5 ± 1.0, medium 7.0 ± 1.2). How this finding should be applied into practical conservation is further discussed under the last section ‘Practical implication’. From our field observations of the study sites we noticed a distinguishable difference of the four largest sand pits from the smaller ones. Selleck Nepicastat The large sand pits could all be described as more homogenous in terms of topology and vegetation; with large

plane areas, steep edges and either even-aged young trees or almost no vegetation at all. We believe this difference between sites could explain why no more sand species were found in large sand pits compared to medium-sized ones. Of the two prominent hypotheses for SAR this observation would give more support to the ‘habitat heterogeneity hypothesis’ than the ‘area per se hypothesis’ (Báldi 2008). However, the strong interactions between the features that the two hypotheses are based upon make drawing clear conclusion difficult without further direct studies (Connor and McCoy 1979; Kallimanis et al. 2008). The rate of increase JPH203 in species number with area were illustrated by the log–log SA-curves of the power function (Fig. 2) and showed a more rapid increase in species number for carabids (z = 0.25) than for all beetles (z = 0.12). According

to Connor and McCoy (1979), z values regularly fall between 0.20 and 0.40, and according to a review by Drakare et al. (2006), the average z value obtained in investigations using independent sampling schemes (among 794 SAR studies considered) was 0.24. Whether the z value has any further biological significance has been debated, often with scepticism (Connor and McCoy 1979; He and Legender 1996; Martin 1981).

However, Drakare et al. (2006) detected apparent systematic correlations between z values and latitude (negative), organism size (negative; explained by the Metalloexopeptidase higher dispersal ability of small organisms) and habitat (lower in non-forested habitats). As this study examined relationships of small organisms dwelling in non-forested habitats at high latitude we should expect low z values, which was true for beetles (0.12), but not for carabids (for which the value was close to the average cited above; 0.25). YH25448 influence of the surrounding matrix In contrast to the sand species, no SAR was found when all species (irrespective of habitat-preferences) were included in the analysis. The same pattern has been observed for other terrestrial habitat islands, in which positive SARs have only been found for the habitat-specific species (Lövei et al. 2006; Magura et al. 2001; Vries de et al. 1996). This can be explained by an influence of the surrounding matrix where matrix species invade the habitat island resulting in an increase of species richness along the edges (Cook et al. 2002; Ewers and Didham 2006; Magura 2002; Niemelä 2001). This edge effect then counteracts the area effect because of the greater edge:area ratio in smaller patches (Lövei et al.