The Ecology of Plants Jessica Gurevitch State University of New York at Stony Brook Samuel M. Scheiner Arizona State University Gordon A. Fox University of South Florida Sinauer Associates, Inc., Publishers Sunderland, Massachusetts U.S.A.CHAPTER Herbivory and Plant-Pathogen Interactions reen plants are the foundation of almost al terrestrial food webs. All animals (including humans, of course) ultimately depend on plants for their existence. Yet casual observation seems to reveal a green world teeming with (uneaten) plants. Why is the world so green? Conversely, what are the consequences of herbivory from the plants’ perspective? Herbivory is the consumption of all or part ofa living plant, Some ecolo- gists use the term “predation” when an herbivore eats and kills an indi ual. Seed predators, or granivores, are herbivores that consume seeds or grains, killing the individual within. Grazers are herbivores that eat grass and other ground-growing plants, while browsers eat leaves from trees or shrubs. Fru givores are herbivores that consume fruits, sometimes without damage to the seeds. Plants are consumed by organisms from a variety of kingdoms: animals, fungi, bacteria, and even other plants. Herbivory can have ecological effects at the level of the individual plant, the population, the community and landscape (eg, patterns of coexistence of plant species), and the ecosystem (eg, nutri- ent cycling). Herbivores can also influence the evolution of plants. Ecologi- cally and evolutionarily, some of the most important herbivores are grazing, ‘mammals and insects. However, other types of herbivores, such as birds, mol- Jusks, and nematodes, can be very important in particular systems. In this chapter we look at the consequences of herbivory and of plant-her- bivore interactions for plant population dynamics and for the structure of plant communities. Can herbivory affect the trajectory of a plant population? Can the presence of herbivores—or their exclusion—have a decisive role in deter- mining what plant species are able to coexist in a community? We begin by examining the effects of herbivory on individuals, populations, and commu- nities, and briefly touch on its landscape effects. In addition, we look at how plants defend themselves against herbivores and at how those defenses evolved. Then we turn to the evolutionary consequences of herbivory and the responses of plants. Finally, we take a look at what is known about the role of plant-pathogen interactions in plant communities. Herbivory at the Level of Individuals What do herbivores do to individual plants? Herbivores can consume an entire plant, causing the death of the individual. By eating seeds, granivores—such as some ants, rodents, and birds—kill individual plants. Alternatively, herbi-214 Chapter 11 vores can eat (or manipulate) only some of the parts of plant, damaging, removing, or destroying those parts, but not necessarily immediately killing the plant. Deer, for example, frequently restrict their consumption to only the newest leaves or parts of shoots. Herbivores can also live on or within a plant and consume some of the plant's resources. Insects called aphids, for example, extract dissolved sugars and other nutrients directly from the phloem; pathogenic microorganisms can par- asitize a plant, depleting its resources over time. Some plants are parasites on other plants, tapping water, sug- ars, proteins, and other resources for their own use. The effects of herbivory on the plant depend, among other things, on what parts of the plant are consumed. Removal of or damage to roots can reduce or prevent the plant’s uptake of water and mineral nutrients and can make the plant more vulnerable to being toppled by wind, flooding, or soil erosion. Consumption of leaves reduces the photosynthetic surface area, while removal of phloem sap may reduce the energy and materials available for growth and reproduction. Consumption of leaves, stems, and twigs may alter the competitive rela- tionships among neighboring plants. Removal of meris- tems may alter the growth form of the plant. Consump- tion of flowers, fruits, and seeds may reduce the potential contribution of the plant to the next generation. OF course, to the new individual in each seed, con- sumption means death. Alternatively, fruits are often consumed without damage to the seeds, in which case the frugivore may disperse the seeds to potentially favorable locations. ‘Also important isthe life history stage at which the plant is attacked or damaged. Seedlings are particularly vulnerable to herbivores. One mouthful for the herbivore can killa seedling, but hardly affect a more mature plant. Grazing on grasses that have just begun to flower can critically affect their ability to produce seeds, whereas heavier grazing after seeds have been shed may have less of an effect on population dynamics, How much do herbivores eat, on average? It has been estimated that about 10% of the leaves of forest tees are lost every year to herbivores (Coley and Barone 1996). Herbivory is greatest in dry tropical