Partitioning species and environmental diversity in fragmented landscapes: do the alpha, beta and gamma components match?

To understand patterns of alpha, beta and gamma diversities in fragmented landscapes we need to explore the three scale components in relation to potential drivers in a scale-dependent manner. Often, the drivers themselves can be partitioned to alpha, beta and gamma diversities. Thus, one can hypothesize that the scale-components of species diversity and drivers’ diversity match, i.e., that species alpha diversity is mainly explained by drivers’ alpha diversity, beta by beta and gamma by gamma. Here, we explore this ‘scale-matching’ hypothesis for spiders in two fragmented agricultural landscapes. In each landscape, we sampled spiders and their potential prey in 12 patches. Then, we sub-sampled pseudo-landscapes in which we calculated spider alpha, beta and gamma diversities using multiplicative diversity-partitioning. Next, we used variance partitioning analysis to explore the relative contribution of eleven explanatory variables from five thematic groups (sampling intensity, area, connectivity, habitat diversity and prey diversity), while further partitioning the habitat and prey diversities to their corresponding alpha, beta and gamma diversities. We found considerable evidence for scale-matching, with spiders’ alpha and beta diversities explained mostly by the corresponding alpha and beta diversities (respectively) of prey and/or habitat. We further found a strong effect of connectivity on spider beta diversity, but not on alpha and gamma diversities. For spiders gamma diversity, a cross-scale effect was observed. Our results suggest that multiple drivers from multiple scales interact in structuring patterns of spider alpha, beta and gamma diversities in agro-ecosystems, yet the strongest effects are of those drivers that match in scale.