Ecological Resilience — What Holling Discovered

Most people use the word resilience to mean recovery. If a system bounces back quickly after a shock, people call it resilient. This is the engineering definition of resilience, and it has been useful in structural engineering, materials science, and control theory.

C. S. Holling called it “engineering resilience” and argued that it misses what matters most in ecological systems. He published his alternative definition in 1973, in a paper titled “Resilience and Stability of Ecological Systems.” The paper is one of the most cited in ecology. It changed how ecologists think about disturbance, change, and the boundaries of what a system can absorb.

Holling’s central claim is simple: resilience is not speed. Resilience is capacity. It is the amount of disturbance a system can take before it reorganizes into a different state with different rules.

The difference between resilience and stability

Holling distinguished two concepts that are often treated as synonyms.

Stability describes how a system responds to small perturbations near an equilibrium. A stable system returns to its equilibrium state after a small disturbance. The faster it returns, the more resilient it is — in the engineering sense. Holling called this property persistence.

Resilience describes how much disturbance a system can absorb before it crosses a threshold and reorganizes. When a lake shifts from clear water dominated by plants to turbid water dominated by algae, it has crossed a threshold. It has not been destabilized in the engineering sense. It was already functioning. It has simply changed into a different state.

The distinction matters because the two concepts imply different management strategies. Engineering resilience suggests maintaining conditions near an equilibrium. Ecological resilience suggests maintaining the capacity to absorb disturbance, which may require allowing the system to occupy very different states at different times.

The shatter experiment

Holling developed his ideas while studying insect population cycles, specifically the spruce budworm, which defoliates forests across northern North America. The budworm population is usually low. Every few decades, it explodes. The forest responds by defoliating, recovering, and then cycling again.

The pattern did not fit the standard models of population dynamics. The budworm was not driven by external climate fluctuations. It was driven by internal feedbacks between the insect, the forest, and the birds that eat it. The system had multiple equilibria — a low budworm state and a high budworm state — and it could jump between them when conditions changed enough.

Holling realized that the question was not how quickly the budworm population returned to its low state. The question was what conditions allowed the low state to persist and what conditions pushed the system past the point of no return.

He illustrated the idea with an analogy that has become standard in the literature. Imagine a landscape of hills and valleys. A ball sitting in a valley will return to the bottom if you nudge it. The size of the valley determines how far you can push the ball before it rolls over the hill into the next valley. The size of the valley is the resilience of that state. The height of the hill between valleys is the magnitude of the disturbance required to trigger a regime shift.

Alternative stable states

The concept of alternative stable states followed directly from this picture. A single ecosystem can support multiple configurations, each stable within its own basin of attraction. A coral reef and a macroalgae-dominated reef can occupy the same physical space under different conditions. A clear lake and a turbid lake can exist at the same nutrient level, depending on history.

Each state is stable. Each is resilient to a certain range of disturbances. But the resilience of each state is finite. Cross the threshold, and the system reorganizes. The reorganization is often irreversible at the scale of human management. Restoring a turbid lake to clarity requires more than reducing nutrients. It may require removing fish, replanting vegetation, and waiting for feedbacks to rebuild the basin of attraction.

This was a fundamental shift from the equilibrium thinking that dominated ecology before the 1970s. Most ecological models assumed a single equilibrium and treated disturbances as noise. Holling showed that disturbance can be the mechanism that pushes a system into a different equilibrium, and that the boundaries between equilibria are as important as the equilibria themselves.

The adaptive cycle

Holling and his collaborators later extended the idea into the adaptive cycle, a four-phase model of how ecosystems and human systems develop over time. The phases are r (growth), K (maturity), O (release), and P (reorganization).

The r phase is rapid growth. Resources are abundant. The system expands. The K phase is conservation. Resources are locked up in structure. The system is efficient but rigid. The O phase is release. A disturbance — fire, storm, harvest — breaks down the accumulated structure. The P phase is reorganization. New configurations form from the debris.

The adaptive cycle is not a theory of recovery. It is a theory of transformation. The release phase is not a failure of resilience. It is a necessary part of the cycle. Systems that resist release accumulate vulnerabilities. Systems that allow release create the conditions for new configurations.

The model has been applied to forests, fisheries, urban systems, and organizational dynamics. It has also been criticized for being too abstract, for lacking precise predictive power, and for inviting analogy without mechanism. Holling himself treated it as a heuristic, not a law.

Thresholds and tipping points

The practical implication of Holling’s work is that management should focus on thresholds, not equilibria. If the goal is to prevent a regime shift, the relevant metric is not how far the system is from equilibrium but how far it is from the threshold that would trigger reorganization.

Forests, lakes, and fisheries all have thresholds. A lake does not become turbid because nutrients increase linearly. It becomes turbid because a threshold is crossed, and the system flips. The flip is often abrupt. The recovery is often gradual. The asymmetry is the defining feature of ecological resilience.

