What a Forest Is — and What It Is Not — in Light of the Universal Declaration of Tree Rights

A Global Classification of Forest Ecosystems, Protection of Primary Forests, and a Program for the Recovery of Naturalness

The word “forest” now encompasses such different realities that it may refer both to a primary forest arising from centuries or millennia of ecological continuity and to a regularly spaced plantation established for industrial harvesting. This equivalence facilitates global inventories, but it becomes inadequate whenever naturalness, integrity, biodiversity, ecological continuity, or the irreplaceable value of an ecosystem must be assessed.

In light of the Universal Declaration of Tree Rights, this article advances a global doctrine founded on an essential distinction: not every area covered by trees is necessarily a forest. It establishes a multidimensional classification, distinguishes genuine forest ecosystems from production-oriented tree-growing systems, places primary forests under a principle of non-substitution, and opens a multigenerational and multi-century horizon for the recovery of naturalness.

This proposal does not replace FAO statistical definitions, national legal categories, or existing scientific typologies. It constitutes an ecological and interpretive framework specific to the Declaration, intended to inform public policy, voluntary commitments, ecological restoration, silviculture, and the intergenerational continuity of forests.

Originally prepared on August 1, 2026; scientifically consolidated on August 4, 2026 following an in-depth study of thirteen publications transmitted by Professors Robin L. Chazdon and Brendan Mackey; further revised on August 10, 2026 following detailed critical comments and bibliographical recommendations provided by Professor Robin L. Chazdon, together with the institutional and methodological sources cited below. Further expanded on August 12, 2026 to establish the methodological foundations for the progressive development of a Global Scientific Annex and Global Forest Atlas. Revised on August 14, 2026 to clarify and strengthen the doctrinal definition of primary forest, its qualification criteria, the treatment of natural disturbances, and its relationship to a regime of strict and permanent protection, and to clarify the scientific, methodological, and institutional autonomy of the Global Forest Atlas. Further revised on August 18, 2026 following examination of the third edition of the Society for Ecological Restoration’s International Principles and Standards for the Practice of Ecological Restoration (Gann et al., 2026), to clarify the distinction between ecological restoration and ecosystem recovery, the concept of the reference model, natural and assisted natural regeneration, and the alignment of recovery indicators with international restoration standards. Further revised on August 19, 2026 following detailed scientific comments and additional scientific references transmitted by Professor Robin L. Chazdon, including Degradation and Recovery in Changing Forest Landscapes: A Multiscale Conceptual Framework (Ghazoul and Chazdon, 2017), to clarify that expected natural regeneration, assisted natural regeneration, or future restoration cannot justify destruction or conversion of an existing primary forest; to distinguish structural and aboveground-biomass recovery from recovery of native biodiversity and community composition; to recognize the importance of even small primary-forest remnants as sources of propagules, habitat, and landscape recovery; to incorporate the concept of recovery debt; and to refine the assessment of resilience, regeneration, degradation, and ecological continuity. Further doctrinal supplement on August 19, 2026 to make explicit, before the classification begins, the progression of the Declaration’s three articles—characterize, substantiate, impose obligations—and their scientific, ecological, and legally delimited translation into the forest doctrine. Further revised on August 24, 2026 following critical scientific comments and two publications transmitted by Dr. Dominick A. DellaSala, to delimit more precisely the scientific scope of Articles I and II; strengthen the definition of forest degradation as loss of ecological integrity relative to appropriate reference conditions; distinguish severe natural disturbance from anthropogenic or climate-amplified degradation; clarify the role of biological legacies and the meaning of recovery according to disturbance origin; reinforce the multi-scale Principle–Criteria–Indicators–Verifiers approach, gross-loss accounting, road and fragmentation indicators, and post-disturbance safeguards; and incorporate the Tongass case study as an illustration of shifting timber supply away from remaining old-growth forests where ecologically and socially appropriate. Further revised on August 25, 2026 after Dr. Dominick A. DellaSala corrected the earlier attachment and transmitted the intended Forest Biome: Trees of Life chapter, followed by five additional materials concerning complex early-seral forests, fire-mediated biological legacies, primary-forest policy, ecological justice, and the role of primary forests and large old trees. This revision distinguishes natural post-disturbance succession and seral states from reference-dependent ecological recovery; makes explicit that severe natural disturbance can generate high-integrity complex early-seral forest rather than a degraded state; strengthens the treatment of biological legacies as continuity structures linking seral stages; clarifies that recovery debt presupposes a demonstrated reference-dependent ecological loss rather than mere difference from a pre-disturbance appearance; and introduces a narrowly delimited ecological-justice framing under Article III without turning it into a scientific Atlas assessment layer. Further revised on September 21, 2026 following specialist scientific comments provided by Professor Leho Tedersoo, Professor of Mycorrhizal Studies at the University of Tartu, concerning the mycorrhizal and below-ground dimensions of global forest assessment. This revision delimits the extent to which fungal and mycorrhizal evidence can currently be operationalized across spatial scales; distinguishes biome- and region-specific interpretation from cautiously comparable global variables; strengthens requirements concerning sampling representativeness, detection limits, and non-detection; recognizes the continuum between arbuscular- and ectomycorrhizal dominance; and expressly guards against causal or mechanistic interpretation of correlative relationships without independent supporting evidence. Further revised on October 25, 2026 following the full review of the doctrine by Dr. Barbara L. Zimmerman, tropical ecologist and Senior Advisor to the Kayapo Project, and examination of two scientific publications she transmitted concerning large-scale conservation of primary forests in Mẽbêngôkre-Kayapó territories. This revision strengthens the separation between ecological condition, tenure, governance, and effective territorial control; clarifies that no Indigenous, community, or legal status can, by itself, be presumed to establish forest ecological integrity, primary status, degradation, or sustainability; more explicitly distinguishes ecological variables from the institutional, economic, and territorial conditions that may contribute to their maintenance or alteration; and clarifies that human presence, traditional uses, or stewardship likewise do not, by themselves, constitute evidence of degradation or loss of primary continuity. Further revised on October 26, 2026 following additional scientific exchanges with Dr. Barbara L. Zimmerman and examination of four additional publications concerning selective logging, persistent forest degradation, unmapped roads, illegal forest activities, and biological indicators of integrity. This revision strengthens the distinction between visible structure and biological integrity; clarifies that defaunation, depletion of faunal or functional assemblages, and depletion of high-value timber species may provide complementary evidence when calibrated to biome and region; establishes that non-detection by remote sensing, a GIS database, or an inventory does not constitute evidence of absence; more explicitly distinguishes observations, anthropogenic pressures, spatial correlates, and ecological diagnosis; and clarifies the detection limits of selective logging according to intervention intensity, spatial and temporal resolution, analytical method, and canopy response.

The use of these publications and the incorporation of scientific comments or bibliographical recommendations do not imply personal or institutional endorsement of this article, its classifications, or its programmatic proposals by the authors, reviewers, or their affiliated institutions.

From the Declaration’s Three Articles to an Ecological Doctrine of Forests

The classification advanced here is not an artificial addition to the Universal Declaration of Tree Rights. It addresses a prior question: when the Declaration’s three articles are used to inform forest protection, what ecological reality must first be recognized, how can scientific knowledge characterize that reality, and what human responsibilities can reasonably be derived from it?

The Declaration’s legal foundations describe a three-stage progression: characterize, substantiate, impose obligations. Article I characterizes the Tree from its living reality; Article II identifies the scientific and ecological basis for the dependence of terrestrial life-support systems on trees and the functions in which they participate; Article III draws from that reality a human responsibility. Applied to forests, this method does not automatically transform a forest into a new legal person or legal subject. It first requires us to determine what is actually being protected and to prevent the single word “forest” from making ecologically non-equivalent realities appear legally or politically interchangeable.

Article I — Characterizing the Tree as Living Leads to Characterizing the Forest as a Living System

Article I of the Universal Declaration of Tree Rights

“The Tree, a sentient living being and source of Life, is a common good of humanity.”

