Jasmonate-mediated plant defenses affect the strength of fear-induced trophic cascades

Background/Question/Methods

Density-mediated trophic cascades whereby predators suppress herbivores and thus indirectly benefit plants have long been reported in the literature. Yet recent studies suggest that predators can also reduce plant damage by eliciting anti-predator behavioral responses in their prey (i.e., trait-mediated or fear-induced trophic cascades). However, we lack a mechanistic understanding of the ecological factors that mediate the strength of such non-consumptive effects in tri-trophic communities. We hypothesize that plant defenses are likely to underlie much of the variation in nonlethal top-down effects of natural enemies on phytophagous insects and their host plants. To test this prediction, we experimentally manipulated both plant chemical defenses and predation risk. Plant defense was altered by using three genetic lines of tomato to both increase and decrease plant quality: wild-type (with inducible resistance intact), jasmonate-insensitive (non-inducible, highest quality), and jasmonate over-expressers (constitutively-induced, lowest quality). The jasmonate response is a biochemical pathway induced by herbivory that confers resistance to a wide range of herbivores, including leaf-chewing insects. Predation risk was manipulated with stinkbugs (Podisus maculiventris) whose stylets were surgically shortened to create ‘risk predators’ that can hunt and scare prey but can not kill. Therefore, the only effect of risk predators on herbivores could be a change in their traits, such as behavior. The interactive effects of plant defense (using mutant tomato lines) and predation risk (using predators with modified mouthparts) were assessed on a diverse assemblage of leaf-chewing insects associated with tomato plants (hornworms – Manduca sexta, armyworms – Spodoptera exigua, and Colorado potato beetle – Leptinotarsa decemlineata). These tri-trophic effects were examined in short-term behavioral assays, as well as in large-scale field enclosures.
Results/Conclusions

Our experimental results were largely consistent across spatial/temporal scales and across herbivore species. The presence of predaceous stinkbugs elicited a ~50% decrease in overall levels of insect defoliation on jasmonate-insensitive (i.e., low defense) plants, but herbivores did not alter their feeding behavior on jasmonate over-expressing (i.e., high defense) tomato lines.  Importantly, the contribution of the consumptive and non-consumptive components varied according to plant type.  On the wild-type plants the consumptive and non-consumptive components contributed approximately equally. On the high quality plants, the non-consumptive component was much stronger than the consumptive component. Thus, an elevation in plant chemical defense dampens the magnitude of fear-based trophic cascades.