Dealing with mutualists and antagonists: Specificity of plant-mediated interactions between herbivores and flower visitors, and consequences for plant fitness

Plants need to deal with antagonists, such as herbivores, while maintaining interactions with mutualists, such as pollinators that help plants to maximize their reproductive output. Although many plant species have inducible defences to save metabolic costs of defence in the absence of herbivores, plant responses induced by herbivore attack can have ecological costs. For example, herbivore-induced responses can affect flower traits and alter interactions with flower visitors. Such plant-mediated interactions between herbivores and flower visitors can affect plant reproductive output. Current knowledge on the generality and specificity of plant-mediated herbivore-flower-visitor interactions and its consequences for plant fitness is limited. In this study, we investigated whether a broad range of herbivores feeding on the annual plant Brassica nigra affect interactions with flower visitors, whether the direction of interactions is predicted by the feeding modes (chewing and sap-feeding) and sites (above- and belowground) of the herbivores, and whether it results in fitness consequences for the plant. Our results show that attack of B. nigra by a range of different herbivores influenced plant interactions with mutualist pollinators and an antagonist florivore, the pollen beetle Meligethes aeneus. Pollinator community composition was affected by herbivory, whereas overall pollinator attraction was maintained. Pollinator community composition of uninfested plants differed from that of chewing and root herbivore-infested plants. Main responders in the pollinator communities to changes induced by herbivory were syrphid flies, bumblebees, and solitary bees. Although the preference of pollen beetle adults was not affected by herbivory, beetle larvae performed best on plants infested with the nematode Heterodera schachtii. The changes in pollinator community composition and syrphid fly visitation can explain the observed increase in seed set of root herbivore-infested plants. Interactions of flowering B. nigra plants with mutualist and antagonist insects are well integrated and conflicting interactions do not reduce reproductive output. Our results suggest some degree of specificity in herbivore-flower-visitor interactions with consequences for plant fitness. Specificity of plant responses were determined at the species level as well as the herbivore functional group level, and differed depending on the flower visitor. Because plant reproduction was affected by indirect plant-mediated interactions, these can potentially result in selection on plant strategies to optimize growth, defence and reproduction.