The importance of heterogeneity revisited from a multiscale and multitaxa approach

The importance of spatial scale for p-diversity has been shown in several studies, but it is unclear how spatial diversity patterns correlate among different organismic groups. We studied spatial diversity organization of plants and several trophic guilds of beetles in beech-dominated forests in two regions of Germany to test whether different trophic guilds are organized independently in space. We applied multiplicative diversity partitioning using a nested hierarchical design of four increasingly broader spatial levels (subplot, plot, forest class, region) and tested for correlations among trophic guilds by using Pearson product moment correlations and Mantel-tests. We observed similar general diversity patterns at different trophic guilds showing a high contribution of beta-diversity to total gamma-diversity and found beta-diversity to be higher at different spatial scales and alpha-diversity to be lower than expected by random distributions of individuals. Results, however, partly depended on the weighting of rare and abundant species. Beta-diversity in our study was caused mainly by species spatial turnover rather than by nestedness. Correlations of alpha-diversity between trophic guilds were low whereas correlations of beta-diversity above subplot level were high. Importantly, more strongly connected trophic guilds revealed not generally stronger relationships than less strongly connected guilds. Three important implications for conservation can be deduced from our results: (1) heterogeneity of beech forests at different spatial scales should be supported in conservation strategies to enhance biodiversity and related functions; (2) the observed high importance of spatial turnover in relation to nestedness implies a concentration of conservation efforts to a large number of not necessarily the richest sites, and (3) recommendation for particular conservation strategies (e.g. selection of priority sites for conservation at regional scale) based on single indicator taxa or functional guild is difficult because of the varied response of the species in our study. (C) 2013 Elsevier Ltd. All rights reserved.