Reframing mine rehabilitation: a systematic review of adaptive ecological restoration, with special reference to the implications for South Africa's mined landscapes

Introduction

Mining activity continues to expand globally, and it is widely recognized for its significant environmental consequences, which most notably include the contamination and degradation of water resources (Hernandez-Santin et al., 2024; Silliman et al., 2024). As countries increasingly pursue sustainable development, the rehabilitation of mine-affected landscapes has become a global environmental priority, with growing recognition that conventional rehabilitation approaches often struggle to address ecological uncertainty and long-term environmental change. In South Africa, mining is among the major contributors to the national economy, yet it is simultaneously associated with profound and enduring environmental risks when not properly regulated (Pagouni et al., 2024). Recognizing this duality, the country has established a comprehensive framework for mine rehabilitation to ensure that mining activities do not culminate in irreversible ecological degradation (Cole et al., 2025). Mine rehabilitation encompasses the set of interventions designed to return disturbed landscapes to a condition that is safe, stable, and aligned with desired post-mining land uses (Cole et al., 2025). This increasingly involves not only soil and vegetation recovery, but also broader ecological and social considerations aligned with contemporary sustainability standards (Hernandez-Santin et al., 2024). These objectives are also closely aligned with the United Nations Sustainable Development Goals (SDGs), particularly SDG 6 (Clean Water and Sanitation), SDG 13 (Climate Action), SDG 15 (Life on Land), and SDG 17 (Partnerships for the Goals), which collectively promote ecosystem restoration, sustainable resource management, climate resilience, and collaborative environmental governance.

Ecological restoration is considered as the intentional activity that initiates or accelerates the recovery of an ecosystem with respect to its health, integrity and sustainability (SER, 2004). Broadly, the concept can be thought of as a process of assisting the recovery of an ecosystem that has been degraded, damaged, or destroyed (Silliman et al., 2024). Silliman et al. (2024) contend that ecological restoration can increase the health and resilience of both nature and humanity. Despite widespread enthusiasm, most ecosystem restoration initiatives have underperformed, failed, or been burdened by high costs that prevent upscaling. If restoration is to emerge as a core and widely deployable sustainability strategy, both its effectiveness and cost-efficiency must improve substantially (Silliman et al., 2024). Nevertheless, the concept of ecological restoration has gained significant prominence within contemporary mine rehabilitation research and policy, signaling a broader paradigm shift toward ecologically grounded and sustainability-oriented post-mining land uses (Silliman et al., 2024; Allison, 2023).

Although terms such as rehabilitation and reclamation persist in the mining industry discourse, often reflecting narrower, engineering-based approaches, the scientific literature and international best-practice frameworks increasingly favor the term ecological restoration due to its emphasis on re-establishing functional ecosystems and recovering biodiversity (Cross et al., 2020). This shift is further reinforced by global policy agendas, most notably the UN Decade on Ecosystem Restoration (2021–2030), which positions restoration as a central strategy for addressing climate change, land degradation, and biodiversity loss (Convention on Biological Diversity, 2020). However, even in the face of this apparent wide use, critical knowledge on effective ways to restore degraded mining areas remains relatively limited (Tedesco et al., 2023).

Although the literature presents a broad spectrum of ecological restoration approaches, one emerging paradigm appears particularly promising: adaptive restoration (Raupp et al., 2024). Rooted in theoretical efforts to navigate uncertainty and complexity, adaptive restoration has evolved as a framework for responding dynamically to the challenges of rehabilitating degraded mining landscapes (Hernandez-Santin et al., 2024). Recent scientific contributions on adaptive restoration in post-mining landscapes encompass empirical case studies from sand extraction sites, bauxite mining operations, and broader landscape rehabilitation programs (Conradi et al., 2021; Littles et al., 2022; Kimuyu et al., 2018). Central to these studies is the principle of iterative learning, whereby systematic monitoring and feedback loops are employed to evaluate ecological responses to remedial actions and subsequently recalibrate restoration strategies (Allison, 2023). This adaptive process underscores restoration not as a fixed intervention, but as a dynamic, evidence-driven practice capable of responding to contextual variability and emergent system behavior (Silliman et al., 2024; Pettorelli and Bullock, 2023).

In this article, we aim to provide a chronology of the major scholarship on adaptive restoration and its development in the literature on mining rehabilitation. Over the past decade, the term adaptive restoration has gained increasing prominence in restoration scholarship; however, the literature reveals considerable conceptual inconsistency and an absence of a clearly articulated research agenda. A synthesis of the adaptive restoration literature is warranted for two key reasons: first, to respond to the need for restoration models capable of addressing the dynamic and large-scale degradation associated with mining landscapes; and second, to clarify the growing conceptual and practical interest in adaptive ecological restoration as a viable framework. Therefore, the objective of this systematic literature review is to develop and offer a clearer and more integrated conceptual framework that specifies the drivers, key components, and practical pathways for adaptive ecological restoration of post-mining landscapes. In particular, we synthesize the evolution, conceptual foundations, and application of adaptive restoration in ecological restoration and mine rehabilitation research. This review answers three (3) research questions: (i) how has adaptive restoration evolved conceptually within ecological restoration scholarship? (ii) how is adaptive restoration applied in mine rehabilitation and post-mining landscape management? (iii) what core principles, governance mechanisms, and resilience pathways underpin adaptive restoration frameworks?

