Healthy plants growing under optimal environmental conditions should be capable of maintaining their full array of metabolic processes, and may provide greater nutritional value to insects capable of countering plant defenses. Such plants may allocate more re- sources to growth, relative to defenses, thereby compensating for losses to herbivores (Glynn et al. 2003, Trumble et al. 1993, see Chapter 12). By contrast, unhealthy plants, or plants growing under adverse environmental conditions (such as water or nutrient limi- tation) may sacrifice some metabolic pathways in order to maintain those which are the most critical to survival (e.g., Herms and Mattson 1992, Lorio 1993, Mattson and Haack 1987, Mopper et al. 2004, Tuomi et al. 1984, Wang et al. 2001, R. Waring and Pitman 1983).
In particular, stressed plants often reduce their production of defensive chemicals in order to maximize the allocation of limited resources to maintenance pathways. They thereby become increasingly vulnerable to herbivores (Fig. 3.10).
Spatial and temporal variation in plant defensive capability creates a mosaic of food quality for herbivores (L. Brower et al. 1968). In turn, herbivore employment of plant de- fenses affects their vulnerability to predators (L. Brower et al. 1968, Malcolm 1992, Stamp et al. 1997, Traugott and Stamp 1996). Herbivore feeding strategies represent a trade-off between maximizing food quality and minimizing their vulnerability to predators (e.g., Schultz 1983, see below).
The frequent association of insect outbreaks with stressed plants (e.g., V.C. Brown 1995, Heliửvaara 1986, Heliửvaara and Vọisọnen 1986, 1993, W. Smith 1981) led T. White (1969, 1976, 1984) to propose the Plant Stress Hypothesis, i.e., that stressed plants are more suit- able hosts for herbivores. However, some herbivore species prefer more vigorously-growing, apparently non-stressed plants (G. Waring and Price 1990), leading Price (1991) to propose the alternative Plant Vigor Hypothesis. Reviews by Koricheva et al. (1998) and G. Waring
and Cobb (1992) indicated that response to plant condition varies widely among herbivore species. Schowalter et al. (1999) manipulated water supply to creosotebushes, L. tridentata, in New Mexico and found positive, negative, non-linear and non-significant responses to moisture availability among the assemblage of herbivore and predator species on this single plant species, demonstrating that, in some cases, both hypotheses are supported.
Regardless of the direction of response, water and nutrient subsidy or limitation clear- ly affect herbivore–plant interactions (Coley et al. 1985, M.D. Hunter and Schultz 1995, Mattson and Haack 1987). Therefore, resource acquisition by insects is moderated, at least in part, by ecosystem processes or environmental changes that affect the availability of water and nutrients for plants (Chapter 11).
Some plant species respond to increased atmospheric concentrations of CO2 (carbon di- oxide) by allocating more carbon to defenses, such as phenolics or terpenoids, especially if fIG. 3.10 The density of mountain pine beetle attacks necessary to kill lodgepole pine increases with increasing host vigor, measured as growth efficiency. The solid portion of circles represents the degree of tree mortality. The solid line indicates the attack level predicted to kill trees of a specified growth efficiency (index of radial growth); the dotted line indicates the threshold above which beetle attacks are unlikely to cause mortality. From R. Waring and Pitman (1983) with permission from John Wiley & Sons.
75 I. rESoUrCE qUALITy
other critical nutrients remain limiting (e.g., Arnone et al. 1995, Chapin et al. 1987, Grime et al. 1996, Kinney et al. 1997, Roth and Lindroth 1994). However, the way in which a plant responds to CO2 enrichment does vary considerably among species, and will also vary as a result of environmental conditions such as light, water and nutrient availability (Bazzaz 1990, Dudt and Shure 1994, P. Edwards 1989, M. Hall et al. 2005, Niesenbaum 1992), with equally varied responses among herbivore species (e.g., Bezemer and Jones 1998, M. Hall et al. 2005, Salt et al. 1996, Watt et al. 1995). Zavala et al. (2008) demon- strated that elevated atmospheric CO2 resulted in down-regulation of gene expression for defense-signaling compounds and, consequently, proteinase inhibitors. Such complexity of responses precludes general prediction of effects of CO2 enrichment on insect–plant interactions (Bazzaz 1990, Watt et al. 1995).
