Abstract:Influential ecological research in the 1980s, elucidating that local biodiversity (LB) is a function of local ecological factors and the size of the regional species pool (γ-diversity), has prompted numerous investigations on the local and regional origins of LB. These investigations, however, have been mostly limited to single scales and target groups and centered exclusively on γ-diversity. Here we developed a unified framework including scale, environmental factors (heterogeneity and ambient levels), and metacommunity properties (intraspecific spatial aggregation, regional evenness, and γ-diversity) as hierarchical predictors of LB. We tested this framework with variance partitioning and structural equation modeling using subcontinental data on stream diatoms, insects, and fish as well as local physicochemistry, climate, and land use. Pure aggregation + regional evenness outperformed pure γ-diversity in explaining LB across groups. The covariance of the environment with aggregation + regional evenness rather than with γ-diversity generally explained a much greater proportion of the variance in diatom and insect LB, especially at smaller scales. Thus, disregarding aggregation and regional evenness, as commonly done, may lead to gross underestimation of the pure metacommunity effects and the indirect environmental effects on LB. We examined the shape of the local-regional species richness relationship, which has been widely used to infer local vs. regional effects on LB. We showed that this shape has an ecological basis, but its interpretation is not straightforward. Therefore, we advocate that the variance partitioning analysis under the proposed framework is adopted instead. In diatoms, metacommunity properties had the greatest total effects on LB, while in insects and fish, it was the environment, suggesting that larger organisms are more strongly controlled by the environment. Broader use of our framework may lead to novel biogeographical insights into the drivers of LB and improved projections of its trends along current and future environmental gradients.

A new perspective on the spatial, environmental, and metacommunity controls of local biodiversity