Predation and Consumer-Resource Interactions
A lynx crouches in the snow, waits, and lunges at a snowshoe hare. Roughly ten years later, hare numbers crash. Lynx numbers crash with them. Why should killing individual hares produce a population-wide, decade-long rhythm? The answer lies not in the drama of a single chase. It lies in the deeper structure of consumer-resource interactions. These interactions are the currency in which energy and individuals move from one trophic level to the next. Predation is often pictured narrowly as one animal eating another. But ecologists use it as the anchor case for a much broader family of interactions. In these, one organism, the consumer, gains energy and matter by consuming another, the resource. This consumption reduces the resource's fitness. This single logical structure recurs across nature: consumer takes from resource, resource population is reduced, consumer population is fed. It appears in forms that look superficially very different. A wolf kills an elk. A ladybird beetle eats aphids. A tapeworm drains a host. A caterpillar defoliates a plant. A whale filters krill from seawater. Understanding predation matters because consumer-resource interactions are one of the primary forces structuring communities. They set population sizes. They drive coevolutionary arms races between attack and defense. They determine which species can coexist. They also channel the flow of energy that ultimately derives from photosynthesis up through trophic levels. Community ecology cannot be understood without first understanding how consumers and their resources are linked, numerically and mechanistically. This article builds the conceptual scaffolding. It covers what predation is and how it differs from other consumer categories. It covers the mechanistic steps of a predation event. It covers the classic experiments that revealed how strongly predators shape communities. It also covers the quantitative logic that later articles will formalize into population models such as the Lotka-Volterra Predator-Prey Model.
- Predation is one form of a general consumer-resource interaction: a consumer gains energy/fitness while reducing resource fitness, usually by killing and consuming it.
- Consumer-resource interactions are classified by whether the resource is killed (predation, herbivory to a point) or debilitated without necessarily dying (parasitism, Herbivory).
- A predation event has a repeatable sequence: detection → pursuit → attack → subjugation → consumption — and defenses evolve to block each stage.
- Functional response (per-capita kill rate vs. prey density) and numerical response (predator population growth vs. prey density) together determine total impact on prey.
- Classic manipulative experiments — Paine's sea star removals, Huffaker's mite microcosms, Gause's Paramecium trials — proved predators can control prey abundance and community structure, not just track it.
- These concepts set up the population-level dynamics (cycles, stability, coexistence) developed in Predator-Prey Population Dynamics and the Lotka-Volterra Predator-Prey Model.
Predation is a consumer-resource interaction in which a consumer's fitness rises and a resource's fitness falls through consumption — and the rate of that consumption, not just its occurrence, determines whether predators regulate, merely track, or barely affect their prey populations.
The core idea
A consumer-resource interaction is any relationship in which one organism (the consumer) obtains energy, nutrients, or other resources by exploiting another organism (the resource), with a negative effect on the resource's survival or reproduction and a positive effect on the consumer's. This is the master category; ecologists subdivide it based on two questions: does the resource die as a direct, immediate consequence of a single interaction event, and how much of the resource organism is consumed?
Predation is the subtype in which a free-living consumer (the predator) kills and consumes another free-living organism (the prey), typically consuming most or all of the individual in one act — a wolf and an elk, a spider and a fly, a whale and a school of krill. Herbivory is consumption of live plant (or algal) tissue, usually without killing the whole plant, since the plant's modular body often survives partial removal (this is covered as its own topic, but note it shares predation's basic energy-transfer logic). Parasitism involves a consumer that lives in or on a single host individual for an extended period, draining resources gradually and often not killing the host, at least not quickly. Parasitoidism (chiefly certain wasps) is a hybrid: the consumer develops inside or on one host individual, as parasites do, but that host is invariably killed, as in predation. These categories form a continuum rather than sharp boxes, but the terminology lets ecologists talk precisely about mechanism, timing, and typical population consequences. Two paired measurements describe how strongly a consumer affects a resource population. The functional response is the relationship between an individual predator's per-capita kill (or consumption) rate and prey density — how many prey one predator eats per unit time as prey become more or less abundant. The numerical response is the relationship between predator population growth (via reproduction or aggregation/immigration) and prey density — how the predator population itself grows when prey are abundant. Multiplying functional response by predator numbers gives total prey mortality from predation; multiplying numerical response effects over generations gives the predator population trajectory. These two responses are the mechanistic bridge between individual foraging behavior and population-level dynamics, and they are exactly what the Lotka-Volterra framework will later formalize mathematically.
