoes in nestedness. Significant differences among taxa in nestedness were consistent across the basin and appear to be related to differences in basic biology, natural history and ecology. Both nestedness and species richness were significantly related to area but not isolation. Most islands in the Great Lakes are not highly isolated and ample evidence exists that dispersal plays a role in community assembly. The strong area effect can be explained by the differences among taxa in their biotic characteristics and habitat requirements. Our results indicate that preserving large islands instead of equivalent areas of small islands would be a more effective conservation strategy and that results of studies conducted on Great Lakes islands may be applicable elsewhere in the basin.