Methodology

A systematic literature review approach was adopted to ensure transparency, replicability, and methodological rigor when identifying, screening, and synthesizing relevant academic contributions. The review focused on peer-reviewed literature addressing adaptive restoration, ecological restoration paradigms, adaptive management, resilience-based restoration, and post-mining rehabilitation. Given that adaptive restoration is an emerging and conceptually diverse field, the review prioritized empirical as well as theoretical studies that contributed to defining, refining, or applying the concept in restoration and rehabilitation contexts. This focused scope maintained conceptual framework comparability and allowed for the integration of interdisciplinary perspectives, while also acknowledging that the field is indirectly informed by broader sustainability and environmental governance literature.

The review followed PRISMA 2020 guidelines and applied a systematic and transparent workflow (Figure 1). The first step was to identify relevant keywords and create a comprehensive search strategy to capture material that answered the research questions. Keywords were refined iteratively and merged across multiple configurations to ensure broad yet targeted retrieval of relevant scholarship.

Literature search strategy

A literature search was undertaken on February 24, 2024, using the Web of Science and Google Scholar databases. A focused concept-driven search was used. The literature review process followed in this study is illustrated in Figure 2.

The first cluster captured restoration scholarship using search terms: “ecological restoration” OR “ecosystem restoration” OR “rehabilitation” OR “restoration ecology.” The second cluster captured adaptive and resilience-oriented approaches using: “adaptive restoration” OR “adaptive management” OR “resilience-based restoration” OR “social-ecological resilience” OR “iterative restoration.” The third cluster captured mining and post-disturbance contexts using: “mine rehabilitation” OR “post-mining restoration” OR “mine closure” OR “mining landscapes” OR “degraded landscapes.” The three clusters were combined using AND (restoration terms AND adaptive terms AND mining terms), while synonymous terms within each cluster were combined using OR. This ensured that retrieved records contained restoration-related scholarship alongside adaptive or resilience-oriented approaches within mining or degraded landscape contexts. In addition, a title-specific topic search for “adaptive restoration” was conducted in Web of Science to analyse publication trends and citation patterns over time (Figure 3).

Inclusion and exclusion criteria

Records retrieved from database searches were screened according to predefined eligibility criteria (Table 1).

Table 1. Criteria for inclusion and exclusion.

Study screening and selection

All retrieved records were exported to Mendeley Desktop for reference management and duplicate removal. The remaining records were screened in two stages: (i) Title and abstract screening to exclude clearly irrelevant studies. (ii) Full-text assessment to confirm eligibility against the inclusion and exclusion criteria. During full-text screening, reasons for exclusion were documented (e.g., non-restoration focus, absence of adaptive frameworks, and unrelated sectoral focus). The number of records retained at each stage-identification, screening, eligibility, and inclusion-is presented in the PRISMA flow diagram (Figure 2).

Data extraction and synthesis

For each eligible study, data were extracted using a structured review template. Extracted fields included: Bibliographic details (authors, year, and publication source), study type (theoretical, conceptual, empirical, and case study), definitions and conceptualizations of adaptive restoration, core principles and frameworks proposed, links to adaptive management and resilience theory, application context (ecological restoration, mining rehabilitation, and socio-ecological systems), governance and institutional dimensions, reported challenges, limitations, and research gaps. A combined chronological and thematic synthesis approach was employed. Chronological analysis mapped the evolution of adaptive restoration scholarship and identified foundational contributions shaping the concept's development (Table 2). Thematic synthesis grouped findings into key domains, including: conceptual foundations of adaptive restoration, links between restoration and resilience thinking, adaptive governance and institutional frameworks, socio-ecological integration in restoration practice, applications in mine rehabilitation and post-mining transitions. This dual synthesis approach enabled the development of an integrated conceptual framework identifying key drivers, components, and pathways for operationalizing adaptive restoration in mining rehabilitation contexts.

Table 2. Chronological overview of foundational references defining the concept of adaptive restoration.

To preserve the methodological integrity of the systematic literature review, the South African practice vignettes presented later in the Discussion are not treated as part of the systematic review dataset. Rather, they are included as anonymized illustrative applications that demonstrate how the conceptual principles identified in the peer-reviewed literature may translate into mine rehabilitation and closure contexts. This distinction is necessary because the review dataset was limited to peer-reviewed academic literature, whereas the practice vignettes draw on anonymized professional rehabilitation, closure and nature-risk assessment experience.

Results and discussion

Chronological evolution of adaptive restoration scholarship

The exact phrase “adaptive restoration” has only recently gained visibility in the scientific literature. A Web of Science topic search shows that publications explicitly using the term began appearing from 2015 onwards (Figure 3), indicating that the concept has emerged as a distinct area of research only in the past decade. Although the number of publications remains relatively low, citation patterns suggest growing scholarly interest in adaptive restoration. It is important to note, however, that the concept itself predates the use of the specific term. Foundational contributions, such as Raupp et al. (2024), established the principles of adaptive restoration well before the terminology became widely adopted, as summarized in Table 2.