Mechanism
Step 1 — Detection. A predator must first locate prey using vision, olfaction, hearing, vibration sensing, echolocation, or chemoreception, depending on the taxon. Search can be active (a cheetah scanning open grassland) or passive (an ambush predator like a mantis waiting at a flower). Prey counter this stage with camouflage, cryptic behavior, and reduced activity when predation risk is high.
Step 2 — Pursuit or encounter. Once detected, the predator closes the distance, or, in sit-and-wait strategies, the prey wanders into range. Encounter rate scales with the density of both predator and prey, and with their respective movement speeds and ranges — this is the biological basis of the "mass-action" assumption used in later population models, where encounter frequency is treated as proportional to the product of predator and prey densities.
Step 3 — Attack. The predator commits to a strike: a lunge, a chase, a strike of venom, an ambush. Attack success depends on prey escape speed, group vigilance (many eyes detect an attacker sooner), and startle/flush responses. Many prey species evolve traits specifically to disrupt this stage — flash coloration, erratic flight, schooling that confuses a predator's targeting.
Step 4 — Subjugation. The predator must overpower or immobilize the prey: venom injection, constriction, crushing bite force, or pack cooperation to bring down large prey. Prey defenses at this stage include armor (shells, spines, exoskeletons), toxins, and sheer size (an adult elephant is essentially immune to subjugation by any single predator).
Step 5 — Consumption. The predator ingests and digests the prey, converting captured biomass into its own growth, maintenance, and reproduction. Digestive efficiency and gut passage time set an upper limit on how much prey biomass a predator can process, which caps the functional response at high prey densities — this ceiling is what produces the saturating (Type II) functional response curves seen throughout nature. Each stage is a checkpoint at which the interaction can fail — most predation attempts, across taxa, actually end in failure at detection, pursuit, or attack, not consumption. This is why coevolution between predator attack traits and prey defense traits (an "arms race") shapes morphology and behavior on both sides simultaneously. The functional response describes how per-capita consumption rate changes as prey density rises, and ecologist C.S. Holling (1959) classified it into three canonical shapes:
Type I rises linearly then plateaus abruptly (typical of filter feeders limited by processing rate). Type II rises steeply at low prey density then decelerates toward a plateau set by handling time — this is the most common pattern and it means predators take a disproportionately large share of prey when prey are scarce, which can be destabilizing. Type III is sigmoid: kill rate is low at very low prey density (prey are hard to find, or the predator hasn't "learned" to specialize on them, or switches to alternative prey), rises steeply at intermediate density, then saturates — this pattern is stabilizing because it gives rare prey a low-mortality refuge.
The evidence
The idea that predators merely "harvest a surplus" without shaping community structure was widespread until manipulative field experiments in the mid-20th century showed otherwise.
Robert Paine's sea star removal experiment (Makah Bay, Washington, 1963–1969) is the canonical demonstration that a single predator can control an entire rocky intertidal community. Paine selected stretches of shoreline and physically removed the predatory sea star Pisaster ochraceus — pulling individuals off the rocks and flinging them out to sea — while leaving adjacent control plots untouched. Pisaster normally preys on mussels (Mytilus californianus), barnacles, limpets, and chitons. Within the removal plots, mussels — freed from predation — multiplied and spread, physically smothering and outcompeting the smaller barnacles, limpets, algae, and chitons for space. Species richness in removal plots fell from roughly 15 species to fewer than 8 within about a year, while control plots remained diverse. Paine coined the term keystone species for Pisaster: a predator whose impact on community structure is disproportionately large relative to its biomass. This experiment proved predation is not a passive "top layer" of the food web but an active force determining which competitors can coexist below it.
