Toward a mechanistic understanding of the effects of nitrogen and phosphorus additions on grassland diversity

Recently, it has been shown that nitrogen (N) and phosphorus (P) additions lead to different effects on grassland diversity. While N additions decrease species richness, the effects of P additions are mainly neutral. This contribution focuses on the mechanisms behind this observed pattern. First, I elaborate on three classical hypotheses explaining the diversity response to nutrient enrichment. These hypotheses highlight the roles of above ground competition ('the light asymmetry hypothesis'), above and below ground interactions ('the total competition hypothesis') and litter production ('the litter hypothesis'). I test whether these hypotheses are able to explain the differences between N and P additions using data from the nutrient network, a globally distributed experiment (including 45 sites from five continents). I show that N addition leads to species loss (c.13%) but P addition has no net effect on species richness. Furthermore, the effects of N and P additions on richness are mediated by biomass, litter and light availability as proposed by the general hypotheses. Nonetheless, there are additional (biomass independent) effects of N addition on species richness. These effects elucidate the main limitation of the general hypotheses that do not explicitly consider the differences between N and P additions. Hence, I review the current knowledge on the differences between N and P additions by integrating the specific characteristics of N and P with the potential mechanisms proposed by ecological theory. The review focuses on mechanisms mediated by soil chemistry and stoichiometry, variation in light acquisition among species as well as species pool size of different functional groups. These mechanisms are incorporated into a detailed graphical model that allows testing the old and new hypotheses presented in this contribution.