2. Mechanisms causing nontrophic effects
3. The nature of indirect effects in communities
4. The nature of indirect effects in ecosystems
5. Direct and indirect effects in context
Species in ecological communities interact directly with another species through consumer-resource, competitive, or mutualistic interactions. Whenever three or more species are engaged in such interactions, we see the emergence of indirect effects in which one species affects another through a shared, intermediary species. Indirect effects can reinforce or counter direct effects and lead to interesting emergent properties. This chapter explores some of the myriad ways that indirect effects emerge in communities and ecosystems. Through the use of selected examples, it shows why consideration of indirect effects is critical to a complete understanding of species interactions in ecological systems.
direct effect. The immediate impact of one species on another’s chance of survival and reproduction through a physical interaction such as predation or interference
food chain. A descriptor of an ecological system in terms of the feeding linkages and energy and materials flows among major groups of species (plants, herbivores, decomposers, carnivores)
indirect effect. The impact of one species on another’s chance of survival and reproduction mediated through direct interactions with a mutual third-party species
nontrophic interaction. A direct interaction that changes the behavior, morphology, or chemical composition of a species in response to the threat of being consumed
trophic interaction. A direct interaction involving the consumption of a resource species by a consumer species
Imagine a herd of wildebeest grazing on a Serengeti plain. Imagine now that a prowling lion—a threat to their life—comes into their vicinity. This causes them all to stop feeding and look up in vigilance to see what the approaching predator will do. The wildebeest are nervous and tense, ready to flee at any sign of attack. Yet they are reluctant to flee because that would mean giving up feeding in a highly nutritious patch of forage, one of a few such high-quality patches currently available within a vast landscape. The resources in the patch are especially favored because they will enable the wildebeest to maximize their resource intake for growth, survival, and reproduction. The wildebeest face a critical choice: do they flee from the predator and give up the valuable food resource or do they stay and risk being captured? This choice is faced by individuals of every species of animal during the course of their daily existence. Nevertheless, the fear factor motivating this choice surely must be short-lived. After all, things will go back to normal once the predator has left or it has subdued the one victim out of the many comprising the herd, right? But the reality is, “No, not exactly.”
The critical question here is: What is considered “normal”? Often the presumption is that once the predator has left, the threat disappears, and animals can resume feeding on their resources with little worry. But ecological science has revealed that this tends not to be the normal case. Instead, many individuals live in a chronic state of vigilance brought about by the fear of being captured. Ambush predators can lie in wait for long periods of time. Individuals that let down their guard and move within the vicinity of any predator lying in wait have a high likelihood of being the predator’s next victim. Prowling predators can hunt in groups, so foraging individuals that do not regularly scan their surrounding environment may find themselves trapped. Normal, in many cases, means living in situations that pose continuous risks of being a predator’s next victim.
The above vignette of wildebeest daily life on the plain encapsulates several key ecological concepts. First, because wildebeest are both consumers of their plant resources and at the same time resources for other consumers—their predators—they are inherently part of an ecological food chain. Their role in that food chain is identified by the kind of consumptive interaction, or technically trophic interaction, in which they are engaged. Because they eat plants, they belong to the herbivore trophic group. Their predators, because they eat herbivore prey, belong to the carnivore trophic group. By extension, species that consume mineralized nutrients and CO2 in order to photosynthesize carbohydrates belong to the plant trophic group. To the victim (i.e., plants fed on by herbivore; herbivores fed on by carnivores), these direct trophic interactions are detrimental because it means loss of tissue or life.
But fascinating things happen when one considers how the trophic interactions play themselves out along the full length of the food chain. For instance, carnivores can lower the population abundance of herbivores through direct trophic interactions. This in turn means that there are fewer herbivores feeding on plants than in cases where carnivores are absent. Fewer herbivores mean more plants. Thus, by feeding on herbivores, carnivores provide a benefit to the plants. It is, however, an indirect benefit because carnivores do not interact directly with the plants. Rather, their effect is mediated by changes in herbivore abundance. Ecologists call this an indirect effect. Indirect effects emerge in all ecological systems whenever three or more species or trophic groups interact.
