Limited dispersal, priority effects, and nutrient availability determine alternative community states in an old-field grassland succession

Background/Question/Methods

Predicting community dynamics after disturbance is central to both theoretical and applied ecology. Modern successional systems are often dominated by exotic plant species, while recovery of native communities may depend on external subsidies. Under this scenario, differential propagule availability, identity of early colonizers, and soil resources may all interact in affecting successional trajectories toward alternative community states. We assessed the effects of early and late successional exotic species, and nutrient availability on native grass species establishment in de novo assembled communities, in the Inland Pampa of Argentina. We established synthetic plant communities by sowing seed cocktails in 2.5 m² plots (n = 6 blocks), where we manipulated the initial floristic composition (10 different community ensembles) and soil N availability (0 and 20 gr N m^-2 yr^-1). Initial species sowing was performed following a modified target-neighbor additive design, with native perennial grasses as the focal group. Each plot was sowed with the focal group plus 3-9 species from a 'neighbor' functional group typical of early (annual/biennial forbs), mid (annual/biennial grasses) or late (exotic perennial grasses) successional fields. During three years, we monitored changes in abundance of plant functional groups in each community.

Results/Conclusions

Throughout the experiment, early emergence of native perennial grasses was substantially lower than that of forbs, annual, or exotic perennial grasses. Furthermore, local abundance of native grasses was strongly determined by the identity of the functional group with which they were sown. When sown alone on bare soil, perennial native grasses reached over 75% cover after two years of succession. Presence of ruderal forbs and annual grasses delayed the establishment of native grasses and decreased their final cover to 55% and 18%, respectively. The presence of exotic perennial grasses in the initial mix inhibited the establishment of the target native grass group, creating an exotic grass-dominated community. Control plots (not sown) also developed into an exotic perennial grass community. Nitrogen addition favored early successional plant groups; performance of native grasses was diminished in all communities under nutrient enrichment. We show that seed dispersal is the primary limitation to re-establishing native perennial grasses in this system. However, competition with exotic grasses further constrains native grass recovery when dispersal limitation is alleviated by seed addition. The strength of priority effects precluding native grass recovery varied among functional groups, being weak for ruderal forbs, intermediate for annual grasses, and strongest for exotic perennial grasses. Indeed, exotic perennial grasses was a strong community attractor when present in the initial seed mix. Our results provide important insights for devising management strategies aimed at restoring native grass communities in pampean agro-ecosystems.