Oecologia

, 167:1

Herbivore-induced resource sequestration in plants: why bother?

Authors

    • Department of BiologyTufts University
  • Alexandra Thorn
    • Department of BiologyTufts University
  • Sara Gómez
    • Department of BiologyTufts University
    • Department of Biological SciencesUniversity of Rhode Island
Concepts, Reviews and Syntheses

DOI: 10.1007/s00442-011-1968-2

Cite this article as:
Orians, C.M., Thorn, A. & Gómez, S. Oecologia (2011) 167: 1. doi:10.1007/s00442-011-1968-2

Abstract

Herbivores can cause numerous changes in primary plant metabolism. Recent studies using radioisotopes, for example, have found that insect herbivores and related cues can induce faster export from leaves and roots and greater partitioning into tissues inaccessible to foraging herbivores. This process, termed induced resource sequestration, is being proposed as an important response of plants to cope with herbivory. Here, we review the evidence for resource sequestration and suggest that associated allocation and ecological costs may limit the benefit of this response because resources allocated to storage are not immediately available to other plant functions or may be consumed by other enemies. We then present a conceptual model that describes the conditions under which benefits might outweigh costs of induced resource sequestration. Benefits and costs are discussed in the context of differences in plant life-history traits and biotic and abiotic conditions, and new testable hypotheses are presented to guide future research. We predict that intrinsic factors related to life history, ontogeny and phenology will alter patterns of induced sequestration. We also predict that induced sequestration will depend on certain external factors: abiotic conditions, types of herbivores, and trophic interactions. We hope the concepts presented here will stimulate more focused research on the ecological and evolutionary costs and benefits of herbivore-induced resource sequestration.

Keywords

DefensePlant–herbivore interactionsStorageToleranceResource allocation

Introduction

For all organisms, allocation of resources to the primary functions of growth and reproduction must be balanced with the various secondary functions required to survive and deal with abiotic and biotic stresses. Plants rely on physiological, chemical, biomechanical, and developmental processes to deal with stress. Stresses that vary in time, such as drought and attack by herbivores, require constant adjustment and these adjustments are essential to growth and survival of plants (Mooney and Winner 1991; Karban and Baldwin 1997; Agrawal and Karban 1999). In response to herbivory, for example, plants can employ two general strategies: production of chemical and morphological defense traits to deter herbivores (“resistance”), and mobilization of storage reserves for regrowth and reproduction after leaf loss (“tolerance”) (Karban and Baldwin 1997). Tolerance mechanisms are often linked to regrowth processes after a herbivory event (Tschaplinksi and Blake 1989a, b; Tiffin 2000). Less appreciated, however, is the herbivore-induced change in resource allocation and physiology that can increase plant tolerance to herbivory (Schwachtje and Baldwin 2008). This phenomenon, termed induced resource sequestration, refers to rapid herbivore-induced changes in resource allocation patterns that result in an increase in export of existing or newly acquired resources from attacked tissues (and/or systemic tissues with vascular connections) into storage organs. These resources are thus temporarily sequested (unavailable) for growth, defense or storage in the tissues from which they were exported.

Evidence for induced resource sequestration

There is growing evidence that herbivore feeding, herbivore cues and signal molecules associated with herbivory cause changes in resource export and allocation to storage tissue (Holland et al. 1996; Schwachtje et al. 2006; Babst et al. 2008; Kaplan et al. 2008). Holland et al. (1996), for example, found that feeding by grasshoppers causes 14C to accumulate in roots. Kaplan et al. (2008) showed a similar pattern in tobacco roots after folivory by chewing herbivores using 13C. Recent studies have adopted the use of short-lived radioisotopes (such as 11CO2), which allows quantification of resource dynamics in vivo and the comparison of allocation patterns before and after treatment. This is possible because of the rapid decay of 11CO2 (t1/2 = 20.4 min). This approach has been used to document an increase in leaf photosynthate export to stems and/or roots within hours of treatment with jasmonic acid (Babst et al. 2005), caterpillar regurgitant (Schwachtje et al. 2006), or feeding by gypsy moth larvae (Babst et al. 2008). In addition to Babst et al. (2005), other studies have also used jasmonates to study induced sequestration. Methyl jasmonate increases photosynthate export from treated leaves (Gómez et al. 2010) and treatment of roots with jasmonates causes photosynthate to be diverted away from the treated roots and into untreated roots (Henkes et al. 2008). Interestingly, silencing the jasmonate pathway in wild tobacco does not prevent induced export and partitioning (Schwachtje et al. 2006), suggesting that other signaling pathways may also be involved.

