Associations between higher plants and their gall-forming arthropod parasites are among the most specific, intimate and tightly co-evolved interspecific interactions known in macroscopic organisms. Galling insects, by definition, do not merely feed upon their host plants, but induce their hosts to form a bewildering array of highly specific tumor-like structures via biochemical interactions with plant hormones and regulatory mechanisms of gene expression (Fig. 1; Price et al., 1987). Recent research has demonstrated that in addition to inducing gross morphological changes, these gall-forming parasites often employ a diversity of more subtle manipulations of plant physiology and chemistry. These include the induction of sugary secretions (Fernandes et al., 1999), increased concentrations of defensive compounds in external gall tissues (e.g. Hartley, 1998), decreased concentrations of defensive chemicals and increased concentrations of nutrients (e.g. amino acids and lipids) in gall-nutritive tissues (Koyama et al., 2004), and suppression of the release of volatile organic compounds (VOCs) (Tooker & De Moraes, 2007).
In this issue of New Phytologist, John Tooker and colleagues (pp. 657–671) provide evidence that the manipulation of plants by galling insects may be even more subtle and complex than previously appreciated. Using a combination of laboratory and field experiments, Tooker et al. demonstrated that not only does the stem-galling tephritid fly Eurosta solidaginis fail to elicit a significant local volatile response in its perennial host, Solidago altissima (Tall Goldenrod), it also appears to systemically inhibit the emission of VOCs typically induced in response to a generalist folivorous caterpillar, Heliothis virescens.
‘... the intimate relationships of gall-formers with plants and their immobility may have selected for strategies to avoid detection by natural enemies; strategies that may include the manipulation of VOC release.’
Volatile organic compounds as indirect defenses?
Volatile phytochemicals induced by herbivore feeding have been shown to be attractive to natural enemies of the herbivores, particularly parasitoids, for a variety of specific plant–herbivore associations (e.g. De Moraes et al., 1998). Whether VOC release attracts natural enemies remains to be determined for this particular system, but it is probable given the widespread reliance of insect natural enemies on plant-derived VOCs (e.g. De Moraes et al., 1998). It is unclear in these systems whether the interactions between natural enemies and volatile chemical cues produced by plants is an evolved response to herbivore pressure; however, the conclusions of Tooker et al. do not rest upon the assumption that the emission of VOCs by S. altissima represents an adaptive indirect defense, only that such chemical cues may alter the mortality rates of herbivores by their enemies.
Insights from a comparative approach
A highlight of the study carried out by Tooker et al. is its comparative approach. By comparing volatile release among S. altissima plants with different herbivores, the authors were able to demonstrate that (1) not all galling herbivores inhibit volatile release in response to the exophytic caterpillar, and (2) the observed response is unlikely to occur as a result of resource depletion associated with feeding, as a xylem-feeding herbivore (spittlebug) failed to suppress volatile release. There has been relatively little study of volatile induction in response to gallers (but see Tooker & De Moraes, 2007), but the intimate relationships of gall-formers with plants and their immobility may have selected for strategies to avoid detection by natural enemies – strategies that may include the manipulation of VOC release. The failure of the gall-forming Gnorimoschema moth to suppress volatile release may be a result of its feeding mechanism (chewing mandibles rather than sucking or rasping), its phylogenetic proximity to the exophytic herbivore, or because, in contrast to Eurosta, its dominant enemies attack early in development before the gall is fully formed (Heard et al., 2006).
Mechanisms of suppression of induced indirect defenses
Tooker et al. suggest that alteration in the accumulation of salicylate may be part of the biochemical mechanism that suppresses induced VOC release. Induced volatile production in maize by a generalist folivorous caterpillar is associated with jasmonate accumulation in wounded tissues (e.g. Schmelz et al., 2003), but it remains to be demonstrated whether the same is true in S. altissima. If it is, salicylate accumulation in the tissues galled by Eurosta could be involved in the suppression of volatiles, given its known antagonism of the induction of jasmonate-mediated defenses (Cipollini et al., 2004). There is a need to determine where volatiles are made in the stem, which is important in order to know whether altered accumulation of salicylate in inner vs outer tissues of the gall could be part of the suppressive mechanism. Moreover, salicylate and jasmonate contents in leaves attacked by the exophytic herbivore, where volatile production was systemically suppressed by Eurosta galls in the stem, need to be assessed to determine the involvement of these defense hormones there. Alternatively, Eurosta may use other mechanisms, including salivary or fecal components, to silence volatile production, at least locally in the gall. Some herbivores can silence direct defenses of plants through the use of glucose oxidase in the saliva, although such mechanisms may ultimately rely on salicylate accumulation in the plant for the actual mechanism of suppression (Musser et al., 2002).