Carl Huffaker's mite microcosm experiments (1958) tackled a different question: can a predator-prey pair persist together, or does the predator inevitably drive the prey (and then itself) extinct? Huffaker used the herbivorous mite Eotetranychus sexmaculatus as prey and the predatory mite Typhlodromus occidentalis as predator, living on arrays of oranges (as food/habitat patches) arranged on trays. In simple, homogeneous setups, the predator mite quickly found and ate all the prey, then starved — both populations crashed within a few generations, exactly as naive intuition (and simple Lotka-Volterra math without spatial structure) predicts. But when Huffaker increased the spatial complexity of the system — scattering oranges farther apart, adding barriers to slow predator dispersal (vaseline barriers), and building in refuges and dispersal corridors — prey could colonize new patches faster than predators could follow, generating three full oscillation cycles of predator and prey abundance before the experiment ended. This showed that spatial heterogeneity and refuges are essential to stable predator-prey coexistence, a foundational insight for Predator-Prey Population Dynamics.
G.F. Gause's Paramecium-*Didinium experiments (1934)</strong> were an earlier, simpler demonstration with the same message. Gause cultured the ciliate Paramecium caudatum as prey with the predatory ciliate Didinium nasutum in simple test-tube microcosms. In homogeneous medium, Didinium invariably found and consumed all Paramecium, then itself starved and died — total extinction of both, every time, showing that without refuge or immigration, predation could be locally destabilizing rather than a self-correcting balance. When Gause added structure — a small amount of sediment where Paramecium* could hide, or periodically added a few immigrant individuals of each species — the populations oscillated instead of collapsing. This was among the first experimental confirmations that the qualitative behavior predicted by the newly developed Lotka-Volterra equations (oscillation, or collapse without refuge) could actually appear in living systems, and that spatial/temporal heterogeneity is what converts an unstable interaction into a persistent one. Together, these experiments established two separate but complementary truths: predators can be strong, community-structuring forces (Paine), and predator-prey persistence depends critically on spatial refuges and heterogeneity, not just on the raw biology of eating and being eaten (Huffaker, Gause).
Worked examples
Example 1 — Functional response arithmetic. A field study of a mantid predator preying on aphids records the following per-capita consumption rates at different aphid densities, fit to a Holling Type II curve:
where c is kills per predator per day, N is aphid density (per m²), a is the attack rate constant, and h is handling time per aphid (days). Suppose a = 0.05 m² day⁻¹ and h = 0.02 days per aphid consumed.
- At low density, N = 20 aphids/m²:
c = (0.05 × 20) / (1 + 0.05 × 0.02 × 20) = 1.0 / (1 + 0.02) = 0.98 aphids/day
- At high density, N = 500 aphids/m²:
c = (0.05 × 500) / (1 + 0.05 × 0.02 × 500) = 25 / (1 + 0.5) = 16.7 aphids/day
- At very high density, N = 5000 aphids/m²:
c = (0.05 × 5000) / (1 + 0.05 × 0.02 × 5000) = 250 / (1 + 5) = 41.7 aphids/day Notice the asymptote: as N → ∞, c → 1/h = 1/0.02 = 50 aphids/day. No matter how dense the aphids get, one mantid cannot physically catch and consume more than 50 per day because handling time (the minimum time to subdue, kill, and ingest each captured aphid) hard-caps the rate. This ceiling — not lack of hunger — is exactly why Type II functional responses saturate, and it is the mechanistic reason a single predator species is rarely able to fully suppress an outbreaking herbivore population once densities get very high; the per-predator impact plateaus even as prey keep multiplying.
Example 2 — Reading a keystone-predator removal experiment. Suppose you run a Paine-style removal experiment on a rocky shore with a whelk that eats barnacles and small mussels, and you record percent cover after 18 months:
| Plot type | Barnacle cover | Mussel cover | Total species richness |
|---|---|---|---|
| Control (whelk present) | 25% | 15% | 12 species |
| Whelk removed | 4% | 68% | 5 species |
How do you reason through this? First, mussel cover jumped roughly 4.5-fold (15% → 68%) once the whelk predator was removed — the direct, expected consequence of releasing mussels from predation mortality. Second, barnacle cover fell (25% → 4%) even though barnacles were not the whelk's exclusive target, revealing an indirect effect: mussels, unchecked by predation, outcompeted barnacles for attachment space, so the whelk was previously keeping mussels in check and thereby indirectly protecting barnacle habitat. Third, total richness dropped by more than half (12 → 5 species), consistent with a keystone-predator effect: one competitively dominant space-holder (mussel) monopolized the substrate once its controller was gone, squeezing out the algae, limpets, and other invertebrates that depended on bare or lightly-occupied rock. The correct ecological conclusion is not "the whelk eats mussels" (a trivial restatement of diet) but "the whelk's predation rate on the competitively superior species maintains the coexistence of many competitively inferior species" — precisely Paine's original finding, and a template for predicting the community consequences of removing (or reintroducing) a keystone predator, as is done in many modern conservation reintroduction programs.