Changes in resource allocation in response to herbivory are not limited to photosynthate. For example, Frost and Hunter (2008) did not observe increased carbon accumulation in storage tissues of oaks following herbivory, but did observe an increase in nitrogen within these tissues. When methyl jasmonate is applied to leaves of tomato, it increases nitrogen (13N) export and partitioning to roots (Gómez et al. 2010), and when applied to roots of alfalfa it increases nitrogen storage within the tap root (Meuriot et al. 2004).

Relatively little is known about the long-term consequences of induced sequestration. Beardmore et al. (2000) showed that chronic exposure of leaves to methyl jasmonate can increase protein concentrations in storage tissues in poplar. Schwachtje et al. (2006) found that induced root partitioning in wild tobacco extends flowering time and thereby increases fitness. Given the increasing evidence for induced sequestration, there is a need to examine the conditions likely to favor this strategy.

Storage: benefits and costs

It is well known that constitutive storage represents an important buffer against abiotic and biotic stresses (Trumble et al. 1993; Kobe et al. 2010). For example, drought often triggers a change in the distribution of starch and sugar reserves and frequently results in greater transport to roots or to young developing leaves (Geiger and Servaites 1991). Similarly, defoliation is well known to result in the mobilization of starch reserves to fuel new plant growth (Tschaplinksi and Blake 1989a, b; Kosola et al. 2001), and the presence of these storage reserves is a key factor determining post-defoliation survival (Canham et al. 1999). The buffering capacity of storage implies that herbivore-induced storage could be adaptive.

There can be costs of storage. Although studies of wild and cultivated species have shown that species with higher rates of storage are more likely to survive stressors such as shading (Kobe 1997; Myers and Kitajima 2007), drought or nutrient stress (Shaw et al. 2002; Paula and Pausas 2011), and defoliation (Anten et al. 2003; Myers and Kitajima 2007), these same studies show that these species grow more slowly. Induced storage may provide a buffering mechanism without the long-term growth costs of constitutive storage.

We note that costs may be transient by varying with plant ontogeny and phenology (Boege and Marquis 2005; Boege et al. 2007; Orians et al. 2010; Van Dam et al. 2001). In particular, deflection from growth may be most costly during periods of rapid growth, including periods of leaf production and fruit maturation. A recent study that focused on growth–defense tradeoffs in willow illustrates this concept (Orians et al. 2010). This study found evidence for a trade-off between allocation to roots and defense in younger seedlings but a positive correlation in older seedlings, a result consistent with the observation that larger plants often grow faster and produce higher concentrations of chemical defenses (e.g., Briggs and Schultz 1990; Orians et al. 2003).