Article I begins by characterizing the Tree. The expression “sentient living being” is retained as the official wording of the Declaration. It should not be presented as a scientific conclusion that trees possess subjective experience, phenomenal consciousness, animal-like feeling, emotion, or pain. Current plant science does document the capacity of trees to detect relevant changes in internal and external conditions, transduce and transmit signals, mount physiological and chemical responses, alter growth and form in response to resources and stress, and participate in complex biological interactions. These capacities support a scientific characterization of trees as responsive living organisms, but they do not resolve the contested question of sentience in the phenomenal or animal sense. The term therefore belongs to the Declaration’s doctrinal vocabulary; the scientific analysis used here is limited to biological capacities that can be evidenced without claiming more than current knowledge supports. See the dedicated scientific and terminological analysis.

That requirement immediately changes the scale of analysis when moving from the Tree to the forest. A forest tree does not exist in isolation: its development depends upon soils, water, microclimate, fungi, microorganisms, fauna, other plants, and processes of dispersal, pollination, decomposition, and regeneration. This is why the doctrine later defines a forest as a living community structured by trees, rather than as the mere addition of woody individuals or the presence of a specified percentage of canopy cover.

The concept of a common good of humanity does not, by itself, call property rights into question and does not automatically transfer ownership or public authority. Within the doctrine, it expresses that a living reality may possess a value extending beyond its private or economic use. Ecological assessment should therefore not be reduced to title, timber volume, or market value. It must document separately ecological origin, integrity, naturalness, continuity, and trajectory.

Article II — Life’s Dependence Substantiates the Need to Distinguish Functions, Continuities, and Irreplaceability

Article II of the Universal Declaration of Tree Rights

“Life on Earth depends upon the existence of the Tree.”

Article II supplies the ecological foundation of this progression. It refers to the role of trees and forest systems in water, soil, carbon and nutrient cycles, climate regulation, habitat, biodiversity, and ecosystem resilience. The Declaration’s legal foundations expressly connect this article to established scientific knowledge and to the conditions that make Earth habitable. Protecting the living world is therefore not merely a landscape preference; it concerns ecological conditions upon which human societies and other living communities depend.

The scientific scope of this proposition must nevertheless remain exact. Article II is not treated here as an experimentally established counterfactual claim that the disappearance of every tree would, by itself and in isolation, mechanically cause the extinction of every form of life on Earth. A total-loss scenario would necessarily involve cascading planetary changes and causal conditions that cannot be reduced to a single variable. Its scientific foundation is narrower and stronger: trees and forest ecosystems are major components of terrestrial life-support systems and participate in hydrological, carbon and nutrient cycles, soil processes, climate regulation, habitat formation, biodiversity, and ecosystem resilience. The foundational wording of Article II should therefore not be confused with a claim of literal experimental proof beyond what ecological science can establish. This life-support framing is consistent with the primary-forest literature, which emphasizes the role of primary forests in planetary life-support systems and their irreplaceable contributions to biodiversity, climate regulation, and ecosystem functioning. Mackey et al. (2015).

This proposition does not mean that all trees are ecologically identical or that every tree-covered area has the same conservation value. It requires the opposite: if the functions and continuities that sustain Life differ according to a system’s history, composition, structure, soils, biological relationships, and trajectory, then those systems must be distinguished before they are counted, protected, managed, or restored. This is the basis for moving from tree cover alone to a multidimensional ecological classification.

Article II therefore helps explain the priority given to conservation of existing ecological continuity. A primary forest is not simply an assemblage of older trees. It belongs to an ecological continuity in which biological legacies, soils, interactions, processes, and regenerative dynamics have developed over long periods, even where some aboveground structures are young following a natural disturbance. This temporal and relational reality is central to ecological irreplaceability.

The scientific consequence is decisive: the return of trees, canopy cover, or aboveground biomass does not demonstrate the return of the same ecosystem. Native species composition, soils, large old trees, faunal communities, microbial communities, ecological interactions, and long-developed structural attributes may recover at very different rates. The existing English doctrinal analysis of primary tropical forests, existing carbon, and restoration applies the same logic: protection must precede attempts to compensate for or rebuild losses that may be rapid, uncertain, and ecologically non-equivalent.

This is why the doctrine develops a principle of non-substitution: a plantation, future restoration project, assisted natural regeneration pathway, or expected recovery trajectory may generate important ecological gains, but none can be treated as an ecological equivalent that retroactively makes the avoidable destruction of an existing primary forest acceptable.

Article III — “Act with the Tree” Translates Knowledge into Human Responsibility

Article III of the Universal Declaration of Tree Rights

“Human beings, endowed with reason and conscience, must act with the Tree in a spirit of fraternity and solidarity.”

Article III marks the transition from characterization and scientific foundation to responsibility. The decisive expression is “act with the Tree”. Acting with the Tree does not mean renouncing every form of human intervention. It means that intervention must be designed from the reality of the living system—its own capacities, limits, relationships, and temporalities—rather than from a presumption that ecological processes can always be replaced by technical action.

That responsibility has a concrete scientific translation: assess self-regeneration before planting; allow natural succession to operate where conditions permit; remove limiting pressures where assisted natural regeneration or ecological restoration is needed; restore soils and hydrology rather than treating canopy alone; maintain dispersers and pollinators; conserve biological legacies; measure trajectories over decades; and recognize that some old-growth attributes require centuries. It also means not diagnosing a naturally disturbed forest as needing ecological “repair” merely because it has moved into a different seral state: where disturbance remains within the characteristic regime and ecological continuity persists, natural succession may itself be the appropriate process. In this sense, the ecological meaning of solidarity begins with protecting existing living continuity before relying upon future repair.

The legal scope must nevertheless remain exact. The word “must” expresses the Declaration’s normative dimension, but the Declaration does not, by itself, constitute a statute, regulation, judicial decision, or international treaty currently in force. It does not automatically confer legal personality, alter ownership, or transfer public authority. The Official Framework of the Universal Declaration of Tree Rights defines this institutional and legal scope. The forest doctrine developed here is therefore an ecological, interpretive, and programmatic translation of the three articles; any binding legal consequence must be adopted by the competent authority under the applicable legal system, as explained further in What the Declaration Does Not Say and the Legal and Methodological Note.

At the normative level, this responsibility also intersects with ecological justice. Ecological justice is used here as an ethical lens extending consideration beyond the distribution of environmental benefits and harms among human communities to the continued integrity, functioning, and existence of non-human communities and ecosystems. It does not replace environmental justice, social justice, human rights, Indigenous rights, or applicable law, and it is not introduced as a scientific classification variable. DellaSala’s discussion of ecocentric conservation and ecojustice is relevant to this distinction, particularly where a naturally disturbed forest may be treated as “damaged” because it no longer conforms to a preferred human visual or productive state. DellaSala (2020). Within the Global Forest Atlas, ecological justice remains a doctrinal and ethical consideration; it is not a substitute for evidentiary ecological assessment and does not become an independent scientific Atlas layer.

The relationship between the three articles and the classification can therefore be stated simply: if the Tree must be recognized as a living reality, if Life depends upon the functions and relationships in which it participates, and if human beings must act with the Tree rather than treat it as interchangeable material, then tree-covered realities must be scientifically distinguished before deciding how they should be protected, managed, restored, or passed on. The classification that follows is the instrument proposed by the Declaration for making that distinction.

I. Why the Words “Tree,” “Plantation,” and “Forest” Are Not Interchangeable

A tree is a living organism. A group of trees is a stand. An area on which trees form a canopy is tree-covered land. A plantation is an area in which trees have been deliberately established for a defined purpose. A forest, by contrast, is a biological ecosystem in which trees structure a living community that extends far beyond the trees themselves. A natural forest is not created by people, although it may have been modified, inhabited, stewarded, or managed by people without losing its natural ecological origin.

These categories may overlap, but they are not identical. A forest contains trees, but the presence of trees alone does not necessarily create a forest ecosystem. A plantation may develop into a complex forest ecosystem, but it does not necessarily possess that quality when it is established. Conversely, a natural forest may be relatively species-poor without ceasing to be a forest: some boreal, montane, swamp, or naturally monodominant forests contain few tree species while retaining fully functioning forest soils, ecological cycles, interactions, and regeneration.

The distinction cannot rest on a single criterion. The number of tree species, regular alignment, planted origin, mode of establishment, or economic purpose are indicators. What truly characterizes a forest ecosystem is their interaction with the site’s history and ecological legacies, soil condition and land-use history, diversity of organisms, capacity for self-regeneration, trophic and symbiotic networks, the presence of multiple phases of life and death, and the ecosystem’s capacity to continue its own dynamics.

Fundamental Terminological Principle

A group of trees may constitute tree cover without constituting a forest ecosystem.

