Multiple Mechanisms Required to Predict Grass Community Composition

Catford, J.A., Graham, L., Shepherd, H.E.R., Hauser, C.E., Munro, N.T., Wintle, B.A., Donohue, J.G., Tilman, D., & Clark, A.T. (2026). Multiple Mechanisms Required to Predict Grass Community Composition. Ecology Letters 29 (4) e70358. 10.1111/ele.70358.

Preview

Text Ecology Letters - 2026 - Catford - Multiple Mechanisms Required to Predict Grass Community Composition.pdf - Published Version Available under License Creative Commons Attribution. Download (2MB) | Preview

Abstract

Accurate prediction of community assembly is a central goal in ecology but is challenging because assembly is governed by numerous mechanisms. Few theoretical models explicitly incorporate or test multiple mechanisms at once. We empirically tested the predictive performance of a plant community assembly model built using all possible combinations of four ‘mechanisms’ (soil resource competition, dispersal and colonisation, spatiotemporal niche differentiation, population growth rates) and 11 underlying ‘attributes’ based on measured traits (e.g., fecundity, phenology). The full model accurately predicted out‐of‐sample biomass observations of five grasses sown in mixture along a soil nitrogen gradient (overall R 2 = 0.65). Alternative model variants, parameterised using subsets of the mechanisms and their nested attributes, still retained high explanatory power if the model included at least three of the four mechanisms. Our results suggest that plant community composition is determined by simultaneous effects of multiple mechanisms, and simpler theories have much lower predictive abilities.

Actions (login required)

View Item