The relationship between biodiversity and ecosystem functioning has become a prominent topic in the ecological literature. However, the contemporary approach that species diversity controls primary productivity contrasts with the historical perspective that species diversity responds to productivity. Moreover, previous experimental results have not been consistent with the patterns observed in nature. To resolve these questions, the multivariate productivity–diversity (MPD) hypothesis proposes a bidirectional relationship between diversity and productivity. It predicts that the resource supply, expressed in terms of resource availability and imbalance, establishes the number of species that can locally coexist. Simultaneously, the resource supply also indirectly affects biomass production, determining the form and cause of the effects of species richness on resource use and biomass. To test the MPD hypothesis, we conducted three field experiments with a subtidal marine macroalgal community using a seasonal upwelling process as a driver of distinct levels of nutrient supply. Seasonally, macroalgal species richness and biomass were assessed and experimental manipulations conducted to investigate the relative importance of species richness and identity effects on biomass production and the mechanisms underlying these. Changes in macroalgal biomass and species richness were observed in response to the nutrient supply. Stronger effects of species identity were detected for all periods investigated, although species richness effects also occurred to some extent. The magnitudes of the net biodiversity and of the complementarity effects were a unimodal function of nutrient supply, whereas a concave-up curve was observed for selection effects. The nutrient supply directly affected the number of species that dominated the local community and, consequently, determined the efficiency with which resources were exploited and converted to biomass. Our results provide evidence consistent with the MPD hypothesis and aids in explaining the discrepancies between experimental results and natural patterns through the merging of two contrasting perspectives in ecology.