Human activities have increased the amount of available nitrogen (N) globally. Increased N-availability can change plant community structure and function, and lead to diversity loss. Species traits associated with differential resource limitation may predict how plant communities will respond to N-enrichment across ecosystems. We focused on specific leaf area (SLA), leaf area per unit leaf mass, as a candidate trait because it is correlated with high relative growth rates, photosynthetic rates, and leaf N-concentrations. Species with high SLA are often associated with high-N, light-limited environments. Thus, we expected that N-enrichment will cause 1) a shift in favor of species with higher SLA, and 2) stronger diversity declines in communities with low SLA. Additionally, we compared the predictive power of SLA to a suite of functional traits associated with lifespan, life form, clonality, and origin. In unfertilized conditions, we measured SLA of 10 individuals of 260 species comprising the top 80% cover of 35 long-term N-enrichment experiments within the United States. These experiments were located in 10 herbaceous communities, ranging from tundra to salt marshes to grasslands. Using relative abundance data from control and N-fertilization treatments we examined if SLA could predict changes in abundance or diversity with N-enrichment within and across sites.
Across all experiments, species with lower SLA were more likely to be lost with N-enrichment. For species that were not lost, SLA weakly predicted changes in abundance in response to N-enrichment. This relationship was statistically significant in only 3 of 35 experiments, and only 2 of 10 ecosystems. However, SLA predicted changes in abundance better than any one functional trait. Community-aggregated SLA was not related to the effect of N-enrichment on species richness. Our findings suggest that SLA in control conditions can predict species loss due to N-enrichment, but can only weakly predict changes in abundance of the species that persist in both environments. Additionally, our findings suggest that SLA is better than course functional groupings for predicting species abundance responses to N-enrichment. However, SLA likely does not approximate the mechanism by which species change in abundance due to N-enrichment very well. These results suggest mechanisms other than a change in light environment are more important in understanding species abundance responses to N-enrichment.