Common beginnings of adaptive restoration

Adaptive restoration has its roots in early critiques of conventional restoration ecology and mine rehabilitation approaches that presumed predictable, linear recovery strategies (Moyo et al., 2021). However, experience with extensively degraded systems indicates that such assumptions rarely hold, especially in landscapes affected by intensive disturbance (Birhanu et al., 2024). Adaptive restoration is a dynamic and iterative approach to repairing and revitalizing degraded ecosystems. It recognizes the inherent uncertainty, non-linearity, and context dependence of ecological recovery processes (Raimundo et al., 2018). As a result, restoration is recognized as an ongoing process of adjustment, learning, and responsiveness to ecological feedbacks over time, rather than the accomplishment of a specific endpoint.

Scholars working in severely disturbed environments, such as mined lands, industrial wastelands, and modified river basins, were among the first to highlight the limitations of static restoration targets and rigid success criteria. These early contributions revealed that fixed closure targets often fail to account for unexpected system responses, temporal lags, and cumulative effects. This proposed restoration practice emphasized experimentation, systematic monitoring, and iterative decision-making as critical mechanisms for navigating uncertainty, laying the conceptual foundation for adaptive restoration as a learning-based approach rather than a purely technical intervention.

Building on these findings, Conrado-Silva, de Niemeyer, and Pires argue that adaptive restoration should move beyond prescriptive models and actively incorporate learning into restoration design and implementation. They suggest that adaptive restoration should: (1) Embrace uncertainty; (2) integrate continuous monitoring and feedback; (3) Allow flexible goal-setting; and (4) Support iterative decision-making.

Situating adaptive restoration in resilience and restoration

Adaptive restoration is closely aligned with resilience thinking, particularly the view of ecosystems as dynamic systems capable of multiple stable states. To understand adaptive restoration, we need to understand its revolution as well. “Evolutionary ecological” is firstly introduced by Soulé and Wilcox (1980). Since the 1980s, a lot has changed and will continue to change, in terms of how ecosystem are perceived. Connecting the goals and how success of restoration is measured, not all activities are under the mandate of “restoration.” Another approach to environmental problem-solving emerged from ecological theories of diversity and stability developed in the 1950s (MacArthur, 1955) and 1960s (Lewontin, 1969), as well as from the holistic thinking of early ecologists, such as Aldo Leopold and his predecessors. This approach focused on the concept of resilience.

Situating adaptive restoration within resilience thinking requires depth recognition and understanding of the fact that perspectives on ecological restoration have evolved alongside advances in ecological theory. Over the years, restoration have evolved from efforts to recreate past ecosystem states toward approaches that enhance ecosystem function, adaptation, and long-term resilience in the face of changing environmental and social conditions. This shift reflects broader changes in ecological thinking, especially the developing knowledge that ecosystems are dynamic, can change in non-linear patterns, and may exist in more than one stable state. Adaptive restoration stems from this conceptual shift, which recognizes uncertainty and continuous change rather than fixed ecological objectives. Figure 4 shows the evolution of restoration perspectives, emphasizing the conceptual transition from past-focused restoration, and more recently, to function-focused restoration and resilience-based restoration approaches.

Research conducted by Young et al. (2022) suggests that human activities in many ecosystems have significantly changed native plant and animal populations, contributing to the introduction of invasive species, transformed natural communities through extractive land uses such as agriculture and mining, degraded water quality, and reduced soil integrity. As such, the mandate of early approaches to ecological restoration was to correct or emulate pre-disturbance or historical ecosystem conditions. The success of restoration was determined by how closely a restored ecosystem matched the reference condition. This approach was assumed on the idea that ecosystems were stable and the past conditions could be clearly identified and recreated. Although this strategy is important in shaping early restoration practice, it was limited by its dependence on fixed historical baselines and did not fully consider ongoing ecosystem change.

Hirsch (2020) argue that Ecological restorations often focus on restoring communities while ignoring ecosystem functioning, or on ecosystem functioning without regard to communities. The shift into the function-focused restoration era: as ecological knowledge also improved, restoration efforts began to focus on restoring how ecosystems operate rather than recreating past ecosystem conditions (Silliman et al., 2024). O'Brien et al. (2022) empathize that restoring ecosystem functions requires multiple species to function. Each species has a unique characteristics that contribute to the overall health and functionality of the ecosystem. Ecological restoration should not be undertaken in isolation; rather, it should be approached within the broader social–ecological system, where connectivity, cross-scale interactions, and ongoing change are recognized and prioritized in order to enhance long-term resilience and adaptive capacity.

Resilience-based restoration is a modern shift of ecological restoration, whereby the relationship between human and nature takes center stage. Moreover, we argue that within ecological restoration, there is a well-recognized need to transform resilience from a theoretical concept to its practical application, as this allows for a broader understanding of how resilience functions in different scenarios and how restoration actions can be implemented to effectively strengthen ecosystem resilience. However, Ren and Coffman (2023) argue that one of several reasons why this transition is so difficult is because of a limited amount of detailed guidance on how to actually undertake actions to build resilience.

Concurrent conversations in restoration and mine rehabilitation

The discussion on ecological restoration and mine restoration has evolved, however human disturbances outweighs the recovery of the ecosystem. Ecological restoration and mine rehabilitation have shifted from past approaches that dealt with restoring previously affected ecosystems, toward concepts that promote the functionality of the ecosystem, long-term sustainability, and resilience to ongoing environmental change (Aerts and Honnay, 2011). However, report by Research Program on Water, Land and Ecosystems reveal that the complex interaction of human activities and environmental factors often makes it challenging to restore degraded landscapes to their original ecological condition. While restoration efforts can result in significant improvements, they may not always be able to restore the land's full ecological potential, thereby requiring adaptive, function-oriented restoration strategies (Fabricius et al., 2022).

