The mechanisms by which plant diversity influences consumers are largely unexplored. Plant diversity reduces among-plant competition, and as a result, may have cascading bottom-up effects through altered resource quantity and quality. Less understood are parallel effects of resource heterogeneity on consumer foraging behaviors, and the consequences of such effects for trophic interactions and feedbacks on plant performance. Here, we asked whether genotypic diversity in the herb Ruellia nudiflora influences seed predator (SP) and parasitoid functional responses, and if such effects in turn influence selection on the plant. We established plots with plants of one or five genetic families (pool ¼ 14) and measured fruit, seed predator, and parasitoid abundance. By eliminating direct plantplant interactions, this experiment explicitly tested for plant diversity effects occurring through altered consumer behaviors (vs. effects through changes in plant quantity or quality). We compared observed plant fitness (under three trophic levels) to projected fitness in the absence of parasitoids (two trophic levels) and in the absence of seed predation (one trophic level) by computing the number of seeds consumed by the SP and the number of seeds rescued from consumption by parasitoids. We then compared the strength and mode of selection on fruit number between levels of diversity, separately under each trophic scenario. Plant diversity did not influence fruit, seed predator, or parasitoid abundance, but did alter plant-SP interactions. SP recruitment resembled a Type II functional response but saturated weakly in polycultures, with fewer SPs at intermediate fruit abundance and more SPs at high fruit abundance relative to monoculture. Parasitoid recruitment was weakly positively density dependent and unaffected by diversity. Importantly, we found that under the bi-trophic scenario, the effect of diversity on SP recruitment altered the mode of selection on fruit number, from directional selection at low diversity to nonlinear (stabilizing) selection at high diversity. In contrast, diversity did not alter selection under mono-or tri-trophic scenarios. Therefore, diversity effects on SP functional responses fed back to alter selection on fruit number, but parasitoids eliminated this linkage by weakening herbivore selection. Collectively, these findings provide novel evidence for the mechanistic basis of eco-evolutionary feedbacks between plant diversity and consumers.