This formulation does not diminish the value of planted trees. Every tree remains a living being and performs biological functions. Nor does it deny the economic usefulness of timber production, the climate-related functions of certain stands, or the landscape value of a park. It simply rejects any claim that systems with radically different histories and modes of establishment, levels of biological or ecological complexity, and functions are ecologically equivalent.

II. The Limits of Current Statistical Definitions

FAO defines forest as land spanning more than 0.5 hectares with trees higher than 5 meters and a canopy cover of more than 10 percent, or trees able to reach those thresholds in situ, excluding land that is predominantly under agricultural or urban land use. This definition performs an indispensable function: it enables States to describe land use, track change over time, and compare highly diverse national data. Consult FAO’s FRA 2025 Terms and Definitions.

That inventory and accounting function is not, however, a comprehensive definition of an ecosystem. Two areas meeting precisely the same thresholds of height, area, and canopy cover may exhibit opposite levels of naturalness, diversity, integrity, continuity, and ecological autonomy.

FAO itself recognizes this heterogeneity. It distinguishes naturally regenerating forests from planted forests and, within the latter, forest plantations that are intensively managed and meet, at planting and stand maturity, the criteria of one or two species, even age class, and regular spacing, from other planted forests that may eventually resemble naturally regenerating forests. Consult the official definitions of planted forests and forest plantations.

The issue, therefore, is not that the FAO definition is incorrect. It answers a different question. It asks: what area must be reported within an internationally standardized land-use category? The doctrine advanced under the Declaration asks: what kind of living system should be recognized, protected, restored, and preserved for future generations as a forest ecosystem?

The Declaration therefore proposes the simultaneous use of two vocabularies without setting them against one another: definitional and accounting language, which is indispensable to international reporting and cooperation, and ecological language, which is indispensable to protecting the living world.

Methodological Clarification

Consistent with Chazdon et al. (2016), this doctrine does not claim that a single definition of forest can serve every statistical, legal, productive, climatic, cultural, and ecological purpose. It is a purpose-built ecological and interpretive framework designed specifically to assess protection, restoration, naturalness, integrity, ecological trajectory, and non-substitution under the Universal Declaration of Tree Rights. It should therefore complement, rather than displace, international statistical definitions, national legal classifications, and scientific ecosystem typologies. Chazdon et al. (2016).

III. The Definition of a Forest in Light of the Declaration

In light of the Universal Declaration of Tree Rights, a forest may be defined as a terrestrial, coastal, or seasonally or periodically inundated ecosystem in which trees constitute a structural component of a biological community capable of self-organization, renewal, and evolution. Human management may occur within such a system, but management intensity, purpose, and ecological effects must be assessed separately from the ecosystem’s origin and condition.

That community cannot be reduced to visible trunks, standing timber volume, or aboveground biomass. Its biotic components include trees, understory flora, animals, fungi, bacteria, other microorganisms, decomposers, roots, seeds, and mycorrhizal networks. Its physical resources and substrates include soils, organic and mineral horizons, water, and the forest microclimate. Its processes include self-regeneration, succession, decomposition, nutrient cycling, characteristic disturbance regimes, and, where a reference-dependent ecological loss has occurred, recovery. Its ecological interactions include pollination, herbivory, predation, dispersal, symbiosis, competition, and facilitation. Deadwood, cavities, young trees, mature and senescent trees, and retained biological legacies are integral parts of this system.

1. Constitutive Characteristics

A forest, in the ecological sense advanced by the Declaration, is identified through a combination of characteristics:

• a biological community, comprising multiple trophic, symbiotic, microbial, faunal, floral, and functional levels;
• an ecological structure, through which trees organize light, humidity, temperature, soils, and habitats;
• an internal dynamic, consisting of growth, competition, cooperation, mortality, decomposition, regeneration, and succession;
• material and biological continuity and ecological legacies, especially in soil structure and condition, propagules, root networks, fauna, microorganisms, nutrient pools, and retained organic matter;
• a capacity for self-regeneration, whether natural or progressively restored, that does not depend exclusively on complete human re-establishment at every cycle;
• functional diversity, which may exist even where tree-species diversity is naturally low;
• territorial embeddedness, connecting the stand to water, landform, climate, neighboring ecosystems, and species movements;
• a history and mode of establishment, because a forest ecosystem is also the product of ecological continuity, past land use, disturbance, management, and ecological trajectory.

2. Operational Meaning of Ecological Naturalness

For the purposes of this doctrine, ecological naturalness does not mean “untouched,” “pristine,” or “free of people,” and it is not used in the commercial sense in which the word “natural” may be applied to products. It denotes the degree to which the composition, structure, ecological processes, disturbance regimes, self-regeneration, and trajectory of a forest are governed by the ecosystem’s own dynamics in relation to an appropriate ecological reference, while recognizing compatible Indigenous, traditional, community, or other low-impact human uses.

Ecological naturalness must therefore be operationalized through explicit criteria and indicators rather than inferred from appearance or from the mere absence of planting. It should be assessed alongside integrity, origin, management regime, land-use history, and trajectory.

3. A Definition Based on Trajectory

Classification cannot be treated as static. A degraded area may recover forest functions. Where planting is used for ecological restoration or rehabilitation, the planted intervention may assist a trajectory toward a self-regenerating forest ecosystem; the initial act of planting does not by itself define the later ecological condition. Conversely, a naturally regenerating forest may lose its structure, fauna, soils, and capacity for renewal without immediately disappearing from forest statistics.

The proposed doctrine therefore assesses both present condition and trajectory. It does not ask only, “What do we see today?” It also asks, “Where did this area come from? Which processes remain active? What successional, seral, recovery, or degradation pathway is it following? Is intervention necessary, or should natural processes be allowed to proceed?”

4. Definition of “Natural Forest” for the Purposes of This Article

For the purposes of this article, a “natural forest” is a forest whose establishment or renewal relies primarily on natural regeneration processes and whose functioning is not organized around complete artificial re-establishment at each cycle. A natural forest may nevertheless be subject to compatible management, including selective silvicultural interventions, customary use, non-timber forest-product management, or wildlife-oriented management; those management regimes must be described separately.

This classification describes the stand’s predominant origin and trajectory. It does not necessarily mean that the forest is primary, long-continuity, intact, or composed exclusively of native species. Its integrity, composition, degree of degradation, and historical continuity should be assessed separately. Consult FAO’s FRA 2025 Terms and Definitions.

5. The Place of Human Intervention

Human presence is not, in itself, incompatible with a forest. Indigenous Peoples and local communities have lived in, cultivated, harvested from, protected, and governed forest environments for generations without necessarily interrupting their major ecological processes. The relevant boundary therefore does not separate nature “without people” from nature “with people.” It distinguishes uses that maintain ecosystem continuity, processes, composition, structure, and regenerative capacity from interventions that simplify, convert, or durably reorganize the ecosystem.

This rule must be applied symmetrically. The Indigenous or community status of a territory, the existence of customary rights, a long history of occupation, or the presence of local governance do not, by themselves, constitute evidence of ecological integrity, primary continuity, or sustainable use. Conversely, human presence, stewardship, traditional harvesting, or certain customary uses do not, by themselves, constitute evidence of degradation or loss of primary status. Effects must be established through evidence specific to the territory, its history, and the practices actually observed.

Research in the Mẽbêngôkre-Kayapó territories of the southeastern Amazon demonstrates concretely why this analytical separation is necessary. Large areas of primary forest have been maintained within territories inhabited over long periods, while sectors belonging to the same broader cultural and legal setting have followed different trajectories under distinct extractive, economic, and institutional pressures. Population identity and tenure regime therefore cannot, by themselves, determine the ecological condition of a territory. Zimmerman et al. (2020); Garcia et al. (2024).

Ownership, tenure, land-use rights, customary rights, governance, institutional capacity, and effective territorial control must therefore be identified separately. These dimensions may help explain an ecological trajectory, but they must not be transformed into substitutes for ecological observation itself.

Proposed Doctrinal Definition

A forest is a living community structured by trees, rooted in soil and a water regime, and capable of regeneration, evolution, and ecological relationships extending beyond the production of biomass alone.

