Loading field map
Restoration ecology is built on a deceptively simple question: when we repair a damaged ecosystem, what should we be trying to recreate? The answer has never been obvious. A forest cleared for agriculture, a wetland drained for development, a grassland invaded by exotic species—each site presents a choice about which past state, which set of functions, or which future trajectory counts as "restored." Over the past century, ecologists have proposed very different answers, and the frameworks they built still shape how restoration projects are designed, funded, and judged.
The earliest systematic answer came from Clementsian Succession, the view that plant communities develop through predictable stages toward a single stable endpoint called the climax. For restoration, this framework offered a clear target: identify the climax community for a region and restore the site until succession reaches that endpoint. The practitioner's job was to remove barriers to natural succession—planting key species, controlling fire, or reintroducing soil biota—and then let deterministic processes take over. This approach dominated early restoration projects in the first half of the twentieth century, especially in North America, where it guided prairie and forest reconstructions.
But even as Clementsian ideas shaped practice, a rival framework was emerging. The Individualistic Concept of Plant Association, proposed by Henry Gleason in the 1920s, argued that plant communities are not superorganisms marching toward a fixed climax. Instead, each species responds individually to environmental gradients, and the assemblages we observe are contingent, overlapping distributions rather than tightly integrated units. For restoration, this was a radical shift. If communities are not predetermined wholes, then there is no single correct endpoint to restore to. The Individualistic Concept did not immediately replace Clementsian thinking—both frameworks coexisted in ecology for decades—but it planted a seed of doubt about historical fidelity as the only goal.
By the 1980s, restoration ecology had become a self-conscious field, and it needed operational tools. The Reference Ecosystem Model answered that need directly. It proposed that every restoration project should be guided by a specific reference site—a relatively undisturbed ecosystem of the same type, chosen as a benchmark for species composition, structure, and function. This framework made historical fidelity measurable: you could count species, measure canopy cover, and compare your restored site against the reference. It gave practitioners a concrete target and a way to demonstrate success to funders and regulators. The Reference Model became the default framework for most restoration projects through the 1990s.
Running alongside the Reference Model, but with a different emphasis, was the Ecosystem Concept. Where the Reference Model focused on composition and structure, the Ecosystem Concept directed attention to flows of energy and nutrients. A restored ecosystem, from this perspective, was one that cycled carbon, nitrogen, and water at rates comparable to an intact system—regardless of whether the exact same species were present. The two frameworks did not directly compete; they coexisted and often complemented each other. A project could use a reference site to guide species selection while also monitoring ecosystem function. But the tension was real: the Reference Model implied that history matters deeply, while the Ecosystem Concept suggested that function could substitute for composition. That tension would become central to later debates.
By the early 2000s, a growing number of restoration ecologists began to question whether historical baselines were even attainable. Many degraded sites had experienced irreversible changes: species extinctions, soil transformations, climate shifts, and invasions by non-native organisms that could not be removed. The Novel Ecosystems framework emerged from this recognition. It argued that some ecosystems have crossed thresholds into new configurations that have no historical analogue. For these sites, aiming to restore a past state is futile or even counterproductive. Instead, practitioners should accept the new species combinations and manage for desired functions—carbon storage, water filtration, habitat for target species—rather than historical fidelity.
Novel Ecosystems directly challenged the Reference Ecosystem Model. If a site is truly novel, then no reference site can guide its restoration. The framework did not reject all use of history—historical knowledge could still inform which functions to prioritize—but it dethroned historical fidelity as the automatic goal. This shift was controversial. Critics worried that the Novel Ecosystems label could be used to justify abandoning restoration efforts on difficult sites. Proponents countered that it was a realistic response to a changing planet. The debate remains unresolved, and it has forced the field to articulate more clearly what counts as success.
