Exclusion of dominant grasses from a prairie seed mix has long-term impacts on species richness and ecosystem properties

Abstract

Establishing and maintaining high species diversity in ecological restorations has benefits for species conservation and the provisioning of ecosystem services. However, restored systems are often characterized by lower diversity than their remnant counterparts. In tallgrass prairie restorations, a few species can become increasingly dominant over time even in high-diversity plantings. Increasing dominance can result in lower overall species diversity compared to remnant prairies, with consequences for ecosystem characteristics and community dynamics. Several methods have been proposed to reduce these impacts of dominant species, including reduced seed quantity of the dominant species at the time of planting and selective removal of dominant species biomass. However, fully excluding dominant species from the seed mix has rarely been considered as a restoration approach and the long-term community and ecosystem impacts of this strategy are unknown. To assess the effects of dominant grass exclusion on the succession of tallgrass prairie restorations, we established high-diversity plantings (~85 species in the seed mix) that experimentally included or excluded two codominant C₄ grasses, Andropogon gerardii and Sorghastrum nutans, during initial seeding. Plots without seeded dominants had higher species richness than plots with seeded dominants 13 years after planting. Plots seeded without dominant grasses also maintained distinct ecosystem characteristics related to carbon and nutrient cycling, including higher light availability at the soil surface, lower root biomass, and lower soil carbon concentrations, respiration rates, and soil moisture content. These differences in community composition and ecosystem characteristics were maintained despite dominant grass dispersal (primarily by S. nutans) into plots where dominants were not planted. Dominant grass colonization of control plots changed the relative aboveground biomass of plant groups, but did not decrease the biomass of planted forbs or subdominant grasses. Rather, the introduction of S. nutans was correlated with a decline in abundance of the weedy native unplanted forb, Solidago altissima. Thus, seeding tallgrass prairie restorations without dominant grasses may produce higher diversity plantings for at least a decade after establishment, with no additional cost or management effort required.