Print PageAboveground net primary productivity (ANPP) is an important metric of ecosystem function in grasslands that is generally positively correlated with precipitation. Ecotypes of a species may develop as a result of local adaptation, such as to the precipitation cline that in part shaped the tallgrass prairie. We hypothesized that ecotypic variation exists in dominant grasses and will scale to affect ANPP in restored prairie. A reciprocal common garden restoration experiment was established in four sites (southern IL, SIL; eastern KS, EKS; central KS, CKS; western KS, WKS) across a steep precipitation gradient in order to test the effect of potential ecotypic variation in dominant C4 grasses (Andropogon gerardii and Sorgastrum nutans ) on community ANPP. Each site contained 16 plots (4 plots/block) seeded to one of 3 grass source populations (SIL, EKS and CKS) or a mixture of these sources. All plots at all sites were also seeded with the same species and seeding rates of forbs. ANPP was determined by clipping, drying and weighing four 0.1 m2 quadrats in each plot at the end of the growing season. We analyzed 5 categories of ANPP for main effects and interaction of the source and site using ANOVA.
Results/Conclusions
Total ANPP varied among sites after two years of reassembly and did not increase across the precipitation gradient as predicted. Total ANPP was highest in WKS and SIL (P<.01). In WKS forbs comprised a larger proportion of total ANPP relative to other sites (P<.001). In SIL total ANPP was comprised of more planted forbs and volunteer grasses relative to other sites (P<0.001). The EKS restorations exhibited the lowest total ANPP (P=.010) and establishment of planted forbs was intermediate of the most western and eastern restoration sites (P<.001). Planted grasses were most productive in ESK, less so in CKS and WKS, and least productive in SIL (P<.001). Ecotypic variation (a source effect) in ANPP of A. gerardii was not evident, although ANPP of S. nutans did vary among sources but not consistently across sites (P<.001). Inconsistent stochastic influences on community composition across sites from the regional and restored species pools (i.e., high establishment of weedy volunteer species in WKS and planted forb and volunteer grass species in SIL, respectively) resulted in ANPP patterns that were not predicted by the precipitation gradient and limited our ability to detect ecotypic variation in dominant grass ANPP in a restoration context that included interspecific interactions.
