Divergent responses of β‐diversity among organism groups to a strong environmental gradient

A limited understanding of how variation in the species composition among communities (i.e., β‐diversity) changes along natural environmental gradients, and the mechanisms responsible, inhibits our ability to understand large‐scale biodiversity change resulting from either natural or anthropogenic drivers. Therefore, our aim was to test key drivers of β‐diversity patterns along a strong, natural environmental gradient for seven widely different organisms groups, that is, root‐associated fungi, litter fungi, soil nematodes, vascular plants, epiphytic lichens, beetles, and spiders. Using previously published community‐level data from boreal‐forested islands, we calculated α‐diversity and β‐diversity for each of the seven organism groups. Out of several available environmental variables, we identified four variables, that is, ecosystem age, total C storage, net primary productivity ( NPP ), and N‐to‐P ratio, as potential predictors of variation in β‐diversity. We found that ecosystem age was the variable with the highest overall importance. We then used two different methods to quantify the relative importance of stochastic and deterministic processes underlying patterns in β‐diversity along the ecosystem age gradient, and our detailed knowledge based on prior data collection in the study system to mechanistically explain among‐group differences in these patterns. We found divergent responses in β‐diversity along the age gradient for the seven different organism groups, due to among‐group differences in the relative importance of deterministic vs. stochastic community assembly, and attributed these results to reliance on resources from different energy channels that are not always related to NPP . Our results highlight the necessity to consider the importance of taxon‐specific resources, and not only NPP , to obtain an understanding of β‐diversity patterns among organism groups and ecosystems, as well as large‐scale patterns in biodiversity. They therefore also suggest that management and protection of β‐biodiversity in the landscape requires explicit consideration of a wide range of habitats.