IV. A Multidimensional Global Classification by Ecological Origin, Condition, Management, and Trajectory

A simple list of “forest types” is scientifically inadequate. The same area may simultaneously be tropical, montane, humid, naturally regenerated, old-growth, fragmented, managed, peat-forming, and governed under Indigenous, community, public, private, or shared tenure. The proposed doctrine therefore uses a multidimensional classification rather than forcing every forest into a single mutually exclusive class.

Ecological naturalness remains an important axis, but it is not a moral scale and it is not a synonym for absence of people. It must be read together with ecological origin and history, present condition and integrity, management regime and objective, tenure and governance, spatial context, and temporal trajectory.

Primary Forests

For the purposes of this doctrine, a primary forest is a naturally regenerated forest ecosystem dominated by native species, in which natural ecological and evolutionary processes, biological legacies, characteristic disturbance regimes, and ecological continuity remain predominant, and whose composition, structure, functions, ecological relationships, and trajectory have not been fundamentally reorganized by industrial intervention or other intensive human intervention. Primary status does not require the absence of human presence and remains compatible with Indigenous, traditional, cultural, or subsistence uses that maintain ecological continuity and integrity. This formulation builds on scientific work concerning primary forests and on existing international definitions without being confined to their statistical reporting function. Research conducted in Mẽbêngôkre-Kayapó territories further shows that long-standing human presence and Indigenous uses can coexist with extensive areas of primary forest where ecological continuity and forest processes remain predominant. See Kormos et al. (2017); Zimmerman et al. (2020); and FAO FRA 2025 Terms and Definitions.

A primary forest or primary-forest complex may include young, intermediate, mature, and old seral stages created by natural fire, windthrow, flooding, insect outbreaks, or other disturbances operating within the ecosystem’s characteristic regime. These phases commonly occur as a patch or gap mosaic within the larger forest system. Natural disturbances and the seral stages resulting from them do not, by themselves, terminate primary status where they remain compatible with the characteristic disturbance regime and ecological continuity of the ecosystem. Primary status cannot be excluded solely on the basis of forest age, canopy openness, fragment size, absence of a legal protection status, the existence of a management plan, or historical low-intensity human use. The term primary forest should not be read as “virgin” or wholly untouched. DellaSala et al. (2014); DellaSala (2020).

Seral stage is therefore an orthogonal descriptor, not a substitute for ecological origin. A forest may be primary by continuity and simultaneously occupy a complex early-seral state after a characteristic natural disturbance. Conversely, a young stand created by clear-cutting may also appear early-seral while possessing a different causal history, a different legacy structure, and a different integrity trajectory. DellaSala et al. (2014) show that complex early-seral forests generated by natural stand-replacing disturbance can retain abundant living and dead biological legacies and support distinctive, highly diverse communities that are not ecologically equivalent to early-seral conditions produced by commercial forestry. The classification must therefore record origin, disturbance regime, biological legacies, seral state, and trajectory separately.

Primary-forest qualification must rest on a converging body of evidence concerning, in particular, natural regeneration, ecological continuity, biological legacies, characteristic disturbance regimes, the history of interventions and land use, present ecological attributes, and landscape context. No single threshold, canopy appearance, age class, statistical category, or administrative listing should be treated as conclusive by itself. The absence of a forest from a national or international inventory or map does not constitute evidence that it is not primary.

Across primary forests—and, where relevant, across the other ecological categories below—the assessment should separately disclose integrity, fragmentation, protection, ownership or tenure, land-use rights, and governance. For primary forests, useful descriptors include:

• large, intact primary forest;
• fragmented primary forest;
• primary forest subject to compatible traditional uses;
• primary forest under threat or undergoing degradation;
• primary forest legally protected and effectively conserved;
• primary forest protected in law but inadequately protected in practice;
• primary forest without a specific legal protection status.

Forests of Very High Ecological Naturalness

This category covers forests that retain nearly all attributes of a primary forest, but whose history includes ancient, limited, or unknown human intervention. It may include some long-continuity forests, mature or late-successional forests, forests subjected to ancient selective harvesting, second-growth forests that have regained complex structure, and traditionally managed forests that have not undergone major alteration.

Five concepts should remain distinct:

• primary forest: natural regeneration, predominance of natural ecological and evolutionary processes, biological legacies, characteristic disturbance regimes, and ecological continuity, without fundamental reorganization of composition, structure, functions, ecological relationships, or trajectory by industrial intervention or other intensive human intervention;
• long-continuity forest: historical continuity of forest cover, without that fact alone determining current integrity;
• old-growth or old forest: advanced structural development, including old trees, large diameters, cavities, deadwood, and long-developed microhabitats;
• mature or late-successional forest: an advanced developmental stage that does not, by itself, establish primary continuity;
• high-integrity forest: composition, structure, ecological processes, connectivity, stability, regenerative capacity, and other ecological attributes close to an appropriate ecological reference; biocultural continuity, tenure, and governance are documented separately.

Naturally Regenerating Second-Growth (Secondary) Forests

A second-growth (secondary) forest re-establishes after a substantial break in ecological continuity, prior land use, or human intervention that has materially reorganized the system—for example abandoned agriculture, significant logging, partial conversion, severe human-caused fire, conflict, or another major transformation. As natural regeneration, ecological interactions, and forest dynamics progressively recover, the area gradually regains the attributes of a forest ecosystem; no single canopy or age threshold should be treated as an instantaneous proof of recovery. A natural disturbance operating within the characteristic regime of a primary forest does not, by itself, transform that forest into secondary forest where primary ecological continuity persists.

The category should distinguish young second-growth forest, intermediate second-growth forest, mature second-growth forest, post-agricultural forest, post-logging forest, and other forest-recovery trajectories following a documented break in prior continuity. Post-fire or post-storm situations must be classified in light of site history and disturbance regime: the event alone is not sufficient to establish secondary origin. Their value cannot be measured by age alone: proximity to seed sources, the effects of previous land use on soil structure, nutrients and soil carbon, hunting, grazing, invasive species, hydrology, and fragmentation determine their trajectory.

Degraded Natural Forests

A forest may remain visible in satellite imagery while having lost much of its integrity. Repeated removal of the largest trees, drainage, roads, human-caused fires, hunting that eliminates seed dispersers, erosion, pollution, or introduced species may profoundly simplify the ecosystem without immediately removing its canopy.

Degradation should therefore be distinguished from deforestation. Deforestation converts forest to another land use. Degradation may retain a superficial appearance of forest integrity, or remain within a statistical forest category, while reducing diversity, structure, functioning, resilience, or regenerative capacity.

For the purposes of this doctrine, forest degradation is an anthropogenic reduction of ecological integrity relative to an appropriate reference condition, expressed through deterioration in native composition, keystone structures, ecological processes, soils, hydrology, regeneration, connectivity, resilience, or other ecosystem attributes, without necessarily causing conversion to a non-forest land use. Degradation is therefore best understood as a continuum of integrity loss operating from individual trees and stands to landscapes, rather than as a binary category. DellaSala et al. (2025).

Where primary or old-growth forests remain, they can provide particularly important reference conditions for assessing loss of integrity. Where such forests no longer exist, reference conditions may be reconstructed from mature or near-natural forests, historical evidence, naturally regenerating structurally complex forests, and other ecologically appropriate evidence. Reference selection must be explicit, biome- and region-specific, and accompanied by uncertainty where historical or ecological reconstruction is incomplete. DellaSala et al. (2025).

Forests Under Ecological Restoration

This category includes areas in which deliberate actions have been taken to assist ecosystem recovery: removing pressures, assisted natural regeneration, hydrological restoration, reconnection, re-establishment of native species, soil restoration, targeted control of invasive species, or setting land aside for natural processes. Ecological restoration is used here in the more specific sense of the process of assisting the recovery of a native ecosystem and should be distinguished from the broader umbrella of ecosystem restoration, which includes a wider continuum of restorative activities.

Ecological restoration denotes the process of assisting the recovery of a native ecosystem; it does not itself constitute the ecological outcome sought. Recovery denotes the outcome sought or achieved, and a project should be evaluated in relation to the recovery trajectory it initiates or supports. A project, budget, or planting operation therefore does not prove that the ecosystem is recovering. Taxonomic diversity, structure, ecological functions, soils, hydrology, and faunal or microbial communities may follow different recovery trajectories and progress at different rates. Assessment should be based on an appropriate reference model of a native ecosystem, measurable ecological goals and indicators, transparent monitoring, and adaptive management. The reference model should not be understood as a frozen reproduction of a historical condition: it describes a high-integrity condition and an appropriate range of ecological variability while accounting for past and anticipated environmental change. See Gann et al. (2026), International Principles and Standards for the Practice of Ecological Restoration, Third Edition.

