Individual plants interact through a variety of mechanisms creating plant-soil feedbacks, in which a plant affects the environment, and this change feeds back to influence the performance of that plant and other members of the community. Feedbacks can have consequences at the community and ecosystem levels; however, despite the large body of work on component processes, how measured interactions among individuals actually affect large-scale patterns of species composition, diversity, and invasion remains largely untested. In this dissertation, I use a combination of 1) spatiotemporal surveys and modeling and 2) measurements of interactions in field experiments, to test the mechanisms through which plants interact and the importance of these interactions in driving community structure and dynamics in two systems, temperate wetlands invaded by hybrid cattail and native dry grasslands. Positive feedbacks are predicted to be important for explaining dominance of invasive species, because modification of the environment to their own benefit would further their invasion. Negative feedbacks are predicted to dominate in native systems where they lead to limitation of conspecific growth, promoting coexistence. In the invaded system, experiments suggest that hybrid cattail (Typha x glauca) produces positive feedbacks: it increases nitrogen cycling twofold and decreases light through high litter production, an environment in which cattail performs well but native species decline. These positive feedbacks could contribute to the pattern found in field surveys that T. x glauca was associated with locally high soil nutrients, low light, and large amounts of litter, and that native diversity was highest in areas of shallow litter depth. In the native grassland system, both transplant experiments and fitting models to survey data suggest that negative feedbacks are common: conspecifics inhibit the individual and population growth of each of the dominant species more than heterospecifics. The intermediaries in these negative feedbacks include soil nitrate and light reduction, however other unmeasured soil properties, such as pathogens or mycorrhizae, also likely play a role. Overall, this suggests that the balance of interactions may shift from negative feedbacks in native systems to positive in invasive systems, which contributes to the coexistence among natives and dominance of invasives.