Literature on ecological restoration highlights the importance of integrating social-ecological systems as approaches to resolve environmental degradation, recognizing the role of human activities, governance systems, and community needs in shaping restoration outcomes (Zolak Poljašević et al., 2025). Conversations in restoration depict that the success of restoration is shaped not only by ecological processes but also by social factors such as land-use decisions, governance structures, institutional capacity, and community values. We believe that mine rehabilitation has similar considerations as ecological restoration, whereby a successful rehabilitation planning is determined by the integration of post-mining land use planning, stakeholder engagement, and socio-economic recovery. Both restoration ecology and mine rehabilitation are blended on adaptive, resilience-based strategies that prioritize ecosystem function, social relevance, and long-term landscape sustainability rather than fixed historical replication.

Characterizing adaptive restoration

The role of adaptive restoration

Adaptive management as a principal that provides the practical through which restoration actions are implemented, monitored, and continuously refined in response to ecological feedback, are the foundation of adaptive restoration. Traditional restoration approaches often dependent on fixed targets based on historical reference conditions; however, in seriously impacted and rapidly changing environments, ecological responses to restoration interventions are frequently unexpected and non-linear. In this context, Thom (2000) and Moyo et al. (2021) emphasizes that continual monitoring plays a crucial part in determining whether restoration interventions are successful in steering ecosystems toward intended functional or resilience-oriented targets and informing needed adjustments to restoration strategies over time. Traditional restoration approaches depending on fixed targets are addressed by adaptive management through the process of treating restoration as continuous processes in which management actions function as experiments, outcomes are systematically evaluated, and subsequent decisions are adjusted based on emerging evidence (Ellison et al., 2020).

Adaptive restoration builds on the principles of adaptive management by incorporating structured learning into ecological restoration practice. Furthermore, research by Roux et al. (2025) states that adaptive restoration is a continuous learning process that uses systematic and purposeful learning about ecosystem conditions and their responses to restoration interventions, disturbances, and ongoing environmental change to continuously refine and improve restoration outcomes. Figure 5 illustrates adaptive restoration as an iterative learning process.

Rist et al. (2013) argue that for restoration to be successful, clear restoration goals must be defined in collaboration with key stakeholders, linked to measurable objectives, and translated into strategic interventions. Continuous monitoring of ecosystem is required to effectively respond to the management interventions. The evaluation of these interventions to assess whether they are steering the system toward desired ecological functions, resilience outcomes, or social-ecological objectives. The feedback from monitoring is used to adjust management strategies in a core plan to keep restoration responsive to new information and changing environmental conditions (Rist et al., 2013). This cycle hones learning, flexibility, and reflexivity as core components of restoration practices.

The role of scale: aligning restoration actions with social and ecological systems

According to (Lynch et al., 2022), scaling restoration needs more than just implementing individual site-level projects; it also needs larger systemic changes in land management practices, distribution of resources, and community engagement. Restoration effectiveness may be affected when governance systems function at scales that do not line up with ecological processes, resulting in fragmented interventions and limited long-term outcomes (Frietsch et al., 2024; Edlmann and Grobbelaar, 2021). Consequently, it is essential to achieve coherence between the spatial and institutional scales used to undertake restoration actions and the scales utilized by social and ecological systems. Aligning these scales allows restoration programs to account for interconnected ecosystem processes and governance structures, promoting the long-term sustainability of both social and ecological systems.

Decentralized, diversity, and flexibility in restoration systems

Governance and institutional frameworks are becoming more prevalent as essential enablers of decentralization, diversity, and flexibility in restoration systems. These principles lead to improved restoration outcomes by redistributing decision-making authority, fostering stakeholder participation, and supporting governance systems that can respond to ecological uncertainty (Noulèkoun et al., 2021). Decentralized governance shifts authority from centralized institutions to local actors, allowing restoration efforts to be shaped by site-specific ecological conditions and socioeconomic aims.

Such decentralization introduces redundancy in governance systems, increasing restoration systems' ability to maintain functionality in the face of disruptions or unforeseen environmental change. Institutional diversity, which includes governmental, private, and community-based actors, enhances adaptive restoration by integrating multiple knowledge systems and supporting collaborative learning processes. The flexibility of governance frameworks allows restoration strategies to be adapted in response to monitoring feedback and evolving ecological dynamics. Together, decentralized, diversified, and adaptable restoration systems increase institutional capacity and allow restoration programs to respond dynamically to emergent issues related to the environment. These governance qualities promote iterative learning and contribute to ensure the long-term sustainability of adaptive restoration projects (Asiimwe et al., 2025).

Disentangling the role of the desired post-mining state

Life after mining has increasingly attracted attention across research, policy, and practice. This increased attention is influenced by the increasing number of mines that have closed and many more that are expected to close in the next decades (Measham et al., 2024). As a result, mine closure is no longer viewed as a discrete endpoint but as part of a broader transition process that unfolds over time (Figure 6). Reframing closure as transition requires a shift in perspective-from viewing rehabilitation as a final technical stage to understanding it as a long-term social-ecological transformation. This transition requires the recognition of the mining system as integrated in regional economies and interconnected social, institutional, and environmental networks (Boggs et al., 2023). Understanding these interdependencies is critical for developing adaptive restoration strategies that transcend beyond site-level rehabilitation and toward sustainable post-mining futures.