Atmospheric deposition of nutrients which are typically limited, especially nitrogen, also affects insect–plant interactions, although the mechanisms involved are not clear. In general, nitrogen deposition increases growth and survival of individual insect herbivores, and promotes population growth (Throop and Lerdau 2004). Such enrichment may per- mit plants to allocate more carbon to growth, and reduce production of non-nitrogenous defenses, making plants more vulnerable to herbivores, as predicted by the Carbon/nu- trient Balance Hypothesis (Holopainen et al. 1995). M. Jones et al. (2004) reported that nitrogen deposition increased bark beetle activity and pine tree mortality. Zehnder and Hunter (2008) found that experimental simulation of nitrogen deposition in milkweed, Asclepias tuberosa, significantly increased foliar nitrogen concentration, plant biomass and per capita aphid, A. nerii, population growth, up to a point. However, increasing di- etary nitrogen does not improve insect performance (Zehnder and Hunter 2009). Joern and Behmer (1998) reported that two grasshopper species differed in their growth and reproduction on diets varying in carbohydrate and nitrogen contents. For Melanoplus sanguinipes, reproductive rate showed a significant negative linear response to increas- ing carbohydrate and a significant quadratic response to increasing nitrogen, with a peak in egg production at 4% nitrogen. Phoetaliotes nebrascensis, on the other hand, showed a much weaker response to increasing nitrogen and no response to increasing carbohydrate.
Experimental fertilization has produced apparently contradictory results (Kytử et al.
1996, G. Waring and Cobb 1992). In some cases, this inconsistency may reflect non-linear responses of insects to increasing nitrogen in plant tissue (Joern and Behmer 1998, Zehnder and Hunter 2009) or different feeding strategies relative to plant allocation of subsidized nutrients (Kytử et al. 1996, Schowalter et al. 1999). In other cases, the conflicting results may reflect changes in nutrient balances, i.e., which nutrients were most limiting (Behmer 2009, Elser and Urabe 1999, Elser et al. 1996, Sterner and Elser 2002). Furthermore, plants differ in their allocation of subsidized nutrients, e.g., to increased production of N-based defenses vs. increased protein content. Other associated species also may influence insect response to subsidized nutrients. Kytử et al. (1996) found that positive responses to N fertilization at the individual insect level often were associated with negative responses at the population level, perhaps indicating indirect effects of fertilization on attraction of predators and parasites.
f. mechanisms for exploiting variable Resource Quality
Although plant defensive chemistry clearly affects insect performance, insects are still ca- pable of feeding on defended hosts. Feeding preferences for less-defended hosts reflect one mechanism for avoiding defenses. However, insects exhibit a variety of mechanisms for improving plant suitability and/or avoiding, circumventing or detoxifying host defenses.
Gall-forming insects control gall formation and the chemical composition of colonized plant tissues, to the benefit of the insect (Saltzmann et al. 2008). Gall formation in the plant apparently is induced by salivary compounds, rather than by mechanical injury (So- pow et al. 2003), and reflects the relationship between shoot length and the dose of gall induction stimulus (Flaherty and Quiring 2008). Gall chemistry returns to that of sur- rounding tissues if the gall-former is killed (Hartley 1998). The inner lining of galls is nutritive tissue that is rich in free amino acids (Price et al. 1987, Saltzmann et al. 2008), but gall tissues outside this lining often are lower in nitrogen and higher in phenolics than are ungalled tissues (Hartley 1998). Y. Koyama et al. (2004) reported that the amount of amino acids exuding from leaves galled by the aphid, Sorbaphis chaetosiphon, was five times that from ungalled leaves. Furthermore, galls retained high amino acid concentra- tions throughout April, whereas amino acid concentrations declined rapidly during this period in ungalled leaves. Y. Koyama et al. (2004) also compared growth and reproduc- tion of another aphid, Rhopalosiphum insertum, which can displace gall aphids or colo- nize ungalled leaves. Growth and reproduction by this aphid were significantly higher for colonies experimentally established in galls, compared to colonies established on ungalled leaves, indicating a positive effect of gall formation.
Some insects vector plant pathogens that induce favorable nutritional conditions, or inhibit host defense (e.g., Bridges 1983). However, not all insects that vector plant patho- gens benefit from host infection (Kluth et al. 2002).
Insects that exploit nutritionally-poor resources require extended periods (several years to decades) of larval feeding, or other adaptations, in order to concentrate sufficient nutrients (especially N and P) to complete their development. Many have obligate asso- ciations with microorganisms that provide, or increase access to, limiting nutrients. Ter- mites host mutualistic gut bacteria or protozoa that catabolize cellulose, fix nitrogen, and concentrate or synthesize other nutrients and vitamins needed by the insect (Mankowski et al. 1998). Termites and some other detritivores feed on feces (coprophagy) after suf- ficient incubation time for microbial digestion and enhancement of nutritive quality. If coprophagy is prevented, these insects often compensate by increasing consumption of detritus (McBrayer 1975). Aphids also may rely on endosymbiotic bacteria to provide requisite amino acids, vitamins or proteins necessary for normal development and repro- duction (Baumann et al. 1995).