In the example of the Serengeti plain, we also see two different mechanisms causing the indirect effect of predators on plants. By engaging in a direct trophic or consumptive interaction with wildebeest, lions lower the numbers of wildebeest feeding on the plants. By scaring the wildebeest away from the resource, they also lower the number of wildebeest feeding on the plants. Moreover, by posing a constant threat that causes wildebeest to remain vigilant, they alter the rate at which wildebeest consume plants. This latter interaction between lion and wildebeest is called a nontrophic or nonconsumptive effect. Counterintutitively, by scaring prey within any given time period, predators can have a greater beneficial effect on plants through nontrophic interactions with their prey than through trophic interactions. By scaring all individuals within a herd, all herbivores stop feeding. In contrast, by directly killing prey, predators may only lower the number of individuals feeding on plants by the small fraction that is subdued within a given time period. Clearly, predators influence their prey populations through both trophic and nontrophic interactions, but a recent synthesis by Preiser et al. (2005) shows that nontrophic interactions can often have the stronger effect in ecological systems.
One reason why nontrophic effects may be highly important in ecological systems is that, unlike trophic effects, which simply involve capturing and subduing prey, they come about by a variety of mechanisms involving changes in any of the morphological, behavioral, and chemical traits.
Prey species may undergo defensive morphological changes that are induced by persistent cues of predation risk. For example, water fleas (Daphnia) that are exposed to a persistent predation threat develop spurs on their head and long, sharp tail spines. When tadpoles of many amphibian species are exposed to predation cues, they develop thick muscular tails that often are conspicuously colored. Thicker tails allow for greater acceleration to evade predator attacks, and bright tail coloration deflects the predator’s attack from the vital head region of the prey to more expendable body parts. Mussels are often preyed on by snails that penetrate their shell by drilling through it. Cues from predaceous snails thus cause the mussels to develop thicker shells.
Predators often home in on their prey by looking for signs of prey activity. Vigilance and avoidance of risky habitat by prey are two behavioral mechanisms that can lower the risk of being captured. Becoming vigilant and decreasing the speed of movement decreases conspicuousness to predators. In addition, prey may switch their habitat use to areas devoid of predators or areas that afford greater cover. For example, in aquatic systems, species of mayflies avoid their fish predators by crawling off the surface of rocks that are covered with food resources and hiding under rocks. Snails vulnerable to crayfish predators that hunt on pond bottoms crawl up and feed on emergent vegetation in the water column. In grasslands, grasshoppers facing hunting spider predators reduce feeding on nutritious but highly risky grass and switch to leafier herbs that are less nutritious but serve as refuges.
When fed on by herbivores, plants are often induced to produce chemicals aimed at deterring herbivore impacts. These chemicals can be quite volatile and hence be diffused into the air to attract species that are enemies of the herbivores. They can also chemically signal to neighboring plants that they have been impacted, thereby causing the neighbor plants to induce the production of chemicals that are nauseating or toxic to herbivores as a preemptive measure.
These morphological, behavioral, and chemical changes, however, come with costs. In a world with finite resources, reducing resource intake because of vigilance or allocating valuable resources toward defenses means that fewer resources are available for life-cycle development, growth, and reproduction. As a consequence reproductive output may be diminished or eliminated altogether if individuals fail to develop fully in ways that overcome seasonal environmental bottlenecks. For example, tadpoles need to develop into legged frogs or salamanders to escape their natal pond environment before it dries up in summer. Investing resources into thick tails to facilitate burst swimming to escape predators may delay body growth.
Inasmuch as these traits are properties of individuals, then, understanding indirect effects in ecological systems necessarily requires scaling from the level of the individual to the level of communities and ecosystems. And just as evolutionary history shapes individuals’ ability to flexibly change these traits to balance the trade-offs, understanding indirect effects in ecological systems necessarily requires blending principles of evolutionary ecology with community and ecosystem ecology.
