Gall insects and indirect plant defenses

Abstract

Many plants can defend themselves against insect herbivory by attracting natural enemies that kill feeding herbivores and limit the damage they inflict. Such “indirect defenses” can be induced by insects feeding on different plant tissues and using a variety of feeding styles. However, we have recently shown that gall-inducing insect species can avoid the indirect defenses of their host plant species and even alter volatile emissions following subsequent herbivory. One of the species we studied, Eurosta solidaginis, induces galls on goldenrod (Solidago altissima) and appears to exert a unique influence over the indirect defenses of its host plant that is not readily explained by levels of defense-related phytohormones, gall formation or resource depletion. Our evidence suggests that this gall-insect species may be able to manipulate its host plant species to avoid and/or modify its defensive responses. The results also provide insight into gall induction because the gall-insect species that we screened did not increase levels of jasmonic acid, which, in addition to triggering volatile emissions, is a powerful growth regulator that could prevent the cell growth and division that leads to gall formation.

In a classic review Price and colleagues argued that plants can influence interactions between herbivorous insects and their natural enemies.¹ Now, nearly 30 years later, a strong body of literature has developed, confirming this notion and documenting that plants can dynamically mediate these interactions. For example, many plant species can actively protect themselves against insect attack using “indirect defenses.” These herbivore-induced defenses are typically mediated by increases in extrafloral nectar or volatile plant compounds that attract predators or parasitoids, which can kill feeding herbivores.² Because the killing action of natural enemies can limit the damage herbivores inflict upon host plants, such indirect defenses can benefit plant reproduction, providing plants that possess such traits with a fitness advantage.^3–6 Indirect defenses have been shown to be induced by the feeding of chewing insects (e.g., caterpillars or beetle larvae), sucking insects (e.g., aphids), and piercing-sucking insects (e.g., thrips) or mites.^7–12 Moreover, insects feeding on leaves or roots can trigger these defenses.^7,11 Unlike these ectophytic feeders, the indirect defensive response of plants to endophytic feeders has not been well-studied.¹³

Gall insects, one class of endophytic herbivores, force their host plants to produce tumor-like growths that provide the insects with food and shelter at the expense of the host plant. As accomplished manipulators of plant physiology, gall-inducing insects provide an opportunity to explore indirect plant defenses from a unique perspective. At least one gall-inducing species induces changes in the volatile emissions from its host plant, and these altered volatiles attract natural enemies and serve as mate location cues for the gall wasp themselves.^6,14,15 Other gall-insect species, however, do not appear to significantly increase volatile cues.¹⁶ For example, we have reported that some gall-insect species appear to avoid indirect defenses and even alter volatile releases induced by other insect species.¹⁷ Similarly, some gall-wasp species can manipulate their host plants to produce galls that secrete “honeydew,” attracting ants that protect the galls from predators and parasitoids.¹⁸ Such phenomena demonstrate that the profound influence gall insects can exert over their host plants can extend to manipulating plant indirect defenses to the benefit of the galling insect.

Using field and laboratory experiments, our most recent work addressed the volatile response of tall goldenrod (Solidago altissima) to two specialized gall-inducing species: the tephritid fly Eurosta solidaginis and the gelechiid moth caterpillar Gnorimoschema gallae-solidaginis (Fig. 1). We compared the volatile response induced by these two gall inducers to those triggered by two polyphagous herbivore species that served as quasi-controls. The generalist caterpillar Heliothis virescens confirmed that plants were capable of mounting a volatile defensive response, while the meadow spittlebug Philaenus spumarius controlled for the sapping influence of parasitism because P. spumarius, like galling insects, can have a strong negative influence on growth of S. altissima.^19,20

Figure 1.

Galls of (A) Eurosta solidaginis and (B) Gnorimoschema gallaesolidaginis induced on stems of Solidago altissima.

The two gall-inducing insect species did not induce any significant changes in the volatile profiles of S. altissima. Moreover, neither gall-inducing species altered levels of jasmonic acid (JA), a product of the octadecanoid pathway that can trigger volatile release, or salicylic acid, which can antagonize the action of JA. In contrast, H. virescens feeding greatly increased JA levels (unpublished data) and the amount of volatiles released by S. altissima. These results suggest that E. solidaginis and G. gallaesolidaginis somehow avoid triggering a typical octadecanoid cascade that can result in increased JA levels, which after an initial burst can remain elevated with continued herbivory for weeks and even months.^21,22 We also found that when H. virescens fed upon E. solidaginis-galled plants, the quantity of volatile compounds released by S. altissima was significantly decreased, revealing that E. solidaginis exerted a unique degree of control over the volatiles of its host plant. The gall-inducing caterpillar G. gallaesolidaginis and meadow spittlebugs did not similarly decrease volatile releases when they were paired with H. virescens suggesting that gall induction per se and resource depletion may not explain the decrease in H. virescens-induced volatiles caused by E. solidaginis. Moreover, E. solidaginis appeared to selectively downregulate structurally similar volatile compounds. For example, of the monocyclic monoterpenes, limonene was strongly downregulated following H. virescens damage to E. solidaginis-galled plants while β-phellandrene remained unchanged between the H. virescens and H. virescens/E. solidaginis treatments. If levels of compounds were changing as a result of resource depletion or some other passive process, related compounds might be influenced similarly. This was not the case, suggesting a possible degree of active control of plant volatile emissions by E. solidaginis.

Our results with S. altissima corroborate our previous work with wheat, which showed that Hessian fly larvae failed to induce indirect defenses and also altered volatile release following caterpillar herbivory.¹⁷ Other work has shown that some insect species do not trigger indirect plant defenses,²³ but our results in disparate monocot and dicot systems go one step further and suggest that insect species may be able to wield active control over the indirect defenses of host plants. This level of control may not be surprising for gall insects, which can have unique influence over host-plant physiology, including the distribution of secondary metabolites.²⁴

Our findings may also provide some insight into gall induction because few, if any, other studies have measured JA levels on the interior of galls. In addition to triggering volatile releases in plants, JA is also a powerful plant growth regulator that has been shown to inhibit cytokinins²⁵ and auxin,²⁶ the two phytohormone classes commonly linked to gall formation.^27,28 Neither E. solidaginis, G. gallaesolidaginis, nor Hessian fly larvae significantly increased levels of JA of their host plants as observed with other herbivore species (unpublished data). Therefore, gall insects would seem to be under strong selection pressure to avoid increasing levels of JA, not only to avoid indirect defenses that could attract natural enemies, but also to stimulate the hyperplasia and hypertrophy necessary to form their galls.

Addendum to: Tooker JF, Rohr JR, Abrahamson WG, De Moraes CM. Gall insects can evade and alter indirect plant defenses. New Phytol. 178:657–671. doi: 10.1111/j.1469-8137.2008.02392.x.