In contrast to the cost of constitutive storage, the costs of herbivore-induced resource sequestration has received little attention. It may result in allocation costs (fewer resources for growth or reproduction), or ecological costs (higher performance of enemies that consume storage tissues). Evidence for allocation costs comes from a study on wild tobacco (Schwachtje et al. 2006). They found that induction of wild-type plants increased carbon allocation to roots (10%) and resulted in smaller plants that exhibited delayed reproduction. Moreover, a transformed genotype that had constitutively greater allocation to roots was shorter and produced fewer reproductive capsules. Interestingly, the transformed genotype mobilized the root reserves after elicitation and this resulted in greater flower production later in the season. Clearly there are costs and potential benefits of induced sequestration. We expect that costs might be even larger in other plants since photosynthate allocation to stems and/or roots has been shown to be as high as 25% (Babst et al. 2008). Moreover, many studies have shown that plants grow more slowly and exhibit reduced fitness after simulated attack (Baldwin et al. 1998; Zavala et al. 2004; Walls et al. 2005; reviewed by Cipollini et al. 2003). While this is usually attributed to the cost of resistance, an increase in induced sequestration could contribute to this difference.

There are also potential ecological costs (Kaplan et al. 2008). Herbivore-induced resource sequestration and potential subsequent exudation into the rhizosphere can alter or create new interactions between plants and other organisms in the soil (Bardgett et al. 1998; Henry et al. 2008). In some cases, induced allocation changes can lead to positive interactions by promoting the colonization by mutualists such as mycorrhizae (Tejeda-Sartorius et al. 2008), but in other cases induced sequestration might incur ecological costs if it attracts or improves the performance of consumers of the organs where resources are being stored. For example, Kaplan et al. (2009) showed that aboveground herbivory in tobacco resulted in an increased allocation of carbon to roots and this was linked to an increase in fecundity of a root nematode.

Conceptual model for resource allocation

While it is apparent that trade-offs often exist understanding the ways in which intrinsic and extrinsic factors are expected to affect resource allocation requires a detailed examination of the status of different tissues. In Fig. 1, we present a conceptual model for possible pathways of resource flow within a plant. In general, mature source leaves are responsible for the majority of carbon fixation, and fine roots for the majority of nutrient uptake. These resources may be allocated locally to growth, defense, or storage (Blossey and Hunt-Joshi 2003; Karban and Baldwin 1997), or exported for use elsewhere in the plant. For a plant in vegetative growth phase that is not experiencing herbivory, investment in new leaves represents the best assurance of long-term growth, but this may not be the case when herbivores are present, since these young leaves are often particularly vulnerable to attack even when these young leaves are highly defended (e.g., Nichols-Orians and Schultz 1990). Allocation to stem and root storage may be increasingly favored as the probability of continued defoliation increases, but this depends on the vulnerability of these storage pools to attack by stem and root herbivores or pathogens (see ecological costs above), and intrinsic traits of the plant (see Fig. 2 below).
https://static-content.springer.com/image/art%3A10.1007%2Fs00442-011-1968-2/MediaObjects/442_2011_1968_Fig1_HTML.gif
Fig. 1

Conceptual model for resource flows in plants. The labile resource pool is derived from newly captured pools of carbon and nutrient pools or from remobilized storage reserves. The labile carbon pool is generated from photosynthesis, primarily by mature source leaves. The labile nutrient pool is obtained from roots. The resulting labile resource pool can then be allocated to support the growth of sink tissues (roots, leaves, or reproductive tissues), to defense traits, and to storage tissues. Herbivore-induced export of resources from leaves or from fine roots (dashed arrows) into stems and storage roots functions to sequester resources in tissues inaccessible to the respective herbivores but may incur opportunity costs if resources allocated for storage limit growth and reproduction or ecological costs if other enemies specialize on these storage tissues

https://static-content.springer.com/image/art%3A10.1007%2Fs00442-011-1968-2/MediaObjects/442_2011_1968_Fig2_HTML.gif
Fig. 2

A fulcrum model for predicting herbivore-induced sequestration in plants. Outcomes depend upon the relative strength of intrinsic and extrinsic factors. See text for presentation of specific predictions

Herbivore-induced resource sequestration: a predictive framework

We present a fulcrum model that explores conditions that will tend to favor greater induced sequestration relative to growth and/or defense (Fig. 2). We note that most plants simultaneously grow, defend and allocate to storage so our goal is to highlight conditions that will maximize the extent of induced resource sequestration. First, we evaluate intrinsic factors such as life history, ontogeny and phenology. Second, we review extrinsic factors including the abiotic and biotic environment, including resource availability and attributes of the herbivores themselves.

