Mutualism is ubiquitous in nature, and nursery pollination mutualisms provide a system well suited to quantifying the benefits and costs of symbiotic interactions. In nursery pollination mutualisms, pollinators reproduce within the inflorescence they pollinate, with benefits and costs being measured in the numbers of pollinator offspring and seeds produced. This type of mutualism is also typically exploited by seed-consuming nonpollinators that obtain resources from plants without providing pollination services. Theory predicts that the rate at which pollen-bearing “foundresses” visit a plant will strongly affect the plant's production of pollinator offspring, non-pollinator offspring, and seeds. Spatially aggregated plants are predicted to have high rates of foundress visitation, increasing pollinator and seed production, and decreasing non-pollinator production; very high foundress visitation may also decrease seed production indirectly through the production of pollinators. Working with a nursery mutualism comprised of the Sonoran Desert rock fig, Ficus petiolaris, and host-specific pollinating and nonpollinating fig wasps, we use linear models to evaluate four hypotheses linking species interactions to benefits and costs: 1) foundress density increases with host-tree connectivity, 2) pollinator production increases with foundress density, and 3) nonpollinator production and 4) seed production decrease with pollinator production. We also directly test how tree connectivity affects non-pollinator production. We find strong support for our four hypotheses, and we conclude that tree connectivity is a key driver of foundress visitation, thereby strongly affecting spatial distributions in the F. petiolaris community. We also find that foundress visitation decreases at the northernmost edge of the F. petiolaris range. Finally, we find species-specific effects of tree connectivity on non-pollinators to be strongly correlated with previously estimated non-pollinator dispersal abilities. We conclude that plant connectivity is highly important for predicting plant-pollinator-exploiter dynamics, and discuss the implications of our results for species coexistence and adaptation.