A comparison of pitfall trapping and quadrat sampling of Carabidae (Coleoptera) on river banks

A comparison was made betweeen the "open" pitfall trap method and quadrat sample method in microhabitats on river banks. The latter method was found to give reliable estimates of the absolute abundance of carabid beetle species. In pitfall traps individuals of the carabid beetle Bembidion schuppelii were over-represented in a sub-optimal microhabitat compared with in optimal ones. The niche-overlap between this species and B. bipunctatum was 0.19 in quadrat samples and 0.59 in pitfall traps. There was no correlation between the relative abundance of individuals of species in quadrat samples and in pitfall traps. Individuals of larger carabid species were highly over-represented compared to smaller ones in pitfall traps, an effect that was independent of diurnal rhythm and life cycle category. Biomass of larger carabid species was also overestimated in pitfall traps compared to that of smaller ones. Imagines were more efficiently captured in pitfall traps than larvae. Pitfall traps had a higher number of species than quadrat samples in all investigated sites. In four of five cases the Shannon-Wiener index gave higher species diversity in pitfall traps than in quadrat samples. It is concluded that the "open" pitfall trap method is inappropriate to study dominance ratio, the ecological role of species and perhaps the species diversity within communities. The general insight into the community structure of carabid beetles may therefore be very biased