Common misconceptions & traps
| Misconception | Why it's wrong |
|---|---|
| "Predators reduce prey numbers, so predation is always bad for the prey species as a whole." | Predation can stabilize prey populations by preventing boom-bust resource overshoot, and can increase prey genetic fitness over evolutionary time by selecting for effective defenses; effects on individuals ≠ effects on the population or lineage. |
| "Predators always eat the most abundant, easiest prey." | Functional response shape matters: Type III predators disproportionately ignore rare prey and switch to whatever is locally abundant, but many specialist predators track a preferred prey species regardless of relative abundance. |
| "Removing predators is harmless if prey have 'enough food' otherwise." | Paine's experiment shows the dominant effect of predator removal is often indirect, reshuffling competitive hierarchies among prey/competitors, not simply "more prey survive." |
| "Predation and competition are unrelated forces." | They interact constantly: a predator can determine which of two competing prey species wins (apparent competition), and competition among predators shapes their combined impact on shared prey. |
| "A high kill rate means the predator controls (regulates) the prey population." | Control requires the kill rate to increase with prey density in a way that damps population growth; a constant or saturating kill rate at high density may fail to regulate a prey outbreak at all (see Example 1). |
The scientist behind it
Robert T. Paine (1933–2016) was an American zoologist at the University of Washington who spent decades doing hands-on manipulative fieldwork on the wave-battered rocky shores of the Pacific Northwest, at a time when most community ecology relied on correlational observation rather than direct experimental removal of species. Paine's insight was deceptively simple: instead of just describing which species co-occur, physically remove one species from real plots in the field and watch what happens to everyone else. This experimental ethos — treat the community itself as the unit of manipulation — became a template for an entire generation of field ecologists.
Origins · 1 minRobert Treat Paine III and how this was discovered
Paine's work, alongside Holling's mechanistic dissection of functional responses and Huffaker's and Gause's microcosm demonstrations, shifted community ecology from a largely descriptive science into an experimental one, where hypotheses about consumer-resource control could be tested directly in nature rather than only inferred from static abundance patterns.
Where this matters today
Biological control in agriculture relies directly on functional and numerical response theory: releasing predatory mites, parasitoid wasps, or ladybird beetles against crop pests only works if the natural enemy's functional response is steep enough (ideally Type III at low pest density) to suppress outbreaks before they cause economic damage — this is the same logic tested in Huffaker's oranges.
Fisheries and marine conservation use keystone-predator logic when managing sea otters, whose predation on sea urchins prevents urchins from deforesting kelp forests; removing otters (historically via the fur trade) triggers an urchin-dominated "barren" state closely analogous to Paine's mussel-dominated shore.
Rewilding and predator reintroduction — the return of wolves to Yellowstone National Park beginning in 1995 is widely discussed as a potential keystone-predator effect, with proposed (and debated) cascading changes to elk browsing behavior, willow and aspen regeneration, and beaver activity, directly extending Paine's original conceptual framework to terrestrial systems.
Public health and vector control treat mosquito-eating fish or dragonfly nymphs as predators whose introduction into standing water can suppress larval mosquito densities, informed by the same density-dependent functional response calculations used for agricultural pests, now applied to controlling disease vectors for malaria and dengue.
Invasive species management often hinges on the fact that many invaders escape their coevolved predators/parasites entirely; understanding native consumer-resource structure predicts which introduced species will boom uncontrolled versus which face effective natural enemies in the new range, guiding both prevention and biocontrol strategy.
Ask the tutor: How does a Type III functional response actually stabilize a predator-prey system compared to Type II?
