EN
Life form spectrum is an informative parameter in a comparative analysis of ecological structure of Collembola communities, and a valuable indicator of conditions of soil environment. This suggests a search of adequate methods of its determination. In the present study, two methods of evaluation of life form spectra in springtail communities are compared. A usual approach is a determination of the species composition and abundance, with a subsequent formal distribution of the species present among the life form categories of a chosen system, e.g. among epedaphic, hemiedaphic and euedaphic life forms considered as dwellers of litter surface and upper horizon, of litter depth and topsoil, and of soil horizon, respectively (Hopkin 1997). By this method (A) all the specimens of a given species are attributed to the same life form. However, species populations contain juveniles with smaller body size, thinner cuticle and weaker pigmentation, which tend to be more sensitive to environmental constraints and to dwell deeper in soil than adults. Therefore the attribution of juveniles and adults of a given species to the same life form may not be accurate. Another approach is a direct attribution of a life form category to each collembolan individual during the determination process, according to the visual assessment of its morphological characters (body size and development of pigmentation, number and pigmentation of ommatidia). In that case representatives of a given species might be attributed to more than one life form category, e.g. adults to epedaphic whereas juveniles to hemiedaphic life form, or to hemi- and euedaphic life form, respectively (method B). To compare both methods, mass collembolan material of a microcosm experiment carried out in a beech-oak forest at Mikolajki Research Station (NE Poland) and aimed at investigation of mesofauna/ earthworm interactions, was used. The comparison showed a statistically significant bias between the spectra of Collembola life forms estimated by methods A and B: the latter spectrum was shifted to the favour of euedaphic forms. The range of the bias was larger in soil than in litter horizon. Method B is suggested to be closer to a real life form distribution in a collembolan community and may be used for corrections of the data obtained by the traditional taxonomic method A.