Predation, competition, and mutualism are often considered to be fundamental direct interactions that determine the structure and functioning of ecological communities. But whenever more than two species are linked together by such interactions, we see the emergence of indirect effects in which the middle species mediates the nature and strength of the indirect effect of the first species on the third species.
Perhaps the most familiar indirect effect is the one described above for the Serengeti food chain. In this kind of system, the two predators (herbivores feeding on plants, carnivores feeding on herbivores) can have negative direct effects on their respective prey through direct consumptive effects. But, once being linked in a feeding chain, the top predators have an indirect positive effect on plants that counteracts the herbivore effect by virtue of lowering the abundance of herbivore prey.
Two species may also have a negative indirect effect on each other’s abundance by interacting with an intermediary species. For example, in systems in which species share a common resource but never interact directly with each other for access to that resource, their own trophic interaction with the resource reduces the availability of the resource for the other species. Here, both species are competitors (they have mutually negative effects on each other), but their effect on each other is indirect.
Three species may also compete directly with each other by preempting each others’ access to space or to important resources through direct physical struggles or territorial interactions. Here, direct competitors can cause indirect interactions that are again opposite in sign to their direct effects. One species, by competing directly with a second species, relieves competitive pressure of the second species on a third species. In this case, the first species will have a positive indirect effect on the abundance of the third species.
The nature of indirect effects can also be quite different even within the same kind of ecological system, depending on whether trophic or nontrophic interactions are dominant.
In a rocky intertidal system, barnacles, mussels, and algae compete directly for space on rock surfaces to which they affix themselves. Barnacles tend to be competitively dominant to mussels and thus usurp most of the space. This enables algae to fill in the interstices between the barnacles. However, a species of whelk (snail predator) prefers to prey on the barnacles and thus opens up space for mussels to become established. The mussels in turn exclude the algae. In other words, barnacles, mussels, and algae are direct competitors for space, and barnacles have a beneficial indirect effect on algae by precluding mussels from becoming established. The whelk in turn has an indirect beneficial effect on mussels and hence an indirect negative effect on algae by directly consuming barnacles. However, this outcome occurs only when snails feed on patches containing adult barnacles that are fully developed. The predatory snails have an alternative effect when they try to feed on patches of younger barnacles. Here, young barnacles develop a predation-resistant morphology when faced with predation cues. This in turn reduces the ability of the predator to suppress the barnacle’s competitive dominance. This nontrophic effect leads to predator indirect effects on mussels and algae that are opposite in sign to those found when trophic interactions dominate.
Some insect species lay their eggs in autumn on the ends of plant shoots. This causes the shoots to die back because of mechanical damage. In the following spring, the plant compensates by produce longer shoots that tend to be very leafy. This change in plant morphology in turn provides new habitat promoting the population sizes of many caterpillar species that would normally not exist on the plant. The caterpillars eat the plants, but they also roll leaves to form shelters. Once the caterpillars abandon the shelters to develop into adults, the leaf rolls are inhabited by aphids and three species of ants that tend the aphids for their honeydew production and in turn protect the aphids from predators. Here, short-term plant damage can induce plant morphological responses that lead to nontrophic indirect effects that enhance the diversity and abundance of a series of insect species.
In an old-field meadow community, a species of generalist grasshopper faces spider predator species with different hunting modes. It turns out that predator hunting mode has important implications for the nature and sign of the indirect effects on plants. In the absence of predators, grasshoppers prefer to consume grass. Mortality from a species of predator with a sit-and-wait ambush hunting mode is comparatively low, but mortality risk caused by chronic predator presence in the upper vegetation canopy induces grasshoppers to switch from feeding on grass to seeking refuge in and foraging on a less nutritious goldenrod species. Thus, the sit-and-wait predator has a net positive indirect effect on abundance of grass and a net negative indirect effect on the abundance of the goldenrod induced by a nontrophic (habitat shift) interaction with the grasshopper. This happens because the spider predator presents a persistent point-source cue of presence within the habitat. The outcome is much different when the grasshopper faces an actively hunting spider species that wanders widely throughout the vegetation and thus presents a diffuse cue of presence. In this case, grasshoppers respond only to imminent predation risk when they directly encounter the predator. Actively hunting predators tend to capture many grasshoppers and thus have a strong effect on the numerical abundance of grasshoppers that overrides the nontrophic effect. This translates into a greatly reduced total numbers of grasshoppers feeding on both grasses and herbs than in the absence of predators. Such a trophic interaction leads to positive indirect effects on both grass and goldenrod.
