A large body of research has addressed how biological diversity mediates the functioning of ecosystems, and most has focused on single processes, notably primary productivity. Yet ecosystems function via highly complex interactions that create numerous natural products and services important to human society. Both this natural complexity and the multitude of benefits human society expects from ecosystems demand an expanded, multivariate concept of ecosystem functioning. The multifunctionality of ecosystems is inherently related to food-web processes because adding trophic levels rapidly increases both the number and types of direct and indirect interactions among organisms and their ecosystem-level consequences.
Results/Conclusions

Recognizing the multivariate nature of ecosystem functioning changes our understanding of the functional importance of biodiversity in several ways: (1) the sampling effect, which considers diversity-function correlations to be statistical artifacts of individual species dominance, becomes less important or disappears; (2) the number of species required to maintain ecosystem (multi)functioning increases; and (3) the slope of the diversity-(multi)functioning relationship becomes shallower, i.e., requires more species to reach an asymptote. The shift in focus to a more realistic, multivariate concept of ecosystem functioning renders several persistent criticisms of biodiversity-function research less persuasive or even irrelevant. It also suggests, soberingly, that previous experiments are more likely to have underestimated than overestimated the importance of biodiversity to ecosystem functioning and service provision in nature.