Co-occurrence of prairie and barrens butterflies: applications to ecosystem conservation

Abundance indices from transect surveys for species restricted to prairies and barrens were correlated to test the degree to which these species of conservation concern co-occur. In 56 pair-wise tests among 15 species by subregion, 29 (52%) correlated significantly, all but one positively, and only 17 (37%) of 46 tests involving samples of >60 individuals for each species were non-significant. The species producing the most significant interspecific correlations in prairie were Speyeria idalia (five of ten tests) and Atrytone arogos (four of six tests); in barrens, Euchloe olympia (seven of eight tests) and Hesperia l. leonardus (six of eight tests). Lycaeides melissa samuelis, the butterfly receiving the most conservation attention in these habitats, produced few significant correlations. To explicate these patterns of co-occurrence, case histories were compiled for these species at sites of comparable vegetation by subregion. No management type was clearly favourable for all specialists of a given habitat, although some managements were favourable for more species than others (e.g. haying vs. burning in prairie). Analysis of variance of management at these sites produced more results with significant effects than did correlations of the species' abundance indices with habitat patch size. These results were inconsistent with prevailing ecological theory about the natural maintenance of these habitats; conversely, a single unified (alternative) theory of ecosystem management could not be inferred from these results. These patterns of butterfly co-occurrence suggest an alternative approach to ecosystem conservation that focuses on subsets of species native to a particular ecosystem. These smaller species assemblages significantly co-occur in range, habitat, and management tolerance, and may be amenable to monitoring with indicator species.