Planted Establishment Within Ecological Restoration or Rehabilitation Pathways

Planting should not be treated here as a separate final ecological class. It is a mode of establishment or restoration intervention whose significance depends on its purpose, ecological context, subsequent management, and trajectory. Planting may be used for ecological restoration or rehabilitation where seed sources, soils, hydrology, or succession conditions have been severely altered; it may also take the form of limited enrichment planting after selective harvesting. In such cases, the objective is not merely to establish trees, but to assist recovery toward a diversified, ecologically coherent, increasingly self-regenerating forest ecosystem.

Classification therefore depends not on planting alone but on the objective and ecological trajectory: whether the intervention is commercial, restorative, conservational, or rehabilitative; whether management progressively reduces dependence on intervention; whether deadwood and understory develop; whether multiple age classes and native or ecologically coherent communities establish; whether fauna and microbial processes recover; whether self-regeneration becomes possible; and whether water and soil cycles are restored. A planted intervention that remains permanently organized around recurrent production cycles belongs in the production-plantation category rather than in a restoration trajectory.

Example. On former agricultural land where native seed sources and dispersal are severely depleted, a mixed planting of locally appropriate native species may be used to create recruitment nuclei, restore shade and microclimate, and facilitate the return of dispersers and spontaneous regeneration. The planted phase is then reported as a restoration intervention. If, over time, the system develops self-regeneration, multiple age classes, ecological interactions, soil and hydrological recovery, and decreasing dependence on human re-establishment, its later ecological condition is assessed on those attributes rather than permanently defined by the fact that planting occurred at the beginning.

V. Production-Oriented Tree Plantations: Why They Must Be Classified Separately from Forest Ecosystems under the Declaration

For the ecological and interpretive purposes of the Declaration, production-oriented tree plantations should be classified separately from forest ecosystems where their predominant purpose is the planned production of raw materials and their management persistently limits ecological autonomy, structural diversity, natural regeneration, and forest dynamics. This separate classification does not alter FAO statistical categories or national legal designations. Its purpose is to prevent production systems from being treated as ecologically equivalent to natural, regenerating, or ecologically restored forest ecosystems, and to prevent the establishment of plantations that replace natural forest ecosystems from being reported as an ecological gain. DellaSala’s forest-biome synthesis likewise argues that tree-cover categories should be complemented by ecological-integrity and forest-quality distinctions and that primary forests provide an important reference condition for comparison with other forest types. DellaSala (2020), Forest Biome: Trees of Life.

Depending on how they are actually established and managed, this category may include plantations of conifers, eucalyptus, poplar, rubber trees, or teak; short-rotation coppice; industrial tree-growing systems; energy plantations; pulpwood plantations; and certain offset plantations established primarily to generate carbon credits or offset claims.

1. Identification Criteria

No single criterion is sufficient. Classification results from a converging set of indicators:

• deliberate establishment and mechanical site preparation;
• one or two dominant tree species selected for production performance;
• trees of the same age or within very narrow age ranges;
• regular rows and spacing;
• genetic selection directed toward uniform output;
• short or predetermined production cycles;
• scheduled thinning, inputs, treatments, or harvesting;
• limited representation of senescent stages and deadwood;
• simultaneous harvest or clear-cutting at stand scale;
• low capacity for self-regeneration and the need for re-establishment at every cycle;
• a production function that clearly predominates over other functions;
• persistent simplification of soils, understory, fauna, or ecological networks.

2. A Classification That Does Not Condemn Timber Production

Wood may provide a renewable material resource when it is produced under ecologically sustainable conditions and used efficiently, particularly in long-lived products. Production plantations may reduce pressure on some natural forests, provide employment, and supply timber and fiber value chains. This does not imply that the combustion of forest biomass is carbon neutral, that wood extraction is ecologically interchangeable with forest protection, or that a plantation replaces the ecosystem it may help preserve.

Terminological transparency also protects the legitimacy of production. A well-designed tree plantation that is clearly reported and integrated into a landscape mosaic can be assessed on its own merits: productivity, soils, water, pesticide use, genetic diversity, fire risk, connectivity, social conditions, and its actual contribution to reducing pressure on natural ecosystems.

3. Other Tree-Based Systems

The following should likewise be distinguished from forests without being devalued:

• orchards and fruit-tree crops;
• olive groves and oil-palm plantations;
• agroforests, agroforestry, and silvopastoral systems, which should be reported separately from orchards, parks, and other tree-based systems because of their distinctive combination of trees, crops or pasture, management, livelihoods, and food-security functions;
• hedgerow-based bocage landscapes, hedgerows, and windbreaks;
• urban parks, gardens, and tree-covered cemeteries;
• roadside rows of trees and isolated trees;
• nurseries and Christmas-tree crops;
• ornamental or heritage plantations.

Agroforestry requires a distinct reporting category. Agroforests and silvopastoral systems can combine substantial ecological value with food production, livelihoods, soil protection, shade, carbon storage, and landscape connectivity. Their mixed agricultural–tree character is precisely why they should not be collapsed either into forest categories or into a residual category with orchards, gardens, and parks.

V bis. Agroforestry and Agroforestry Systems Worldwide: Ecological Classification, Biocultural Diversity, Integrity, Territorial Trajectories, and Atlas Methodology

Proposed scientific and methodological framework — submitted for expert review. The eight system families and the associated criteria are working proposals of the UDTR Global Forest Atlas. They are not an internationally adopted or independently validated nomenclature.

1. Recognizing Agroforestry as a Socioecological System in Its Own Right

Agroforestry occupies a distinctive position in the relationships between human societies, trees, agriculture, and forest ecosystems. It encompasses land-use practices and systems in which perennial woody plants—trees, shrubs, palms, or bamboo—are intentionally combined with agricultural crops, livestock, or pastoral activities, in specified spatial arrangements or temporal sequences.

This general definition corresponds to the terminological foundations used by the Food and Agriculture Organization of the United Nations (FAO). It entails ecological and socioeconomic interactions among the system’s components. Agroforestry cannot therefore be reduced either to the incidental occurrence of trees on farmland or to a percentage of tree canopy cover. FAO — Agroforestry Overview.

Agroforestry is not a single vegetation type. It embraces systems whose structures, origins, functions, biological composition, management intensity, and historical trajectories may differ profoundly.

A complex multistrata tropical agroforest, a shaded cocoa plantation, a Sahelian agroforestry parkland, a Mediterranean wooded pasture, a hedgerow-based agricultural landscape, and an arable farm combining rows of trees with annual crops have neither identical ecological characteristics nor identical relationships with forests. Their shared designation as agroforestry does not establish equivalent ecological value.

The present doctrine proposes recognizing agroforestry as an autonomous category of analysis, describing the organization of land use and biological interactions without prejudging the ecological identity of the environments concerned. This distinction requires four separate, interconnected dimensions:

• Ecological identity describes living communities, habitats, processes, naturalness, and the history of a place.
• Agroforestry use describes the intentional organization of agricultural or pastoral activities together with woody plants.
• Ecological integrity assesses biological and functional characteristics against appropriate reference conditions.
• Temporal trajectory documents previous transformations, observed changes, and possible developments.

These dimensions must first be assessed separately and then related to one another. The presence of an agroforestry system proves neither that a natural forest currently exists nor that a forest did not previously exist, and neither ecological improvement nor degradation. Each conclusion requires evidence specific to the territory.

2. Distinguishing Agroforestry from Forests Without Overlooking Their Ecological Relationships

International statistical definitions of forest rely in particular on area, tree height, canopy cover, and predominant land use. Under FAO Global Forest Resources Assessment definitions, land predominantly used for agriculture is generally excluded from the statistical “forest” category, even where significant tree cover remains. Particular arrangements involving temporary crops within a forest rotation, notably Taungya-type systems, receive specific treatment.

This distinction is necessary for statistical comparability, but it cannot, on its own, settle all questions of forest ecology. At landscape scale and in historical perspective, at least five situations must be distinguished.

First: agroforestry established on pre-existing agricultural land. Trees are introduced or retained among crops or pasture already in place. The resulting system may improve certain ecological functions compared with the previous agricultural condition without thereby becoming a natural forest.

