Does productivity drive diversity or vice versa? A test of the multivariate productivity-diversity hypothesis in streams

The idea that productivity regulates species diversity is deeply ingrained in the field of ecology. Yet, over the past few decades, an increasing number of experiments have shown that species diversity controls, rather than simply responds to, biomass production. These contrasting perspectives have led to a seeming paradox: Is diversity the cause or the consequence of biological production? Here we present empirical evidence for the multivariate productivity-diversity (MPD) hypothesis, which argues that differing perspectives on productivity-diversity relationships can be resolved by recognizing that historical research has focused on how resource supply regulates both the productivity and richness of local competitors, whereas more recent studies have focused on how the richness of a colonist pool regulates the efficiency by which resources are captured and converted into new tissue. The MPD hypothesis predicts that three pathways operate concurrently to generate productivity-diversity relationships in nature: (1) resource supply directly limits the standing biomass and/or rate of new production by primary producers, (2) producer biomass is directly influenced by the richness of species that locally compete for resources, and (3) resource supply rate indirectly affects producer biomass by influencing the fraction of species from a colonist pool that locally coexist. To examine whether this set of pathways explains covariation between productivity and diversity in natural streams, we used nutrient-diffusing agar "patches'' to manipulate resource supply rates in 20 streams throughout the Sierra Nevada mountain range, California, USA. We then measured the fraction of periphyton species from the stream colonist pool co-occurring on each nutrient patch, as well as the standing biomass and rates of primary production. Natural patterns of covariation agreed with predictions of the MPD hypothesis. Algal biomass was an increasing function of nutrient supply, and an increasing function of local richness. The fraction of species from the colonist pool found co-occurring on a patch was a concave-down function of nutrient supply, causing nutrients to indirectly affect biomass via control over local richness. These results suggest that the MPD hypothesis is a viable explanation of patterns of diversity and productivity in natural stream ecosystems, and that it has potential to merge the historical view that productivity drives diversity with a parallel view that diversity drives productivity.