Intrinsic factors

To understand how patterns of resource allocation change in response to herbivory, it is necessary to characterize the status of the plant prior to herbivory. This status depends on a range of factors, but centrally on aspects of plant life history, plant ontogeny and phenology.

Life history

Plant species diverge greatly in their inherent patterns of allocation to storage. Many plants constitutively allocate large amounts of resources to rhizome and root storage. This is true for cultivated crops such as beets, carrots and potato, for biennial and perennial wild plants such as Alliaria petiolata (biennial) and Pastinaca sativa (biennial to perennial; Sosnová and Klimešová 2009), and for species that have a high capacity for resprouting (Paula and Pausas 2011). For species with high constitutive storage, the amount of carbon that can be sequestered during a folivore attack is likely to be a small fraction of the total storage pool. Induced carbon sequestration may provide little benefit in these cases. In contrast, for species with little root storage or for species that maintain high storage pools in their leaves during the growing season, induced export is predicted to be beneficial. In these plants, resources deflected from growth and into storage during an attack may provide a critical pool of resources necessary for regrowth.

Although these expectations have not been explicitly tested, a few studies fit the predictions. Photosynthate export to roots in the annual Nicotiana attenuata increased only 10% (Schwachtje et al. 2006). In contrast, induced sequestration of photosynthate was close to 25% in Populus (Babst et al. 2008), a woody perennial with high concentrations of starch in its leaves (Babst et al. 2005). Red oak, however, exhibited no induced sequestration of photosynthate (Frost and Hunter 2008). Compared to other species, oaks have a large root system and readily resprout following cutting (Abrams 2003). The lack of induced sequestration is consistent with the prediction that induced sequestration would be low in species with high constitutive storage. Clearly further research explicitly comparing species with different intrinsic traits is warranted.

Other life-history traits may also be important. Grime (2001) classified species as being ruderal (short-lived weedy species), competitive (long-lived dominant species), or stress tolerators (species adapted to stressful environments). These differences are likely to influence a plant’s relative resource allocation to growth, defense and storage. Rapid growth is a characteristic of ruderal species, and individuals that do not prioritize growth are likely to be overgrown, making the opportunity costs of induced storage very high. Only after establishment might we expect induced sequestration in these species. In contrast, both competitive and stress tolerant species are expected to invest significant resources in storage as a way to buffer against environmental fluctuations (Kobe et al. 2010; Paula and Pausas 2011). While induced resource sequestration may be more common in these species, we expect it to be negatively correlated with constitutive levels of storage as species with more constitutive storage already have the capacity to recover from tissue loss.

Ontogeny

Herbivore-induced sequestration is also postulated to vary with plant ontogeny (Fig. 2). Young plants and their young tissues are particularly prone to attack due to a higher nitrogen content and less developed physical properties (McKey 1974; Coley and Barone 1996; Fenner et al. 1999; Wainhouse et al. 2009). Their small size also makes it likely that herbivores can rapidly remove most if not all of the leaf area (e.g., Fritz et al. 2001). This may limit the benefits of induced sequestration and favor defense and growth. In contrast, older plants may benefit from induced export of resources to short-term storage pools prior to reproduction or to late-season sequestration of resources for overwintering (perennials only).

Phenology

Both leaf and reproductive phenology are predicted to influence patterns of induced sequestration (Fig. 2). At leaf flush, young expanding leaves are generally highly susceptible to herbivores (Coley and Barone 1996), making rapid maturation a key defensive trait (Aide 1988). We expect minimal induced sequestration during periods of leaf expansion; rather, the production of new leaf tissue and the defense of existing tissue is likely to be particularly important to both young annuals and first-year perennial plants as predicted by the Optimal Defense Hypothesis (McKey 1974; van Dam et al. 1996; de Boer 1999). Once expanded, induced sequestration rates are predicted to be higher. We also expect higher rates of induced sequestration prior to reproduction. During seed and fruit maturation, however, we expect minimal induced sequestration (especially in annuals) since reproductive tissues are strong sinks.

