II.10

Interactions between Plants and Herbivores

Rebecca J. Morris

OUTLINE

1. The diversity of herbivores

2. Herbivore–plant population dynamics

3. The impact of herbivores on plant populations

4. Plant responses to herbivory

5. The impact of plants on herbivore populations

6. Herbivore–plant interactions at the community level

Herbivores are animals that feed on living plants. Herbivory is one of most common ecological interactions and is exhibited by species ranging from microscopic mites to giant pandas. Herbivore–plant interactions have features in common with all other consumer–resource interactions, although there are significant differences. Notably, plants do not necessarily die when they have been attacked by herbivores. Although there is no compulsory link between herbivore and plant dynamics, herbivores can affect the population dynamics of the plants on which they feed, and plants can affect herbivore population dynamics. Herbivore–plant interactions have been studied through a combination of observational time series data, mathematical modeling, and experimentation, and here a variety of examples are discussed.

GLOSSARY

functional response. Results from switching behavior when the herbivore alters the composition of its diet as a result of short-term changes in relative food availability

herbivore. An animal that feeds solely on living plant tissue

herbivory. The consumption of living plant material

host plant. The plant on which an insect herbivore feeds

numerical response. Acts by dispersal with mobile herbivores aggregating in regions of high food availability, or in the longer term by increasing reproductive success

population cycles. Changes in the numbers of individuals in a population repeatedly oscillating between periods of high and low density

population dynamics. The variation in time and space in the size and density of a population

resource. An environmental factor that is directly used by an organism and that potentially influences individual fitness; plants are a resource for herbivores

1. THE DIVERSITY OF HERBIVORES

Herbivores are animals that feed solely on living plant material. They are taxonomically and ecologically diverse and range from single-celled zooplankton to wildebeest, and from leaf-mining moths to marine iguanas. They can be found in terrestrial, marine, and freshwater ecosystems. Insects and mammals are the most well-known groups of herbivores and have been studied most intensively, but there are many other types of herbivore including some species of birds, fish, reptiles, crustaceans, and molluscs.

Herbivores can feed on all the different types of living plant tissue including leaves, fruits, pollen, flowers, and seeds. Each herbivore, however, tends to specialize on a particular type of plant tissue. Herbivores exhibit a variety of feeding methods including chewing, sucking, boring, and galling. Folivores, which feed on leaves, are some of the most common herbivores and include mammals such as deer and insects such as grasshoppers. Frugivores are fruit eaters ranging from monkeys to wasps; and granivores are the seed eaters, or seed predators, including squirrels and weevils. Herbivores remove approximately 10% of net primary production, at least in terrestrial ecosystems. Herbivory typically does not kill the plant but influences the fitness of plants by reducing growth and reproduction and potentially increasing mortality. However, seed eaters (and some species that feed on seedlings) do have a significant effect on seed abundance and can directly influence plant populations, assuming that the plant is seed limited.

There is such a wide variety of herbivore species that it is useful to consider the differences between the two most studied groups of herbivores: insects and mammals. Insect herbivores differ from mammalian herbivores in their size, metabolic rate, population density, numbers, and the kinds of damage they cause. Insects tend to be more specialized than mammalian herbivores and are more likely to have an intimate lifelong association with their host plant. Mammalian herbivores are likely to have a more immediate and, in the long term, more profound impact on plant populations than invertebrates because of their greater body size, polyphagy, individual bite size, mobility, and tolerance of starvation. A relatively high proportion of mammalian herbivore populations are food limited, whereas a comparatively high proportion of insect herbivore populations are regulated by predators, parasites, and diseases.

2. HERBIVORE–PLANT POPULATION DYNAMICS

The population dynamics of herbivore–plant systems shares features exhibited by all consumer–resource interactions. Consequently, the mathematical models for these systems have the same logical foundations, largely based on the Lotka-Volterra model and its variations. However, herbivore–plant systems differ from other consumer–resource relationships, for example, predator–prey relationships, in several important ways. Classifying consumer–resource relationships according to the closeness and duration of the relationship (intimacy) and the probability that the interactions will result in the death of the organism concerned (lethality) can highlight these differences. Many herbivores, in particular grazers, score low on both intimacy and lethality. Other herbivores, such as sapsuckers like aphids, are functional parasites and score highly on the intimacy scale but are still low on the lethality scale.

