Evidence suggests that insects whose existence depends on fruit production may serve as good bioindicators, especially for pollination services. In this study we examined this hypothesis, focusing on one of Israel's rarest butterflies, Tomares nesimachus (Rhopalocera; Lycaenidae). The butterfly is affiliated with open grasslands in the north of Israel, where its sole host plant, Astragalus macrocarpus, occurs. In a two-year field study, we mapped the distribution of the butterfly in Israel and tried to evaluate the causes of its rarity and decline. We found that its abundance in habitat patches was correlated with several factors, including (a) fruit production of the host plants in each patch, (b) the number of other patches in the vicinity, (c) the characteristics of the "matrix" outside the patches within a distance of up to 2 km, (d) the richness of large bee species (the plant pollinators), and (e) the grazing management of the butterfly's habitat. The characteristics of the matrix further affected the richness of bees in the butterfly's habitat. Since butterfly abundance is correlated with species richness of pollinators, it can serve as an indicator of diverse, species-rich butterfly habitats (grasslands). We deduce that species that depend on fruit production and by themselves do not contribute to their host-plant pollination, may almost necessarily be affected by multiple ecological factors and therefore can indeed serve as good indicators. Our results further indicate that the butterfly's population dynamics are affected by large-scale processes, and hence its conservation requires the preservation of large open landscapes. Consequently, T. nesimachus may also serve as an umbrella species.
Butterflies serve as a focal group in conservation worldwide, not only in terms of the efforts toward their protection, but also in terms of their wide use as bioindicators for identifying ecological trends and for advancing conservation theory. Further, significant involvement of scientists in applied-conservation is aided by the public's interest in butterflies. In this introductory paper to the compendium we delineate several areas where the vast knowledge on butterflies, or their charisma, can be used to advance conservation. These include: (a) integrating small-scale and large-scale approaches; (b) expanding the extent of scientific networks and standardizing conservation approaches; (c) integrating butterflies into global change scenarios; (d) extending the use of models; and (e) strengthening the link between science and applied conservation. A main aim of this compendium it to advance the latter.
Butterflies may serve as indicators of biodiversity trends, but for this purpose reliable methods of monitoring their distribution and abundance are essential. We discuss advantages and disadvantages of the currently used methods and suggest potential refinements, based on methodological advances achieved in other organisms. While assessing butterfly distribution, it is vital to account for imperfect species detection at investigated sites. This can be achieved through conducting repeated presence-absence surveys within a single season, and analyzing data with statistical models that estimate detection probability and site occupancy by a species. Transect counts, predominantly used for assessing butterfly abundance in monitoring programs, are cost-effective and easy to implement, but less reliable than mark-release-recapture sampling frequently applied for the same purpose in research studies. Deficiencies of transect counts stem from the fact that they do not account for individual detection probability and temporal fragmentation of butterfly populations, i.e., the situation in which just a small fraction of individuals belonging to a single generation is present on any day of a season. Consequently, transect counts can only yield relative abundance indices, which presumably correlate well with daily butterfly numbers, but not necessarily with their seasonal population sizes. Possible refinements to transect counts that would allow the estimation of individual detection probability include double observer or double zone approaches. In contrast, finding an effective way to estimate longevity (a measure of temporal fragmentation) with transect counts seems impossible. Instead, efforts should be made to evaluate how variation in longevity affects transect-count results.