COMMUNITY ECOLOGY Tobias Purtauf Æ Jens Dauber Æ Volkmar Wolters The response of carabids to landscape simplification differs between trophic groups Received: 25 June 2004 / Accepted: 22 September 2004 / Published online: 14 October 2004 Ó Springer-Verlag 2004 Abstract We studied the response of carabid species richness and density to landscape simplification (mea- sured as percentage cover of non-crop habitat sur- rounding each study site) in 36 wheat fields using pitfall traps. Carabids were divided to trophic groups following the literature. The number of species from different trophic groups declined with increasing landscape simplification in the order: carni- vores > phytophages > omnivores. Density compen- sation of both carnivores and phytophages suggests that species decline is caused by the loss of specific resources rather than by an overall reduction in food availability. Increasing evenness indicates that a greater share of phytophagous species contributes to density compensation at poorer sites. A comparison with data from complementing studies shows that marked dif- ferences in species numbers (carnivores > omni- vores > phytophages) are due to a different sensitivity of trophic groups to agricultural management. Since our findings seem to be partly due to increasing sen- sitivity to landscape changes with trophic rank, and partly to decreasing sensitivity of depauperate com- munities to local environmental stress, species loss can best be explained by the co-action of factors at local and regional scales. Species richness decline might significantly alter the role of carabids as biocontrol agents. Keywords Carabidae Æ Trophic rank Æ Species richness Æ Density compensation Æ Landscape composition Introduction Predicting the effect of spatial changes on populations and communities is a major challenge to biodiversity research (Dunning et al. 1992; Bestelmeyer et al. 2003). One essential prerequisite for doing so would be to establish quantitative links between spatial patterns and biodiversity (Gustafson 1998; Tischendorf 2001; Turner et al. 2001). This is, however, very difficult, because each taxon of the enormously rich terrestrial community potentially responds differently to the spatial dynamics and the heterogeneity of its environment (Kareiva 1994; Dauber et al. 2003). One possibility for simplifying this problem is provided by the hypothesis that species-area effects should become stronger as trophic rank increases (Holt et al. 1999). Empirical evidence either challenges (Borrvall et al. 2000; Woodward and Hildrew 2002) or supports (Kruess 2003; Thies et al. 2003) this hypothesis. However, aggregation into functional groups not only makes it possible to detect spatial influences across broad ecological groups, rather than being reliant on the presence of a few key indicator species, but may also help in predicting associated influences on ecosystem functioning (Cole et al. 2002). Species at higher trophic ranks depend on spatial composition at larger scales than local patches, since they are more likely to have large home ranges (e.g. Ritchie and Olff 1999). This suggests that landscape ef- fects are different for different trophic groups. We have tested this hypothesis by investigating the impact of landscape simplification on carabid communities inhabiting wheat fields. Landscape simplification should lead to a significant decline in the effective area, since carabids critically depend upon a variety of habitats within their movement range for the provision of alter- nate food sources, as well as refuge and overwintering sites (Bommarco 1999; Fournier and Loreau 2002). Though most carabids are unspecialised carnivores, several species are either phytophagous or omnivorous (Thiele 1977). Species belonging to the omnivorous T. Purtauf (&) Æ J. Dauber Æ V. Wolters IFZ, Department of Animal Ecology, Justus-Liebig-University of Giessen, Heinrich-Buff-Ring 26-32, 35392 Giessen, Germany E-mail: tobias.purtauf@allzool.bio.uni-giessen.de Tel.: +49-641-9935711 Fax: +49-641-9935709 Oecologia (2005) 142: 458–464 DOI 10.1007/s00442-004-1740-y