An herbivore-induced plant volatile reduces parasitoid attraction by changing the smell of caterpillars

Abstract

Herbivore-induced plant volatiles (HIPVs) can mediate tritrophic interactions by attracting natural enemies of insect herbivores such as predators and parasitoids. Whether HIPVs can also mediate tritrophic interactions by influencing the attractiveness of the herbivores themselves remains unknown. We explored this question by studying the role of indole, a common HIPV in the plant kingdom. We found that herbivory-induced indole increases the recruitment of the solitary endoparasitoid Microplitis rufiventris to maize plants that are induced by Spodoptera littoralis caterpillars. Surprisingly, however, indole reduces parasitoid recruitment when the caterpillars themselves are present on the plants. Further experiments revealed that indole exposure renders S. littoralis caterpillars unattractive to M. rufiventris, leading to an overall reduction in attractiveness of plant-herbivore complexes. Furthermore, indole increases S. littoralis resistance and decreases M. rufiventris parasitization success. S. littoralis caterpillars are repelled by indole in the absence of M. rufiventris but specifically stop avoiding the volatile in the presence of the parasitoid. Our study shows how an HIPV can undermine tritrophic interactions by reducing the suitability and attractiveness of caterpillars to parasitoids.

Fig. 1. Effect of indole on parasitoid attraction to plant-herbivore complexes.

(A to F) Female M. rufiventris parasitoids were allowed to choose between odors from wild-type (WT) and indole-deficient igl-mutant plants that were induced by wounding and caterpillar (S. littoralis) oral secretions (n = 6). (A) Synthetic indole dispensers (50 ng hour⁻¹) and empty dispensers (control; n = 5). (B) WT and igl-mutant plants that were infested by S. littoralis caterpillars (n = 8). (C) S. littoralis–infested igl-mutant plants and S. littoralis–infested igl-mutant plants supplemented with an indole dispenser (n = 6). (D) S. littoralis–induced WT plants and S. littoralis–induced igl-mutant plants with the larvae removed from the plants (n = 8). (E) S. littoralis–induced WT plants and S. littoralis–induced igl-mutant plants with the larvae added back into the systems in small-mesh cages (n = 8). (F) Bars represent the percentages of parasitoids choosing either of the odor sources. Pie charts indicate the proportion of parasitoids that made a choice. False discovery rate (FDR)–corrected P values are given for treatment comparisons [generalized linear model (family, quasi-Poisson)], followed by least squares means (LSM) pairwise comparisons.

Fig. 2. Indole renders caterpillars unattractive to parasitoids.

(A to C) Female M. rufiventris parasitoids were allowed to choose between odors from S. littoralis caterpillars that had been feeding on WT or indole-deficient igl-mutant plants (n = 8). (A) S. littoralis caterpillars that had been feeding on artificial diet in the presence of an indole-releasing dispenser (50 ng hour⁻¹) or control dispenser (n = 5). (B) Surface extracts of S. littoralis caterpillars that had been feeding on artificial diet in the presence of an indole-releasing or control dispenser (n = 6). (C) Bars represent the percentages of parasitoids choosing either of the odor sources. Pie charts indicate the proportion of parasitoids that made a choice or not. FDR-corrected P values are given for treatment comparisons [generalized linear model (family, quasi-Poisson)], followed by pairwise comparisons of LSM.

Fig. 3. Indole increases caterpillar resistance toward parasitoids.

(A) Survival of S. littoralis larvae (±SE, n = 7) under control conditions (−) and after parasitism by M. rufiventris (+). Caterpillars were exposed to empty control dispensers (green bars) or indole-releasing dispensers (50 ng hour⁻¹; blue bars). (B) Survival of S. littoralis larvae (±SE, n = 12) that fed on WT and indole-deficient igl-mutant plants after parasitism by M. rufiventris. (C) Number of M. rufiventris parasitoids (±SE, n = 7) developing successfully within S. littoralis caterpillars in the presence of indole or control dispensers. (D) Number of M. rufiventris parasitoids (±SE, n = 12) developing successfully within S. littoralis caterpillars that reared on WT and igl-mutant plants. (E) M. rufiventris attack rates (±SE, n = 12) of S. littoralis larvae that fed on WT or igl-mutant plants. P values are given for treatment comparisons (Student’s t tests).

Fig. 4. Parasitoid exposure suppresses the repellent effect of indole toward caterpillars.

S. littoralis larvae were allowed to choose between odors from synthetic indole dispensers (50 ng hour⁻¹) and empty dispensers (control). (A and B) Naïve caterpillars (A) and scared caterpillars that were in proximity to M. rufiventris females for 30 min without direct contact (B) were tested separately (n > 4). Bars represent the percentages of caterpillars choosing either of the odor sources. Pie charts indicate the proportion of parasitoids that made a choice or not. FDR-corrected P values are given for treatment comparisons [generalized linear model (family, Poisson)], followed by pairwise comparisons of LSM.