Processes determining plant species diversity in restored tallgrass prairie
A fundamental goal of restoration ecology is to restore biological diversity in degraded or fragmented environments. In practice, restorations often have lower diversity than reference communities, highlighting a need to identify theoretical and practical barriers to the restoration of native diversity. North American tallgrass prairie is an ideal system in which to study the restoration of plant diversity. Because remnant prairies are rare, prairie restoration projects are now very common, and the size and longevity of grassland plants makes them easy to measure and manipulate. Here, we present the first synthetic study to test the relative importance of soil characteristics, management actions, seed mix design, and site characteristics for predicting prairie restoration success. We found that across many restorations, invasion by exotic plants was the best predictor of outcomes, significantly reducing beta-, and site-level plant diversity. We also found that seeding more species reduced exotic species to increase diversity. Mowing also tended to increase diversity. We examined how plant diversity, richness, grass-forb ratios, and abundance of milkweeds (an important forb assemblage) differ between restored and remnant prairies and found that remnants are more diverse, less grassy, and have considerably higher milkweed abundances. Within restored prairies, the common milkweed was the most abundant species, and abundances of all milkweeds were correlated with higher soil pH, and variables associated with disturbance, including lower soil density and habitats with more edge. In a new prairie restoration in north-central Iowa, we tested whether the use of cover crops or additions of target prairie seeds were effective methods for increasing plant species diversity. We predicted that cover crop treatments seeded the fall before a diverse prairie mix would suppress weeds, and indirectly facilitate better recruitment of prairie species. Instead, we found that cover crops tended to reduce weeds slightly, but they also reduced or had no effect on prairie establishment. In a seed addition study, we added diverse seed mixes in the 1st, 2nd, or 3rd year or annually after initial seeding. We found that these addition treatments did not increase total species richness or diversity, indicating low seed limitation on diversity. Together, the results of this work add to the mounting evidence that restorations tend to be compositionally different, and less diverse than remnants, that exotic species play an important role in community assembly during grassland restoration, and that assembly processes that occur early in a restoration are the most important and can have long-lasting effects.