Considering the current rate of extinctions, it is crucial to understand the consequences of these losses of biodiversity for the functioning of ecosystems. Grasslands proved a very suitable ecosystem for biodiversity-ecosystem functioning research.
In earlier experiments, nitrogen-fixing legumes dominated the effects of plant diversity, but in our experiment without legumes we also found that diverse plant communities have higher productivity than species-poor ones. We identified complementarity in nutrient uptake and a novel mechanism - more efficient nutrient use at high diversity – as the most important driving forces behind the positive effects of diversity. In contrast to primary productivity, invasibility of the community was mainly affected by the presence of two particular plant species. Plant diversity per se only played a minor role.
We found that the diversity of nematodes, being a very important component of the soil food web, increased with plant diversity. Specific associations between plants and nematodes resulted in increased nematode diversity at high plant diversity.
Of course, higher trophic level organisms do not only respond to plant diversity, they may also regulate plant diversity. Earlier studies have reported strong effects of herbivores, pathogens and mutualists on plant communities, but most of these studies investigated a single group of organisms in isolation from other driving forces. We show that the separate effects of these organisms on plant diversity may change dramatically when studied in combination. A neutral effect of aboveground insect herbivores on plant diversity and a positive effect of belowground insect herbivores turned strongly negative when both herbivores were present in the plant community.
It is argued that a combined approach, using interactions between changes in plant diversity and multiple ‘drivers of plant diversity’ (both biotic and abiotic) is needed to fully understand plant community dynamics. To be able to predict the consequences of biodiversity loss worldwide it is crucial to integrate the results of this approach over different ecosystems and ecosystem processes across large spatial and temporal scales.