{
  "id": "restoration-ecology",
  "title": "Restoration Ecology",
  "lang": "en",
  "source_lang": "en",
  "summary": "Restoration ecology investigates how degraded ecosystems recover and develops the scientific foundations for assisting and evaluating their recovery.",
  "category": "nature",
  "subcategory": "ecology",
  "type": "concept",
  "aliases": [
    "Ecological restoration science"
  ],
  "generated_at": "2026-11-04",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/restoration-ecology/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/restoration-ecology/",
      "json": "https://aiwiki.page/zh/restoration-ecology.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Restoration ecology is the scientific study of the recovery of degraded, damaged, or destroyed ecosystems. It investigates how ecological communities and processes respond when disturbances are reduced or corrective interventions are introduced, and why some recovery trajectories succeed while others stall. It provides the scientific foundation for ecological restoration, the practical activity of assisting ecosystem recovery, while also using restoration projects as experiments that test ecological theory.\n\nThe field connects ecology, environmental history, soil science, hydrology, genetics, and social research. Its subjects range from restoring native vegetation on a former mine to reconnecting a river with its floodplain or rebuilding an oyster reef. Recovery is assessed through multiple attributes, including species composition, physical conditions, ecological interactions, and the capacity to persist without indefinite corrective intervention. Restoring one attribute does not necessarily restore the others.\n\nRestoration ecology has become increasingly important as land conversion, pollution, biological invasions, and climate change affect ecological systems worldwide. Its central questions concern both possibility and limits: which processes can be re-established, how past conditions should inform future goals, how long recovery takes, and how ecological outcomes interact with land use, livelihoods, and environmental governance."
    },
    {
      "heading": "Historical development and disciplinary identity",
      "text": "Practices that repair damaged landscapes long predate restoration ecology as a formal discipline. Communities have maintained productive landscapes through controlled burning, water management, rotational harvesting, and the protection of culturally important species. Such practices are not automatically ecological restoration in the modern sense, but they demonstrate that human intervention and ecological persistence are not necessarily opposites. Historical research also shows that many landscapes once described as untouched were shaped by long-standing human management.\n\nA frequently cited precursor of modern experimental restoration is the prairie work begun at the University of Wisconsin–Madison Arboretum in the 1930s. Aldo Leopold and colleagues sought to reconstruct native ecological communities on former agricultural land rather than merely establish attractive vegetation. Work on what became Curtis Prairie began in 1935. Experiments involving native planting and prescribed fire helped demonstrate that the recovery of tallgrass prairie depended on disturbance regimes and species interactions, not simply the presence of an initial seed mixture.\n\nDuring the second half of the twentieth century, mine reclamation, wetland mitigation, river rehabilitation, and endangered-species management generated additional practical knowledge. These activities differed in purpose. Mine reclamation often emphasized stable landforms, erosion control, and vegetation cover; ecological restoration increasingly sought recovery of native communities and their processes. Recognizing that distinction helped define a research field concerned with ecological integrity rather than vegetation establishment alone.\n\nInstitutional consolidation accelerated in the late 1980s and early 1990s. The Society for Ecological Restoration was founded in 1988, and the journal Restoration Ecology began publication in 1993. The society’s 2004 primer and later international standards helped establish a shared vocabulary for reference ecosystems, recovery trajectories, and evaluation. The second edition of its international principles and standards appeared in 2019.\n\nRestoration ecology developed alongside conservation biology but addresses a somewhat different set of questions. Conservation commonly emphasizes preventing losses and maintaining existing populations and habitats; restoration investigates how to recover attributes already lost or impaired. The two are complementary because restored areas depend on surviving populations, source habitats, and protected ecological processes. Restoration also has an unusual experimental role: attempts to rebuild communities expose gaps in explanations of how those communities assemble and function."
