Multitrophic functional diversity predicts ecosystem functioning in experimental assemblages of estuarine consumers
The use of functional traits to explain how biodiversity affects ecosystem functioning has attracted intense interest, yet few studies have a priori manipulated functional diversity, especially in multitrophic communities. Here, we manipulated multivariate functional diversity of estuarine grazers and predators within two levels of species richness to test how species richness and functional diversity predicted ecosystem functioning in a multitrophic food web. Community functional diversity was a better predictor than species richness for the majority of ecosystem properties, based on general linear mixed effects models. Combining inferences from 8 traits into a single multivariate index increased prediction accuracy of these properties relative to any individual trait. Structural equation modeling revealed that functional diversity of both grazers and predators was important in driving final biomass within and between trophic levels, with stronger effects observed for predators. We also show that different species drove different ecosystem responses, with evidence for both sampling effects and complementarity. Our study extends experimental investigations of functional trait diversity to a multilevel food web, and demonstrates that functional diversity can be more accurate and effective than species richness in predicting community biomass in a food web context.