forests, some- ‘what less in tropical rainforests, and least in temperate forests. Young leaves tend to be eaten more readily than ‘mature leaves in the Tropics. As you might expect, there is tremendous variation among species, locations, and. years in the degree of damage caused by herbivores. Can herbivory ever actually help plants to grow or reproduce? In the 1980s and early 1990s, a group of scientists postulated the existence of overcompensation, in which plants purportedly respond to herbivory by ‘growing more (McNaughton 1983). ([Fa plant's hypoth- sized extra growth in response to herbivory resulted in no net difference between grazed and ungrazed individuals, it would be called compensation.) The researchers suggested that overcompensation was due to the coevolution of plants and herbivores, particular- ly grasses and mammalian herbivores. Buffalo saliva and urine, for example, were thought to contain growth stimulants for grasses (Detling et al. 1980). These ideas were highly controversial and received a great deal of attention; it seemed difficult for many ecologists to believe that being eaten could actually be a good thing for plants. "While there was some experimental evidence for these ideas, when taken as a whole, they were not well supported by the available data (Belsky 1986; Belsky et al. 1993). One of the possible explanations for reported overcompensation was that researchers had measured only aboveground plant parts, while underground reserves may have been depleted to stimulate the observed aboveground growth. Long-term herbivory might result in significant depletion of these under- ground reserves, ultimately harming the ability of the plant to recover from subsequent bouts of herbivory. Another explanation was that in dense plant stands, if herbivores eat only shaded, unproductive understory leaves, there may indeed be no reduction in the photo- synthetic capacity of the plant, and thus no negative effects of herbivory. Joy Belsky and her associates (Bel- sky etal. 1993) argued that overcompensation seemed to occur mainly in the experimental treatments most favorable to plant growth, such as the combination of high nutrient availability and reduced competition. Fcologists now generally believe that herbivory usu ally harms individual plants (Hawkes and Sullivan 2001) While there are cases of overcompensation, on average, herbivory reduces both growth and reproductive output, resource addition increases both, and there isno evidence ‘of an interaction between the two effects Figure 11.1).One interesting observation is that plant responses to her- bivory depend to some degree on their growth forms and phylogenies: for example, monocots with basal meristems (euch as grasses) have more regrowth after herbivory under high-resource conditions, while dicot herbs and woody plants regrow more after herbivory under low- resource conditions. Unfortunately, most studies have used only short-term measures of plant response to her- ivory, not measures of lifetime fitness. Herbivory and Plant Populations ‘The extent to which herbivores affect plant population dynamics is a highly controversial and unresolved question. Two different explanations have been put forward for why herbivores should not be important. ‘The “top-down” school of thought argues that herbi- vores are maintained at such low densities by their own predators that they rarely exert negative effects onand therefore untestable, Paul Ehrlich and L. Charles Birch (1967) argued that the HSS hypotheses hin upon the belief that because species persist, they mus be regulated at or near an equilibrium—a “balance of nature.” Ehrlich and Birch argued that, to the contrary most popstlations experience so much random envi ronmental variation that their populations undergo large fluctuations, Much research and debate has been stimulated by these ideas in the ensuing decades; to do justice to this body of work would require another book! Confining ourselves to the question with which we began, how- ever—whether herbivory is important for plant com- munities—the answer seems clear. In a recent meta- analysis of a large number of experimental studies on herbivory, David Bigger and Michelle Marvier (1998) concluded that on average, herbivory causes a substan- tial reduction in plant biomass in natural plant com- munities. Contrary to the assumptions of many ecolo- gists, invertebrates such as insects have a much greater effect than vertebrates Plant Defenses Because plants cannot move, most plants have to “sit there and take it” from herbivores, resulting in natural selection for plants that are tougher, less palatable, and Figure 119 A) A haze of trich enrod, Asteraceae). (Photograph © J. Wexler Visus es covers the stem and leaves of this Solidago pl limited.) (B) Scanning ant-Pathogen Interactions 223 erally better defended. There are a great many dif- ferent types of plant defenses, just as there are a great many different ways that plants can be attacked. Plant Physical Defenses Physical and mechanical defenses include obvious struc- tures such as