Extrinsic factors

Abiotic

Light and soil nutrient availability have large effects on patterns of allocation to roots and to storage. In response to light limitation, both the ratio of roots to shoot (Mooney and Winner 1991) and the concentration of storage compounds are much lower (Nichols-Orians 1991), suggesting that induced resource sequestration will be constrained by light availability. In contrast, root:shoot ratios increase in response to soil nutrient limitation, and a recent study by Kobe et al. (2010) showed that investment in non-structural carbohydrates within roots contributed to this pattern. This leads us to predict that the capacity for induced sequestration may be greater for plants experiencing low nutrient conditions.

Biotic

The extent of damage, herbivore specialization, mobility, feeding guild, and gregariousness are all likely to affect the likelihood of induced resource sequestration (Fig. 2). Evolutionarily, species typically consumed by large browsing mammals, for example, may maintain high constitutive storage and exhibit little induced storage. Insect herbivores, whose populations fluctuate widely from year to year, represent a more variable selective pressure (Hunter 1991), and could select for an induced sequestration response. Even within insect herbivores, the benefits of induced sequestration are likely to vary and this could lead to the evolution of specific plant responses (Agrawal 2000). To date, the evidence for induced sequestration comes from plant responses to herbivorous insects. Below, we examine how the extent of damage and characteristics of the herbivorous insect are both likely to affect patterns of resource sequestration.

Extent of damage

Ecologically, under mild or moderate herbivore defoliation, allocation of storage should be costly since stored resources are unavailable for investment in new tissues. A smaller leaf area not only limits growth rates during herbivory but would also be expected to limit regrowth potential. In contrast, if leaf area loss eventually leads to complete defoliation, the growth rate of leaves during herbivore attack is irrelevant (all leaves are removed), and the increase in stored carbon pools from induced storage would be expected to increase the regrowth potential. Similarly, induced storage during a mild attack is expected to be beneficial if early season herbivory predicts more severe future attack or if late season defoliation is common for and is predicted by early-season herbivory. In particular, if complete defoliation later in the growing season is likely, induced storage in response to prior attack would be beneficial. Moreover, since young tissues are often more vulnerable to subsequent damage than mature tissues (Denno and McClure 1983; Nichols-Orians and Schultz 1990), allocation to new growth could result in higher total leaf removal. This could shift the balance to favor storage over continued production of new leaves.

Specialist versus generalist species

We suggest that induced sequestration may be a more effective strategy against specialist herbivores than induced chemical defenses (Fig. 2). Many specialist herbivores have evolved effective detoxification mechanisms, and even use the toxic chemicals as feeding or oviposition cues (Macel and Vrieling 2003; Müller-Schärer et al. 2004; Hopkins et al. 2009). The failure of many chemical defenses to deter specialist herbivores leads to the prediction that induced sequestration would be prevalent in response to specialists. In contrast, induced chemical defenses are quite effective against generalists, and therefore we predict plants to allocate more resources to defense than to storage when attacked by generalists.

Sedentary versus mobile species

For sessile herbivores, larvae develop in the tissue on which the adult females lay their eggs. Unless other ovipositing females are present, the risk of attack to uninfested leaves is zero. Moreover, it is not uncommon for these sedentary species to aggregate (Whitham 1983; Orians and Björkman 2009). We therefore predict that damage by sedentary herbivores will favor export of resources from the attacked leaves (Fig. 2), although the opposite pattern may be observed if herbivores are able to hormonally manipulate the plant (Giron et al. 2007). Indeed, high densities of leaf miners are known to trigger early leaf abscission (Bultman and Faeth 1986). In contrast, mobile herbivores often move between leaves. We therefore expect mobile herbivores to induce export both locally and systemically, as has been observed for gypsy moths on Populus (Babst et al. 2008).

