What is a Trophic Cascade?

Predators eat prey. By so doing, predators can impact both prey abundance and behavior (e.g., prey get scared when predators are around and hide or move away). When the impact of a predator on its prey's ecology trickles down one more feeding level to affect the density and/or behavior of the prey's prey, ecologists term this interaction a feeding, or trophic cascade (Figure 1). In this situation, by controlling densities and/or behavior of their prey, predators indirectly benefit and increase the abundance of their prey's prey (Figure 1). Trophic cascades by definition must occur across a minimum of three feeding levels. Indeed, this is how they most commonly occur, although evidence of 4- and 5-level trophic cascades have been shown in nature, but are far less common.

The History of the Concept

Prevailing ecological theory has long held that climate and local resource pools ultimately control the distribution of species and primary productivity of ecosystems. Of these ecosystems, those dominated by plants (e.g., tundra, deciduous and coniferous forest, rain forest, grasslands, seagrass beds, mangroves, and salt marshes) are typically green in appearance, and consumer control in these communities was considered insignificant or subtle — potentially affecting species composition, productivity, and other ecosystem properties, but not capable of regulating the long-term persistence and structure of these systems.

In 1960, Hairston and colleagues proposed an opposing view that consumers play a much more important role in structuring plant-dominated ecosystems (Hairston et al. 1960). They argued that the world is green because higher trophic levels control herbivore abundance. In essence, they suggested that grazers must not be limited by food given the abundance of green material on Earth but limited instead by predators that keep their populations and negative impacts on plants in check. Thus, the world is kept green by a tri-trophic interaction, where predators control grazers that would otherwise overgraze and eliminate vegetation. In response, others pointed out that what is green is not necessarily edible or of sufficient quality to allow increases in herbivore populations. This chemically mediated, bottom-up view suggests that most plants have "won" the predator-prey arms race and are heavily defended and therefore free from significant enemy attack. This debate is ongoing, but the dominant view of ecologists remains that consumers impact many aspects of plant ecology, but are not key drivers of the productivity of entire autotrophic ecosystems.

In recent decades, however, examples of conspicuous, grazer control of entire ecosystems have emerged. These studies show that runaway herbivory can replace more subtle community effects with obvious overgrazing, converting green ecosystems to barrens. In several cases, foundation plant species (sensu Dayton 1972) are eliminated by grazers and replaced with functionally inferior species or unvegetated substratum. In terrestrial systems, both invertebrates (e.g., native or introduced beetles and moths) and vertebrates (e.g., introduced possums) can denude entire forest canopies (Hard et al. 1983, Owen & Norton 1995), insects can defoliate mangrove stands (Anderson & Lee 1995, Feller 2002), and ungulates and elephants can convert savannas to sandy deserts. Similar examples exist in marine systems, where urchins can convert kelp forests and seagrass beds to barren rock-bed and sandflats, respectively (Estes & Palmisano 1974, Rose et al. 1999). These examples of run-away consumption are considered undesirable from a conservation and management perspective because overgrazed systems tend to have lowered biodiversity and productivity, and less economic and aesthetic value. From an ecological perspective, these examples warn that whole-ecosystem regulation by consumers maybe much more prevalent and potent than generally recognized.

Trophic Cascades: Consumer Control Trickles Up the Ladder

Since the world is generally green, run-away grazer effects have been considered exceptions and viewed as relatively unimportant. Over the past 50 years, however, it has become clear that consumers can be important drivers of ecosystems, even when they are overwhelmingly green. Interestingly, such appreciation for grazer effects has developed by focusing on predators. When ecosystems are green, predators are often holding grazers in check, while, when they are overgrazed, predator loss or removal is often responsible for elevated grazer densities and plant loss. This tri-trophic interaction, where predators benefit plants by controlling grazer populations, is known as a trophic cascade. Hairston et al. (1960) first hypothesized that the world is green because predators control grazers, and Carpenter and colleagues further developed the trophic cascade concept with experimental studies in lakes, demonstrating that fish can control zooplankton abundance which, in turn, controls phytoplankton levels (Carpenter et al. 1985).