    },
    {
      "heading": "Recovery mechanisms, succession, and ecological thresholds",
      "text": "An early expectation in restoration was that removing a damaging activity would allow a predictable sequence of recovery. Research on ecological succession showed why this can occur, but also why it cannot be assumed. Abandoned farmland may develop toward woodland where seed sources, suitable soils, and sufficient moisture remain. Elsewhere, persistent grasses, altered fire regimes, or a lack of seed dispersers may prevent trees from establishing. Similar starting conditions can therefore produce different outcomes.\n\nRecovery is shaped by both abiotic filters and biological interactions. Soil acidity, salinity, water availability, and nutrient concentrations determine which organisms can survive. Competition, herbivory, disease, and mutualism then influence establishment and persistence. Planting native trees into compacted soil, for example, may produce little recovery if roots cannot penetrate the substrate. Conversely, correcting physical conditions may be sufficient where native species can recolonize naturally.\n\nThe order in which organisms arrive can have lasting consequences. These priority effects occur when early colonists alter resources, microclimate, or competitive conditions for later arrivals. Dense exotic grasses can intercept light and water before native seedlings establish. Early native vegetation may instead suppress weeds, attract dispersers, and create conditions favorable to additional species. Restoration experiments manipulate sowing order, planting density, and arrival timing to distinguish these mechanisms.\n\nPositive feedbacks can also maintain degraded conditions. Loss of vegetation may increase erosion, which removes fine soil and organic matter, further reducing plant establishment. In shallow lakes, suspended sediment and algal growth can reduce light enough to prevent submerged vegetation from returning even after nutrient inputs decline. These situations illustrate ecological thresholds: reversing the original pressure may not immediately reverse the resulting state. Additional intervention may be needed to overcome the feedback.\n\nEcological resilience concerns the capacity of a system to absorb disturbance without losing its organizing processes and identity. Restoration may seek to recover that capacity by rebuilding functional groups, spatial heterogeneity, or disturbance regimes. Yet degraded systems can also be resilient in an undesirable state. Resilience is therefore not synonymous with ecological quality. Research distinguishes resistance to disturbance, the rate of recovery afterward, and the persistence of particular community configurations.\n\nThese mechanisms explain the contrast between passive and active restoration. Passive approaches remove pressures and allow recovery; active approaches manipulate organisms or physical conditions directly. The distinction is a continuum rather than a strict division. Excluding livestock, controlling fire, or restoring a water supply can require substantial management while leaving species assembly largely to natural processes."
    },
    {
      "heading": "Reference ecosystems and the definition of recovery",
      "text": "Restoration requires an explicit account of what recovery means. A reference ecosystem provides that account by representing the ecological attributes expected in the absence of the relevant degradation, while allowing for environmental change. It is not necessarily a single untouched site. A reference model may combine several relatively intact sites with historical documents, aerial photographs, soil evidence, species records, and knowledge held by local and Indigenous communities.\n\nReference conditions must distinguish human-caused degradation from natural variation. A river may naturally alternate between channels, floodplain wetlands, and exposed sediment bars. A fire-maintained woodland may periodically lose much of its aboveground biomass without losing its ecological identity. Defining either system from one photograph or survey risks mistaking a temporary state for a permanent target. Reference models therefore often specify ranges, spatial mosaics, or recurring processes rather than one fixed species list.\n\nHistorical baselines are particularly difficult where ecological change preceded systematic observation. Pollen and charcoal records, tree remains, archaeological evidence, and early land surveys can reveal former vegetation and disturbance regimes. Each source has limitations: pollen production varies among species, written accounts can be selective, and surviving reference sites may occupy unusually inaccessible or unproductive locations. Combining independent evidence reduces dependence on any one imperfect record.\n\nRecovery is usually multidimensional. A restored woodland may regain canopy cover before it recovers understory plants, cavity-bearing trees, or decomposer communities. A wetland may resemble a reference site in vegetation while retaining altered soil chemistry. The Society for Ecological Restoration’s standards organize assessment around attributes including the absence of threats, physical conditions, species composition, structural diversity, ecosystem function, and external exchanges. These categories help prevent a single convenient indicator from standing in for the whole ecosystem.\n\nTargets also differ among ecological restoration, rehabilitation, reclamation, and habitat creation. Rehabilitation may recover useful ecological functions without seeking the full composition of a reference ecosystem. Reclamation may emphasize safe and stable land after industrial use. Creating a treatment wetland can deliver valuable water-quality benefits without recreating the wetland type that previously occupied the site. Clear terminology makes these differences visible rather than treating all environmental improvements as equivalent.\n\nChanging environmental conditions make reference models dynamic but do not make them arbitrary. Historical evidence still identifies native species pools, ecological interactions, and causes of degradation. Future climate and land-use projections help assess which attributes remain feasible. The resulting target is a reasoned ecological model with stated uncertainties, not simply a preference for a particular landscape appearance."
