Fitness differences override variation-dependent coexistence mechanisms in California grasslands

Andrew J. Muehleisen, Institute of Ecology and Evolution, University of Oregon, Eugene, OR, Lauren Shoemaker, Department of Botany, University of Wyoming, Laramie, WY, Caitlin T. White and Katharine N. Suding, Institute of Arctic and Alpine Research, University of Colorado, Boulder, CO, E. Ashley Shaw, Biology Department, University of Oregon, Eugene, OR, Lauren M. Hallett, Environmental Studies Program and Biology Department, University of Oregon, Eugene, OR

Background/Question/Methods
Environmental variability is considered a key source of stable coexistence between species, but the extent to which it can offset fitness differences, particularly among multiple species, is still being understood. California grasslands are an ecosystem defined by environmental variability, with variation in early-season rainfall thought to drive shifts among species. Such variability presumably promotes coexistence between species in this system, but tests of coexistence in the context of environmental variability in other systems often fail to find stable coexistence. Here we assess coexistence and its interaction with rainfall variability among species in a California grassland. Using cold-frame rainout shelters with retractable roofs we manipulated rainfall at the start of the growing season (early-season drought versus an unmanipulated control, replicated eight times in a random-block design). Within each rainfall treatment we manipulated the absolute and relative abundance of six focal species selected for their functional diversity (two native annual forbs, one non-native forb, and three non-native grass species). We measured seed production and used these data to parameterize competition models between species under each rainfall condition. Using these models, we performed pairwise invasion analyses to determine the potential for coexistence among these species and partitioned their low-density growth rates (LDGR) to quantify the impact of rainfall variability on competitive outcomes.
Results/Conclusions
Very little potential for coexistence was observed among focal species; competitive exclusion was common, with competition dominated by Avena fatua in both treatments and Vulpia myuros in dry falls (both non-native grasses). Variation-independent mechanisms had a strong, negative effect on LDGR, suggesting that large intrinsic fitness differences precluded coexistence among these species. Rainfall variability had a net stabilizing effect on coexistence, driven by relative nonlinearity in per-capita competition and seed production. Contrary to historical assumptions, the temporal storage effect had little influence. Taken together, these results suggest that California grasslands are sensitive to priority effects, indicated by mutual negative LDGR among species pairs. Rainfall variability was not sufficiently stabilizing to offset fitness differences between species, but did dictate competitive outcomes for some species. However, potent competitors like A. fatua will likely dominate irrespective of rainfall conditions.