More recently, ecologists have differentiated between population- and community-level trophic cascades. In population-level cascades, predator removal leads to overgrazing and local extinction of a dominant plant species, but a less palatable species replaces the first, keeping the ecosystem functionally intact despite the shift in species composition. In contrast, in a community-level cascade, predator removal leads to overgrazing of the entire plant community with the concomitant loss of associated ecosystem services.

When and Where Do Community-Level Trophic Cascades Occur?

In aquatic systems, control of plant community structure via trophic cascades has been demonstrated in a variety of habitats, including lakes (Carpenter et al. 1985), rivers (Power 1992) and intertidal (Wootton 1992) and subtidal (Estes & Palmisano 1974) marine systems. In these systems, the standing crop of the plant community is reduced wholesale, and the substrate completely denuded, when predators do not suppress one or a few species of potent grazers. Most of these communities are characterized by simple food webs with little redundancy of consumers, and producers that are single-celled phytoplankton, diatoms, or macroalgae. Because of these factors, Strong (1992) has suggested that top-down control of primary production via trophic cascades may be an idiosyncratic attribute of simple, aquatic systems that are not buffered from run-away consumer effects by multiple predators, and are characterized by weedy, poorly defended primary producers. In general, Strong (1992) suggested that trophic cascades tend to be more important in aquatic vs. terrestrial systems, in simple vs. complex food webs, in homogenous vs. heterogeneous systems, in communities dominated by nonvascular plants (i.e., algae), and in systems where impalatable plants don't replace those that have been overgrazed.

Recent evidence from other systems, however, has suggested that communities dominated by more heavily defended vascular plants are also susceptible to cascading consumer effects. For example, Jackson (1997) has argued that, before humans colonized the Caribbean, large turtle and manatee populations exerted strong, top-down control on seagrasses, which are highly chemically-defended vascular plants, in shallow-water habitats. Consequently, the seagrass beds that have dominated the Caribbean this century may represent recent, human-induced release from consumer control. In terrestrial habitats, others have suggested that large grazing mammals, released from predation, can control vascular plant assemblages in a similar manner (Pace et al. 1999). Examples like these warn that the potential for top-down control of plant community structure may be more pervasive than currently envisioned, especially in those systems thought to be relatively unsusceptible to cascading consumer effects because of plant type and quality (e.g., impalatable, vascular plants, Strong 1992).

Trophic Cascades Across Diverse Ecosystems

Perhaps the best recognized example of a tri-trophic cascade comes from the Aleutian Islands and southeast Alaska, where sea otters (Enhydra lutris), invertebrate herbivores (i.e., sea urchins) and macroalgae demonstrate spatial and temporal density patterns suggesting widespread predator facilitation of plant persistence (Figure 2, Estes & Duggins 1995). Although a historically common component of near shore marine communities, sea otters were hunted to near extinction in the early part of the 20th century for their luxurious pelts and now patchily persist along the coast. Where sea otter populations have lingered, Estes & Duggins showed that they suppress the density and biomass of hold-fast grazing urchins and thus have a strong, indirect positive effect on the abundance of macroalgae (kelp). In contrast, at sites where sea otters have long been absent, sea urchin populations have swollen to high densities and maintain extensive urchin barrens characterized by low coverage of kelp. As sea otter populations have expanded into new sites in recent decades, predictable changes in the density of sea urchins, kelp, and the organisms that utilize the habitat created by healthy kelp beds, have been observed, demonstrating the potential for whole-ecosystem recovery with the reinstatement of predator populations (Estes & Duggins 1995).

Since the 1920s, the spatial extent of trembling aspen forests (Populus tremuloides) has decreased significantly, and the age distribution of remnant trees shifted toward older individuals, indicating widespread recruitment failure of this critical habitat-forming species in mountain forests of the western US (Ripple & Larson 2000). The loss of aspen has been correlatively linked with the extirpation of grey wolves (Canus lupus) from Yellowstone National Park from the 1880s-1920s and subsequent increase in browsing pressure on aspen suckers, the clonal offspring that maintain aspen stands, exerted by herds of elk, Cervus elaphus (Figure 3). Over the past decade, grey wolves have been re-introduced to this region, affording scientists the opportunity to study the mechanisms underlying this apparent trophic cascade that mediates tree dynamics in these mountainous landscapes. Using GPS trackers to monitor movement patterns of elk, a recent study proposed that elk avoid ‘risky' areas with high densities of grey wolves and spend more time in alternate habitats where wolf densities are lower, thereby relaxing local browsing pressure on aspen suckers in high-risk foraging areas (Fortin et al. 2005). Although this suggestion of wolf led trophic cascade is controversial and has been challenged recently (Kauffmann et al. 2010), these findings point to the presence of a behaviorally-mediated trophic cascade in which grey wolves have an indirect positive effect on aspen survivorship and growth by altering the foraging patterns of elk grazers. More research, however, is needed to resolve these food web connections!