Herbivores rarely kill the individuals on which they feed and throughout their lives will eat parts of many individuals. Whereas predators tend to kill their prey, grazers consume only part of a resource individual, perhaps concentrating on the young leaves or the flowers. This has important implications for theoretical models. Simple theoretical models employ a logistic model for vegetative regrowth after herbivory, but in fact, a linear initial regrowth model is more apt because the plant biomass is not usually reduced to near zero. The primary productivity of plant communities is also an important factor affecting herbivore–vegetation dynamics. Models of plant growth include the logistic hyperbolic functional response, in which the dynamics of the system becomes increasingly less stable as plant standing biomass increases, and the globally stable regrowth–herbivory–regrowth model.

The functional response of herbivores is measured as units of plant biomass removed, whereas for predators, it is the number of individuals eaten. The functional response of herbivores is difficult to quantify because it must take into account the parts of the plant that are available to the herbivores as well as differences in nutritional quality. A hyperbolic functional response is used as a reasonable approximation of herbivore foraging, at least for grazers. The spatial immobility of plants also sets them apart from other resources such as prey. The spatial arrangements of plants can influence herbivores by affecting their average density of bites and average bite size.

Another factor that distinguishes herbivory from predation is that dynamic changes in resource quality are much more likely. Plant quality may change if herbivores consume the better-quality resources first, thus decreasing the average quality of the remaining vegetation; or it may alter if the plant increases defense of its remaining biomass or increases the amount of new biomass produced. Consequently, herbivory can modify the frequency distribution of plant qualities.

Despite the distinguishing features of herbivore–plant dynamics, there is no necessary link between fluctuations in the numbers of herbivores and the plants on which they feed. Fluctuations in plant populations may have nothing to do with herbivore feeding and may be caused by extrinsic factors such as the weather or by competitive interactions with neighboring plants. Similarly, herbivore numbers may be determined by natural enemies or shortage of breeding sites and may have nothing to do with plant numbers.

Stabilizing Influences

There is an inherent tendency for consumer–resource systems to oscillate. However, the dynamics of herbivore–plant interactions is stabilized by the fact that many plants are long-lived. Plants also have an absolute refuge, their below-ground biomass, which acts as a powerful stabilizing influence. Other mechanisms also stabilize the herbivore–plant systems, for example, the switching behavior of herbivores as well as mechanisms that depend on fluctuations in population densities and environmental factors in space and time. The latter can generate heterogeneity in the plant or herbivore distributions, resulting in refuges. Territorial behavior, for example, can have a stabilizing effect by affecting the population densities of the herbivores in space and thus providing refuges for the plant.

3. THE IMPACT OF HERBIVORES ON PLANT POPULATIONS

It is well accepted that herbivores can have a negative impact on the growth, reproductive output, and survival of many plant species. However a strong herbivore effect on plant performance does not automatically imply an equally intense effect on population dynamics. Even if herbivores reduce plant abundance, it does not necessarily follow that this mortality will lead to a reduction in the number of individuals in subsequent generations. For example, if herbivore-induced mortality of plants reduces plant density, the survival or fecundity of plants that escape herbivory may be increased as a result of reduced intraspecific competition. Herbivores can have a big effect on plant populations if they have no natural enemies or other limiting resources. The impact of herbivory on plant abundance is determined in part by whether or not the herbivore is capable of mounting a response to changes in plant abundance. The response must be large and rapid enough to check or reverse change in plant abundance and could be a functional or a numerical response. In the following sections, a mixture of observational data and experiments are described to explore the effects of herbivores on plant populations.

Insect Outbreaks: The Spruce Budworm

Insect herbivore outbreaks provide good evidence that herbivores can have a significant impact on plant population dynamics, with the best examples coming from forest pests. When herbivore and plant populations interact, there can be multiple equilibria or alternative stable states. These result in a large change in abundance (an outbreak and subsequent crash) as a result of either a small change in carrying capacity or a small environmental perturbation when carrying capacity is close to a critical threshold value. Eruptive or cyclic outbreak can be caused by a sudden increase in availability of high-quality food. For example, the Spruce budworm (Choristoneura fumerana) feeds on balsam fir and white spruce in Canada and experiences outbreaks correlated with a dramatic increase in the availability of high-quality food. In the 1940s, it killed these two tree species over an area of 52,000 km2 because of a dramatic increase in high-quality food, as a large area of balsam fir in monoculture matured at the same time.