What's the difference between predation and herbivory?
Both are consumer-resource interactions, but predation typically involves a consumer that kills a free-living animal prey and consumes most or all of the individual in one act (a hawk and a mouse). Herbivory involves consumption of live plant or algal tissue, and because plants are modular (they have many repeated parts — leaves, shoots, roots), herbivores often remove only part of an individual without killing the whole plant. A caterpillar eating one leaf is herbivory, not predation, even though the underlying logic — a consumer gains energy while reducing the resource's fitness — is identical. The distinction is about typical damage pattern and whether the resource is a single, killable unit or a modular organism that can survive partial loss.
Is parasitism a form of predation?
No, though the two are related consumer-resource categories. Predation involves a free-living consumer that typically kills its prey quickly and consumes it largely in one act. Parasitism involves a consumer (the parasite) that lives in or on a single host individual over an extended period, draining resources gradually, and frequently does not kill the host, at least not rapidly — the parasite depends on the host staying alive for much of the interaction. Parasitoids (many wasp species) blur the line: like parasites, they develop on or in a single host individual, but like predators, they invariably kill that host by the time development is complete.
What exactly is a "keystone species," and does every predator qualify?
A keystone species is one whose impact on community structure (species composition, diversity, or dynamics) is disproportionately large relative to its abundance or biomass. Robert Paine coined the term after removing the sea star Pisaster ochraceus and watching species richness collapse as mussels overran the shore. Not every predator is a keystone species — many predators have effects roughly proportional to their abundance, or their removal causes only modest, local changes. A species earns the "keystone" label specifically when its removal triggers disproportionate cascading changes, often because it suppresses a competitively dominant prey species that would otherwise monopolize a limiting resource like space.
How do functional response and numerical response combine to determine total predation impact?
Functional response tells you how many prey one predator eats per unit time at a given prey density; numerical response tells you how the predator population itself grows (via reproduction or immigration) as prey density rises. Multiplying an individual predator's kill rate (functional response) by the number of predators present (shaped by the numerical response) gives the total predation rate on the prey population. If both responses rise steeply with prey density, predators can strongly suppress prey outbreaks; if either response saturates or lags, prey can escape control, which is exactly what happens during many real insect pest outbreaks.
Why did Gause's and Huffaker's simple lab systems always end in extinction without added structure?
In a homogeneous container with no refuges, a predator eventually finds and consumes essentially every prey individual, because there is nowhere for prey to hide and no barrier slowing the predator's search. Once prey are gone, the predator has no food and starves. This isn't a flaw in the organisms — it reflects a genuine mathematical instability inherent in the simplest predator-prey models (like unmodified Lotka-Volterra dynamics) when there's no spatial refuge, migration, or prey aggregation to let some prey escape detection. Real, persisting predator-prey systems in nature almost always have such structure, which is why adding oranges spaced apart, barriers, or occasional immigrants let Huffaker's mites cycle for multiple generations instead of collapsing.
Does predation always lower total prey biomass, or can it sometimes increase ecosystem productivity?
Predation reduces the fitness and typically the abundance of the immediate prey individuals consumed, but its net effect on standing prey biomass or on the wider ecosystem is not always negative. Moderate predation (or grazing, in the herbivory case) can increase prey population turnover, remove weaker or diseased individuals, and free resources (space, light, nutrients) for remaining prey to use more efficiently, sometimes increasing net primary or secondary productivity relative to an unpredated, resource-saturated population. This is one reason predator or grazer removal experiments (like Paine's) frequently reveal counterintuitive outcomes rather than simple "fewer predators, more prey" logic.
How does this topic connect to the Lotka-Volterra model I'll see next?
Everything mechanistic covered here — encounter rates, functional response shape, numerical response, refuges — becomes the biological justification behind the terms in the Lotka-Volterra predator-prey equations. The basic model assumes a Type I (linear, unsaturating) functional response and mass-action encounters, which is why it predicts perpetual, undamped oscillations; real systems deviate from this because of Type II/III saturation, spatial refuges (as Huffaker showed), and density-dependent numerical responses, all of which modify the simple math toward damped oscillations, stable equilibria, or extinction. Understanding the mechanism first makes the equations interpretable rather than abstract.