    },
    {
      "heading": "Diagnosing degradation and designing interventions",
      "text": "Effective restoration begins with a diagnosis of the processes preventing recovery. Low native plant cover is an observation, not an explanation. Its causes might include seed limitation, excessive grazing, unsuitable hydrology, nutrient enrichment, recurrent fire, or soil toxicity. These mechanisms require different interventions. Adding seeds where seedlings are continually eaten can waste resources; excluding herbivores where no viable propagules remain may produce little change.\n\nDiagnosis operates across spatial scales. At the site level, investigators may measure soil bulk density, salinity, water-table depth, and contaminant concentrations. At the landscape level, they examine nearby source populations, barriers to movement, upstream pollution, and the distribution of land uses. Dispersal limitation is especially important where restored sites are isolated from surviving native communities. Suitable habitat does not guarantee recolonization if organisms cannot reach it.\n\nThe sequence of intervention often matters as much as the intervention itself. Earthworks or contaminant treatment generally precede biological establishment because later disturbance can destroy planted communities. In a drained wetland, restoring water levels may be necessary before wetland vegetation can persist. In a nutrient-enriched grassland, reducing nutrient availability or exporting biomass may be necessary before slow-growing native plants can compete successfully. Some constraints must be addressed together because they reinforce one another.\n\nProject design converts diagnosis into testable objectives. An objective might specify a seasonal groundwater range, recruitment of a target tree species, or reduction of invasive cover within a defined period. It differs from an activity such as planting 10,000 seedlings. The latter records effort; the former describes an ecological outcome. Explicit objectives make it possible to distinguish implementation failure, such as blocked water-control structures, from a mistaken ecological hypothesis.\n\nRestoration increasingly uses adaptive management, in which interventions are treated as opportunities to learn. Managers specify alternative explanations, predict responses, monitor outcomes, and revise actions accordingly. This process is strongest when competing treatments are replicated and compared rather than simply altered through informal trial and error. For example, a factorial experiment combining seed addition with herbivore exclusion can reveal whether establishment is limited by propagules, grazing, or both.\n\nDesign must also consider unintended effects. Soil disturbance can expose buried weed seeds, herbicide treatment can affect non-target plants, and new channels can redistribute flood risk. Small pilot trials and staged implementation can reveal such effects before interventions are expanded. Not every restoration project permits tightly controlled experiments, but documenting initial conditions, treatment details, and deviations from the plan makes its results more interpretable."
    },
    {
      "heading": "Terrestrial restoration: soils, vegetation, and disturbance",
      "text": "On land, recovery often begins below ground. Mining, cultivation, and heavy machinery can remove topsoil, compact remaining material, and disrupt soil organisms. These changes affect infiltration, aeration, root growth, and nutrient cycling. Restoration treatments may include reshaping unstable slopes, relieving compaction, replacing topsoil, or adding organic material. Their effects depend on context: nutrient-rich amendments can stimulate vegetation on severely depleted substrates but favor competitive weeds in naturally nutrient-poor ecosystems.\n\nSoil organic matter influences water retention, aggregation, and nutrient supply, yet rebuilding it can take much longer than establishing plant cover. Roots, litter, and microbial activity gradually generate new organic inputs. Fungi and bacteria mediate many of these transformations, while soil animals redistribute organic material and alter pore structure. Inoculation with soil organisms is sometimes tested where these communities have been severely depleted, but outcomes depend on compatibility, substrate conditions, and the organisms introduced.\n\nVegetation restoration ranges from natural regeneration to direct seeding and planting. Natural regeneration can be effective where remnant trees, seed banks, and animal dispersers remain. Assisted regeneration may protect recruits, suppress competing vegetation, or create sheltered establishment sites. Planting becomes more important where desired species are absent or dispersal is too slow to meet project goals. Different methods can be combined within one landscape rather than applied uniformly.\n\nThe source and diversity of planting material influence both establishment and future adaptability. Locally collected seed may reflect adaptation to local soils and seasonal conditions, but small source populations can contain limited genetic variation. Moving material from elsewhere can increase diversity while introducing risks of maladaptation or disruption of locally adapted gene combinations. Restoration genetics therefore examines environmental matching, population structure, and the scale over which seed transfer is appropriate.\n\nDisturbance regimes are essential in many terrestrial ecosystems. Grasslands and savannas may depend on particular combinations of fire, grazing, and seasonal drought. Excluding all disturbance can favor woody encroachment or a few dominant grasses rather than recover the intended community. Conversely, frequent intense burning can prevent woodland regeneration. Prescribed disturbance is consequently designed around timing, frequency, intensity, and spatial pattern, not merely whether disturbance occurs.\n\nTerrestrial restoration cannot be reduced to tree planting. Afforesting an ancient grassland can damage its native biodiversity, alter water use, and eliminate open-habitat species. In forests, dense single-species plantations may establish tree cover without recovering the structure and composition of native woodland. Recovery of old-growth forest attributes, such as large cavities, substantial deadwood, and complex vertical structure, may require many decades or centuries even where young trees grow rapidly."