thos and spines, which probably serve best to discourage mammalian browsers and birds, but do little to deter insects. Single-celled plant hait chomes—serve many functions, including protection (Figure 11.94). Insects are deterred by leaf hairiness and can be impaled by some trichomes. Other trichomes secrete noxious compounds that can deter vertebrates or sticky substances that impede insects. Species of Urtica (stinging nettles, Urticaceae) have brittle, elongate tri- ‘chomes that break off when brushed, leaving a pointed fragment that pierces the skin and injects a painfully irsi- tating fluid (Figure 11.98). Trichomes of this kind have ® (gold. electron micrograph of trichomes on the leaf of Urtica dioica (stinging nettle, Urticaceae). The long trichome in the center isa stinging hair; when its tip s broken off by some action (such as the touch of a human hand or foo, it easily penetrates the skin and injects the neurotransmitter acetylcholine along with histamines that produce allergic reactions. The shorter trichomes do not have hese protective chemicals. (Photograph © A. Syred/Science Photo Library.)224 Chapter 11 evolved independently in four plant families: Urticaceae, Euphorbiaceae, Loasaceae, and Boraginaceze. Other physical defenses include materials that make entry into the plant difficult, such as thick bark on trunks fr roots or the tough coats that protect seeds and some fruits, such as nuts. Various cells and tissues that make up the plant body probably also serve a defensive func- tion, such as the cap of thick selerenchyma (a plant tis- sue with lignified cell walls) that surrounds the vascu- Jar bundles carrying food and water in young stems, and the waxy cuticle on the surface of leaves, which both reduces water loss and protects against fungal attack. Leaf toughness presents an important mechanical barrier to chewing insects, mammalian browsers, and other herbivores, as well as interfering with tissue digestibility. Toughness is primarily a result of the con- tent, type, and placement of sclerenchyma fibers (elon- gated cells), although short sclereids (sclerenchyma cells with thick, lignified, much-pitted walls) may be responsible for the hardiness of some structures. Tough- ness may also depend on other thick-walled cell types, such as xylem and collenchyma (supporting tissue made of living elongated cells with walls thickened irregularly) Newly formed, expanding leaves are con- siderably less tough than older ones because the struc- tural tissues involved in making leaves tough interfer® with the leaf’s ability to grow to full mature size. For this reason, new leaves are vulnerable to attack, and ‘must depend on alternative means of protection, such, as chemical defenses. Grasses are unpalatable to most generalist herbi- ‘vores because they sequester large amounts of silica, which makes them difficult to chew and difficult to digest (picture chomping down on ground-up bits of “ “Ambient COs Flevated COz Figure 11.10 ‘An experiment comparing the per- formance of leaf miners of the genera Stigmella, Cameraria, and Stilbosis on (Querous murifota(rayrtle oak, Fagaceae) grown under ambient ver- sus entiched CO, conditions. (A) The insects removed more tissue from the plans grown under enriched CO, conditions (error bars are+ 1 standard error) (B) Nevertheless, the insects feeding on those plans had higher mortality rates from a number of causes (error bars are + 1 standard deviation). (After Stiing et al. 1999) © SigmetlaComeraria Stone Leaf miner species glass, or large amounts of sand). The silica is contained in specialized epidermal cells as well as in other plant parts. A few other taxa also use silica as a deterrent tp herbivory, including horsetails (Equisetaceae) and palms, (Arecaceae). ‘Another characteristic that has been suggested to act as a defense against herbivores is the nutritional qual ty of leaf tissue. Eatirig leaves with a low nitrogen and water content results in poor growth and survival for herbivores. Leaves with a higher nitrogen and water content are usually preferred by herbivores over those with a lower content (all else being equal). This fact poses an evolutionary dilemma for plants: The metabolic enzymes responsible for growth and photosynthesis contain nitrogen. Reducing the concentration of nitro- xgen-containing compounds in the leaves to discourage herbivory could result in reduced photosynthesis and ‘Nitrogen is almost always limiting to plants, but it is even more limiting to herbivores. Animals need much ‘more nitrogen to function than plants do, and they main- tain much higher concentrations of nitrogen in their bod- ies, Plants growing in soils with low available nitrogen may have higher concentrations of carbon relative to nitrogen in their tissues, reducing their nutritive value. ‘As global CO, concentrations increase because of human ‘emissions (see Chapter 22), many herbivore populations are expected to decline because of the relative decline in carbon to nitrogen ratios, and thus the nutritional value, of many plants. Recent studies by Peter