No abstract

The BMRH described the average response of litter mixtures. The decrease in variance and the convergence to the predicted values based on CWMs was not due to the 'idiosyncratic annulment' of species interactions but was a mathematical consequence of CWMs being sums of random variables. Since convergence occurs with increasing SR and since SR increases with increasing spatial scale, the spatial scale will be a determinant in the prediction of ecosystem processes, such as litter decomposition rates.

Community-weighted mean values provided good predictions of mixed-species litter decomposition, converging to the predicted values with increasing species richness and in climates less favourable to decomposition. Under a context of climate change, abiotic variability would be important to take into account when predicting ecosystem processes.

We tested the hypothesis that interactions between plant species during the process of mixed-species leaf litter decomposition increases with increasing functional diversity of leaves within the mixtures; specifically, there is a positive correlation between functional dispersion and the deviations from Grime's biomass-ratio hypothesis, with a null intercept. We measured decomposition rates (mg g 1 d 1 ) of mixed-species leaf litter from two experimental designs: 1) a microcosm experiment with litterbags of species mixtures combining six tree species, alone and in 42 combinations, and 2) an in situ litterbag experiment with all possible mixture combinations of four herb species (from one to four species). Interaction strengths and directions were measured as deviations from community-weighted means (CWM) of monoculture decomposition values, following the biomass-ratio hypothesis (BRH). Functional diversity was measured as Laliberté and Legendre's functional dispersion (FDis), using leaf dry matter content (LDMC), leaf nitrogen and carbon contents, and proportions of water soluble compounds, cellulose, hemicellulose and lignin. Correlations between FDis and deviations from BRH varied strongly, depending upon the combination of functional traits, the plant type or the environmental conditions, and the way in which prediction error was expressed (absolute or actual deviation). For tree species, FDis that was based on a combination of water soluble compounds, hemicellulose concentration, and LDMC was negatively correlated with interaction strength but positively with its absolute value. For herbs, interaction strength (absolute or actual) decreased as FDis of the mixtures increased, based on cellulose and lignin contents. There was no positive correlation between functional dispersion and the deviations from Grime's biomass-ratio hypothesis, with a null intercept. Despite a relationship between litter interactions and functional divergence, this relationship was not generalisable. Other functional traits that were missing in our study might have played an important role.

No abstract