2007; Fletcher et al. 2007). By only including species specialized to the habitat studied, the habitat island will more likely resemble an actual island and hence better follow island biogeography theory (MacArthur and Wilson 1967). Our findings strengthen the notion that only species specialized to the habitat studied should be included when applying SAR in terrestrial habitat

patches. Our notion that sand pits are influenced by species from the surrounding matrix is further strengthened as the species assemblages in the sand pits were related to CT99021 datasheet the surrounding edge habitat. When surrounded by forests there was a higher proportion of forest species in the patches, and when surrounded by open areas the proportion of open ground

species was higher. For the proportion of sand species there was no relationship with the type of edge habitat. These patterns combined strongly suggest that there are edge effects mainly affecting our small sand pits (0.02–0.23 ha). Species composition The area of the sand pit was the major factor influencing species composition. The main difference PD0332991 research buy in species composition was between small sand pits and medium/large ones where most sand species were associated with the medium/large sand pits (Fig. 3). The composition of carabids was in addition influenced by the proportion of sand material. This variable differentiates between the coarseness of the ground material (either sand or gravel) hence some species seem to have preference for one or the other soil type. Effect of environmental variables The proportion of sand material had a positive influence on species number of all beetles, whereas the influence was not significant for sand species. Also, the number of forest species increased with an increase in proportion of sand material (when the type of edge habitat was accounted for). We would have expected a connection between sand species and proportion of

sand material but why total species number and forest species would be affected is puzzling to us and thus we keep from speculation about its reasons. The proportion of sand species was positively influenced by tree cover. The influence of tree cover is puzzling and CYTH4 we can only speculate of its function. It might work as a wind shelter improving the microclimate or it could be due to that boreal sand species have evolved to use habitats produced by ground fires in forests, where a lot of trees are retained. Carabids as indicators The value of carabids as indicators of total beetle species diversity in sand pits lies almost solely in their high representation among the sampled species. The analyses including all beetles gave similar results to those including only carabids for the SAR and species composition (CCA), but not for the other environmental variables tested. Thus, we cannot fully support carabids as useful indicators of other beetles in sand pits.