The influence of species diversity on ecosystem productivity: how, where, and why?

The effect of species diversity on ecosystem productivity is controversial, in large part because field experiments investigating this relationship have been fraught with difficulties. Unfortunately, there are few guidelines to aid researchers who must overcome these difficulties and determine whether global species losses seriously threaten the ecological and economic bases of terrestrial ecosystems. In response, I offer a set of hypotheses that describe how diversity might influence productivity in plant communities based on three well‐known mechanisms: complementarity, facilitation, and the sampling effect. Emphasis on these mechanisms reveals the sensitivity of any diversity‐productivity relationship to ecological context (i.e., where this relationship should be found); ecological context includes characteristics of the surrounding environment, temporal and spatial scales of observation, and the intensity of human management. In particular, the legitimacy of the sampling effect as a mechanism of productivity enhancement is dependent upon the degree to which stochastic events influence immigration and extinction processes in a given ecosystem. A mechanistic approach also requires that the three mechanisms be separated and quantified in diversity experiments, and I examine the most appropriate analyses for doing so, focusing on the overyielding technique. Finally, I question why productivity per se is a relevant management concern in non‐agricultural systems once relationships among diversity, productivity, and the qualities of the surrounding environment are considered.