Reconciling results of biodiversity-productivity analyses in natural and experimental systems

Background/Question/Methods

Experiments on the effect of species richness on productivity of plants consistently report positive relationships, while observational studies found no or negative relationships. Consequently, it remains debatable that the invoked processes underlying the positive relationship, e.g., complementarity, are relevant in natural systems where species richness is not maintained by weeding. By using an approach where mixtures of different plant species richness were sown and afterwards left open for invading plants, we show that these discrepant results can be reconciled. 

Results/Conclusions

We found an overall significant negative slope between species richness and aboveground biomass, as well as a very strong effect of plant species composition. By separating the experimental plots into clusters according to compositional similarity, we observed that the slopes are very variable within clusters, from positive to negative. Using a measure describing the potential for niche complementarity of a composition, we find that the slopes are positively related to this metric. Consequently, negative and positive biodiversity-productivity relationships are not incompatible results, but can be placed along a continuum controlled by the potential for niche complementarity of the plant community.