Stiling and his associates (Stling etal. 1999) found that several species of leaf miners consumed more tissue from oak leaves grown under enriched CO; conditions, but nevertheless hhad greater rates of mortality (Figure 11.10). © 5 © Died in mine Preyed on Parastized FateF a more Plant Secondary Chemistry Plants can marshal an arsenal of chemical weapons in response to herbivores. Chemical ecologists distinguish between primary metabolites and secondary chemicals, (or secondary metabolites). Primary metabolites (such ‘as sugars, amino acids, and DNA) are compounds nec- essary for the basic functioning of the plant. Secondary chemicals constitute a broad group of compounds that servea wide variety of purposes, including defense and. attraction of pollinators, rather than primary functions such as photosynthesis, Secondary compounds are gen- erally found only in particular species or groups of species, and often only in specific organs or tissues. The distinction between primary and secondary compounds is somewhat arbitrary; some primary metabolites, for example, are also used in defense. The terms are to some extent a holdover from an earlier era when secondary chemicals were believed to be waste products by plant physiologists and biochemists unaware of their ecolog- ical functions. ‘The three major categories of defensive secondary chemicals are phenolics, alkaloids, and terpenes. These Herbivory and Plant-Pathogen Interactions 225 contain subunits of more than one type. Additional defensive compounds include toxic proteins and amino acids, protease inhibitors, and cyanogenic compounds. Phenolics include a large variety of chemicals consisting, of an aromatic ring with an attached hydroxyl group, —OH (Figure 11.11). Probably the most important defen- sive phenolic compounds in angiosperms and gym- nosperms are the tannins, which reduce the digestibili- ty of plant tissues. They are packaged in cell vacuoles and are present in high concentrations in the leaves of many woody plants, such as those in the Fabaceae, Fagaceae, Myrtaceae, and Polygonaceae. Another impor- tant group of phenolic compounds are the lignins, which impregnate secondary (woody) cell walls, giving, them structural strength as well as providing a barrier to attack by herbivores and pathogens. Other pheno- lics include poisonous saponins as well as the flavonoids and anthocyanins, pigments that also give flowers and. fruits their colors. seni a a te 3 - ao" 2 HQ HO COOH ie wea BES So me (Galore) (Catechin) R Flavan-3-ol oud Ry=R; = OH: Luteotin Re Ho. Figure 1.11 oe Defensive phenolic compounds ‘include tannins and flavonoids. Examples of specific ones are given below the structures. (After Larcher 1995.) ae! Eee) "ox chon Ba Dees Basic skeleton:226 Chapter 11 The alkaloids are another broad group of com: pounds and include many that are used as pharmaceu- ticals Figure 11.12). Some 10,000 alkaloids have been iso- lated and their structures analyzed. Alkaloids are relatively small molecules that contain nitrogen. They havea bitter taste, and many are toxic to herbivores. Alkaloids are generally highly specific to the plant species or group of species in which they are found. ‘They are often effective in small quantities—as in the cases of cocaine, nicotine, and caffeine—although in ‘seme cases particular plants or plant parts produce high concentrations of alkaloids. ‘Terpenes are found in all plants, and an individual may contain many different terpenes (Figure 11.13). Ter- penes play a wide variety of functional roles in plants. ‘They are composed of multiple units of the hydrocarbon isoprene (C,H,), and they may be large or small, depending on how many isoprene units they contain ‘The compound isoprene is emitted in sunlight by the leaves of some plant species (such as Eucalyptus), some times in large quantities, and may serve to protect them from heat damage. The latex (milky sap) found in mem- bers ofthe spurge (Euphorbiaceae) and dogbane (Apoc- ynaceae) families contains toxic terpenoid defenses. The oils that are responsible for the characteristic flavors and. scents ofthe mints (Lamiaceae) are terpenes; they deter herbivores and reduce the growth of bacteria and fungi. Cardiac glycosides can cause heart damage in verte- brates and are poisonous to many insects; these terpenes are produced by many unrelated plant species, includ- ing species in the Scrophulariaceae (such as Digitalis, fox- ¢gloves) and the Apocynaceae (such as Asclepias, milk- ‘weeds, and Apocymuun, dogbanes). Phytoecdysones are terpenes that mimic insect molting hormones and alter the development of insect larvae; they are produced by a number of taxa, including ferns, cycads, and some angiosperms. Plants in the Bravsicaceae (mustard family) contain characteristic secondary compounds, chiefly gluco: lates, which are effective repellents of most generalist herbivores, These mustard-oil precursors (responsible for the “bite” of the many cultivated species in this fam- ily) almost completely deter mammals and nonadapted insects. Glucosinolates are not as effective against