Humans have introduced hundreds of exotic species to grasslands of the U.S., and their abundance and impact varies greatly among species, with a few being very abundant and many being highly rare. Exotic species can have multiple effects on ecosystems at a variety of scales. Exotic species can differ from natives on average, and mean responses across species can be used to model their effects. However, incorporating abundances into mean responses can alter these predictions. For example, if a given environment favors species with lower than average decomposition, than community shifts towards these species will lead to decreases in decomposition that can run counter to the average differences observed across species. Ecosystem models seldom include community shifts, and native vs. exotic comparisons are usually made with mean responses from monocultures without species interactions included. Here we compared the average and community feedback effects of exotic species on decomposition in grasslands. Litter bags were placed in the field for 4, 16 and 64 weeks with litter from all native vs. all exotic species mixtures with all species equally represented (no community shift treatment), or litter with species relative abundances matching field values. Mass loss was used as a measure of decomposition.
Results/Conclusions
Considering species compositional shifts completely changed the conclusions on whether exotics had higher or lower decomposition rates than natives (litter treatment x origin[Natives vs. Exotics] interaction, P < 0.001). In bags that did not include species shifts, decomposition rates were 15-26% higher in exotic species than native species. However, in bags that had realistic abundances that matched the litter actually falling in the field, decomposition rates were 12-19% higher in native than exotic species. N mineralization was also higher under native than exotic plots. In our southern Plains system, native dominated areas contain a mixture of C3 and C4 species, whereas community shifts in exotic communities leads to strong dominance by C4 grasses, which have a lower decomposition rate than C3 species. Without considering this effect, we would have falsely concluded that exotic species in this system had higher decomposition rates than natives. Our results indicate that 1) we need to better incorporate community shifts into ecosystem and invasive species models, and 2) the replacement of native grasslands by exotic species has reduced nutrient cycling and decomposition rates in the southern Plains.