Plants can also take their defense into their own hands. In some cases, plants produce extrafloral bodies that produce nectar to attract ants. In exchange for this reward from the plants, ants defend the plant against attack by herbivorous insects. Here the plant and ant species engage in a mutualistic interaction in which the plants change their morphology to provide a direct benefit to the ant; and the ant in turn provides an indirect benefit to the plant through either trophic (eating herbivore pests) or nontrophic (scaring pests off the plant) effects. In an arid system, a species of herb (coyote tobacco) is attacked by three herbivores (the hornworm caterpillar, a beetle, and a leaf bug) that either eat plant tissue or suck plant sap. On attack by the herbivores, the plant releases volatile chemicals into the air, and the chemical plumes attract predatory insects that in turn prey on the herbivores. In this case, a trophic effect—herbivores feeding on the plant— induces a chemical response by the plant that leads to a nontrophic effect—attraction of predators—that in turn precipitates an indirect positive effect of predators on the plants.
An ecosystem is a conceptualization of nature that considers the communities of species comprising a location, the rate and efficiency of energy and materials transfer among species within the community, and vital ecosystem processes such as plant production. Ecosystems are often viewed as being organized into chains of feeding dependencies, comprised of at least three trophic levels. There are fundamentally two kinds of food chains that determine the pathway of energy and material flow throughout a system. The plant-based chain involves live plant biomass, herbivores, and carnivores. The detritus-based chain involves nonliving plant matter, decomposers, and carnivores. In both cases, plants draw up water and nutrients from the soil and carbon dioxide from the air and are stimulated by sunlight to convert those different chemicals into tissue. In the plant-based chain, herbivores eat that plant tissue and are themselves eaten by their predators. In the detritus-based chain, decomposers eat the dead plant matter and are themselves eaten by their predators. In both chains, old individuals die, and the chemical constituents of their body are also broken down by decomposers and are recycled back through the system by nourishing plants, etc.
The multilevel trophic structure of the plant-based and detritus-based chain also means that indirect effects can propagate within an ecosystem. These newly discovered indirect effects involve a combination of trophic and nontrophic interactions and influence not only the abundance of plants and plant matter but also the rate and efficiency of material cycling.
A deeply ingrained view in ecology is that herbivores have direct negative effects on plant abundance and production by consuming plants. However, this view has been challenged in light of some observations that modest levels of herbivory might indeed enhance plant production. Such a direct and mutually beneficial effect of herbivory to both herbivores and plants is known as the grazing optimization hypothesis. It turns out, however, that this direct mutualism may only be apparent. Instead, the enhanced production may be driven by an indirect interaction in which herbivores alter cycling of nutrients that are essential for plant growth. If herbivores return nutrients back to the soil in the vicinity of their grazing locations, through urination and defecation, then higher levels of grazing may translate into proportionately higher rates at which herbivores return those nutrients to the soil than lower levels of grazing. Thus, it is the indirect effect of herbivores on plants mediated by nutrient cycling that enables plants to compensate better for loss of tissue to herbivores at intermediate levels of herbivory than at lower levels of herbivory.