The Release of Insect Biological Control Agents—Cactoblastis and the Prickly Pear Cacti

The best observational evidence of the impact of insect herbivores on plant populations comes from the release of specialist insect biological control agents against target weeds. Biological control involves the introduction of herbivores outside their native range, where they may be unrestricted by natural enemies and competitors, and consequently they may exhibit stronger effects than those on plant populations in natural systems. For example, the South American moth Cactoblastis cactorum was introduced to control the prickly pear cacti, Oputia inermis and O. stricta, in Australia. Before the introduction of the moth, vast areas of Australia were covered in the cacti, but the moth wiped out the cacti, which has since remained at low population levels and only in small patches. Several factors have affected the success of Cactoblastis in Australia. First, heterogeneity in the distribution of the herbivores may exert a regulatory effect. The refuge for the cacti lies in the highly aggregated egg-laying behavior of the moth. If larval densities are too high, the insects will not survive and will perish along with the plant itself, thus stabilizing the interaction at low densities. Second, cacti growing under conditions of water or nutrient stress have thick mucilaginous segments that suppress the development of the Cactoblastis population in certain areas, such as the coastal areas of Queensland. Finally, high temperatures are also regarded as being of major importance in reducing the fecundity of the summer generation of Cactoblastis. Cactoblastis is now considered a pest and a serious threat to the high diversity of Opuntia species, in particular in the southwestern United States and Mexico.

Experimental Exclusion of Insect Herbivores

The best experimental evidence for the effect of herbivory on plant dynamics comes from exclusion experiments using insecticides in natural communities. Using insecticides to remove insects has its drawbacks, but it is the best way of doing so experimentally. A classic experiment from the 1960s studied the effects of insect exclusion using insecticides on the hemiparasitic woodland herb cow wheat (Malampyrum lineare), demonstrating a dramatic increase in plant abundance. More recently, the exclusion of flower-head-feeding insects using insecticides on thistle (Cirsium canescens) in Nebraska has shown higher densities of seedlings on sprayed sites as well as higher mature plant densities in the next generation.

Introducing or Excluding Mammalian Herbivores

Observational evidence of mammalian herbivores affecting plant populations comes from the profound effect of introduced mammals on native vegetation, particularly on islands or island continents. Rabbits introduced to Australia have reduced plant biomass, altered plant communities, and caused erosion, and introduced red deer (Cervus elaphus) browsing native forests have had significant effects in New Zealand. Exclusion experiments for mammals are easier to carry out than those on insects and typically involve fencing plots. There have been many large herbivore exclusion experiments showing that the abundance of plants is affected by removing herbivores. For example, the gazelle (Gazella dorcas) feeds on the bulbs, leaves, and flowers of the desert lily (Pancratium sickenbergeri)in the Israeli desert. In exclosure experiments, fenced populations of lilies had twice as many plants as unfenced, showing a significant negative impact of herbivory on the plant population. More recently, an effect of ungulates on the population dynamics of two montane herbs has been demonstrated during a 7-year experiment in Spain. Spanish ibex (Capra pyrenaica) and domestic sheep (Ovis aires) had a significant negative effect on the population dynamics of Erysimum mediohispanicum and E. baeticum. As well as having a direct impact on the abundance of plants, mammalian herbivores [including black-tailed prairie dogs (Cynomys ludovicianus) in North American savanna and mule deer (Odocoileus hemionus) and elk (Cervus elaphus) in Arizona pine forest] can also prevent the transition from grassland to woodland.

