When do propagules matter? The role of ecological filters and regeneration dynamics during community assembly in tallgrass prairie restorations

Ecological restoration aims to augment and steer the composition and contribution of propagules for community regeneration in degraded environments. Three studies were conducted to elucidate the role of regeneration dynamics and dominant species on community assembly during tallgrass prairie restoration. In the first study, patterns in the abundance, richness, and diversity of seed and bud banks were quantified across an 11-year chronosequence of restored prairies and in prairie remnants to elucidate the degree to which the germinable seed bank, emerged seedlings, belowground buds, and emerged ramets were related to community regeneration. There were no directional patterns in the abundance, richness, or diversity of the germinable seed bank across the chronosequence. Emerged seedling abundance of sown species decreased during restoration, whereas richness and diversity of all emerged seedlings and non-sown emerged seedling species decreased across the chronosequence. Conversely, abundance and richness of belowground buds increased with restoration age and belowground bud diversity of sown species increased across the chronosequence. Numbers of emerged ramets also increased across the chronosequence and was driven primarily by the number of graminoid ramets. There were no temporal changes in abundance and richness of sown and non-sown emerged ramets, but diversity of sown emerged ramets increased across the chronosequence. This study demonstrates that after initial seeding, plant community structure in restored prairies increasingly reflects the composition of the bud bank. In the second study, abundance and richness of ramets, emerged seedlings, seed rain, and the soil seed bank were measured in a restoration experiment consisting of a split plot design with population source of dominant grasses (cultivar vs. local ecotype) and sown subordinate species (three unique pools of non-dominant species) as the subplot factor, respectively. Different sown species pools were included to assess whether any observed differences in propagule abundance or richness between the dominant species sources was generalizable across varying interspecific interactions. Abundance of emerged ramets was similar between communities sown with cultivar and local ecotypes of the dominant grasses but differed among sown species pools in prairie restored with cultivars but not local ecotypes. Number of emerged seedlings also differed among species pools, but only in communities sown with local ecotypes of the dominant grasses. There was also higher seedling emergence in communities sown with local ecotypes relative to cultivars of the dominant grasses in one species pool. Richness of the seed rain was influenced by an interaction between dominant grass population source and sown species pool, resulting from (1) higher richness in prairie restored with local ecotypes than cultivars of the native grasses in one species pool and (2) differences in richness among species pools that occurred only in prairie restored with the local ecotype grass source. Abundance and richness of the seed bank was not affected dominant grass population source. This study addressed a poorly understood potential effect of using cultivars in ecological restoration, specifically on the abundance and supply of propagules for community assembly. These results suggest that if both local ecotype and cultivar sources are available for restoration, using local ecotypes could result in more seedling germination and richness in the seed rain. One of the central concepts of ecology is to understand the processes that influence species diversity, and how the resulting diversity affects ecosystem functioning. Diversity has been hypothesized to be responsible for long-term community stability, contrasted by the idea that dominant species regulate temporal stability (mass ratio hypothesis). In the third study, community metrics (total plant cover, forb cover, C4 grass cover, richness, and diversity) were measured in a restoration experiment consisting of a split plot design with sown dominant grasses (Andropogon gerardii, Schizachyrium scoparium, and Sorghastrum nutans) and subordinate species (three unique pools of non-dominant species) as the subplot factor, with treatment (control vs. suppression of dominant grasses) as the sub-subplot factor, respectively. Dominant grass suppression had little effect on forb cover, richness, and diversity, but influenced total and C4 grass cover. Propagule addition increased community richness and diversity in year of sowing and year after sowing, but contributed little to total cover. Dominant grass suppression had an effect on new species recruitment in one of two species pools, with suppression of all dominant grasses having the greatest influence on total cover and richness of new species. These results suggest that dominant species collectively are responsible for modulating stable species composition during community assembly and can act as a biotic filter to the recruitment of new species, but diverse subordinate species assemblages are more important for temporal stability