    },
    {
      "heading": "Freshwater, coastal, and marine restoration",
      "text": "Aquatic restoration is strongly governed by water movement and connectivity. A wetland’s hydroperiod—the timing, duration, and frequency of inundation—helps determine its vegetation, oxygen conditions, and nutrient transformations. Rewetting drained land can restore some of these processes, but simply adding water may be insufficient if subsidence has changed surface elevation or if nutrient-rich soils produce persistent algal growth. Water depth, seasonal variability, and exchange with surrounding waters all matter.\n\nRiver restoration has moved beyond reshaping short channel sections toward recovering processes at the scale of the drainage basin. Sediment supply, flood flows, riparian vegetation, and the movement of organisms determine channel form and habitat diversity. Reconnecting floodplains can restore inundation and sediment deposition; removing barriers can reopen migration routes. These measures may remain constrained by upstream dams, water abstraction, or diffuse pollution outside the project boundary.\n\nDam removal provides unusually informative examples because it alters flow, sediment transport, and biological access simultaneously. Removal of the two large dams on Washington State’s Elwha River occurred between 2011 and 2014. The project reopened habitat to migratory fish and released large quantities of stored sediment downstream. It also demonstrated that short-term turbidity and channel adjustment can accompany longer-term recovery, making the timing and choice of monitoring indicators particularly important.\n\nCoastal restoration must account for elevation relative to tides and sediment availability. A salt marsh can persist only within a suitable range of flooding and salinity. Mangrove planting often fails when seedlings are placed on unsuitable shorelines or where tidal exchange remains obstructed. Restoring hydrological access can sometimes allow natural recruitment more effectively than mass planting. Seagrass restoration similarly depends on underwater light, sediment stability, water quality, and the survival of early plantings.\n\nMarine restoration includes oyster-reef construction, kelp recovery, seagrass transplantation, and coral reef restoration. Techniques differ substantially because the underlying constraints differ. Oyster restoration may require stable settlement substrate and surviving larval sources. Kelp forest recovery may depend on controlling intense sea-urchin grazing or restoring predator relationships. Coral transplantation can increase local coral cover, but it does not remove regional thermal stress, pollution, or disease exposure.\n\nThese systems illustrate the importance of restoring causes rather than appearances. Artificial structures can provide attachment surfaces without reproducing the full biological functions of a reef. A reconstructed river channel can look natural while remaining disconnected from its floodplain. Aquatic restoration research therefore tests exchanges of water, sediment, nutrients, and organisms alongside visible habitat structure."
    },
    {
      "heading": "Measuring recovery and establishing causal evidence",
      "text": "Restoration monitoring asks whether ecological change occurred, whether the intervention caused it, and whether the change corresponds to the stated goals. These are separate questions. Vegetation may increase after planting because several unusually wet years favored establishment. Fish numbers may rise in a restored reach because fish moved from adjacent habitat rather than because the river supports a larger population. Evaluation requires an appropriate baseline and comparison, not just a favorable observation after construction.\n\nBefore–after–control–impact designs compare change at treated sites with change at untreated sites over the same period. Random allocation and replication strengthen causal inference where feasible. Large projects, however, often lack equivalent controls, and nearby sites may influence one another through dispersal or water movement. Researchers then use multiple reference sites, repeated measurements, phased implementation, and statistical models to separate intervention effects from background variation. No design eliminates the need to understand the ecology of the system.\n\nIndicators can describe composition, structure, function, and persistence. Species richness is useful but cannot distinguish recovery of native specialists from an increase in widespread weeds. Functional diversity records differences in ecological traits, while recruitment, survival, and reproduction indicate whether populations can sustain themselves. Measurements of decomposition, nutrient retention, and soil development assess processes that may respond differently from vegetation cover. Monitoring across several dimensions reduces the risk of declaring success prematurely.\n\nA simple normalized indicator for a variable expected to increase during recovery is\n\n\\[R_t=\\frac{X_t-X_0}{X_{\\mathrm{ref}}-X_0},\\]\n\nwhere \\(X_t\\) is its value at time \\(t\\), \\(X_0\\) is the initial value, and \\(X_{\\mathrm{ref}}\\) is a reference value. This expression can help compare trajectories, but it is not a universal measure of restoration success. Reference values vary, the denominator may be small, and some desirable variables decrease rather than increase. A composite score can also hide trade-offs among attributes.\n\nNew observation methods expand monitoring without replacing field interpretation. Satellite and drone imagery can map canopy development, inundation, and landscape change. Acoustic recording can track calling birds, frogs, and bats. Environmental DNA can detect organisms from genetic material in water or soil, although detection does not necessarily establish local abundance or reproduction. Each method samples particular organisms or properties and has characteristic detection biases.\n\nTime is a major source of uncertainty. Some responses occur within months; soil formation, old-growth structures, and the return of poorly dispersing species can take decades. Monitoring that ends after vegetation establishment may miss later invasion, recruitment failure, or hydrological change. Long-term experiments and consistent reporting of unsuccessful treatments are especially valuable because they reveal recovery limits that short project evaluations cannot detect."
