Understanding the mechanisms through which biodiversity affect community stability, and predicting this effect is an important goal of community ecology. We use a mechanistic model that describes the dynamics of a competitive community driven by a combination of density dependence, environmental forcing and demographic stochasticity to develop an approximate predictor of the temporal variance of community biomass.

Results/Conclusions

Our predictor is exact when species are symmetrical (they have the same carrying capacities and competitive effects of each species on each other are the same), and we show that it is still an excellent approximation when species differ. Our predictor is based on the relative abundance of species in the mixture, the covariances of their environmental responses and an average species intrinsic growth rate. It can be estimated using temporal time-series of individual species monocultures plus the average abundance of each species in a polyculture. Our predictor shows that the asynchrony of species responses to exogenous environmental forcing is the key mechanism that determines community stability. By contrast, we explain why summed variances and summed covariances fail to provide insights into the mechanism that generates the stabilizing effect of species diversity on community properties. We clarify the relationships between summed covariances, interspecific competition, functional compensation, species asynchrony, statistical averaging, and community stability.