In general, food resources do not have the proper proportions of nutritional compo- nents that are required by animals for optimal nutrition. Insects have evolved a variety of strategies that govern the extent (trade-off) to which they will overeat a limiting nutrient and undereat an overabundant nutrient for optimal nutrition. Simpson and Raubenheim- er (1993) pioneered efforts to describe the fundamental variables of nutritional homeo- stasis, i.e., the rules that govern trade-offs when diets have suboptimal nutrient balance.
K.P. Lee et al. (2002, 2003), Raubenheimer and Simpson (1999, 2003) and Simpson et al.
(2002) described nutrient balancing strategies for several grasshopper and caterpillar spe- cies varying in host range.
Behmer (2009) reviewed insect strategies for dealing with nutritional imbalances in food resources. According to the Nutritional Heterogeneity Hypothesis (Fig. 3.11), the amount of nutritionally imbalanced food that is consumed should reflect the probability of encountering food that is equally and oppositely imbalanced. This probability is higher for insects with wide diet breadth, compared to those which specialize on a single food source. Therefore, insects specializing on a particular resource will be unable to compen- sate for nutritional imbalance, and should evolve to use small amounts of imbalanced food most efficiently, rather than suffer fitness costs of overeating imbalanced food (Fig. 3.11b,
77 I. rESoUrCE qUALITy
fIG. 3.11 Four possible protein–carbohydrate regulation strategies revealed by intake array and fitness landscape plots. The thin gray lines are nutritional rails for different foods, and each thick, black dotted line is an intake array, which reveals the optimal strategy under different fitness cost scenarios. An intake array is constructed by connecting the protein–carbohydrate intake points obtained for each food. A fitness landscape corresponding to each nutrient regulation strategy has been fitted over nutrient space, and the red area in each plot indicates the fitness peak that corresponds to the protein–carbohydrate intake target (in these examples a 1:1 ratio). Fitness costs increase as distance from the intake target increases, and this is represented by colors becoming successively cooler. Panel a shows linear fitness costs, and the fitness contours are straight parallel lines in each of the quadrants around the intake target. The intake array corresponds with feeding to either a horizontal or vertical line that passes through the intake target, except where the feeding rail and the fitness contour are coincident. Panel b shows symmetrical quadratic costs.
This regulatory strategy is most often seen in insect specialist herbivores. Panel c shows fitness contours and intake arrays that are ellipses in each of the quadrants around the intake target.
Panel d shows symmetrical quadratic with interaction costs. Here the fitness contours are tilted ellipses, and the intake array is more linear than it is in panel b. This is the regulatory strategy most often seen in generalist insect herbivores. From Behmer (2009) with permission from Annual Reviews, Inc.
closest distance rule). In contrast, mobile generalists should eat as much as possible of imbalanced foods as they are encountered, with a high probability of achieving nutritional balance overall (Fig. 3.11d, fixed proportion rule).
Insects also must balance the fitness costs of ingesting harmful chemicals with the costs of regulating nutritional balance. Behmer’s (2009) analysis indicated that plant defenses may have little effect on insect growth and survivorship when nutrient balance is optimal, but become increasingly deleterious, even fatal, as protein/carbohydrate imbalance increases.
Several mechanisms are employed to avoid or circumvent the defensive chemicals of a host. Some herbivores avoid exposure by moving to new resources in advance of an induced response (Paschold et al. 2007). Others sever the petiole or major leaf veins to inhibit translocation of induced defenses during feeding (Becerra 1994, Karban and Agrawal 2002). Sawflies (Diprionidae) sever the resin canals of their conifer hosts or feed gregariously to consume foliage before defenses can be induced (McCullough and Wagner 1993). Species that feed on plants with photooxidant defenses often feed at night or inside rolled leaves to avoid sunlight (Berenbaum 1987, Karban and Agrawal 2002).
Sequestration and excretion are alternative means of avoiding the effects of host toxins that cannot be detoxified. Sequestered toxins are transported quickly to specialized stor- age tissues (the exoskeleton or protected pouches), whereas remaining toxins are trans- ported to the Malphigian tubules for elimination. Boyd (2009) noted that high-Ni insects had elevated concentrations in Malphigian tubules and exuviae, indicating that elimination was being used as a strategy for feeding on hyperaccumulating plants. Sequestered toxins also may be used in the insect’s own defensive strategy (Blum 1981, 1992, Boyd and Wall 2001, Conner et al. 2000, L. Peterson et al. 2003).
