Background/Question/Methods There is a long running debate regarding the appropriateness of using grass cultivars that have been selected for traits or wild collected seed to restoring or augment native grasslands. A field study was established in March 2006 to test whether differences within multiple dominant species widely used in community re-assembly acts as a filter on community assembly and scales to affect ecosystem function in restored tallgrass prairie. This experimental design contained replicated experimental community assemblages that differ in source (cultivar vs. wild collected) of the dominant grasses. We used this experiment to determine whether population source of the dominant grasses affects the genetic structure of subordinate species. Inter-simple sequence repeats (ISSR) DNA markers were used to characterize differences in population genetic structure of one dominant  grass (Sorghastrum nutans) and two subordinate forbs (Chamaecrista fasciculata, Silphium integrifolium) growing in the matrix of wild collected or cultivars of the dominant prairie grasses.
Results/Conclusions Multiple response permutation procedure (MRPP) analysis indicated differences between Chamaecrista fasciculata (T=-3.13, A=0.05, P =0.01) and Silphium integrifolium (T=-2.55, A=0.01, P =0.01) growing with cultivar grasses versus with wild collected grasses. Principle components analysis (PCA) using frequency data by replicated plots demonstrated C. fasciculata and S. integrifolium plants grown in field plots established with wild collected grasses were more genetically similar to other plants grown in the same source population of grasses than those grown with cultivars of the native grasses. The field plots were established with similar population sources of subordinate species and varied only in terms of dominant grass population source (cultivar vs. wild collected).  Therefore difference in subordinate species genetic signature was a consequence of the dominant grass population source.