Second: agroforestry resulting from the transformation of natural forest. A forest or part of one is converted, thinned, or restructured to accommodate agricultural production. Native trees, habitat, and considerable biodiversity may remain, but the conversion and any associated losses must be established and recorded.

Third: agricultural or agroforestry uses within forest landscapes. Some practices occur under forest canopy or in mosaics containing forest fragments, where the entire territory cannot legitimately be assigned a single ecological category. The relevant units, predominant uses, interventions, and actually retained forest components must be identified.

Fourth: agroforestry on formerly deforested or degraded land. Introducing, retaining, or regenerating trees may contribute to soil rehabilitation, habitat diversification, and ecological connectivity. Those effects must be measured against the previous condition and must not automatically be called restoration of a forest ecosystem.

Fifth: long-established, culturally shaped, ecologically complex agroforestry systems. Some have evolved through practices handed down over generations, combining selection, maintenance, regeneration, cultivation, livestock, and conservation of long-lived trees. Their heritage and biocultural significance may be considerable without automatically making them primary or high-integrity natural forest.

Two symmetrical simplifications must therefore be rejected: treating every agroforest as natural forest, and assuming that no agroforestry system can contain ecologically important forest components. A land-use designation must never erase ecological identity, biological legacies, or the history of territorial transformation.

To prevent double counting, land areas should be assigned to spatial units consistent with the statistical definitions used. Agroforestry practices should also be recorded in a distinct descriptive layer that may overlap other ecological information. An agroforestry designation does not, by itself, determine the statistical land category, ecological identity, or legal status of any mapped unit.

The distinction is one of statistical comparability, not a universal ecological verdict. It is therefore important to retain, as separate findings, the predominant land use, the identity of surviving ecological components, any documented history of forest conversion, and the degree of confidence supported by the available evidence. The same agroforestry label must never be interpreted as evidence that a site either is or has never been a natural forest.

3. Towards a Multidimensional Global Typology

Common functional classifications distinguish three broad categories: agrisilvicultural systems combining trees and crops; silvopastoral systems combining trees, pastures, and/or animals; and agrosilvopastoral systems combining trees, crops, and livestock. These categories recognized in FAO terminology provide an initial framework, but do not capture the full diversity of worldwide agroforestry arrangements.

In their contribution For a universal classification of agroforestry (2022), Emmanuel Torquebiau and Fabien Balaguer proposed five structural families: crops under tree cover, agroforests, linear agroforestry arrangements, agroforestry involving animals, and sequential agroforestry. This proposal illustrates the usefulness of observable structural characteristics while confirming that international agroforestry classification remains open to scientific discussion. Torquebiau and Balaguer (2022).

For the Global Scientific Annex and UDTR Global Forest Atlas, the following eight operational agroforestry descriptors (AF-1 to AF-8) are proposed. They are not mutually exclusive final forest classes or an internationally endorsed nomenclature.

These eight families are an Atlas working proposal, not an adopted or validated international classification. They are descriptive and potentially overlapping: a farm may simultaneously contain crops under tree cover, hedgerows, and wooded pasture. Multiple descriptors may therefore apply to one management unit, while area accounting must explicitly prevent their areas from being summed more than once.

Membership in any family is not an ecological-quality certification. Even a highly simplified agroforestry system should be recognized as agroforestry if its defining criteria are met, while its ecological performance is independently evaluated without favorable presumption. Likewise, an orchard, hedgerow, or wooded pasture is not automatically agroforestry: intentional integration, relevant interactions, actual use, and territorial context must be established.

The proposed AF-1 to AF-8 descriptors describe observable arrangements rather than a hierarchy of ecological worth. Their application must distinguish the agricultural management unit from the area actually occupied by individual tree-based features. If an agroforestry plot also contains a hedgerow and shaded crops, the overlapping descriptions remain useful, but its hectares must not be added several times when estimating territorial extent.

4. The Worldwide Diversity of Agroforestry Systems

Agroforestry cannot be understood through an exclusively tropical, European, or modern agricultural model. Analysis must recognize traditional and contemporary forms, biogeographical differences, land-tenure histories, and the knowledge through which they have developed.

Sahelian and Sudanian Africa

Agroforestry parklands retaining or encouraging Faidherbia albida, Vitellaria paradoxa (shea), or Parkia biglobosa (néré), among other species, maintain trees within cultivated areas. Their performance must be considered in relation to soil fertility, woody regeneration, crop interactions, food uses, grazing pressure, and climate. Classification must also distinguish human-shaped agroforestry landscapes from natural savannas, open woodland, and other native ecosystems, which must not be characterized as incomplete forests merely because tree density is low.

Humid Tropical Africa and Island Regions

Cocoa, coffee, and other multistrata systems differ considerably in tree diversity and management intensity. In northeastern Madagascar, vanilla agroforests offer a particularly informative example. Research shows that their floristic composition and structure depend strongly on land-use and ecological origin. Systems derived from forest fragments and those established on former farmland or fallows do not have equivalent biodiversity characteristics. Tree cover and crop type alone are insufficient to establish ecological value. Osen et al. (2021); Osewold et al. (2022).

Tropical Americas

Coffee and cocoa agroforestry, forest gardens, crop–tree associations, and complex traditional systems represent major configurations. They may contribute to connectivity among remnants of natural habitat, conservation of some species, and diversified production. Their contribution must nevertheless be distinguished from that of adjacent natural forests, especially when an agroforest originated through forest conversion.

Tropical and Subtropical Asia

Systems combining food crops, fruit trees, forest species, commercial crops, and woody production are highly diverse. Some complex agroforests and tree gardens maintain multilayered vegetation; other systems geared toward a dominant commercial crop remain ecologically simplified despite several woody species. Forest uses and activities along forest margins must be documented by separating land use, canopy change, conversion history, and the biological continuity of the ecosystems concerned.

Mediterranean and Temperate Europe

Spanish dehesas, Portuguese montados, wooded grasslands, grazed orchards, certain hedgerow-based landscapes, within-field tree arrangements, and alley cropping reflect the historical and present diversity of European agroforestry. A systematic review published in 2025 covering 232 studies of traditional European agroforestry systems highlights their significance for biodiversity, soil functions, water regulation, and cultural heritage. Maintaining these systems may depend on active management: abandonment, intensification, and simplification can generate different ecological trajectories and must not automatically be interpreted as improvement or deterioration. Le et al. (2025).

Arid and Semi-Arid Regions and Oasis Systems

Some oasis systems combine date palms, fruit trees, low-growing crops, and traditional water-management arrangements. Evaluating sustainability requires attention to the origin and availability of water, salinization, soils, cultivated species, local uses, and climate trajectories. Increased tree cover cannot be considered apart from water consequences or possible effects on naturally open ecosystems.

North America and Other Temperate Regions

Alley cropping, shelterbelts, functionally integrated wooded strips, wooded pasture, and some forms of farming under forest canopy expand the range of agroforestry practices. Depending on location, benefits may involve soils, erosion control, livestock shade, agricultural biodiversity, connectivity, and diversified production. This geographical survey is illustrative rather than exhaustive. A definitive global classification will require documented regional references developed with the relevant scientific expertise and knowledge-holders.

Historical land tenure and the practices of local knowledge holders are essential to interpreting these geographical examples. A regionally recognizable form of agroforestry does not, in itself, establish a uniform ecological origin or comparable biodiversity, management intensity, water use, or history of conversion across sites.

5. Eight Dimensions of Ecological and Territorial Characterization

The doctrine proposes assessing each agroforestry system against eight analytically distinct dimensions:

A. Ecological Origin and Land-Use History

Identify the previous condition of the territory, dates and means of transformation, deforestation where applicable, surviving biological legacies, and the sequence of land uses.

B. Structural Organization

Describe tree density and spatial arrangement, vegetation layers, age classes, diameters, woody composition, associated crops, and spatial relations among components.

C. Composition and Biodiversity

Document native, endemic, introduced, and invasive species; community composition; species dependent on natural habitats; and biological functions actually observed.

D. Ecological Functioning

Assess soils, organic matter, nutrient cycles, water, erosion, microclimate, biological interactions, regeneration, and processes relevant at site and landscape scales.

E. Management Regime and Intensity

Characterize interventions, harvests, inputs, crop protection, irrigation, grazing, cutting, renewal, and production goals.

