Previous attempts to explain the positive relationship between the number of species in, and the productivity of, ecosystems have focused on sampling effects and niche differentiation. Both of these mechanisms explain the behavior of entire ecosystems largely in terms of the "individualistic" properties of their component populations. For example, niche differentiation requires only that each species occupy the part of the ecosystem to which it is best adapted. More holistic alternatives have been identified, such as inter-species facilitation, e.g., by legumes of other plants species. But the relative importance of individualistic vs. holistic mechanisms has not been quantified. Here I quantify their relative importance through re-analysis of published data from the Cedar Creek "big biodiversity" experiments.
Interactions, as opposed to species' individualistic properties, dominate the relationship between species richness and productivity in these experimental plots. However, because niche differentiation relies on habitat heterogeneity, and such heterogeneity increases with spatiotemporal scale, we should expect niche differentiation to play a relatively larger role at scales larger than the plots analyzed here. The North American breeding bird survey could be used to test this prediction, since it allows species richness, ecosystem productivity, and other relevant variables to be estimated for several different scales.