The question of when community assembly is deterministic and converges toward a single endpoint and when it is historically contingent with multiple endpoints has been debated for decades. This study examined how biotic processes interact with environmental conditions to determine the assembly trajectory of plant communities. We tested the hypothesis that species-level negative feedbacks (in this case negative frequency dependence; i.e., when a species performance decreases as it becomes more common) cause community-level convergence. We also hypothesized that positive feedbacks (i.e., when performance increases as a species becomes more common) should lead to divergence in community composition. In addition, we hypothesized that soil nitrogen availability would affect these feedback processes and lead to differing assembly trajectories. We tested these hypotheses in 135 1m² experimental annual grassland communities, each containing 6 species. At three nitrogen levels, we varied the initial seeding abundance of species (i.e. their relative frequency) and followed species abundances through time.

Results/Conclusions

Although the most abundant species exhibited no frequency dependence, overall species population growth rates were more likely to be negatively associated with initial seeding frequency (i.e., negative frequency dependent; increased most when rare) in high nitrogen environments. Also in high nitrogen environments, communities became more similar in composition in comparison to communities at ambient and low nitrogen environments, where the community assembly trajectory diverged (communities became less similar to each other) over time.  We conclude that 1) there is evidence that the degree and direction of species frequency dependence may be important in determining the rate and magnitude of convergence in community assembly and 2) that nitrogen enrichment may increase the likelihood of grassland communities converging in composition, thereby decreasing regional (beta) diversity.