Effects of biological invasions on community diversity vary with spatial scale, but temporal changes in native and exotic diversity over a range of scales are rarely documented. Understanding the cause of localized native species declines is essential to predicting effects of invaders and successfully restoring natives. Abiotic environmental heterogeneity can promote coexistence if different species maximize their growth and reproduction under different abiotic conditions. Biotic interactions can alter the effects of abiotic conditions, however, with the potential to promote either coexistence or exclusion from the community. Thus, explicitly testing the interaction between the abiotic and biotic environment on native species performance is necessary to understand which factors drive coexistence with or exclusion from the invaded community. Long-term monitoring has shown declines in certain native annual forbs over the past 40 years at one California grasslands site, while abundance of certain exotic grasses has increased. These grasses may competitively exclude forbs via direct competition between seedlings or indirect competition through build-up of litter. We established a two-way factorial competition removal experiment to examine the effects of these two factors on the grassland community and six seeded forb species with documented declines.

Results/Conclusions

Herbicide application significantly reduced total grass and increased total forb abundance early in the growing season; litter removal did so to a lesser extent. Combining treatments did not produce a synergistic effect on either grass or forb abundance, however. Seeded species responses were highly idiosyncratic among species, treatments, and spatial location. Generally, some form of competition removal increased abundance of seeded species that also had high germination rates in laboratory trials. Species not previously observed in some locations were able to germinate and survive, suggesting that recruitment limitation is one factor driving these species' distributions across this reserve. High levels of spatial variation in community properties, such as total litter, grass, and forb abundance, were observed overall, such that treatment effects were only observed after accounting for this variability. Thus, effects of management actions involving competition removal will strongly vary and limited resources should either be focused on locations where they will have the greatest impact or, when this is unknown, spread over highly variable sites.