Mechanisms underlying biodiversity-ecosystem functioning relationships are closely linked to mechanisms underlying community assembly. Therefore, to understand the first it is necessary to understand the second. For example, if diversity effects are caused by sampling the best among many species, then a single best species will maintain as much function as a diverse assemblage. On the other hand, complementarity among species in resource use or in hosting specialized enemies may increase community performance and at the same time allow a large number of species to co-exist. We took advantage of a large plant biodiversity experiment to test whether sampling effects led to high abundance of single or few species or whether, instead, several species maintained high abundance. We tested for complementarity by asking whether abundant species were more different from each other than the same number of randomly selected species. Difference was measured as phylogenetic distance between species, because this should be the most comprehensive and neutral measure of both differences in resource and enemy "niches".
Results/Conclusions
We found that experimental communities with initially even species distributions rapidly developed log-normal rank-abundance distributions. However, the most abundant species were generally more different from each other than expected by chance. This effect was independent of initial species and functional group richness. At the same time, diverse communities whose most abundant species were different were also the most productive ones. When invasion into experimental communities was allowed, differences between the most abundant species increased even further because there were greater opportunities for complementarity when all species in the species pool were allowed to colonize the community. Colonization did not always lead to an increase in productivity. These results show that sampling effects or environmental filtering of species with similar traits are less common than the opposite forces of species occupying different niches or limiting similarity, at least for the investigated grassland ecosystem. However, a decoupling of mechanisms promoting diversity and productivity may occur in communities re-assembling by invasion. In this case, species may invade because they find enemy-free niches, yet take resources away from resident species that in the original community might have had higher productivity.