The “desired post-mining state” refers to the long-term ecological, social, and economic conditions that a mined landscape is anticipated to reach after closure. To disentangle its role, distinguish this overarching vision from intermediate technical closure criteria and specific land-use options selected throughout rehabilitation planning. This distinction is crucial because combining a desirable post-mining condition with closure criteria or predefined land uses can lead to rigid planning frameworks that prevent adaptive decision-making. Such conflation risks capturing regions into narrow development pathways or unrealistic restoration efforts, including trying to recreate pristine pre-mining conditions that may no longer be possible.

The ultimate objective of adaptive restoration is to restore ecological integrity and efficient land systems capable of sustaining resilient social-ecological futures. However, progress toward this goal is restricted by ongoing challenges in implementing sustainable mining methods that prioritize environmental responsibility throughout the extraction process (Zine et al., 2024). These constraints underscore the need to separate between long-term restoration visions and short-term technical compliance measures, allowing for more adaptable and context-responsive restoration pathways.

Understanding how adaptive restoration emerges in mine rehabilitation

Environmental and socio-economic imperatives underscore the urgent need for restoration, as unrehabilitated mine lands contribute to widespread of environmental degradation (Tedesco et al., 2023). The discourse around restoration and rehabilitation from scholars suggest that ecosystem rehabilitation stretches beyond the mere establishment of vegetation in affected landscapes, it is rehabilitation that involves the reconstruction of ecosystem (van Wyk and Haagner, 2025). This discussion highlights the importance of ongoing planning, coordinated management, and responsible stewardship to keep rehabilitated land environmentally sustainable and economically productive.

An illustrative example noted by Roberts (2010) and Roberts et al. (2012) is the Buffelsdraai Landfill Site Community Reforestation Project in KwaZulu-Natal Province which serves as an example of adaptive restoration practices that can be used to influence mine rehabilitation approaches. This initiative, launched in 2008 as a collaboration between the eThekwini Municipality (Durban City) and the Wildlands Conservation Trust, intended to restore land that had been utilized for sugarcane agriculture for over a century. The key factor for this intervention was the need to offset some of the city's carbon footprint from hosting the 2010 FIFA World Cup, demonstrating the integration of ecological restoration with socio-environmental objectives (Matiza et al., 2024). The success of these initiatives has been proven to be dependent on adaptive management approaches, particularly constant consultation and participation with local communities to reduce resource exploitation and illegal harvesting activities (Nkambule et al., 2017; Roberts, 2010). This case underscores key elements related to mine rehabilitation: restoration strategies must be flexible, responsive to ecological problems, and incorporate socioeconomic considerations, particularly community engagement (Rosa et al., 2020). Such adaptive frameworks can improve rehabilitated landscapes' ecological resilience while also aligning with broader sustainability goals such ecosystem services.

Illustrative South African practice vignettes for adaptive mine restoration

The preceding discussion shows that adaptive restoration is most useful where restoration outcomes are uncertain, ecological responses are non-linear, and social-ecological systems continue to change after intervention. To extend this argument into the South African mining context, the following anonymized practice vignettes illustrate how adaptive restoration principles may be operationalized in post-mining and mine-affected landscapes. These examples are not presented as part of the systematic review dataset, but as practice-based illustrations of how adaptive restoration can inform landform rehabilitation, wetland and water-pathway management, closure transition planning, ecosystem-service recovery, and monitoring-based governance.

Case vignette 1: successional rehabilitation in a strip-mined mafic ore landscape

A strip-mined mafic ore landscape in South Africa illustrates the limitations of conventional rehabilitation approaches that assume that reshaping, soil amelioration, seeding and passive recovery will automatically produce a stable post-mining ecosystem. In this case, the disturbed landscape consisted of strip-mined areas, exposed substrates, altered drainage pathways, erosion-prone slopes and vegetation communities at different stages of recovery. Some areas had already been treated as rehabilitated, but field observations indicated that surface stability, erosion control and vegetation establishment remained uneven. The restoration challenge was therefore not simply to achieve short-term vegetation cover, but to rebuild landform–soil–vegetation relationships capable of supporting ecological succession.

The adaptive restoration response shifted the rehabilitation logic from a fixed end-state model toward a staged, successional and evidence-led approach. Rather than attempting to impose a mature vegetation state on disturbed pioneer substrates, the approach treated early successional vegetation as part of the recovery pathway. This aligns with Raupp et al. (2024), showing that adaptive restoration in sand-mined landscapes requires experimentation, monitoring and iterative modification of management actions. It also reflects the broader argument by Montemagni et al. (2026) that restoration becomes adaptive when results are analyzed and used to inform subsequent interventions, rather than when restoration is treated as a once-off design exercise.

