Print PageTallgrass prairie is restored across a variable climate using seed mixes that vary in both propagule source and diversity. Restoration practitioners may influence the structure of a restored prairie community by manipulating the biotic filter through selection of seed source (i.e., cultivar or non-cultivar) and diversity of seed mixes (i.e., richness and ratio of grasses to forbs introduced). Dominant grass cultivars are selected for agronomic traits, whereas it is generally assumed that locally collected sources are locally adapted. In addition to propagule source, diversity of seed mixes may directly affect the composition of planted species, and indirectly affect the establishment of volunteer species from the regional species pool. We measured plant community structure during the first 5 years of prairie restoration in plots seeded with either cultivars or non-cultivars of dominant grasses that contained subplots seeded with 5 different levels of diversity (subplot treatment). This field experiment was replicated across a precipitation gradient in Kansas and Illinois. Species composition (percent cover of all species in permanent quadrats within each subplot) data were analyzed according to a split plot design with repeated measures using the mixed model procedure.
Results/Conclusions
Seed source affected key community metrics in Kansas, while these community parameters were unaffected by source in Illinois. In Kansas, after the first growing season, native species richness was greater over all diversity treatments in cultivar plots (P=0.03), whereas native richness was similar in both sources in Illinois (P=0.99). In Kansas, after the first growing season, dominant grass cover was greater in cultivar plots across all seeded diversity treatments (P=0.01). In Illinois, dominant grass cover similarly increased in cultivar and non-cultivar plots (P=0.89). In Kansas cover of the focal forbs was greater in non-cultivar plots by the fifth growing season (P<0.01), whereas cover of the planted forbs was unaffected by grass source in Illinois (P=0.29). In both sites, native cover was lower and volunteer cover was higher in subplots seeded with the lowest diversity (P<0.03). Dominant grass cover was inversely related to subplot seeded diversity only in Illinois (P<0.01). These results demonstrate that the abiotic filter modulates the expression of intraspecific differences between cultivars and non-cultivars. In Kansas, the drier climate may enable cultivars to exploit their agronomic advantages. Manipulation of community structure though seed mix composition can facilitate achieving restoration goals of high native cover and minimal weed cover.
