The local is best strategy for restoration assumes that plant populations are commonly adapted to local site conditions. Yet, that assumption may be increasingly violated in the face of climate change. Here we use open top chambers to simulate global warming on 12 restored prairies sown with local genotypes, genotypes from the upper Midwest, or southern genotypes of four common prairie species (Coreopsis lanceolata, Echinacea purpurea, Chamaecrista fasciculata, and Rudbeckia hirta) to investigate whether warming alters the seed sources most likely to lead to successful establishment in restored prairies.
Results/Conclusions
We detected substantial genetic variation among the three seed sources for all four focal study specie in seedling growth rate (all P < 0.045) and for flowering phenology (F = 42.17, P < 0.0001) for C. fasciculata, the one annual species included in our study. Warming altered plant traits and also altered which population of C. fasciculata and E. purpurea performed best in these restored prairies, although the magnitude of the warming effect varied across sites (warming x source x site interaction for C. fasciculata F = 3.86, P = 0.012; E. purpurea: F = 2.72, P = 0.03). For example, although the magnitude of warming effect varied across sites, on average southern genotypes of E. purpurea substantially outperformed other seed sources under warming, but the three seed sources exhibited similar growth under ambient temperature conditions. These findings suggest that climate change will shift the most effective source populations for use in restoration although these effects may be difficult to predict given substantial variation across sites.