Tooker et al. also suggest that resource limitations caused by galling could play a role in the systemic suppression of volatile responses by Eurosta. Systemic induction of direct defenses in young leaves of some plants relies on the import of carbon from other leaves in a source–sink relationship (Arnold & Schultz, 2002). Because Eurosta galls can alter the patterns of carbohydrate translocation among organs of S. altissima (McCrea et al., 1985), disruption of source–sink relationships between stem and leaves could be part of the mechanism of the systemic suppression of exophytic herbivore-induced volatiles (but presumably not local Eurosta-induced volatiles). The authors address this possibility by comparing the suppressive effects of Eurosta with those of two other herbivores, but found that feeding by the galling moth Gnorimoschema and the xylem-feeding spittlebug failed to suppress VOC release. However, as the authors recognize, these organisms may not be appropriate controls given their different feeding modes and resources. Systemic suppression as a result of resource limitation also seems unlikely because the release of low-molecular-weight volatiles is thought to be among the least costly forms of plant defense (Halitschke et al., 2000). A parallel examination of the induction or suppression of ‘costly’ direct defenses and ‘cheap’ indirect defenses by gallers or associated leaf-feeders would help to clarify this possibility.
Trait-mediated indirect community effects
Although the study of Tooker et al. is preliminary in several respects, the local and systemic suppression of VOC release by E. solidaginis points to potential community-wide trait-mediated indirect effects of this gall-maker. If suppression of VOC release results in lower mortality rates of relatively stationary leaf-chewing insects, as the authors suggest, then this could significantly affect the community of insect herbivores feeding on S. altissima. In contrast to most examples of host-mediated indirect interactions, which tend to be negative (Kaplan & Denno, 2007), the interactions between galling Eurosta flies and the exophytic Heliothis caterpillars may be positive. Similar consequences have been observed with indirect interactions between above-ground and below-ground herbivores in maize, in that attractiveness to enemies is reduced in doubly infected plants as a result of changes in VOC release (Rasmann & Turlings, 2007). Such facilitative interactions among herbivores could result in positively correlated distributions of herbivores, a contagious distribution of herbivore damage and unstable population dynamics via negative density dependence of herbivore mortality. Given the growing evidence that VOCs released from one plant can induce or prime neighboring individuals for the induction of defense, suppression of VOC release in a galled ramet could also increase the susceptibility of neighboring ramets or genets by preventing this priming response.
The study of Tooker et al. represents an initial incursion into a relatively new and potentially insightful area of research involving the tritrophic, plant-mediated facilitation of insect herbivores. Yet, there are many conceptual and empirical gaps in our knowledge that need to be filled before the ecological importance of these interactions can be fully appreciated. For example, in this particular system, it remains to be shown that the VOCs are attractive to enemies and that inhibition of these ‘indirect defenses’ results in lower mortality risk for both endophytic and exophytic herbivores. As outlined above, the physiological mechanisms of VOC induction by exophytic herbivores and suppression by gall-makers need to be clarified, especially with regard to the roles of salicylic acid and jasmonic acid pathways and their potential antagonism. Finally, the effects of these interactions on the community need to be explored in the field to determine whether these tritrophic interactions significantly influence community structure and dynamics. Taking a broader view, these interactions need to be explored in additional tritrophic systems to understand whether such plant-mediated facilitative effects are widespread, and, if so, whether they function by similar or distinct mechanisms. While there has been growing recognition of community-wide effects of structural changes in plants caused by herbivores (e.g. Crawford et al., 2007), the diverse, yet visually inconspicuous, biochemical and physiological interactions between plants and gall-making herbivores may prove to be even more pervasive and ecologically significant.