4. PLANT RESPONSES TO HERBIVORY

The tolerance of plants to herbivory is a major plant strategy, much more important in herbivore–plant relationships than in other consumer–resource interactions. Plants can compensate for attack by herbivores by increasing the amount of new growth or by changing the effectiveness of existing plant parts. For example, partial defoliation may result in a more effective use of existing leaves rather than the production of new leaves. Plants also show a diversity of defensive adaptations against herbivores. These include physical barriers such as tough leaves, thorns, or hairs, and chemical defense (using secondary compounds such as cucurbatacin or nickel, which may be either deterrents or toxic compounds). Other organisms may also be used for defense; for example, in the ant–acacia mutualism, acacia ants (Pseudomyrmex ferruginea) actively defend the acacia plant from herbivores, and in return the ants are supplied with shelter in the form of modified thorns, protein-rich Beltian bodies, and carbohydrates from extrafloral nectaries. Some defenses are already present on the plant, and others are induced after the plant has been attacked. In turn, herbivores have found ways of circumventing plant defenses by changing their behavior through avoiding the plants or disabling the defense or by detoxifying or excreting the compounds. In the case of the chrysomelid beetle (Labidomera clivollis) feeding on milkweed (Asclepias syriaca), disabling the defense involves the beetle biting into lateral leaf veins near the midvein in order to cause the veins to leak before feeding.

5. THE IMPACT OF PLANTS ON HERBIVORE POPULATIONS

Plants are an important food resource for herbivores. In the 1960s, Hairston, Smith, and Slobodkin hypothesized that herbivores could not be resource limited because “the world is green,” and plants are obviously abundant and intact. Their hypothesis was, and still is, criticized because the world is not always green and because green plants are not necessarily edible or of high enough quality for the herbivores to eat. It is now widely accepted that variation in quantity and quality of plant resources can have a fundamental effect on herbivore abundance and population dynamics. Changes in plant quality caused by herbivore feeding also have the potential to feed back and influence subsequent growth, reproduction, and mortality of herbivores. However, predation and parasitism are still thought to be important regulatory agents as well, particularly for insect herbivores.

Tracking Host Plant Abundance: The Cinnabar Moth

The cinnabar moth (Tyria jacobaeae) and its food plant, tansy ragwort (Senecio jacobeae), studied by Dempster in the 1970s, provides a classic example of a food-limited specialist insect herbivore tracking the abundance of its host plant. Larval numbers are dependent on the amount of food present, but plant numbers are not determined by the amount of larval feeding. The moth larvae build up in numbers until the limit in its food supply is reached, and the population then crashes as a result of starvation. The plants survive defoliation, and may in fact multiply as a result, because damaged plants can produce new shoots during the year of defoliation, which develop and produce mature seeds later the same year. Consequently, after the crash, the population of the moth is allowed to rise once more. Increased larval dispersal as a result of food shortage also leads to overexploitation of the host resulting in large population fluctuations. Recent analyses of long-term data sets have shown that the cinnabar moth–ragwort system can show fundamentally different dynamics at two different sites. In coastal dunes in the Netherlands, the moth has a delayed density-dependent effect on plant growth rate, and ragwort has a positive effect on changes in moth density when at low density but inhibits moth population growth rate at high densities. In contrast, in southern England, there is no evidence for either a direct or delayed effect of the moth on ragwort, or vice versa. In southern England, direct density-dependent mortality and mass plant recruitment resulting from soil disturbance (by rabbits) stabilize the plant–herbivore system, but plant and herbivore dynamics are uncoupled. In the Netherlands, time delays, and in particular delayed density dependence acting on the insect, determine the plant–herbivore population dynamics. These differences are caused by differences in the importance of seed limitation in plant recruitment in the two locations.

Experimental Effects of Plant Quality and Quantity on Insect Populations

It is difficult to find experimental studies that focus solely on the bottom-up effects of the plant on insect herbivores and do not investigate the top-down effects of predators concurrently. However, the greater abundance of pests attacking agricultural crops, compared to wild vegetation, provides indirect evidence that an increase in plant quantity and quality (as a result of fertilization) leads to an increase in insect abundance. Manipulative experiments have been used to investigate the effect of plant resources on insect herbivore dynamics, with an assemblage of sap-feeding phytophagous insects inhabiting Atlantic-coast salt marshes. The population density of all herbivores increased when the quality of the plant (in this case leaf nitrogen content) was elevated. Other field experiments have found that the density of a native perennial shrub, Gossypium thurberi, in the United States had a significant effect on survival and cumulative numbers over several generations of its most abundant herbivore, the lepidopteran caterpillar, Bucculatrix thurberiella.

