Functional litter diversity accelerates decomposition: underlying patterns and processes

Human land use increasingly simplifies vegetation, replacing diverse communities with species-poor stands and monocultures. This raises a key question for the global carbon cycle: how does the loss of plant and litter diversity alter both the rate and the stoichiometry of litter decomposition? Here we use the world’s largest biodiversity–ecosystem functioning experiment to link aboveground plant diversity, litter diversity (both traits and species), and the fate of carbon and nitrogen during decomposition. We partitioned litter-diversity effects on mass loss into complementarity and dominance components, further decomposing changes in element mass into weight and concentration effects and contrasting initially high- vs low-N species. Mixed-species litter showed clear positive diversity effects on mass loss, decomposing faster than monocultures across a wide range of vegetation conditions. The additive partitioning revealed that these gains were driven primarily by complementarity among litter species rather than dominance by the fastest-decomposing species, highlighting the importance of functional differences among species. Despite faster mass loss, mixtures retained more carbon, and especially more nitrogen, than expected from monoculture relationships between mass loss and element loss. This “extra” retention arose because microbes immobilized nitrogen onto initially N-poor litter, generating opposite deviations in N retention among partner species and producing relatively recalcitrant, N-enriched residues. Concentration effects, rather than changes in litter mass alone, explained much of the net diversity effect on N retention, indicating that litter functional diversity reshapes stoichiometric trajectories during decay and partially decouples carbon and nitrogen fluxes. Our results show how litter diversity and trait differences among species jointly accelerate decomposition while constraining carbon losses via stoichiometric controls. Incorporating such diversity-dependent controls on both decomposition rates and element retention into Earth system models could improve predictions of soil carbon persistence under ongoing vegetation homogenization.