The rapid loss of biodiversity from natural communities has motivated much research investigating how this loss impacts the provision of ecosystem services. The majority of these studies have concentrated on individual ecosystem functions, and have, for the most part, identified a positive, but decelerating, relationship between biodiversity and ecosystem functioning. Only a small subset of the species in a community is generally important for providing individual ecosystem functions, suggesting that many species are seemingly redundant. Therefore, losing a large proportion of the community may have no effect on the function of the ecosystem. However, natural ecosystems are managed for multiple purposes, and if there is little overlap in the species contributing to different functions, then much higher biodiversity may be necessary. We analysed published data from eight European grassland biodiversity experiments to examine whether different sets of species are important for the provision of different ecosystem functions. We identified the set of species that contributed to each ecosystem function individually. We then determined the degree of overlap between these sets of species, and the number of species required to provide combinations of ecosystem functions.  

Results/Conclusions

Our analyses demonstrated that there were appreciable differences in the sets of species contributing to different ecosystem functions. The average proportional overlap between ecosystem function pairs ranged between 0.2 and 0.5, depending on site.  As a result, there was a positive, decelerating relationship between the number of important species and the number of ecosystem functions considered. This pattern was common to all sites. We conclude that the number of species required for fully functional ecosystems might be far greater than estimated by experiments concentrating on single ecosystem functions.