Solar radiation and functional traits explain the decline of forest primary productivity along a tropical elevation gradient

One of the major challenges in ecology is to understand how ecosystems respond to changes in environmental conditions, and how taxonomic and functional diversity mediate these changes. In this study, we use a trait-spectra and individual-based model, to analyse variation in forest primary productivity along a 3.3 km elevation gradient in the Amazon-Andes. The model accurately predicted the magnitude and trends in forest productivity with elevation, with solar radiation and plant functional traits (leaf dry mass per area, leaf nitrogen and phosphorus concentration, and wood density) collectively accounting for productivity variation. Remarkably, explicit representation of temperature variation with elevation was not required to achieve accurate predictions of forest productivity, as trait variation driven by species turnover appears to capture the effect of temperature. Our semi-mechanistic model suggests that spatial variation in traits can potentially be used to estimate spatial variation in productivity at the landscape scale.