Complex odor from plants under attack: herbivore's enemies react to the whole, not its parts

Abstract

Background: Insect herbivory induces plant odors that attract herbivores' natural enemies. Assuming this attraction emerges from individual compounds, genetic control over odor emission of crops may provide a rationale for manipulating the distribution of predators used for pest control. However, studies on odor perception in vertebrates and invertebrates suggest that olfactory information processing of mixtures results in odor percepts that are a synthetic whole and not a set of components that could function as recognizable individual attractants. Here, we ask if predators respond to herbivore-induced attractants in odor mixtures or to odor mixture as a whole.

Methodology/principal findings: We studied a system consisting of Lima bean, the herbivorous mite Tetranychus urticae and the predatory mite Phytoseiulus persimilis. We found that four herbivore-induced bean volatiles are not attractive in pure form while a fifth, methyl salicylate (MeSA), is. Several reduced mixtures deficient in one component compared to the full spider-mite induced blend were not attractive despite the presence of MeSA indicating that the predators cannot detect this component in these odor mixtures. A mixture of all five HIPV is most attractive, when offered together with the non-induced odor of Lima bean. Odors that elicit no response in their pure form were essential components of the attractive mixture.

Conclusions/significance: We conclude that the predatory mites perceive odors as a synthetic whole and that the hypothesis that predatory mites recognize attractive HIPV in odor mixtures is unsupported.

Figure 1. Response to a mixture of all five spider mite induced volatiles of Lima bean.

The odor of spider-mite-infested Lima bean (IB) was more attractive than the artificial mixture, plus a Lima bean leaf disc (B+M5) if no odor (NO) was the alternative. The artificial mixture plus the odor of non-infested Lima bean (B+M5) was preferred over a Lima bean leaf disc (B) to a similar extent as a spider-mite infested leaf disc (IB) was preferred to B. In a direct test the mites did not differentiate between the artificial odor (M5+B) and the odor of spider-mite-infested Lima bean (IB). The Y-axis represents the preference index (−100 total repellence, +100 total attraction). A star above the bar indicates a choice based on significance of G_p<0.05. Horizontal bracket bars with stars below the bars represent significant differences betwwen the pooled experimental results based on a Chi-square test (P<0.05).

Figure 2. Attraction to individual HIPV, a mixture of all HIPV and mixtures reduced by one component.

The left half of each panel depicts results of choice experiments in absence of Lima bean odor while the right half depicts the same experiments in presence of Lima bean background odor (indicated by a leaf disc). The first red bar represents attraction to the pure compound, whereas the second bar represents attraction to the mixture of all five HIPV (for comparison present in all Figures) and the third red bar attraction to a mixture reduced by one HIPV. The Y-axis gives the preference index (−100 total repellence, +100 total attraction). The abbreviation M5 refers to the mixture of all 5 HIPV, and M4 to a mixture reduced by one component. The letter h in a bar indicates significant heterogeneity (G_h<0.05) among replicates. A star above the bar indicates a choice based on significance of G_p<0.05. Capital letters (A,B) above the bars indicate differences with the mixture (M5) based on ANOVA followed by Dunnett's post-hoc test (P<0.05). Horizontal bracket bars with stars underneath represent significant differences between pooled experimental results, as inferred from Chi-square tests.

Figure 3. The experimental setup.

The choice arena was constructed from a Petri dish (Ø 9 cm) positioned up side down. An insect glue barrier (ig) divided the dish in two compartments that each contained an odor source (o) An opening at the bottom allowed for the connection of a cartridge containing the mites. The cartridge (c) was fitted to a vacuum pump. The vacuum gives rise to a radial airflow over the bottom of the choice arena, thus establishing two odor fields that extended from the odor sources to the cartridge. Arrows indicate air flow direction in the system.