insect herbivores that have become specialized on this family, however, and only high concentrations can offer some protection. Louda and her colleagues have found that glucosinolates not only can affect the interaction between plants and their herbivores, but also can indi- rectly control the distribution and abundance of the plants (Louda and Rodman 1996). Cardamine cordifolia (Gittereress, Brassicaceae) is native to the Rocky Moun- tains of North America, where it grows only in shaded, moist forest edges. Bittercress may be restricted from sunny sites by greater vulnerability to chronic herbivory Morphine ° La onc Oo ree YO Gy Cocaine © Figure 11.12 Examples of various alkaloids used by humans and their sources. (A) Opium comes from the milky sap of seed pods of Papacer soieram (opium poppy, Papaveraceae), mor phine is deived from opium. (Photograph © R. Shiel/ An nals Animals) (B) Cocaine is found inthe leaves of Ery fhroxylum cace (coca, ErySoxylacee). (Photograph © G Dimijian/Photo Researchers Inc) (C) Caffeine is found in coffee “beans,” the fruits of Coffer (coffe, Rubiaceae), as well fs many other plans. (Photograph © K. Fink/ Photo Researchers, Inc) (D) Nicotine comes from the leaves of ‘Nicotiana tabacum (tobacco, Solanaceae) and other species in this genus. (Photograph © G. Grant/Photo Researchers Inc)HAC cH OHO HC ‘COOH wt oi Ho cay on HoRGC Scr Ro (A) Seponin (Hederagenin) ole Row on Ho, Gue ah Gite Gah Ho™ yl (© Steroid alkaloid (Demissin) at those sites. Ina series of experiments, plants grown in sunny sites experienced greater water stress, which reduced their glucosinolate concentrations. Conse~ quently, those plants had more insect herbivores and suf fered greater damage. Thus, herbivory may be control- ling the plant's distribution by causing differential damage among microhabitats. Constitutive versus Induced Defenses Constitutive defenses are those that are present in a plant regardless of herbivore damage. They may be pres- ent throughout the life of a plant or may change as an individual grows and matures. Induced responses are elicited by an attack by herbivores. If these responses serve to protect the plant (whether or not they harm the herbivore), they are called induced defenses; if they have a negative effect on the attacking herbivores (whether or not they defend the plant), they are termed induced resistances (Karban and Baldwin 1997). Both physical and biochemical defenses can be induced. ‘There has been a great deal of discussion regarding the relative benefits of constitutive versus induced defens- es (Karban and Baldwin 1997; Lerdau and Gershenzon 1097). Historically, most research on plant responses to herbivory has focused on constitutive defenses, but beginning in the 1980s, interest in induced defenses ‘reatly increased. Acdassic example of an induced defense is the pro- duction of cyanogenic glycosides by Trifolium repens (white clover, Fabaceae) when the leaves are damaged; this response can also be induced by frost damage. Cyanogenic glycosides are an effective deterrent against snails, which are voracious herbivores in regions with, mild, wet winters, The ability to produce these defensive compounds is controlled by two genes, and populations of white clover in Great Britain are polymorphic for the oH (@) Cardenolide (Calotropagenin) Herbivory and Plant-Pathogen Interactions 227 Figure 11.13 0 Trlterpenes play important roles in plant defenses against herbivores. (A) Hedera helix English ivy, Araliaceae) contains the saponin hederagerin ints leaves and fruits (@) Ascepis (milkweeds, Apocynaceae) con- tains the cardiac plycoside calotropagenin {© Selon denies (potato, Solanacese) produces the steroid alkaloid emissin nits leaves. (D) Arnumber of taxa produce ecdysones, compounds tha mimic insect hormones and interfere with insect meta- ‘morphosis. (After Larcher 1995.) {(D) Sterol (Eedysone) trait (Dirzo and Harper 1982a,b). In the presence of snails, plants that had the ability to produce cyanogenic glyco- sides had higher survival rates than those that did not. In the absence of snails, however, those plants had reduced growth and reproduction compared with acyanogenic (non-cyanide-producing) plants. As pre- dicted by evolutionary theory, areas with high snail den- sities had a higher proportion of cyanogenic individuals, while those with fewer herbivores had a preponderance of acyanogenic individuals. Ina study of Lepidium virginianum (pepper-grass, Brassicaceae), Anurag Agrawal (2000) showed that induced defenses were effective against a generalist her- bivore, but not against an herbivore that specialized on that plant (Figure 11.14). Whether provided with a choice of food items or not, generalist caterpillars did substan- tially more damage to plants whose defenses had not been induced. However, specialist caterpillars produced the same amount of damage in induced and noninduced plants. ‘The production of sap by conifers in response to attack by bark beetles is an example of an induced response. In addition to the physical defense of the sticky, suffocating