Identifying thresholds is difficult. They depend on context. They can shift as conditions change. A threshold that held for a decade may not hold after a drought. This is why Holling’s framework has been associated with adaptive management — iterative learning about thresholds rather than fixed targets.

What the concept cost

Holling’s redefinition of resilience was powerful. It was also imprecise. The metaphor of basins of attraction is geometrically clear but ecologically vague. A basin of attraction has a boundary. The boundary has a size. But ecosystems do not come with measured boundaries. Estimating the size of a basin requires data that is rarely available.

The concept also invites overgeneralization. The adaptive cycle has been applied to everything from software development to economic innovation to organizational learning. Each application borrows the metaphor without the mechanism. The metaphor is appealing because it is general. It is dangerous because it is general.

Holling understood this. He was careful to ground his ideas in empirical systems — the budworm, boreal forests, fire regimes. He did not claim that every system follows the adaptive cycle. He claimed that resilience thinking — focusing on capacity to absorb disturbance rather than speed of recovery — is necessary for understanding complex systems.

Why the redefinition matters

The engineering definition of resilience asks: how fast does the system return? The ecological definition asks: how much can the system absorb before it changes?

The two questions are not interchangeable. A system that returns quickly to equilibrium may have very low resilience in the ecological sense. It may be balanced on a narrow ridge between two deep valleys. A system that takes decades to return may be deeply resilient. It may occupy a wide valley that can absorb massive disturbance.

The distinction is not merely academic. It shapes how managers think about risk. If the goal is engineering resilience, the goal is to maintain conditions near equilibrium and respond quickly when they deviate. If the goal is ecological resilience, the goal is to maintain the diversity of feedbacks, the redundancy of functions, and the capacity for reorganization that allow the system to persist through change.

The two goals can conflict. A fire-suppression policy that maximizes engineering resilience in a fire-adapted forest accumulates fuel until a single ignition triggers a catastrophic release. An adaptive management policy that accepts periodic disturbance may look like failure under the engineering definition. It may be the only way to maintain ecological resilience.

What the concept does not say

Ecological resilience is not a theory of optimality. It does not claim that systems tend toward maximum resilience. It does not claim that more resilience is always better. A system with high resilience may persist in a state that is undesirable from a human perspective. A turbid lake is resilient. A clear lake is resilient. Resilience does not improve the state. It preserves it.

Ecological resilience is not a theory of predictability. Thresholds can shift. Basins of attraction can merge or disappear. A system that appears stable may be approaching a tipping point. A system that appears unstable may be oscillating within a single basin. The framework describes structure, not outcomes.

Ecological resilience is not a prescription for management. It is a lens. It directs attention to the capacity of systems to absorb disturbance, to the existence of multiple equilibria, and to the asymmetry between rapid collapse and gradual recovery. It does not tell managers what to do. It tells managers what to look for.

The legacy

Holling’s 1973 paper was published in Science, a general-interest journal, not a specialized ecology journal. This was unusual and may have contributed to its impact. The paper reached researchers in limnology, forestry, fisheries, and oceanography. It reached economists, political scientists, and systems theorists. It became a reference point for anyone studying complex systems.

The concept of ecological resilience has influenced climate science, where the question of tipping points in the climate system has become urgent. It has influenced conservation biology, where the goal is often to maintain ecosystems in desirable states rather than restore them to historical baselines. It has influenced resource management, where the acceptance of multiple equilibria has changed how quotas, harvest levels, and protected areas are set.

The concept has also been criticized. Some ecologists argue that alternative stable states are rare and that most ecosystems follow continuous trajectories rather than discrete regime shifts. Some critics argue that the metaphor of basins of attraction is too abstract to generate testable predictions. Some argue that the adaptive cycle has been stretched beyond its empirical grounding.

These criticisms are valid. They are also part of the concept’s strength. A theory that invites application and criticism is more useful than a theory that is too precise to challenge. Holling’s contribution was not the final word on resilience. It was the first word that mattered.

What the concept teaches

The engineering definition of resilience is a measure of speed. The ecological definition is a measure of capacity. The two definitions answer different questions. They are not compatible. They are both useful.

The ecological definition is useful because ecosystems do not behave like springs. They do not always return to equilibrium. They reorganize. They cross thresholds. They adopt new configurations. Understanding these behaviors requires a concept of resilience that focuses on what a system can absorb, not how fast it recovers.

The concept is not a theory of everything. It is not a metaphor that explains every system. It is a lens that focuses attention on the capacity of systems to persist through change, on the existence of multiple equilibria, and on the asymmetry between rapid collapse and gradual recovery.

It is a lens that is useful for anyone studying complex systems. It is a lens that is useful for anyone managing resources. It is a lens that is useful for anyone trying to understand why systems change the way they do.