Alteration of nutrient cycling and alteration of plant production have also been observed to occur across the entire food chain. For example, the nontrophic indirect effects of predators in the old-field meadow system described above lead to important and ramifying indirect effects on plant productivity, plant species diversity, and the biophysical properties of the whole ecosystem. In the absence of predators, herbivores have a comparatively weak effect on highly productive goldenrod, which allows it to grow rapidly into tall, dense stands that shade the surrounding soil. In the presence of predators, herbivore consumption both thins goldenrod stands and stunts the height of the remaining stems, thus suppressing the most productive plant species in this ecosystem. This creates a more open and patchy environment, enabling more photosynthetically active solar radiation to reach the soil surface. This in turn facilitates the proliferation of other, less-productive herb species, which are intolerant of shady conditions caused by goldenrod. The altered plant species composition of this ecosystem further causes changes in the rate of nitrogen cycling because dead goldenrod plants are more difficult to decompose than other herb species.
Introduction of foxes onto the Aleutian Island chain has had a hugely transformative effect on some of these arctic island ecosystems because the foxes substantially reduced abundant seabird populations that breed in colonies on these islands. Seabirds normally provide an important nutrient subsidy to the islands by feeding on marine organisms and then excreting nitrogen- and phosphorus-rich guano onto the islands. Nutrient subsidized fox-free islands supported lush, thick plant communities dominated by grasses. Fox-infested islands tended to be composed of less lush low-lying herbs and dwarf shrubs. These different plant communities and their associated productivity supported different compositions and abundances of arthropod species. Thus, foxes indirectly influenced plant productivity and composition, and animal species composition, by directly disrupting a major source of nutrients to the islands. This was achieved through a combination of trophic and nontrophic effects. Devastation of seabird populations through trophic interactions causes the loss of a major vector of nutrients. The threat of future predation (a nontrophic effect) also discourages surviving seabirds from returning to the breeding colonies in later years. Over the long term, this can eliminate the offshore nutrient subsidy altogether.
The detritus-based chain is a major pathway of energy and material flow in a tropical river system. In this system, a detritivorous fish species has major effects on the cycling of carbon, an important building block of living organisms. By consuming detritus, the fish lower the abundance of dead organic matter particles in the river. This in turn indirectly lowers the abundance of algal and bacterial biofilms and enables the establishment of nitrogen-fixing bacteria that contribute to live-biomass production that serves as an important food resource for other species. In addition, consumption of water-borne particulate matter by the fish clears the water column. This enables solar radiation to penetrate deeper into the water column, thereby indirectly enhancing the production of nitrogen-fixing bacterial biomass. The fish, in turn, redistribute organic material more evenly by excreting organic matter throughout the river system, which in turn indirectly enhances the ability of other detritus-consuming organisms to exist within the river system.
Trophic and nontrophic interactions can have qualitatively different effects on the efficiency of energy and nutrient transfer up food chains. In rocky intertidal ecosystems, predatory green crabs influence the behavior and foraging rate of one of its principal prey, a carnivorous snail that feeds on barnacles. When faced with predation risk, the snail becomes increasing vigilant and therefore feeds less. But it also becomes stressed. Such stress, in turn, elevates the snail’s metabolic costs, thus leaving less resource available for growth and development. In other words, the efficiency at which barnacle tissue is converted into snail tissue—called secondary production—becomes diminished relative to conditions in which predation risk is absent. This finding questions the classical view that transfer efficiencies between trophic levels in ecosystems tend to be fixed. In addition, the poorer nutritive quality of snails stressed by predation risk means that the total amount of energy transferred further up the food chain to the snail’s predators will ultimately be reduced. In this situation, predators indirectly harm their own welfare through nontrophic interactions with their prey. The increased attenuation of energy transfer and secondary production caused by nontrophic interactions may bolster the idea that lack of energy flow up food chains is why so many food chains in nature are short.
Species in communities and ecosystems are wholly dependent on other species for their survival and reproduction. These dependencies can be direct, as, for example, a predator capturing and subduing a prey species, or indirect, where a carnivorous predator may benefit a plant species by consuming the plant species’ herbivore enemies. Direct and indirect effects can come about through a variety of mechanisms including trophic interactions and myriad forms of nontrophic interactions. The number of indirect effects in ecological systems rises in direct proportion to the number of species that are directly linked together in a chain of dependencies. These myriad direct and indirect interactions are what contribute to the fascinating complexity of ecological systems. A complete understanding of species interactions in communities and ecosystems therefore requires explicit consideration of direct interactions in tandem with indirect interactions.