Gregarious versus solitary species

We predict that the ability to rapidly sequester resources into storage organs may be an essential response to gregarious herbivores (Fig. 2). In fact, the propensity to aggregate is the one factor repeatedly associated with insect species that commonly reach outbreak densities and cause extensive defoliation (Nothnagle and Schultz 1987; Larsson et al. 1993; reviewed by Hunter 1991). Moreover, gregarious species are often the most frequently observed herbivores on their host plants (Björkman et al. 2000; Carson and Root 2000; Dalin 2006). Solitary species, in contrast, are less likely to become numerically dominant and often exhibit conspecific avoidance and even experience higher mortality rates when aggregated (Jones 1987; Eber 2004). This will tend to limit the magnitude of damage, unless individual herbivores are large (e.g., later instars of some insect species such as Manduca sexta). Induced sequestration may be critical to regrowth following attack by gregarious species whereas for solitary species it would more likely represent an opportunity cost.

Feeding guild

To date, studies documenting induced sequestration have used leaf-chewing herbivores as models, either by releasing herbivores on the plants, inducing them with regurgitants/salivary cues, or by treating plants with jasmonates. To our knowledge, no studies have examined the effects of mammalian browsers or piercing/sucking insects such as aphids, whiteflies, adelgids and scale insects. Several lines of evidence suggest that induced sequestration will be limited in response to both. Browsers can rapidly defoliate an entire plant and thus there would be little time to respond. Although piercing/sucking insects do not cause defoliation, we still expect little induced sequestration. By feeding directly from the phloem, they cause less tissue damage, and thus tend to cause little or no induction (Walling 2000, 2008). Some even silence plant defense responses (Walling 2008). Still other piercing/sucking insect species are able to hormonally manipulate the plant and thus actually increase sink strength within the attacked tissues (Giordanengo et al. 2010).

Communities of attackers: from aboveground to belowground

The adaptive value of induced sequestration of carbohydrates in the roots and stems depends on the security of this pool of storage reserves (Fig. 2). Despite the benefits that resource sequestration can confer to plants in response to herbivory, exporting resources to storage organs can have far-reaching consequences that may not always have a positive effect on plant performance. Plants are simultaneously attacked above- and belowground by a myriad of herbivores and pathogens (Masters et al. 1993; Van der Putten et al. 2001; Blossey and Hunt-Joshi 2003; Dicke 2009; Kaplan et al. 2009). Therefore, the presence of root attackers could represent a major cost to export of material from the leaves by providing additional resources to root herbivores and pathogens (Kaplan et al. 2008). Similarly, stem-borers could also exploit the sequestered resources. Thus, the success of exporting aboveground resources into stems or roots as a strategy to safeguard valuable resources will depend on the herbivore/pathogen pressure on those tissues.

Conclusions

There is increasing evidence showing that plants increase their allocation to storage tissues in response to herbivory. All else being equal, however, allocation to storage represents an allocation cost since investment in new growth would increase plant size and ultimately reproductive potential. Yet certain conditions are more or less likely to favor such a strategy. We have argued that greater attention to the ecological context is needed before testing when and where induced sequestration is likely to be common and to evaluate its adaptive value. In Fig. 2, we have outlined several conditions predicted to favor induced sequestration and other conditions that make such a strategy less likely. The balance of these forces is expected to determine the magnitude of induced storage in a given species or population. We hope this paper stimulates further research into the benefits, costs and mechanisms of this phenomenon.

Acknowledgments

We thank the anonymous reviewers for their valuable comments on the manuscript. The project was supported by the National Research Initiative (or the Agriculture and Food Research Initiative) of the USDA National Institute of Food and Agriculture, grant number # 2007-35302-18351.

Copyright information

© Springer-Verlag 2011