Food-Limited Mammalian Herbivores: Population Cycles of Arctic Lemmings

A strong positive correlation between the biomass of large vertebrate herbivores (migratory African buffalo and Serengeti wildebeest) in African grasslands and the rate of plant productivity suggests that the herbivores may be food limited. However, the current view is that rather than being regulated by resources in general, animal numbers are regulated in a density-dependent manner by the forage available in key resource areas, especially during droughts. Population cycles, where some species exhibit fairly regular density cycles, are often evident in mammalian herbivore populations, and there is evidence that some may be caused by herbivore–plant interactions. For example, arctic lemmings (Lemmis and Dicrostonyx spp.) show population cycles linked to their resources. There are many other examples of population cycles in mammals, but although food may be a contributory factor influencing the population cycling, there are other more important causative factors, including predators. For example, boreal forest populations of snowshoe hares (Lepus americanus) go through 10-year population cycles in Arctic Canada. Heavy exploitation during a period of high-quality food is followed by an extended period when the regrowth foliage is of low quality (because of increased levels of induced defenses). The rate of increase in herbivores becomes positive only once the plant quality recovers its initial high levels. The quality and quantity of food do have important effects on population dynamics but cannot explain cycles completely, and predation is thought to be the most important factor, as confirmed by an 8-year experiment involving the manipulation of supplemental food and predator abundance.

6. HERBIVORE–PLANT INTERACTIONS AT THE COMMUNITY LEVEL

Studying the causes of herbivore population dynamics is complex. It is often difficult to establish whether a species is tracking or causing the population changes in another species. Experiments are difficult to carry out because stopping the cycles is not always possible and because when one causative factor is removed, another may come into play. Searching for single-factor explanations for herbivore or plant dynamics will often be futile, not to mention extremely difficult to establish unambiguously. The huge range of explanations even among rodents, such as lemmings and voles, is testimony to the fact that no single mechanism can explain population cycles in every individual case, especially given that even the same species in different locations can exhibit very different dynamic patterns.

Although many examples of herbivores affecting plant populations, and vice versa, are described here, there are also many studies that have shown no effect at all, strongly suggesting that the effects are system specific. This chapter has focused on simple herbivore–plant interactions between two species, but plant and herbivore populations do not exist in isolation—they exist in multispecies systems. Although understanding the dynamics of herbivore–plant interactions is essential, it is crucial to remember that other factors may also play an important role in the community dynamics. Direct interactions with predators, competitors (plants or herbivores), and the environment, as well as a multitude of indirect interactions, can all affect the interactions between plant and herbivore species.

FURTHER READING

Crawley, Michael J. 1983. Herbivory. The Dynamics of Animal–Plant Interactions. Oxford: Blackwell Scientific Publications. A comprehensive overview of herbivore–plant interactions research up to the early 1980s, but considerable progress has been made since.

Crawley, Michael J. 1997. Herbivore–plant dynamics. In M. J. Crawley, ed. Plant Ecology, 2nd ed. Oxford: Blackwell Science, 401–474. An invaluable and comprehensive review of interactions between herbivores and plants.

Danell, Kjell, Patrick Duncan, Roger Bergstrom, and John Pastor, eds. 2006. Large Herbivore Ecology, Ecosystem Dynamics and Conservation. Cambridge: Cambridge University Press. A recently published book with chapters addressing the impact of large herbivores on plant population dynamics and vice versa.

Herrera, Carlos M., and Olle Pellmyr, eds. 2002. Plant–Animal Interactions. An Evolutionary Approach. Oxford: Blackwell Scientific Publications. Provides a broad and basic background to insect and mammalian herbivory and granivory, including small sections on population dynamics.

Maron, John L., and Elizabeth Crone. 2006. Herbivory: Effects on plant abundance, distribution and population growth. Proceedings of the Royal Society London Series B 273: 2575–2584. A recent review focusing on when and where herbivores have the greatest effect on plant populations and suggesting future directions for the field of herbivore–plant dynamics.

Turchin, Peter. 2003. Complex Population Dynamics: A Theoretical/Empirical Synthesis. Princeton, NJ: Princeton University Press. A detailed synthesis of principles and models of population dynamics considering herbivore–plant interactions as a special case.