The effects of plant genotype and environmental factors on tri-trophic interactions have usually been investigated separately, limiting our ability to compare the relative strength of these effects as well as their potential to interactively shape arthropod communities. We studied the interactions among the herb Ruellia nudiflora, a seed predator, and its parasitoids using 14 maternal plant families grown in a common garden. By fertilizing half of the plants of each family and subsequently recording fruit number, seed predator number, and parasitoid number per plant, we sought to compare the strength of plant genetic effects with those of soil fertility, and determine if these factors interactively shape tri-trophic interactions. Furthermore, we evaluated if these bottom–up factors influenced higher trophic levels through changes in abundance across trophic levels (density-mediated) or changes in the function of species interactions (trait-mediated). Plant genetic effects on seed predators and parasitoids were stronger than fertilization effects. Moreover, we did not find plant genetic variation for fertilization effects on fruit, seed predator, or parasitoid abundance, showing that each factor acted independently on plant resources and higher trophic levels. Both bottom–up forces were transmitted via density-mediated effects where increased fruit number from fertilization and plant genetic effects increased seed predator and parasitoid abundance; however, seed predator attack was density-dependent, while parasitoid attack was density-independent. Importantly, there was evidence (marginally significant in one case) that fertilization modified the function of plant-seed predator and seed predator–parasitoid interactions by increasing the number of seed predators per fruit and decreasing the number of parasitoids per seed predator, respectively. These findings show that plant genetic and soil fertility effects cascaded up this simple food chain, that plant genetic effects were stronger across all trophic levels, and that these effects were transmitted independently and through contrasting mechanisms.