In practical terms, the rehabilitation design included surface profiling, stable slope creation, rough-and-loose surface preparation, woody patch establishment, woody debris placement, drainage protection and indigenous vegetation establishment. Such measures are consistent with ecological restoration standards that define restoration as a process of assisting ecosystem recovery, rather than simply producing a visually rehabilitated surface (Young et al., 2022). A staged trial approach allowed learning from one rehabilitation area to inform later phases. This is important because rehabilitation treatments may perform differently across slope position, substrate condition, runoff behavior, erosion pressure and vegetation establishment zones.

The main lesson from this case is that adaptive restoration should begin at the landform-design stage. In mined landscapes, ecological recovery depends not only on species selection, but also on the physical template created for succession. Success should therefore be assessed through indicators such as surface stability, reduced erosion, rainfall infiltration, pioneer species establishment, woody and herbaceous vegetation development, sediment retention and evidence of self-sustaining ecological succession. In this sense, adaptive restoration converts rehabilitation from a once-off technical activity into a monitored learning process that can be refined over time.

Case vignette 2: adaptive restoration of a mine-influenced pan–wetland complex in a mature goldfield

A mature goldfield landscape provides a second example of why adaptive restoration is necessary where mine closure, water management and wetland recovery are inseparable. The landscape included tailings storage facilities, evaporation and return-water infrastructure, seepage pathways, dust-generating surfaces, wetland and pan receptors, altered hydrological connectivity and post-closure water-transition risks. In such settings, the ecological target cannot simply be a return to a pristine pre-mining wetland condition. The system has become a hybrid, mine-influenced receptor landscape shaped by decades of altered hydrology, sediment movement, salinity risk, nutrient loading, dust deposition and closure uncertainty.

Adaptive restoration in this context requires a shift from site-based rehabilitation to pathway-based restoration. The key question is not only what should be rehabilitated, but which source–pathway–receptor linkages must be stabilized to reduce long-term ecological and closure risk. This approach aligns with Measham et al. (2024), who argue that mine closure should be understood as a broader post-mining transition rather than a discrete technical endpoint. It is also consistent with the ICMM integrated mine-closure approach, which frames closure planning as a dynamic and iterative process that should consider environmental, social and economic outcomes throughout the life of mine (Brock, 2021).

In practical terms, the restoration response would prioritize wetland and pan margins, seepage-affected zones, sediment pathways, disturbed grassland–wetland ecotones, evaporation-dam interfaces and hydrologically connected receptor areas. Potential interventions may include seepage interception, wetland-buffer enhancement, passive or semi-passive polishing systems, erosion-control structures, dust-source stabilization, vegetation-based sediment retention and staged rehabilitation of disturbed margins. Monitoring would need to track water quality, hydrological behavior, sediment accumulation, vegetation recovery, wetland-edge condition, dust deposition and ecological integrity.

This case demonstrates that adaptive restoration can function as a closure-readiness tool. The aim is not only to close infrastructure, but to guide a changing mine-influenced landscape toward lower residual risk, stronger ecosystem-service delivery and more credible post-closure performance. This is particularly important in goldfield landscapes, where water-related liabilities, wetland degradation and ecological uncertainty may persist long after mining has ceased.

Case vignette 3: legacy goldfield catchment restoration using source-pathway-receptor logic

A third anonymized case concerns a legacy goldfield catchment where tailings facilities, disturbed landforms, mine-water systems, exposed surfaces, seepage zones, wetlands, dams, streams and community-facing land uses interact through multiple environmental pathways. In this type of landscape, the restoration problem is not confined to a single disturbed site. Instead, ecological degradation is distributed across a catchment-scale system in which water, dust, sediment, vegetation condition, community trust and closure liability are functionally connected.

Adaptive restoration in such a setting should be organized around source–pathway–receptor logic. Sources may include tailings storage facilities, exposed soils, residue deposits, roads, disturbed surfaces, water-management infrastructure and rehabilitating landforms. Pathways include runoff, seepage, sediment transport, dust deposition, altered hydroperiods, dust-to-water transfer and closure-related hydrological change. Receptors include wetlands, seeps, riparian systems, dams, grasslands, agricultural land, downstream water users and nearby communities.

This approach is supported by restoration and mine-closure literature that emphasizes the need to plan restoration through ecosystem services, stakeholder relevance and social-ecological outcomes (Rosa et al., 2020). It is also consistent with adaptive management theory, which treats management actions as structured interventions that must be monitored, evaluated and adjusted as evidence emerges (Ellison et al., 2020). In the South African restoration context, Moyo et al. (2021) similarly highlight the importance of adaptive management, reflection and stakeholder engagement in restoration initiatives.

In practical terms, this type of legacy goldfield catchment could be managed through Living Lab-style restoration pilots linked to monitoring, reporting and verification. Potential interventions include grassland rehabilitation trials, seep and wetland restoration pilots, hybrid wetland-polishing systems, vegetative dust suppression, slope-stabilization trials, soil-recovery experiments and aquatic-receptor monitoring. Rather than applying a uniform rehabilitation prescription across the entire footprint, the site would test interventions under real conditions, compare their performance across receptor types and refine the approach over time.

The key lesson is that adaptive restoration can convert fragmented compliance management into an integrated closure and environmental governance system. Success indicators should therefore extend beyond hectares rehabilitated. They should include reduced exposed-surface risk, improved wetland condition, reduced sediment mobilization, stabilized seepage interfaces, improved vegetation establishment, stronger ecosystem-service delivery, reduced community-facing environmental risk and demonstrable reduction in closure uncertainty.