pitch, this response has a biochemical function. Coniferous sap contains substantial amounts of various terpenes (including monoterpenes and sesquiterpenes) as well as phenolics and other com- pounds that are particularly toxic to the symbiotic fungi that attack with their beetle partners. Raffa and col- leagues (1987, 1991) found that wounding a tree mechan- ically caused the production of small amounts of monoterpenes, while trees attacked with living fungi responded with massive amounts of monoterpenes. ‘Trees responded mast strongly to those particular fungi associated with the bark beetle species that ordinarily attack them.228 Chapter 11 Generalist B Control | z l Saeco e ‘Choice "No choice a ee 3 &E £2 x ae E 1 . ‘Choke "No choice Induced defenses in Leitivom virginionum (pepper-grass, Brassicaceae) were effective against caterpillars of (A) gener- alist herbivores (noctuid moths), but not against caterpillars @ ofthe specialist butterfly Pieris rapa (Pieridae). The defenses were induced by allowing a fixed number of larvae to feed on plants several days before the experimental trials. ‘The amount of damage was the same whether or not the ‘caterpillars had a choice of induced or noninduced plants. Error bars are + 1 standard error, (After Agrawal 2000.) ‘Many different classes of chemicals have been found, to act as induced defenses. lan Baldwin and his associ- ates (Baldwin 1988, 1991; Karban and Baldwin 1997) hhave done much to increase our understanding of the [process of induction. In some cases chemical defenses are induced by highly specific cues, such as caterpillar saliva. In other cases any mechanical damage is suffi- cient. The toxic alkaloid nicotine, for example, is pro- duced in response to mechanical damage to the leaves of Nicotiana (tobacco, Solanaceae). Nicotine mimics the neurotransmitter acetylcholine, blocking the receptor for that neurotransmitter in both insects and vertebrates. At low doses, nicotine stimulates the nervous system, but at higher doses it is a depressant, ultimately capable of causing paralysis and death. Smokers become addict- ced to the excitatory effects of low doses of this potent nerve toxin; caterpillars or grazing mammals may be paralyzed and killed by the same toxin. Small amounts of nicotine are present in undamaged tobacco plants. After leaf damage occurs, nicotine levels can increase four to ten times, reaching concentrations that ean kill an herbivore after a single meal. [Nicotine is synthesized in the roots of tobacco plants, in response to leaf damage and transported via the xylem to the leaves. The signal indicating that the leaves have been damaged, cuing the roots to produce nicotine, is apparently a hormone transported from leaves to 00s through the phloem. Greater leaf damage induces more nicotine production. Because of the long-distance nature of this response and the time needed for biosynthesis, it takes from ten hours to several days for nicotine to increase to its maximum levels. What might be the advantages of belowground synthesis of nicotine, given the vulnerability of the plant during this prolonged wait? If nicotine production occurred largely in the shoots, herbivores consuming the leaves might severe- ly compromise the plant’s ability to defend itself from further attack. The site of the synthesis of this defen- sive compound is thus protected belowground from leaf-eating herbivores. Even if large amounts of the aboveground parts of the plant are damaged or destroyed, the plant can continue to defend itself ‘Some plants increase physical defenses in response to herbivory. Grass plants can increase their silica con- centrations. Opuntia (Cactaceae) can secrete additional mucilage and fiber, effectively walling off insect herbi- Evolutionary Consequences of Plant-Herbivore Interactions ‘You may be wondering, after reading about the arsenal of defenses possessed by plants, why all herbivores do not simply starve to death. There are several reasons that they do not. First, fortunately for the rest ofthe bio- logical world, plants are not uniformly well defended. Species, populations, individuals, life stages, and plant parts vary in their defenses, and the less well defend- ed are often sought out and depended on for food. Sec- ond, some herbivores have evolved various means of overcoming plant defenses, leading to what has been termed a “coevolutionary arms race” between con- sumed plants and their would-be consumers. The behavior, biochemistry, and morphology of herbivores may contribute to their ability to overcome plant defens- es. Some insects, for example, are capable of carefully avoiding the ducts and canals that carry toxic latex in certain plants. Even more surprisingly, some chrysomelid beetles neatly incise these canals “upstream” before beginning to feed, preventing the flow of the toxic compounds to the leaf they are con- suming. Giraffes, with their tough mouths and phe- nomenally long tongues, freely browse Acacia - (Fabaceae) in the African savanna, undeterred by the plants’ abundant and sharp thorns (Figure 11.15)-In the