F. Food and Socioeconomic Functions

Describe production, dietary diversification, yields, income, working conditions, tenure security, and benefits or constraints for affected populations.

G. Territorial Continuity and Temporal Trajectory

Assess connectivity, canopy changes, ecological gains and losses, climate risks, changes of use, and persistence of system functions.

H. Evidence Quality and Uncertainty

Distinguish direct measurements, satellite observations, historical series, management declarations, local knowledge, models, and inadequately verified hypotheses.

These dimensions must not automatically be collapsed into one overall index. A productive system can retain limited biodiversity; it can store carbon while placing substantial pressure on water; or it can contain many tree species after a recent conversion of natural forest. Multifunctionality neither implies equality of ecological functions nor removes trade-offs.

The eight dimensions are complementary, not interchangeable. A classification must distinguish direct observations from reconstructions, statements supplied by managers, local and Indigenous knowledge, and models. Where an observation cannot support a reliable conclusion about ecological origin, integrity, or causal attribution, that conclusion should remain provisional or undetermined rather than being inferred from canopy cover.

6. Assessing Ecological Functions Without Presuming Benefits

Agroforestry systems may contribute to a range of ecological and social functions, but the magnitude and direction of effects depend on context, composition, history, and management.

Biodiversity

Several vegetation layers, native trees, associated habitats, and landscape connections may support certain biological communities. Species richness alone does not establish integrity. Assessments must consider abundance, community composition, specialist and threatened species, and differences relative to ecologically appropriate references. A species-rich agroforest is not necessarily a functional replacement for primary forest or another natural ecosystem.

Soils and Biogeochemical Cycles

Trees may protect soil physically, contribute organic matter, recycle nutrients, and reduce erosion. Such effects cannot be assumed to apply everywhere: they depend on species, soil, climate, residue management, inputs, removal rates, and interactions with crops. Fungal, microbial, and mycorrhizal observations can enrich assessment where supported by appropriate sampling, explicit detection limits, and scientifically defensible interpretation.

Water and Microclimate

Woody components may change shade, temperature, infiltration, and hydrological flows. Effects require system-specific assessment. In some conditions increased transpiration or competition for water may offset benefits. Tree presence alone is therefore insufficient proof of an improved water balance.

Carbon and Climate

Agroforestry can store carbon in biomass and soils and contribute to climate adaptation. Assessment must distinguish existing stocks, newly accumulated stocks, carbon flows, intervention-related emissions, indirect land-use change, and reversal risks. An agroforestry carbon gain does not automatically demonstrate a biodiversity gain; the two must be evaluated separately.

Production and Food Security

Diversified products may strengthen economic and food-system resilience. Trade-offs can nevertheless arise through yields, light and water competition, labor requirements, disease, market access, and transition costs. Performance should be assessed at the relevant scales of crop, farm, household, territory, and productive cycle. Food and economic value are not direct measures of forest naturalness.

The direction and strength of any effect must be assessed with respect to a specified comparison: an earlier condition on the same land, a comparable agricultural system, an appropriate natural-ecosystem reference, or the surrounding landscape. An improvement relative to simplified agriculture does not imply full recovery of an ecosystem, and a favorable result for one function must not be used to compensate automatically for losses in another.

7. Non-Substitution and Protection of Natural Ecosystems

Agroforestry expansion must not be invoked as automatic ecological justification for destroying, converting, or degrading a natural ecosystem. This applies to primary forests, high-integrity natural forests, wetlands, peatlands, and other habitats whose biological functions cannot be recreated merely by planting trees alongside agricultural production.

Conversion of a natural forest into agroforestry must be recorded as a transformation of the pre-existing ecosystem, even when substantial tree cover remains. Losses in biological composition, structure, continuity, ancient carbon stocks, and ecosystem functions must remain visible in ecological accounts.

Conversely, context-appropriate agroforestry on historically simplified agricultural land may deliver significant benefits if they are demonstrated and do not displace ecological pressure elsewhere. The same principle protects naturally open ecosystems: savannas, natural grasslands, and ecologically valuable arid habitats cannot be treated as land available for tree conversion merely because they have low or no forest cover.

Assessment should separately identify:

1. Benefits demonstrably delivered by improving an existing agricultural system.
2. Ecological losses caused by converting natural habitat.
3. Biological functions and legacies retained in long-established agroforestry.
4. Restoration, rehabilitation, or natural-succession trajectories that are genuinely observed.
5. Risks of agricultural pressure being displaced to other territories.

Non-substitution is a conservation doctrine, not a universally binding legal rule purportedly created by the Declaration itself.

Five distinct matters must remain visible in reporting: documented benefits arising from improvement of pre-existing agricultural land; losses caused by conversion of a natural ecosystem; ecological functions and biological legacies conserved in long-established agroforestry; verified restoration, rehabilitation, or natural-succession trajectories; and the risk that agricultural pressures are displaced to other territories. Non-substitution is a proposed conservation doctrine, not a universally binding legal rule created automatically by the Declaration.

8. Ecological Integrity and Appropriate Reference Conditions

The ecological integrity of an agroforestry system cannot be properly assessed without specifying the reference against which it is compared. Different comparisons answer different questions:

Historical comparison: has the system gained or lost ecological properties relative to its own previous condition?

Agricultural comparison: does it differ from comparable non-agroforestry systems under similar conditions?

Ecosystem comparison: which components of the appropriate natural reference ecosystem remain, have changed, or are missing?

Landscape comparison: does the system sustain habitats, connections, or functions important at territorial scale?

Temporal comparison: are observed properties stable, improving, declining, or at particular risk?

These reference types cannot be interchanged. An agroforestry system may be richer in species than monoculture without approaching the biodiversity of the corresponding natural community. It may retain old trees while certain layers or ecological interactions have been simplified. Likewise, more canopy within farmland may favor some species without demonstrating complete ecological recovery.

References must be adapted to biome, ecoregion, biocultural setting, land-use history, and the scientific question. No universal threshold of tree density, species number, carbon, or canopy cover can alone define integrity or ecological quality across all agroforestry systems.

Comparisons against historical conditions, agricultural alternatives, natural ecosystem references, landscape functions, and temporal change answer different questions. They must not be conflated. A diversified agroforest may outperform a monoculture while remaining very different from a native forest community; an increase in tree cover may benefit some species without proving complete ecological recovery. References must be selected for the relevant biome, ecoregion, biocultural context, land-use history, and evaluation purpose.

9. From Doctrine to Verification: Applying the PCIV Framework to Agroforestry

The Principle–Criteria–Indicators–Verifiers (PCIV) method developed in the general forest doctrine can be adapted to agroforestry. It should connect broad characterization to documented properties and evidence.

Proposed guiding principle: recognize, describe, and monitor agroforestry systems according to their identity, origin, functioning, ecological and socioeconomic functions, and trajectory, without automatically assimilating them to natural forest ecosystems.

These indicators are illustrative; the operational definitions, sampling designs, reference conditions, comparability rules, and any thresholds require specialist scientific protocols. One positive indicator is insufficient to establish overall performance. Biodiversity, soil, water, carbon, and socioeconomic findings must remain analytically separate.

The agroforestry indicators are a specialized application of the Atlas’s general PCIV framework, as established in Section XIX (global indicators) and Section XXVI (Global Scientific Annex and Global Forest Atlas), not an alternative protocol. Agricultural productivity, socioeconomic outcomes, and biocultural significance must be reported within their own dimensions and must not automatically be translated into scores for forest integrity or naturalness.

Where data are inadequate, the Atlas should preserve a provisional or undetermined assessment rather than fabricate precision.

The PCIV approach is a methodological architecture rather than evidence that all listed variables can already be measured everywhere. Indicator definitions, sampling procedures, reference conditions, uncertainty, validation design, and any thresholds require specialist protocols. Agricultural and socioeconomic performance, tenure, and biocultural value should be documented without being converted into an automatic naturalness or forest-integrity score.

10. Mapping and Monitoring Methodology for the UDTR Global Forest Atlas

The Atlas must move beyond representing tree-covered agricultural areas alone. In September 2026 the FAO published Monitoring global agroforestry extent: An analysis of methodological approaches and data sources — Part A, addressing difficulties in global agroforestry measurement and the possibilities afforded by land observation and standardized satellite-image interpretation. Its approach, including the Global Forest Resources Assessment Remote Sensing Survey, supports combining remote sensing, expert interpretation, land use, and repeated observation. It does not prove that all agroforestry types can already be mapped globally, exhaustively, or unambiguously. FAO (2026).

