Models are central to the role of science as a tool to make sense of the natural world. In ecology, coexistence theory fundamentally relies on population models to understand diversity maintenance in biological communities. Models of coexistence historically simplify the biology of diverse systems into equations of pairwise species competition, an abstraction that is often imported into empirical tests of coexistence theory. As a result, empirical coexistence studies often overlook the role of trophic interactions, mutualisms, spatial heterogeneity, and facilitation in their formulations of what is important for diversity maintenance. My dissertation asks if and how overlooking such dynamics matters for understanding plant interactions and thus coexistence. I determine the roles of pollinators and spatial heterogeneity on plant performance in annual plant communities using (1) field tests in a system of sympatric flowering plants in the genus Clarkia (Onagraceae) and (2) simulation models. In Chapter 1, I ask how shared pollinators affect Clarkia interactions using a combination of interaction plots, pollen supplementation, and pollinator behavior experiments. I find the effect of pollinator sharing on co-occurring Clarkia is to amplify the effect of plant interactions over vegetative growth- i.e. sharing pollinators intensifies both facilitative and competitive interactions between Clarkia, implying thatsharing pollinators destabilizes plant coexistence. In Chapter 2, I use a reciprocal transplant experiment to ask how spatial variation in hydrology affects Clarkia performance. I find that hydrological variation corresponds to Clarkia spatial partitioning by driving spatially variable germination rates, resulting in highest population growth rates in the home patches in three of four species. This evidence is the first step in determining if Clarkia coexist via the spatial storage effect, an oft- overlooked coexistence mechanism. In Chapter 3, I use simulation models to ask if plant species facilitate each other by better supporting pollinator populations in two- species communities compared to single-species communities. I find that such facilitation via pollinator support can occur in two-species communities that exhibit low-germination rates and species-specific responses to the environment. My findings indicate that facilitation can be incredibly important for understanding plant diversity maintenance.