Humans are influencing biodiversity in ecosystems worldwide because of a variety of well-characterized global changes such as changes in atmospheric composition (e.g., elevated CO2 and other greenhouse gases), climatic changes resulting from those atmospheric changes, increasing amounts and mobility of key plant nutrients (e.g., nitrogen), changes in land use (and therefore amount of habitat), and enhanced invasions of species. All of these changes are adversely affecting biodiversity. The key question is whether declines in biodiversity will adversely affect the way ecosystems work, in terms, for example, of total plant growth, the availability of nutrients, and trophic relations—who eats who and how much.

Ecosystem processes and the resulting ecosystem services that benefit humanity are greatly influenced by the characteristics of the species present in that ecosystem. Declining biodiversity could influence these processes because key species or functional groups—sets of species with similar effects on processes—are lost from the ecosystem. For example, having more species of plants, each of which can get nutrients from different layers of the soil, or which together create a denser canopy to absorb sunlight, could lead to greater overall rates of plant production at the ecosystem scale.

Theory suggests that this should be the case, and experimental evidence at least partially supports the theory. However, other than a few key examples (e.g., plants with mutualistic associations with bacteria that enable them to use atmospheric nitrogen), we still don’t know exactly which species and how many species partition resources in this way in many ecosystems. It’s also possible that higher diversity affects processes not because of resource partitioning, but because one or two species are fast growing or very effective at taking up nutrients. Higher diversity just means a greater probability of having at least one such effective species included in the ecosystem. Our review suggests that both mechanisms can occur and that future research should attempt to more clearly delineate the different ecological conditions under which each mechanism operates.

Species diversity could influence not just overall process rates, as just discussed, but also the stability of ecosystem processes in response to disturbances (both natural and man-made), damage by pests or diseases, or year-to-year variation in climate. Again, theory, observational studies, and some experimental studies lend support to this prediction. However, understanding the underlying mechanisms is more difficult. Having more species may be important to stability because different species often have different environmental preferences. So, if conditions change such that they are less optimal for one species, other species are around to compensate. Some studies have found, however, that changes in which species are present have a larger effect on process stability than compensation among many species. One of the main points of our paper is that just knowing how many species are present is less powerful than knowing which species are present, because it is the functional traits of those species—not something special about diversity itself—that influence ecosystem processes.