Background: Plants have inducible defenses to combat attacking organisms. Hence, some herbivores have adapted to suppress these defenses. Suppression of plant defenses has been shown to benefit herbivores by boosting their growth and reproductive performance.
Results: We observed in field-grown tomatoes that spider mites (Tetranychus urticae) establish larger colonies on plants already infested with the tomato russet mite (Aculops lycopersici). Using laboratory assays, we observed that spider mites have a much higher reproductive performance on russet mite-infested plants, similar to their performance on the jasmonic acid (JA)-biosynthesis mutant def-1. Hence, we tested if russet mites suppress JA-responses thereby facilitating spider mites. We found that russet mites manipulate defenses: they induce those mediated by salicylic acid (SA) but suppress those mediated by JA which would otherwise hinder growth. This suppression of JA-defenses occurs downstream of JA-accumulation and is independent from its natural antagonist SA. In contrast, spider mites induced both JA- and SA-responses while plants infested with the two mite species together display strongly reduced JA-responses, yet a doubled SA-response. The spider mite-induced JA-response in the presence of russet mites was restored on transgenic tomatoes unable to accumulate SA (nahG), but russet mites alone still did not induce JA-responses on nahG plants. Thus, indirect facilitation of spider mites by russet mites depends on the antagonistic action of SA on JA while suppression of JA-defenses by russet mites does not. Furthermore, russet mite-induced SA-responses inhibited secondary infection by Pseudomonas syringae (Pst) while not affecting the mite itself. Finally, while facilitating spider mites, russet mites experience reduced population growth.
Conclusions: Our results show that the benefits of suppressing plant defenses may diminish within communities with natural competitors. We show that suppression of defenses via the JA-SA antagonism can be a consequence, rather than the cause, of a primary suppression event and that its overall effect is determined by the presence of competing herbivores and the distinct palette of defenses these induce. Thus, whether or not host-defense manipulation improves an herbivore's fitness depends on interactions with other herbivores via induced-host defenses, implicating bidirectional causation of community structure of herbivores sharing a plant.