Résumé
Brazil is a megadiverse country from which around 10% of all species are known. However, many areas in Brazil have not been adequately studied, particularly for soil animals. This includes the state of Santa Catarina, where only 18 of the approximately 300 known Brazilian earthworm species occur, and where very little is known of the impacts of land use management on earthworm populations (density and diversity). Therefore, the aim of this study was to evaluate earthworm species richness in five different land-use systems (LUS) with increasing anthropogenic impact: native forest (NF), eucalyptus plantation( EP), pasture (PA), integrated crop-livestock (ICL) and no-tillage cropping (NT), in six counties, three each from the Western and Plateau regions of the state. Nine monoliths of 25 cm x 25 cm x 20 cm depth were sampled at each site and worms hand sorted. Qualitative samples were also taken by digging at least 20 holes per site to improve the likelihood of collecting rare species. Samples were taken in the Winter and Summer months of 2011-2012 (July-August 2011, December 2011-January 2012). Contrary to expectations, species richness was higher in LUS with higher (ICL and NT = 15 and 17 spp.) vs. lower anthropogenic impacts (EP, PA, NF = 9-10 spp.), mainly due to the presence of exotic species in the cropping systems. Native species predominated in PA and NF, although natives were also found in highly disturbed ecosystems (NT, ICL) and in the West region all worms collected in NT were native. In total 24 species were identified in all LUS, with 19 native species, including several that were new to science. Several species were collected exclusively in each region, and overall qualitative samples yielded 24 species while quantitative samples only 16. Therefore, qualitative sampling appears to be more effective in determining earthworm species richness at regional levels, although quantitative samples are a useful addition when LUS are compared within a region, and if abundance and diversity index calculations are needed. (C) 2014 Elsevier B.V. All rights reserved.