Identifying suitable genetic stock for restoration often employs a ‘best guess’ approach. Without adaptive variation studies, restoration may be misguided. We test the extent to which climate in central US grasslands exerts selection pressure on a foundation grass big bluestem (Andropogon gerardii), widely used in restorations, and resulting in local adaptation. We seeded three regional ecotypes of A. gerardii in reciprocal transplant garden communities across 1150 km precipitation gradient. We measured ecological responses over several timescales (instantaneous gas exchange, medium-term chlorophyll absorbance, and long-term responses of establishment and cover) in response to climate and biotic factors and tested if ecotypes could expand range. The ecotype from the driest region exhibited greatest cover under low rainfall, suggesting local adaptation under abiotic stress. Unexpectedly, no evidence for cover differences between ecotypes exists at mesic sites where establishment and cover of all ecotypes were low, perhaps due to strong biotic pressures. Expression of adaptive differences is strongly environment specific. Given observed adaptive variation, the most conservative restoration strategy would be to plant the local ecotype, especially in drier locations. With superior performance of the most xeric ecotype under dry conditions and predicted drought, this ecotype may migrate eastward, naturally or with assistance in restorations.

Water availability has driven adaptive variation in leaf structure in populations of A. gerardii, particularly between sand bluestem and big bluestem. Genetically based differences in leaves of A. gerardii indicate adaptive variation and evolutionary forces differentiating sand bluestem from big bluestem. Environmental responses of A. gerardii leaves suggest an ability to adjust to drought, even in populations adapted to mesic home environments.

The phenotype of two Andropogon gerardii subspecies, big bluestem and sand bluestem, varies broadly throughout the Great Plains of North America, giving rise to ecotypes within the species. This study sought to discriminate between genetic and environmental variation of big bluestem and sand bluestem by examining gas exchange and leaf anatomy in common gardens across a climatic gradient of the Great Plains. Thirteen populations of big bluestem and one population of sand bluestem, constituting five ecotypes, were planted in community plots and a single plant plots in a common garden at each of four sites ranging from western Kansas to southern Illinois. Photosynthesis, stomatal conductance, intercellular CO2, transpiration, and intrinsic water use efficiency were measured three times in the 2010 growing season. In addition, leaf thickness, midrib thickness, bulliform cells, interveinal distance, and vein size were assessed by light microscopy. Abundant phenotypic variation exists among ecotypes within community plots. At all planting sites, big bluestem ecotypes from xeric environments had higher photosynthesis, stomatal conductance, and transpiration compared to mesic ecotypes. Single plant plots also had abundant phenotypic variation; ecotypes native to xeric environments also had higher photosynthesis, stomatal conductance, and transpiration, but differences were more distinct. In addition, sand bluestem, which was only planted in single plant plots, had similar photosynthesis, stomatal conductance, and transpiration to the big bluestem ecotype native to the most xeric environment. Sand bluestem also had higher water use efficiency and lower intercellular CO2 than any big bluestem ecotype. Leaf anatomy assessments indicated xeric ecotypes of A. gerardii had thicker leaves and fewer bulliform cells. Environmental variation was as important as genetic variation for gas exchange and leaf anatomy in both community and single plant plots. Compared to xeric sites, mesic sites had higher photosynthesis, stomatal conductance, and water use efficiency and lower intercellular CO2 and transpiration in community and single plant plots. Leaves from mesic sites also had thicker midribs, larger veins, and a greater proportion of bulliform cells. Ecotypes of A. gerardii across the Great Plains are adapted to water availability. Drought-adapted ecotypes of A. gerardii were shorter in stature and had smaller, thicker, narrower leaves, which reduced the evaporative surface area of these plants. Evidently, A. gerardii controls water loss by reducing evaporative surface area more than it does by increasing the proportion of bulliform cells. This allows drought-adapted ecotypes to have higher photosynthetic rates, stomatal conductances, and transpiration rates in both mesic and xeric environments compared to ecotypes native to mesic environments. This study brings to light potential responses of big bluestem ecotypes to climate change. This study also indicates the phenotypic variation among big bluestem could prove useful in the restoration of native prairies.