What is the observed relationship between species richness and productivity?

Understanding the relationship between species richness and productivity is fundamental to the management and preservation of biodiversity. Yet despite years of study and intense theoretical interest, this relationship remains controversial. Here, we present the results of a literature survey in which we examined the relationship between species richness and productivity in 171 published studies. We extracted the raw data from published tables and graphs and subjected these data to a standardized analysis, using ordinary least‐squares (OLS) regression and generalized linear‐model (GLIM) regression to test for significant positive, negative, or curvilinear relationships between productivity and species diversity. If the relationship was curvilinear, we tested whether the maximum (or minimum) of the curve occurred within the range of productivity values observed (i.e., was there evidence of a hump?). A meta‐analysis conducted on the distribution of standardized quadratic regression coefficients showed that the average quadratic coefficient was negative (i.e., the average species richness–productivity relationship was curvilinear and decelerating), and that the distribution of standardized quadratic regression coefficients was significantly heterogeneous (i.e., the studies did not sample the same underlying species richness–productivity relationship). Looking more closely at the patterns of productivity–diversity relationships, we found that, for vascular plants at geographical scales smaller than continents, hump‐shaped relationships occurred most frequently (41–45% of all studies). A positive relationship between productivity and species richness was the next most common pattern, and positive and hump‐shaped relationships co‐dominated at the continental scale. For animals, positive, negative, and hump‐shaped patterns were common at most geographical scales, and no one pattern predominated. For both plants and animals, hump‐shaped curves were relatively more common in studies that crossed community boundaries compared to studies conducted within a community type, and plant studies that crossed community types tended to span a greater range of productivity compared to studies within community types. Sample size and plot size did not affect the probability of finding a particular productivity–diversity relationship (e.g., positive, hump‐shaped, etc.). However, hump‐shaped curves were especially common (65%) in studies of plant diversity that used plant biomass as a measure of productivity, and in studies conducted in aquatic systems.