Ecosystem effects of biodiversity manipulations in European grasslands

We present a multisite analysis of the relationship between plant diversity and ecosystem functioning within the European BIODEPTH network of plant-diversity manipulation experiments. We report results of the analysis of 11 variables addressing several aspects of key ecosystem processes like biomass production, resource use (space, light, and nitrogen), and decomposition, measured across three years in plots of varying plant species richness at eight different European grassland field sites. Differences among sites explained substantial and significant amounts of the variation of most of the ecosystem processes examined. However, against this background of geographic variation, all the aspects of plant diversity and composition we examined (i.e., both numbers and types of species and functional groups) produced significant, mostly positive impacts on ecosystem processes. Analyses using the additive partitioning method revealed that complementarity effects (greater net yields than predicted from monocultures due to resource partitioning, positive interactions, etc.) were stronger and more consistent than selection effects (the covariance between monoculture yield and change in yield in mixtures) caused by dominance of species with particular traits. In general, communities with a higher diversity of species and functional groups were more productive and utilized resources more completely by intercepting more light, taking up more nitrogen, and occupying more of the available space. Diversity had significant effects through both increased vegetation cover and greater nitrogen retention by plants when this resource was more abundant through N-2 fixation by legumes. However, additional positive diversity effects remained even after controlling for differences in vegetation cover and for the presence of legumes in communities. Diversity effects were stronger on above- than belowground processes. In particular, clear diversity effects on decomposition were only observed at one of the eight sites. The ecosystem effects of plant diversity also varied between sites and years. In general, diversity effects were lowest in the first year and stronger later in the experiment, indicating that they were not transitional due to community establishment. These analyses of our complete ecosystem process data set largely reinforce our previous results, and those from comparable biodiversity experiments, and extend the generality of diversity-ecosystem functioning relationships to multiple sites, years, and processes.