Seven years after a manmade impoundment created a network of islands in the midst of a Venezuelan rainforest, John Terborgh and his colleagues took advantage of this natural experiment and compared the abundance of predators of invertebrates (birds, lizards, anurans, and spiders), seed predators (rodents), and herbivores (howler monkeys, iguanas, and leaf-cutter ants) on islands with and without vertebrate predators to assess the potential for a trophic cascade. Compared to islands with predators, they found a hyper-abundance of seed predators and herbivores, and a significant depression in recruitment of seedlings and saplings of canopy-forming trees on predator-free islands (Figure 4, Terborgh et al. 2001). This dramatic shift in community structure revealed that consumer densities are not limited by food availability as historically thought, but rather by strong top-down control. Consequently, even in tropical rainforests characterized by complex food webs and high primary productivity, where trophic cascades were thought to be inconsequential, the removal of top-down regulation can elicit a meltdown in the stability and structure of biological communities (Terborgh et al. 2001).

Salt marshes are among the most productive ecosystems on Earth. For over 60 years, marshes have been viewed as the classic example of a bottom-up regulated system dominated by relatively unpalatable plants controlled by physical conditions and nutrients. Recent research throughout the Western Atlantic has shown that these ecosystems are under strong consumer control by crabs (Holdredge et al. 2009), snails (Figure 5, Silliman & Zieman 2002) and snow geese (Jefferies et al. 1997), and that, in the southern US, plant growth is ultimately controlled by a powerful trophic cascade (Silliman & Bertness 2002). In these southern marshes, experimental manipulation of consumers showed that the most common snail in the salt marsh ecosystem, long thought to be a detritivore specialist, actually controls marsh plant growth by farming fungus on live plant tissue. At high densities, this plant-grazing snail can kill marsh grass and convert salt marshes to exposed mudflats. Blue crabs, by controlling the abundance and distribution of snails, anchor and complete this newly discovered trophic cascade (Figures 5 and 6). When blue crab densities have dropped due to drought stress, snail numbers have increased and mowed down marshes creating a mosaic of vegetated and exposed mudflat patches (Silliman et al. 2005). Combined, these findings reveal that bottom-up and top-down forces are both key determinants of this key coastal plant community.

Conclusions and Conservation Implications

Trophic cascades are powerful interactions that strongly regulate biodiversity and ecosystem function. Trophic cascades were originally thought to be rare, but now we understand that they occur across diverse terrestrial, freshwater and marine ecosystems, and are common features of many green plant communities, including vascular plant assemblages, long thought to be resistant to consumer control. Although we know that trophic cascades can be more powerful under certain conditions, for example climate stress and nutrient enrichment in salt marshes, much research is needed in this area to refine our understanding of when and where trophic cascades will be important.

One of the traditional paradigms of natural resource management is that forcing functions across trophic levels are considered largely bottom-up in nature (Estes & Peterson 2000). Temperate and tropical forests, kelp beds, grasslands, and salt marsh communities, have long been viewed as classic examples of bottom-up regulated systems dominated by relatively unpalatable plants controlled by physical conditions and nutrients. Accordingly, the conservation of these and other plants systems have generally neglected top-down effects in management and conservation efforts for over half of a century. Results from the studies described above call into question the dominance of the bottom-up only paradigm and its wide-scale application to conservation and restoration of plant ecosystems. Managers and ecologists will need to reevaluate their understanding of controls on plant communities and incorporate top-down effects into their conservation plans. Failure to identify and integrate top-down forces may lead to trophic cascades transforming highly diverse and productive plant communities to barren or almost barren flats, with concomitant loss of associated biodiversity and ecosystem function.