Community assembly theory predicts that local community composition will vary depending on its assembly history.  We tested this prediction in a long-term split-plot restoration experiment at two sites.  The experiment involved establishing treatments that varied species arrival order, over-seeding all plots with a common seed mix of 30 tall-grass prairie species, and then sampling the resulting communities to test how treatments affected plant community assembly.  We manipulated the identity of early-arriving tallgrass prairie species (C₃ annual Chamaecrista fasciculata, C₃ perennial Elymus canadensis, C₃ biennial Rudbeckia hirta, C₄ perennial Bouteloua curtipendula, a mixture of all four species, or a control) in 5 x 5 m main plots (n = 60).  Species were selected because they establish early compared to other members of their functional groups.  We applied the following timing and priority effect treatments to 2 x 2 m subplots within each main plot (n = 240): 1) early-spring seeding of both the early-emerging species and the seed mix, 2) early-spring seeding of the early-emerging species with the seed mix added the following year, 3) late-summer seeding of both the early-emerging species and the seed mix, and 4) late-summer seeding of the early-emerging species with the seed mix added the following year.  Species composition, diversity, and exotic-native proportions were sampled over three growing seasons and analyzed with repeated measures ANOVA.  Fire temperatures were estimated in a spring burn in year three.

Results/Conclusions

Early-emerging species did not affect diversity, but E. canadensis had the highest biomass from the prairie mix and lowest proportion of exotics.  Adding the mix in early spring led to diverse native communities, but adding the mix in the late summer, or adding prairie mix the following year led to low diversity communities dominated by exotics.   Relative abundance of species from the prairie mix was 4-times higher and species diversity was 16% higher when seed was added with the early-emerging species in early-spring compared to other treatments.  This spring treatment was native-dominated (68% native), whereas other treatments were exotic-dominated (78-84% exotic).  These species composition differences affected fuel mass and fire temperatures, key ecosystem processes in tall-grass prairie.  Our results support the prediction that communities will reach multiple stable states under similar abiotic conditions under this time frame.  These states were strongly affected by timing of arrival and priority effects.