Stably coexisting communities deliver higher ecosystem multifunctionality
Linking the processes determining species coexistence and effects of biodiversity on ecosystem functioning is important for a mechanistic understanding of how species loss impacts ecosystems1. There are good theoretical reasons to expect stably coexisting communities to deliver higher ecosystem functioning, as niche differences should promote coexistence and complementarity2, however, there is still little empirical evidence to demonstrate this relationship3. We present a novel experimental approach to tackle this question, in which we first measured interaction coefficients and intrinsic growth rates for 12 grassland species4 and used population models to estimate coexistence5. We then assembled 48 three-species communities, predicted to differ in their degree of coexistence and measured seven above and belowground functions. Here we extend findings demonstrating that biodiversity increases multifunctionality6,7 to show that amongst communities with the same initial diversity, multifunctionality is higher when all the species are predicted to coexist. Decomposing coexistence into the influence of niche and fitness differences showed that multifunctionality was high in niche differentiated communities but also in those with a single highly competitive species. We further examined the effect of indirect interactions, which provide an alternative mechanism of coexistence in multispecies communities, and found that they also provided an alternative route to high multifunctionality. Our results point the way to a unification of coexistence and biodiversity theory and to a more mechanistic understanding of the functional effects of biodiversity change.