While the goal-setting frameworks were debating what to restore to, a separate methodological school was gaining traction. Adaptive Management emerged from resource management in the 1990s and was quickly adopted by restoration ecologists. Its core insight is that restoration actions are experiments. Because ecosystems are complex and our knowledge is incomplete, every intervention should be designed to test assumptions, monitored for outcomes, and adjusted based on what is learned. Adaptive Management does not prescribe a specific endpoint; it prescribes a process. It asks practitioners to articulate their model of how the system works, specify measurable predictions, implement treatments as replicates, monitor results, and update the model.
This framework is compatible with almost any goal-setting framework. A project guided by the Reference Model can use Adaptive Management to test which planting methods best achieve the reference composition. A project guided by Novel Ecosystems can use it to compare which functional outcomes are achievable under different management regimes. But Adaptive Management also carries its own demands: it requires long-term monitoring budgets, institutional flexibility, and a tolerance for uncertainty that many restoration programs lack. Its influence has been profound, but uneven—most restoration projects still lack the experimental design and monitoring that Adaptive Management requires.
The most recent frameworks have pushed restoration ecology further away from static endpoints. Resilience Thinking, which entered restoration discourse around 2000, draws on concepts from complex systems theory: adaptive cycles, thresholds, regime shifts, and panarchy. It asks restorationists to think not about restoring a particular state, but about restoring the capacity of an ecosystem to absorb disturbance and reorganize without flipping into a degraded regime. This reframes the goal entirely. Instead of asking "What did this site look like in 1800?", Resilience Thinking asks "What configuration of species and processes gives this system the greatest ability to persist under future shocks?"
Resilience Thinking revisits the old Clementsian interest in succession, but with a non-equilibrium lens. Where Clements saw a single climax, Resilience Thinking sees multiple possible stable states and the thresholds that separate them. Restoration becomes an exercise in identifying critical thresholds—the point at which a grassland becomes shrubland, or a clear lake becomes turbid—and managing to keep the system on the desired side of those thresholds. This framework has been especially influential in coastal, wetland, and rangeland restoration, where regime shifts are well documented.
Social-Ecological Systems thinking extends Resilience Thinking by insisting that human communities are part of the system being restored. It argues that restoration projects cannot succeed if they ignore the social, economic, and political context: who uses the land, what knowledge they hold, what institutions govern access, and how benefits and costs are distributed. This framework emerged from the recognition that many restoration projects failed not because of ecological mistakes, but because they alienated local stakeholders or ignored the drivers of degradation. Social-Ecological Systems thinking does not replace Resilience Thinking; it broadens it. The two frameworks now often travel together, with resilience providing the ecological theory and social-ecological analysis providing the implementation strategy.
The four active frameworks—Adaptive Management, Novel Ecosystems, Resilience Thinking, and Social-Ecological Systems—share a common rejection of static, historically fixed endpoints. All four accept that ecosystems are dynamic, that uncertainty is unavoidable, and that restoration must be adaptive rather than prescriptive. They also agree that restoration goals should be explicit and measurable, even if the measures are functional rather than compositional.
But they disagree on several key points. The deepest disagreement concerns the role of history. Novel Ecosystems argues that historical baselines are often irrelevant; Resilience Thinking treats historical information as useful for understanding thresholds but not as a target; Social-Ecological Systems sees history as one input among many, mediated by human values. A second disagreement concerns the primacy of ecological versus social goals. Resilience Thinking tends to prioritize ecological function and threshold dynamics, while Social-Ecological Systems insists that social outcomes—livelihoods, equity, governance—are equally important. Adaptive Management remains neutral on these questions, offering a method rather than a goal.
These disagreements are not signs of weakness. They reflect a mature field that has moved beyond the search for a single correct answer. Restoration ecology today is a pluralistic discipline, and the choice of framework depends on the site, the stakeholders, and the resources available. The frameworks do not replace each other; they coexist as tools for different problems. A student entering the field today inherits not a settled doctrine, but a productive tension between history and novelty, function and composition, ecology and society—and a set of frameworks for navigating that tension.