Herbivorous insects produce a variety of catalytic enzymes, in particular those associ- ated with cytochrome P-450, to detoxify plant or prey defenses (Feyereisen 1999, Karban and Agrawal 2002, W. Mao et al. 2006, Y. Mao et al. 2007). Some insects produce salivary enzymes that minimize the effectiveness of plant defenses. Salivary enzymes, such as glu- cose oxidase, applied to feeding surfaces by caterpillars, inhibit the expression of genes which are responsible for the activation of induced defenses (Bede et al. 2006, Felton and Eichenseer 1999, Musser et al. 2005, 2006). The saliva of Hemiptera gels into a sheath, which separates the insect’s stylet from plant cells, perhaps reducing induced plant re- sponses (Felton and Eichenseer 1999).
Digestive enzymes responsible for detoxification are typically microsomal monooxy- genases, glutathione S-transferases, and carboxylesterases (Hung et al. 1990). These en- zymes fragment defensive compounds into inert molecules. Microsomal monooxygenases provide a general-purpose detoxification system for most herbivores (Hung et al. 1990). In addition, more specific digestive enzymes are produced by species that encounter particu- lar defenses. Ascorbate is a primary antioxidant found in the gut fluids of foliar-feeding insects, to reduce the effect of phenolic oxidation (Barbehenn et al. 2008). However, plant tissues that contain high concentrations of particularly reactive tannins can overwhelm this antioxidative capacity (Barbehenn et al. 2008). Exposure to plant toxins can induce the production of detoxification enzymes (Karban and Agrawal 2002). For example, cat- erpillars feeding on diets containing proteinase inhibitors showed reduced function of particular proteinases, but responded by producing other proteinases that were relatively insensitive to dietary proteinase inhibitors (Broadway 1995, 1997). The compounds pro- duced through detoxification pathways may be used to meet the insect’s nutritional needs (Bernays and Woodhead 1982), as in the case of the sawfly, Gilpinia hercyniae, which detoxifies and uses the phenolics from its conifer host (Schửpf et al. 1982).
79 II. rESoUrCE ACCEPTABILITy
The ability to detoxify plant defenses may predispose many insects to detoxify synthet- ic insecticides (Feyereisen 1999, Plapp 1976). At least 500 arthropod species are resistant to major insecticides which are used against them, primarily through a limited number of resistance mechanisms that confer cross-resistance to plant defenses and structurally related toxicants, and, in some cases, to chemically unrelated compounds (Hsu et al. 2004, Soderlund and Bloomquist 1990). Le Goff et al. (2003) reported that several cytochrome P-450 genes code for detoxification of DDT, imidacloprid and malathion. In some cases, insect adaptation reflects mutations that reduce binding to, or sensitivity of, target en- zymes (Hsu et al. 2004, 2006, 2008).
Gut pH is a factor that affects the chelation of nitrogenous compounds by tannins.
Some insect species are adapted to digest food at high gut pH, in order to inhibit chela- tion. The insect thus is relatively unaffected by high tannin contents of its food. Examples include the gypsy moth feeding on oak, Quercus spp., and chrysomelid beetles, Paropsis atomaria, feeding on Eucalyptus spp. (Feeny 1969, Fox and Macauley 1977).
Many predaceous insects use their venoms primarily for subduing prey, and secondarily for defense. Venoms produced by predaceous Hemiptera, Diptera, Neuroptera, Coleoptera and Hymenoptera function to paralyze or kill prey (Schmidt 1982), thereby minimizing injury to the predator during prey capture. The carabid beetle, Promecognathus, which is a specialist predator on Harpaphe spp. and other polydesmid millipedes, avoids the cyano- genic secretions of its prey by quickly biting through the ventral nerve cord at the neck, inducing paralysis (G. Parsons et al. 1991). Nevertheless, host defenses increase handling time and risk of injury and mortality for the consumer (Becerra 1994, Schmidt 1982).
Diversion of limited resources to detoxification enzymes or to avoidance efforts all involve metabolic costs (Karban and Agrawal 2002, Kessler and Baldwin 2002). Lindroth et al. (1991) evaluated the effect of several specific nutrient deficiencies on detoxifi- cation enzyme activity in the gypsy moth. They found that larvae on a low-protein diet showed compensatory feeding behavior (although this was not enough to offset their reduced protein intake). Soluble esterase and carbonyl reductase activities increased in response to protein deficiency, but decreased in response to vitamin deficiency. Poly- substrate monooxygenase and glutathione transferase activities showed no significant response. Furthermore, Carrière et al. (2001b) reported that the resistance of the pink boll- worm, Pectinophora gossypiella, to transgenic (Bt) cotton was associated with a reduced percentage emergence from diapause, compared to non-resistant bollworm, indicating the fitness costs of developing resistance strategies.
