Might field experiments also be inadvertent metacommunities?

Metacommunity theory predicts that the composition and diversity of a site depend on its characteristics and those of its neighborhood. Dispersal between plots in a field experiment could link responses observed in a focal plot to both its treatment and those of its neighbors. However, the diversity, composition, and treatments of neighboring plots are rarely included in analyses of experimental treatments. We analyzed a spatially gridded grassland nitrogen addition experiment and found that plant species richness and the composition of focal plots were influenced not just by their nitrogen treatment but also by the number of species in neighboring plots and their abundances. For each additional species in a focal plot's neighborhood, the species richness of the focal plot increased by 0.30 species. Control plots had a significant loss of species, at a rate of similar to 0.23 species per year during the 23-year experiment, but only when their neighborhoods had low species richness. Changes in the abundance of the three dominant species depended both on the nitrogen treatment of a focal plot and on their abundance in adjacent plots. Our analyses suggested that both the experimental nitrogen treatments and metacommunity processes codetermined plant species richness and plant species' abundances. Our findings suggested that analyzing many traditional field experiments with a metacommunity perspective may reveal a confounding of experimental treatments and provide empirical data to test metacommunity theory.