Ecological theory predicts that niche and relative fitness differences among plant species underlie variation in the biodiversity-productivity relationship. However, the mechanisms by which plant species differ are rarely experimentally quantified, leaving causal connections between plant diversity and productivity elusive. Plants in the legume family are infected by beneficial soil bacteria called rhizobia, which fix atmospheric nitrogen (N) in exchange for carbon (C) from their plant hosts. As most plants obtain N from the soil, access to atmospheric N via this mutualism can differentiate legumes from non-legumes, thus influencing the productivity-diversity relationship. Increased atmospheric carbon dioxide (CO2) and soil N will likely influence the effect of legume species on the productivity-diversity relationship by altering the legume-rhizobia mutualism. We are experimentally examining the biotic processes that differentiate four legume species from each other and from non-legumes to understand how these differences cause variation in patterns of primary productivity within the BioCON experiment at Cedar Creek LTER. This research will inform management of ecosystem function under current and future global change scenarios, with important implications for human well-being.