Agrawal, Anurug A. 2001. Phenotypic plasticity in the interactions and evolution of species. Science 294: 321–326. This article presents an important synthesis showing how species traits change flexibly as a consequence of the nature and strength of species interactions in communities.
Hairston, Nelson G., and Nelson G. Hairston Jr. 1993. Cause-effect relationships in energy flow, trophic structure and interspecific interactions. American Naturalist 142: 379–411. This key paper formalizes the concept of ecosystem in terms of species interactions and their influence on energy and materials cycling and transfer efficiencies.
Loreau, M. 2001. Linking community, evolutionary and ecosystem ecology: Another perspective on plant-herbivore interactions. Belgian Journal of Zoology 131: 3–9. This seminal article elaborates the need to link evolutionary ecology with ecosystem ecology and uses the case example of grazing optimization in plant-herbivore interactions to show how indirect effects can control processes that were previously thought to be driven by direct effects.
Maron, John L., James A. Estes, Donald A. Croll, Eric M. Danner, Sarah C. Elmendorf, and Stacey L. Buckelew. 2006. An introduced predator alters Aleutian Island plant communities by thwarting nutrient subsidies. Ecological Monographs 76: 3–24. This article presents a detailed analysis tracing how introduced foxes affect the source and fate of nitrogen and phosphorus and its implications for island ecosystem production.
Ohgushi, Takayuki. 2005. Indirect interaction webs: Herbivore-induced effects through trait changes in plants. Annual Review of Ecology Evolution and Systematics 36: 81–105. This review presents a synthesis of the ways that herbivore attack on plants can propagate morphological changes in plants that then alter the composition and abundance of arthropod species inhabiting the plants.
Preisser, Evan L., Daniel L. Bolnick, and Michael F. Benard. 2005. Scared to death? The effects of intimidation and consumption in predator-prey interactions. Ecology 86: 501–509. This article presents a comprehensive meta-analysis that quantifies the relative strength of trophic and nontrophic effects observed in a host of experimental studies of predator-prey interactions.
Price, Peter W., Carl E. Bouton, Paul Gross, Bruce A. McPheron, John N. Thompson, and Arthur E. Weiss. 1980. Interactions among three trophic levels: Influence of plants on interactions between insect herbivores and natural enemies. Annual Review of Ecology and Systematics 11: 41–65. This classic, foundational paper spells out a working hypothesis for the direct and indirect effects of plant antiherbivore defenses in ecological food chains.
Schmitz, Oswald J. 2006. Predators have large effects on ecosystem properties by changing plant diversity not plant biomass. Ecology 87: 1432–1437. This experimental study traces the direct and indirect effects that top predators have on plant diversity and nitrogen cycling in a meadow ecosystem.
Skelly, David K. 1997. Tadpole communities. American Scientist 85: 36–45. This article presents an empirical synthesis in narrative form that illustrates how amphibian community structure is influenced by trophic and nontrophic effects of predators.
Strauss, Sharon Y. 1991. Indirect effects in community ecology: Their definition, study and importance. Trends in Ecology and Evolution 6: 206–210. This landmark article clarifies and defines direct and indirect effects. The paper stimulated much formal experimental analysis on indirect effects in ecological systems.
Taylor, Brad W., Alexander S. Flecker, and Robert O. Hall. 2006. Loss of harvested fish species disrupts carbon flow in a diverse tropical river. Science 313: 833–836. This experimental study examines how a dominant fish species influences ecosystem structure through consumption and translocation of carbon in plant detritus.
Trussell, Geoffrey C., Patrick J. Ewanchuk, and Catherine M. Matassa. 2006. The fear of being eaten reduces energy transfer in a simple food chain. Ecology 87: 2979–2984. This article discusses the experimental study that quantifies the extent to which nontrophic effects of predators on prey limit secondary production of prey in a rocky intertidal system.