Case vignette 4: pathway-based adaptive restoration in a transformed platinum-belt landscape

A fourth case, from a transformed platinum-belt landscape, demonstrates the value of adaptive restoration where post-mining ecosystems are hybrid, pathway-sensitive and linked to future land-use performance. The landscape included disturbed landforms, residue facilities, water-management infrastructure, wetland and tributary receptors, rehabilitated areas, dust-prone surfaces and land-use expectations such as grazing or low-intensity agricultural use. In this context, the central rehabilitation question is not only whether disturbed areas can be revegetated, but whether the post-mining landscape can remain safe, stable, uncontaminated and ecologically functional without repeated intervention.

The adaptive restoration logic in this case rests on four linked components: field-based testing, monitoring, adaptive decision-making and measurable reduction in nature-related risk. Field-based testing allows restoration and ecological-engineering interventions to be implemented under real site conditions. Monitoring then determines whether these interventions improve vegetation establishment, dust control, soil stability, water quality, wetland condition, sediment retention and ecosystem-service delivery. Management actions are then refined based on observed performance rather than assumed success.

This approach is particularly important in mined landscapes where ecological and engineered systems overlap. Wetland margins, tributary buffers, tailings facility shoulders, drainage low points, rehabilitated slopes and disturbed surfaces may all function as hybrid ecological infrastructure. Depending on how they are managed, these areas may remain long-term liabilities or be improved to provide hydrological attenuation, sediment retention, dust buffering, soil stabilization, habitat support and improved closure performance. Recent mine-rehabilitation research similarly emphasizes the importance of setting restorative goals at regional and landscape scale, because mine rehabilitation outcomes can influence broader landscape connectivity and ecological function (Hernandez-Santin et al., 2024).

The main lesson is that adaptive restoration strengthens closure defensibility by reducing uncertainty before full-scale implementation. It allows practitioners to ask practical questions: which surface treatments reduce dust and sediment loading, which wetland-edge interventions improve stability, which rehabilitation designs can tolerate local runoff and grazing pressure, and which hybrid systems should be restored, engineered, stabilized or monitored over a longer timeframe? This is a more realistic model for South African mined landscapes than treating rehabilitation as a once-off compliance exercise.

Synthesis from the practice vignettes

Collectively, these vignettes show that adaptive restoration in South African mined landscapes is not a single technique, but a governance and learning architecture. Across strip-mined, goldfield and platinum-belt settings, the common requirement is to move beyond static rehabilitation prescriptions toward monitored, evidence-led and pathway-aware restoration. The cases also show that adaptive restoration is most valuable where uncertainty is high: disturbed substrates do not follow predictable successional trajectories; wetlands and hybrid aquatic systems respond to changing water regimes; dust, seepage and sediment pathways connect operational sources to ecological and community receptors; and closure outcomes depend on long-term ecological performance. In this sense, adaptive restoration provides a practical bridge between restoration ecology, ecological engineering, mine closure planning and environmental governance. It allows post-mining landscapes to be managed as dynamic social-ecological systems rather than as static compliance footprints.

These practice vignettes therefore reinforce the need to frame post-mining landscapes as social-ecological systems in which ecological recovery, closure planning, community outcomes and long-term governance are interdependent.

Building resilient post-mining social-ecological systems

The question of how resilience can be evolved as an imperative concept and analytical lens for understanding the dynamic relationship between humans and their environments is a question that needs to be asked in today's adaptive restoration discourse. In the current era of accelerated environmental change, socioeconomic pressures, and climatic uncertainty, resilience thinking provides a framework for exploring how interconnected human-natural systems respond, adapt, and expand in the face of disturbance (Perera et al., 2024). Lerback et al. (2022) further add that its application has spread widely across multiple disciplines, including environmental management, sustainability science, disaster risk reduction, and ecological restoration, where it is used to assess system stability, recovery capacity, and long-term adaptability.

Within this discourse, the mining industry plays a critical role in shaping sustainability trajectories, especially in resource-dependent regions (Sun and Hasi, 2024; Sypień et al., 2025). Research by Wu et al. (2021) support our statement that mining landscapes are a distinct and influential form of land transformation that typically includes large-scale physical changes of terrain, hydrological systems, vegetation cover, and settlement patterns. As such, these changes have complicated and long-lasting social and ecological implications, such as land degradation, biodiversity loss, water contamination, disrupted livelihoods, and increased vulnerability of surrounding communities. As a result, mining zones serve as extremely sensitive social-ecological systems in which environmental integrity and human wellbeing are immensely interconnected.

In the South African context, empirical studies by Deng et al. (2025) and Kafu-Quvane and Mlaba (2024) point out the considerable environmental pressures posed by mining activities, contributing to concerns about ecological degradation and socioeconomic instability in affected communities. These findings underscore the importance of shifting beyond conventional rehabilitation approaches and toward integrated strategies that foster both ecological recovery and community adaptive capacity. Consequently, exploring innovative approaches to promote community resilience within mining ecosystems is critical to achieving sustainable development outcomes. Strengthening resilience not only improves environmental restoration, but also improves local communities' ability to adapt to environmental risks, diversify livelihoods, and participate meaningfully in post-mining transitions (Chipangamate et al., 2023).