    },
    {
      "heading": "Landscape context and a changing climate",
      "text": "Restoration sites function within larger landscapes. Their ecological outcomes depend on surrounding land use, the size and condition of source habitats, and the ease with which organisms and materials move between patches. Habitat fragmentation can leave a restored patch environmentally suitable but inaccessible to some species. Research in landscape ecology therefore examines patch area, isolation, surrounding land cover, and the spatial arrangement of interventions.\n\nImproving landscape connectivity can involve restoring continuous vegetation, stepping-stone habitats, or migration routes. Connectivity is species-specific: a strip usable by a large mammal may be ineffective for a shade-dependent plant or a freshwater fish. It can also transmit undesirable effects, including invasive organisms, disease, or fire. Corridor placement consequently requires attention to the organisms and processes intended to benefit, rather than assuming that every connection is beneficial.\n\nClimate change adds a moving environmental baseline. Increasing temperatures, altered rainfall, more severe droughts, and changing fire weather can make historical species combinations harder to sustain. Coastal ecosystems face sea-level rise, while reefs experience marine heat stress and ocean acidification. Restoration research examines whether recovery targets remain viable over the expected lifetime of the project, not merely under conditions present at establishment.\n\nOne response is to emphasize ecological processes and adaptive capacity alongside historical composition. Greater structural and genetic diversity may spread risks, while topographic and hydrological variation can provide local refuges. Another approach, assisted migration, deliberately moves organisms toward places expected to become climatically suitable. This raises scientific questions about establishment, ecological interactions, and possible invasion, as well as governance questions about movement beyond historical ranges. Its effects cannot be generalized across all taxa or regions.\n\nRestoration can contribute to climate change mitigation, but carbon storage and ecological recovery are not interchangeable. Rewetting drained peatlands can reduce continuing carbon dioxide emissions while increasing methane emissions, particularly during some stages of recovery. Forest regrowth can accumulate carbon, yet fire, drought, and future land conversion affect permanence. Carbon accounting must also distinguish additional storage from what would have occurred without intervention.\n\nClimate-oriented projects can conflict with other ecological objectives when evaluated through a single metric. Fast-growing plantations may store biomass without recovering native biodiversity; poorly placed tree planting may reduce streamflow or damage naturally open ecosystems. Research therefore evaluates carbon, water, habitat, and disturbance responses together, including benefits and costs displaced beyond the project boundary."
    },
    {
      "heading": "Human institutions, knowledge, and restoration economics",
      "text": "Restoration takes place on land and water governed by property systems, customary rights, regulations, and social relationships. These institutions affect which interventions are possible and who maintains them. A technically effective treatment may not persist if it conflicts with grazing arrangements, fishing access, or water allocation. Conversely, secure tenure and durable local institutions can support maintenance long after initial project funding ends.\n\nLocal and Indigenous knowledge can contribute information about historical species distributions, seasonal water behavior, fire regimes, and culturally important ecological relationships. Such knowledge is not simply a substitute for missing scientific data; it can define different questions and priorities. Collaboration involves decisions about authority, consent, attribution, and control over knowledge. Ecological recovery and cultural continuity can be linked, especially where customary management helped maintain the ecosystem being restored.\n\nEconomic analysis distinguishes installation costs from long-term expenditure. Site preparation, seed collection, engineering, and planting may be followed by years of weed control, replacement planting, monitoring, and infrastructure maintenance. Opportunity costs also matter: land allocated to restoration may no longer be available for cultivation or development. Comparing projects solely by cost per hectare ignores differences in ecological condition, recovery probability, and the functions restored.\n\nBenefits may include ecosystem services such as water purification, flood moderation, fisheries support, and recreation. Some can be expressed in monetary terms; others, including cultural relationships and the persistence of rare species, are not readily represented by market prices. The distribution of benefits is as important as their total value. Floodplain restoration, for example, may reduce downstream flood exposure while requiring changes in land use by upstream landholders.\n\nBiodiversity offsets and compensatory mitigation make measurement especially consequential. An offset claims to compensate for residual ecological damage elsewhere after avoidance and minimization measures. Such claims depend on additionality, ecological equivalence, permanence, and the time lag between loss and gain. A young restored habitat is not immediately equivalent to a mature habitat destroyed today, and some ecological features cannot be recreated within relevant planning periods. Uncertainty increases when predicted future gains are exchanged for certain present losses.\n\nGovernance research also examines project accountability. Reporting hectares treated, trees planted, or money spent is relatively straightforward; demonstrating durable recovery is harder. Long-term contracts, independent monitoring, transparent objectives, and clearly assigned maintenance responsibilities can influence ecological outcomes. Restoration ecology consequently treats institutions as part of the causal explanation of success or failure, rather than as external administrative details."
