Island-mainland lepidopteran assemblies: a blue-print for species conservation in fragmented habitats

The theory of island biogeography predicts that species diversity is lower on small islands and on islands far distant from the mainland, and higher on large islands with high habitat heterogeneity, and on islands close to the mainland. In addition, population and species persistence (i.e. their long-term survival) depends on life-history traits, including morphological, ecological and behavioural characteristics. These theories can also be used as basic hypotheses to test species persistence and community assembly of fragmented habitats. Lepidopteran communities living on small and isolated habitats, like small islands located along the northern German coast-line, provide an excellent model system to test for potential effects of species traits on their persistence and on the community assembly on islands and the mainland. Lepidopterans were assessed on the island Hallig Grode and on the adjoining mainland over a period of almost 50 years. About 500 species were recorded and classified according to their morphological, ecological and behavioural characteristics. Our data suggest less changes in the lepidopteran community composition on the island compared with the mainland. Yet, the number of species was considerably lower on the island compared with the mainland. The lepidopteran species community on the island was dominated by generalist species associated with meadows and grasslands. In contrast, lepidopteran species found at the mainland were frequently associated with forest and shrubs, and frequently showed a higher level of ecological specialization. Lepidopteran species caught on the island were equipped with more elongated wings compared with species found at the mainland. Our data suggest that the island situation functions as ecological filter allowing only specific species to persist and thus creates a subset of the mainland species community. Extrapolating our findings to isolated habitat patches across fragmented landscapes, we expect a similar filtering effect for species composition, and the long-term extinction of sedentary specialist species.