Research limitations

This study has several limitations that should be acknowledged. First, although the systematic review employed a comprehensive search strategy across Google Scholar, Scopus, and Web of Science, it was limited to peer-reviewed publications published in English between 2000 and 2025. Consequently, relevant studies published in other languages or within the gray literature may not have been captured. Second, the review relied on published evidence and did not include primary empirical data or stakeholder perspectives from mining practitioners, policymakers, or affected communities. Third, while the study presents illustrative South African practice vignettes to demonstrate the practical application of adaptive ecological restoration, these are intended to provide contextual examples rather than empirical validation of the proposed conceptual framework. Despite these limitations, the systematic synthesis provides a robust overview of the evolution, principles, and application of adaptive ecological restoration within mine rehabilitation and offers a valuable foundation for future empirical and comparative research.

Theoretical and practical implications

Theoretical implications

This review advances the adaptive ecological restoration literature by providing a chronological synthesis of its conceptual evolution and by developing an integrated conceptual framework for adaptive mine restoration. By consolidating diverse strands of scholarship across ecological restoration, adaptive management, resilience thinking, and environmental governance, the study clarifies the core principles underpinning adaptive restoration and addresses existing conceptual inconsistencies within the literature. The proposed framework provides a foundation for future empirical research by identifying the key drivers, governance mechanisms, and adaptive processes necessary to support resilient post-mining landscapes.

Practical implications

The findings have important implications for mine rehabilitation policy and practice. The review demonstrates that adaptive ecological restoration offers a practical framework for managing uncertainty through long-term monitoring, iterative decision-making, stakeholder participation, and integrated environmental governance. For regulators, mine operators, environmental practitioners, and policymakers, the proposed framework can inform the design, implementation, monitoring, and evaluation of mine rehabilitation programes. The South African practice vignettes further illustrate how adaptive restoration principles may be applied to strengthen post-closure planning, improve ecosystem resilience, and support more sustainable and socially responsive mine rehabilitation outcomes.

Conclusion

This review set out to examine the evolution of adaptive restoration scholarship and its relevance for mine rehabilitation, with particular reference to South Africa's mining landscapes. Through a systematic synthesis of the literature, the study demonstrates that adaptive restoration has evolved from a concept rooted primarily in adaptive management and ecological resilience into a broader governance framework that integrates ecological, social, institutional, and landscape-scale considerations. Rather than viewing rehabilitation as the achievement of a predetermined end state, adaptive restoration recognizes restoration as a continuous process of learning, monitoring, evaluation, and adjustment in response to changing environmental conditions and emerging knowledge. The review identified three broad phases in the development of adaptive restoration scholarship. Early research established the theoretical foundations of adaptive management and resilience thinking; subsequent studies expanded the discourse by incorporating social-ecological systems, stakeholder participation, and governance perspectives; while recent scholarship has increasingly emphasized climate adaptation, ecosystem services, landscape connectivity, and long-term post-mining transitions. Together, these developments illustrate a clear shift from static rehabilitation models toward adaptive, evidence-based, and systems-oriented restoration approaches.

Building on these findings, this paper proposed a conceptual framework for adaptive mine restoration that integrates iterative decision-making, ecosystem monitoring, stakeholder engagement, governance, and resilience into a coherent restoration process. The inclusion of illustrative South African practice vignettes further demonstrates how adaptive restoration principles may be operationalized across different mining contexts, including landform rehabilitation, wetland restoration, catchment management, and mine closure planning. Although these vignettes are not part of the systematic review dataset, they provide practical illustrations of how adaptive restoration can strengthen environmental governance and improve restoration outcomes under conditions of uncertainty.

The findings suggest that adaptive restoration offers a valuable pathway for addressing the complex environmental and social challenges associated with mine rehabilitation in South Africa. By embracing continual learning, monitoring, and adaptive decision-making, restoration programes can move beyond compliance-driven rehabilitation toward more resilient post-mining landscapes capable of supporting ecological integrity, ecosystem services, and sustainable livelihoods. This approach is particularly relevant in regions characterized by abandoned mines, legacy pollution, changing climatic conditions, and competing post-mining land-use expectations.

Recommendations for future research and practice

While this review provides a comprehensive synthesis of the existing literature, it is limited to peer-reviewed publications and does not empirically evaluate restoration outcomes in operational mining environments. Future research should therefore examine the implementation of adaptive restoration within active and closed mine sites, develop measurable indicators for assessing adaptive restoration performance, and explore how governance arrangements, community participation, and monitoring systems influence long-term rehabilitation success. Comparative studies across different mining commodities and geographical regions would also contribute to refining adaptive restoration as a practical framework for achieving sustainable mine closure and resilient social-ecological systems.

Ultimately, adaptive restoration should not be viewed simply as a rehabilitation technique, but as a governance and learning framework capable of guiding post-mining landscapes through uncertainty and change. By integrating ecological science, adaptive management, stakeholder collaboration, and long-term monitoring, adaptive restoration provides a scientifically robust and practically relevant foundation for achieving more resilient, sustainable, and socially responsive mine rehabilitation in South Africa and other mining-dependent regions.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Naser Valizadeh, Shiraz University, Iran

Reviewed by: Takunda Shabani, Donga Rural Service Centre, Zimbabwe

Aaron Tettey Tetteh, University of Energy and Natural Resources, Ghana