Some caterpillar species are able to suppress plant induction of defenses by means of prostaglandins in their oral secretions (Schultz and Appel 2004). Schultz and Appel (2004) reported that application of prostaglandin E2 or oral regurgitant from gypsy moth or forest tent caterpillar, Malacosoma disstria, reduced production of tannins by wounded red oak, Quercus rubra, leaves by 30–90%, compared to untreated controls, which in- creased their tannin production by 50–80% in response to wounding.
II. ReSouRce AcceptABIlIty
The variety of resources and their physical and biochemical properties, including their defensive mechanisms, is too great in any ecosystem for any species to exploit all pos- sible resources. The particular physiological and behavioral adaptations of insects, which enable them to obtain sufficient nutrients and to avoid toxic or indigestible materials, determine their feeding preferences, i.e., which resources they can or will exploit. Insects
that are adapted to exploit particular resources often lose their ability to exploit others.
Even species that feed on a wide variety of resource types (e.g., host species) are limited in the range of resources they can exploit. For example, gypsy moths feed on a variety of plant species (representing many plant families) that share primarily phenolic defenses, whereas plants that utilize terpenoid or alkaloid defenses are not exploited (J. Miller and Hansen 1989).
Insects face an evolutionary choice between maximizing the efficiency with which they exploit a particular resource (specialists) or maximizing the range of resources that they are able to exploit (generalists). Specialists maximize the efficiency of exploiting a partic- ular host plant through specific detoxification enzymes or avoidance strategies, thus mini- mizing the effect of host constitutive and induced defenses, but in so doing, they sacrifice the ability to feed on other plant species, which use different defenses (Bowers and Put- tick 1988). By contrast, generalists maximize the range of resources that may be exploited through generalized detoxification or avoidance mechanisms, such as broad-spectrum mi- crosomal monooxygenases. This strategy sacrifices efficiency in exploiting any particular resource, because unique biochemicals reduce digestion or survival (Bowers and Puttick 1988). Plant compounds that provide effective defense against generalists may be large- ly ineffective against specialists, and may even be phagostimulants for adapted species (Shonle and Bergelson 2000). Tallamy et al. (1997) reported that cucurbitacins (bitter trit- erpenes characterizing the Cucurbitaceae) deter feeding and oviposition by non-adapted mandibulate insect herbivores, but stimulate feeding by haustellate insect herbivores.
Generalists may benefit from a mixed diet by optimizing nutrient balances or through dilution of any single host’s defensive compounds (Behmer 2009, Bernays et al. 1994), or by increasing their energetic efficiency on stressed hosts that have sacrificed production of defenses (Kessler et al. 2004). Kessler et al. (2004) demonstrated that when tobacco, N. attenuata, was transformed to silence its induced defense genes, it became suitable for new (non-adapted) herbivores, such as the western cucumber beetle, Diabrotica un- decimpunctata, that fed and reproduced successfully. Generalists may be favored over specialists when host plants are rare or occur inconsistently. Wiklund and Friberg (2009) reported that fitness of a generalist pierid butterfly, Anthocharis cardamines, was increased by its ability to reproduce on any of a variety of host species, each of which varied widely in abundance and suitability over time.
Some generalists, which occur over large geographic areas, may be more specialized at the local level. Parry and Goyer (2004) demonstrated that the forest tent caterpillar is a composite of regionally specialized populations rather than an extreme generalist. In a re- ciprocal transplant experiment, tent caterpillars from Louisiana and Michigan, U.S., and Manitoba, Canada, were reared on the variety of hosts exploited by northern and south- ern populations. Tent caterpillars from northern populations showed greatest growth and survival on trembling aspen, Populus tremuloides, and red oak, Quercus rubra, which are both northern host species, and poorest growth and survival on water tupelo, Nyssa aquatica, which is a southern host species. Tent caterpillars from southern populations showed greatest growth and survival on water tupelo and poorest growth and survival on sugar maple, Acer saccharum, a northern host species. Feeding preferences reflect resource quality, susceptibility and acceptability.
Resource quality, as described above, represents the net nutritional value of the re- source after deducting the energy and resources needed to detoxify or avoid defenses.
Some of the nutrients in any food that is acquired must be allocated to production of detox- ification enzymes, or to energy expended in searching for more suitable food. Although