    },
    {
      "heading": "International targets and unresolved research questions",
      "text": "Restoration has become prominent in international environmental policy. The United Nations Decade on Ecosystem Restoration runs from 2021 to 2030 and is led by the United Nations Environment Programme and the Food and Agriculture Organization. The Bonn Challenge, launched in 2011, established a goal of bringing 150 million hectares of deforested and degraded land into restoration by 2020, later extended to 350 million hectares by 2030. These commitments concern land brought into restoration, not necessarily land that has already achieved ecological recovery.\n\nThe Kunming–Montreal Global Biodiversity Framework, adopted in 2022 under the Convention on Biological Diversity, includes a target to ensure that by 2030 at least 30 percent of areas of degraded terrestrial, inland-water, coastal, and marine ecosystems are under effective restoration. The wording makes both coverage and effectiveness important. It is distinct from the framework’s area-based conservation target, even though protection and restoration can overlap spatially.\n\nLarge targets create a scientific problem of comparability. A hectare of assisted natural regeneration, a hectare of reconstructed wetland, and a hectare containing planted tree seedlings represent different interventions and recovery prospects. Global reporting therefore needs information on initial degradation, ecosystem type, treatment, ecological response, and duration. Area totals alone cannot establish whether native communities, functions, or resilience have improved.\n\nEvidence synthesis generally finds that restoration improves ecological conditions relative to degraded baselines, while full convergence with reference systems is less consistent. A 2009 meta-analysis by José María Rey Benayas and colleagues, covering 89 studies, reported average increases of 44 percent in biodiversity and 25 percent in ecosystem-service measures relative to degraded systems. Those averages did not imply complete recovery or uniform outcomes across ecosystem types. Differences in study duration, indicators, and reference conditions remain central to interpreting such syntheses.\n\nSeveral unresolved questions concern scale and predictability. Results from small experimental plots may not transfer to whole landscapes, where dispersal, fire, herbivore movement, and governance operate differently. Interventions that work in one region can fail elsewhere because species pools, soils, and land-use legacies differ. Research increasingly seeks mechanistic explanations that identify when an approach should work, rather than treating techniques as universally transferable.\n\nAnother research frontier concerns recovery below ground and across trophic levels. Vegetation is often measured because it is visible and comparatively easy to survey, but soil communities, parasites, predators, and decomposers may follow different trajectories. Re-establishing food webs requires recovering interactions as well as species. Long-term experiments that combine demographic measurements, interaction networks, soil processes, and landscape observations are testing whether apparently restored communities can reproduce, reorganize, and persist through subsequent disturbance."
    }
  ],
  "links": [
    "ecosystem",
    "ecological-restoration",
    "ecology",
    "tallgrass-prairie",
    "conservation-biology",
    "mutualism",
    "erosion",
    "soil-organic-matter",
    "fungi",
    "bacteria",
    "drainage-basin",
    "salt-marsh",
    "mangrove",
    "coral-reef",
    "species-richness",
    "environmental-dna",
    "habitat-fragmentation",
    "landscape-ecology",
    "climate-change",
    "sea-level-rise",
    "ocean-acidification",
    "climate-change-mitigation",
    "ecosystem-services",
    "united-nations",
    "united-nations-environment-programme",
    "kunming-montreal-global-biodiversity-framework",
    "convention-on-biological-diversity",
    "food-web"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}