Not just candy: A herbivore-induced defence-related plant protein in honeydew enhances natural enemy fitness

Abstract

Herbivore feeding often increases secondary metabolite production in plants. These herbivore-induced plant proteins might end up in honeydew excreted by phloem-feeding insects. This is important because honeydew is one of the most abundant and accessible carbohydrate sources for natural enemies in many agroecosystems and these proteins can thus mediate many tri-trophic interactions. Here, we hypothesized that defensive metabolites induced in the phloem by herbivory accumulate in the honeydew excreted by phloem-feeding insects and, consequently, affect the fitness of the herbivores' natural enemies that feed on it. We used a tri-trophic system consisting of citrus plants, the mealybug Planococcus citri and its primary parasitoid Anagyrus vladimiri. First, we assessed A. vladimiri fitness when fed on P. citri honeydew. We then collected honeydew of seven phloem-feeding insects, including P. citri, and analysed their protein content. Finally, we analysed the effect of superoxide dismutase (SOD), an antioxidant enzyme associated with plant defences that was commonly found in the analysed honeydews, on A. vladimiri fitness. The fitness of A. vladimiri increased when fed on honeydew compared to a sucrose-based diet, demonstrating that honeydew can contain compounds that benefit natural enemies. Proteomic analyses showed that defence-related plant proteins were present in honeydew of seven phloem-feeding insects analysed. Among these, the enzyme SOD was present in honeydew of all of them. Moreover, the levels of SOD were 10-fold higher in the phloem of plants infested by P. citri than in that of uninfested plants. SOD was also actively excreted in P. citri honeydew, and we proved that it increases the fecundity of the parasitoid A. vladimiri. We conclude that enzymatic proteins induced by herbivory in the phloem and involved in plant defence, accumulate in the honeydew excreted by phloem-feeding insects and, contrary to the current paradigm, at least some of these can have positive effects on the third trophic level.