Environmental effects on plant and animal populations can be mediated via both direct effects on individuals, such as water stress, and/or indirect effects, such as a reduction in the densities of competing species. Quantifying the strength of direct vs. indirect effects is critical to accurately predicting responses to a variable climate. If direct effects are strong, responses in one community will accurately predict responses in another community with different interacting species. If indirect effects are strong, effects in novel communities cannot be accurately predicted because the net effect depends strongly on the community of interacting species. In spite of their importance, a unifying framework for understanding the strengths of direct and indirect effects is lacking for primary producers, particularly for species interactions besides competition, or when multiple indirect effects operate simultaneously. This research will quantify the strengths of direct and indirect effects mediated through three species interactions (competition, and two herbivore guilds) for more than 120 plant species. In addition, this project will fund ~20 undergraduates to engage in a structured mentoring and training program designed to improve quantitative skills, provide training in modeling to up to 75 graduate students, and increase scientific literacy for 200 Kansas residents annually. The project addresses three questions that are central to ecological predictions: (1) What fracti