Melilotus officinalis (yellow sweetclover) causes large changes in community and ecosystem processes in both the presence and absence of a cover crop

Non-native species are hypothesized to decrease native species establishment and cover crops are hypothesized to decrease non-native species abundance. Although many studies have compared invaded to non-invaded habitats, relatively few studies have experimentally added non-native species to directly examine their effects. In a greenhouse mesocosm experiment, we tested the effects of non-native forbs (Melilotus officinalis, Verbascum thapsus, and Lespedeza cuneata), a proposed C-3 grass cover crop (Pascopyrum smithii), and a commonly seeded non-native C-3 grass (Bromus inermis) on the establishment of target native C-4 prairie grass species. All treatments contained the same seed density of target C-4 species and were begun on bare soil collected from the field. The legume M. officinalis strongly decreased the abundance of all other species, species diversity, and light and soil moisture levels. Surprisingly, M. officinalis took up relatively large amounts of labeled nitrogen (N-15) from the soil early in its development, but M. officinalis fixed nitrogen, thus increasing nitrogen in biomass nearly fivefold by the end of the study. We found few effects of either C-3 grass species on non-native forbs or C-4 target species, but seeded P. smithii did increase species diversity. Non-native plants therefore impeded native C-4 grass establishment through long-lasting effects of target species seedbank depletion (death of most target seedlings) and altered nutrient availability. The effects of M. officinalis were not reduced by the presence of a cover crop.