Community invasibility, recruitment limitation, and grassland biodiversity

Plant species composition, species abundances, and species richness were strongly recruitment limited in a 4-yr experiment in which seeds of up to 54 species were added to patches of native grassland. Four field seasons after a one-time addition of seed, many added species were still present and reproducing, with plots seeded at the highest rate having species richness that was 83% greater and total plant cover that was 31% greater than controls. Total plant community cover increased significantly with the number of species added as seed, but total cover of pre-existing species was independent of the number of species added as seed, suggesting that the new species mainly filled previously ''empty'' sites. The proportion of added species that became established was negatively correlated with initial species richness of plots, suggesting that species-rich sites were more resistant to invasion. Plot invasibility also depended on the abundances and species richness of plant functional groups in the plots, but was independent of seed size and of total plant cover. The major functional groups of plants differed in their abilities to invade as seed, with perennial grasses being the poorest invaders and herbaceous legumes being the best. Thus, local biotic interactions and recruitment dynamics jointly determined diversity, species composition, and species abundances in these native grassland communities. This supports a metapopulation-like perspective over a purely interspecific-interaction perspective or a purely regional perspective, suggesting that recruitment limitation may be more important, even on a local scale, than often recognized.