No indication of methodological biases in tullgren and macfadyen extraction of edaphic microarthropods

Separating soil-dwelling microarthropods from their substrate is a fundamental procedure in soil zoological research. Although the earliest methods to automatically extract animals from soil were described more than a century ago, the efficiency of existing techniques is still open to debate. In this study, we mutually compared the performance of four methods of very different operation mode (active and passive extraction, and traps) to inspect for biases of either method. From a temperate grassland and a forest site, we collected the adult oribatid and collembolan fauna using a Tullgren extractor, a Macfadyen canister extractor, heptane flotation and pitfall traps. Between Tullgren and Macfadyen, we did not find substantial differences in group and species abundances, rarified species richness, body length distributions, and collembolan life form. Heptane flotation produced fewer individuals than the two active methods, but comparable patterns of species composition, body lengths and collembolan life form. Pitfall traps strongly underestimated abundance and species richness of Oribatida and of eu- and hemiedaphic Collembola, but caught high numbers of epedaphic individuals, and several epedaphic species that were missed by the other methods. In conclusion, our data indicated that the relative efficiency of the Tullgren and Macfadyen methods was high and estimates of abundance and species richness reliable. We suggest to complement the traditional soil coring with pitfall trapping to adequately represent the epedaphic collembolan species in the data.