Effects of biotic homogenization on community assembly and biomass production in an α × β-diversity grassland experiment.

Various studies have shown that local loss of plant species within patches (i.e., α-diversity loss) hampers plant biomass production, raising concerns about the consequences of biodiversity loss. However, biodiversity loss occurs at multiple spatial scales, and species losses at large spatial scales (γ-diversity loss) can be either caused by α-diversity loss, or by the homogenization of species compositions across patches (β-diversity loss). Yet, we poorly understand the consequences of β-diversity loss for ecosystem functioning. Here, we present the findings of the first study to experimentally manipulate variation across not only α-diversity but also β-diversity. In a grassland experiment, we manipulated plant β-diversity by setting up 56 plots, each divided into four subplots with two different soil types. Across the subplots within a plot, we sowed either the same plant species combinations (low β-diversity treatment), partially overlapping species combinations (intermediate β-diversity), or nonoverlapping species combinations (high β-diversity treatment). We hypothesized that high β-diversity could increase biomass production by: (1) enabling species sorting and (2) maintaining high α-diversity over time. Four years after setting up our experiment, we found that β-diversity increased biomass production and that these positive effects could not be solely explained by having overall high γ-diversity. This was because high sown levels of β-diversity increased realized α-diversity, which in turn increased biomass production. In contrast, species sorting did not result in increased biomass production. Our results have the potential to inform restoration projects, where sowing different species at different sites could have delayed benefits for both biodiversity and ecosystem functioning, although further studies are needed to explore the benefits of β-diversity for ecosystem management.