The Atlas could develop an agroforestry module interoperable with existing international sources under the following methodological requirements.

10.1. Defining Observation Units

Explicitly distinguish farm, parcel, agroforestry arrangement, woody patch, landscape mosaic, and ecological unit. A single agricultural management unit may contain several ecologically different environments. Observation scale must follow the scientific question and the resolution of available evidence.

10.2. Separating Information Layers

Where the data permit, mapping should distinguish:

• Tree cover and structure.
• Predominant land use.
• Agroforestry practices present.
• Ecological identity of observed units.
• Origins and history of transformation.
• Principal compositional and integrity attributes.
• Pressures, risks, and trajectories.
• The confidence attached to each item of information.

These layers may overlap, but their areas must not be summed as independent surfaces. A spatial unit is counted only once in any defined area total, even when it carries multiple agroforestry descriptors.

10.3. Identifying Agroforestry Practices

Detection of trees in an agricultural parcel is a possible clue, not sufficient proof of agroforestry. Classification must examine intentional integration of components, spatial or temporal arrangement, and the relevant ecological or socioeconomic interactions. Where verification is impossible, the provisional label “agricultural tree configuration compatible with agroforestry — verification required” may be used without confusing it with a confirmed classification.

10.4. Historical Reconstruction and Detection of Conversion

Historic maps, aerial photography, older satellite records, ecological surveys, and local knowledge may document previous conditions. Reconstruction must distinguish observation, interpretation, and hypothesis. Without adequate evidence, neither forest nor non-forest origin may be assigned by default. Recent conversions and documented ecological losses must be recorded separately from any subsequent canopy gains.

10.5. Field Validation and Sampling

Remote-sensing methods should be complemented, when needed, by field observations representative of system types, regions, and management gradients. Validation must document classification errors, confusion among systems, spatial uncertainty, and detection limits. Quality control must distinguish the data used to develop a classification from those used to validate it. The Atlas should record confirmed, provisional, contested, and undetermined assessments in accordance with its general evidentiary architecture.

10.6. Temporal Monitoring

Monitoring should expose:

• Establishment or disappearance of woody components.
• Changes in diversity and structure.
• Intensification or other evolution of agricultural practices.
• Changes in production and management.
• Conversion of natural environments.
• Retention or disappearance of inherited trees.
• Regeneration, recruitment, and mortality.
• Changes in measured ecological functions.

Monitoring intervals should reflect the processes assessed: crops, management changes, woody biomass, soils, and biodiversity have different response times. One observation cannot reliably establish an ecological trajectory.

10.7. Comparability and Interoperability

Results should be comparable, where methodologically justified, with FAO terminology, national classifications, and other relevant monitoring frameworks without claiming to supersede them. Every result must identify its reference framework; data produced under different definitions must not be aggregated without explanation. Agricultural tree-cover statistics must not be presented as an exhaustive count of agroforestry. Conversely, sparsely wooded or temporally sequential systems must not automatically be excluded because satellite detection is difficult.

A remotely detected woody feature is only a possible indication of agroforestry. Intentional integration with agricultural activity, spatial or temporal organization, and relevant ecological or socioeconomic interactions require supporting information before the practice can be confirmed. Where confirmation is lacking, an explicitly provisional label—‘agricultural tree configuration potentially compatible with agroforestry; verification required’—is preferable to a definitive classification.

11. Specific Ecological Accounting for Agroforestry Systems

The doctrine proposes a distinct agroforestry descriptive account, integrated with other ecological and statistical accounts. For any given area total, each spatial unit must be counted only once, even where several agroforestry descriptors apply. Overlapping descriptive layers may coexist, but their areas must not be added as though they were separate land surfaces. This rule complements the disaggregated accounting principles in Sections XIX and XX. The account should progressively distinguish:

1. Existing established and maintained agroforestry systems.
2. New systems established on pre-existing agricultural land.
3. Systems arising from documented conversion of natural ecosystems.
4. Systems retaining pre-existing trees or natural forest fragments.
5. Systems undergoing documented ecological rehabilitation or restoration.
6. Systems undergoing simplification, intensification, degradation, or disappearance.
7. Systems whose origin, condition, or trajectory remain insufficiently documented.

Areas, carbon stocks, habitat transformations, production, and ecological effects must not be merged into one performance index. Gross losses of natural forest and other natural ecosystems must remain visible irrespective of agroforestry expansion elsewhere.

Ecological restoration must also be distinguished from agricultural rehabilitation. Ecologically improved agroforestry constitutes ecosystem restoration only insofar as a trajectory consistent with an appropriate reference model and the relevant recovery objectives is demonstrated.

The separate agroforestry account should distinguish existing systems maintained over time; new systems established on pre-existing agricultural land; systems demonstrably derived from natural-ecosystem conversion; systems retaining pre-existing trees or forest fragments; documented ecological rehabilitation or restoration pathways; simplified, intensified, degraded, or abandoned systems; and situations whose origin or trajectory remains inadequately documented. Area, carbon, habitat change, production, and ecological performance must remain analytically separate. In particular, gross natural-forest losses must not be offset within the account by agroforestry expansion elsewhere.

12. Agroforestry Trajectories and Management Responsibilities

Ecological characterization of agroforestry should not be fixed at its establishment date. The Atlas must be able to track several possible trajectories:

Trajectory A — Biocultural continuity. Conservation of long-established systems in which agricultural practices, biological communities, and social uses have evolved together.

Trajectory B — Agricultural diversification. Introduction or development of woody and biological components within pre-existing agriculture, with assessment of benefits and trade-offs.

Trajectory C — Simplification or intensification. Species loss, disappearance of vegetation layers, replacement of native trees, rising inputs, or other changes capable of altering ecological functioning.

Trajectory D — Natural ecosystem conversion. Documented transformation of forest or another natural environment into tree-based agriculture, retaining explicit records of ecological losses and legacies.

Trajectory E — Documented rehabilitation or restoration. Development of degraded land toward explicitly defined ecological objectives, without conflating agronomic improvement with recovery of a natural ecosystem.

Trajectory F — Abandonment, succession, or land-use change. Changed agricultural management that may lead, according to circumstances, to vegetation succession, forest regeneration, altered pastoral organization, or other ecological conditions.

None of these trajectories should automatically be ranked without evidence. Protecting a heritage agroforestry system may require continued human management; natural regeneration on former farmland may instead require reducing intervention. The aim is not to impose one trajectory on all territories, but to document when changes sustain or compromise biodiversity, ecological functions, and community rights.

These trajectories carry no automatic ranking of ecological value. Maintaining an old biocultural system may require continued appropriate human management, whereas allowing natural regeneration on formerly cultivated land may call for reduced intervention. The appropriate assessment concerns the evidence that a trajectory maintains or compromises biodiversity, ecosystem functioning, and the rights of the communities involved, rather than whether management has become more or less intensive in the abstract.

13. Land Rights, Traditional Knowledge, and Biocultural Integrity

Agroforestry systems can reflect knowledge and practices developed across generations. Their assessment cannot be limited to species, area, and yield without considering social institutions, rights of use, land tenure, customary practices, and the actual transmission of knowledge. The Atlas must distinguish ecological observations from information about governance and rights.

Recognition of biocultural value must not be artificially converted into a naturalness score: it is a distinct and complementary dimension. Indigenous Peoples’ and local communities’ knowledge should be used in respect of their rights, institutions, applicable consent requirements, and legitimate disclosure restrictions. Participation must not be reduced to extracting information to populate an externally controlled classification.

Information concerning traditional practices, sensitive locations, exploited resources, or customary territories may require differentiated access controls. An agroforestry scientific classification cannot, by itself, change landownership, create resource-access rights, extinguish customary rights, establish a protected area, or impose a new legal duty. Binding legal consequences require the applicable law and the competent authority.

Ecological observations must be kept distinct from information about tenure, customary uses, institutions, and governance. The recognition of biocultural value is relevant in its own right and must not be artificially converted into a naturalness score. Access to sensitive information about traditional practices, customary territories, resource use, or important locations may need to be restricted. Scientific classification alone does not create, transfer, or abolish property, access, or customary rights, nor does it establish a protected area; binding consequences depend on applicable law and competent authorities.

14. Scientific Programme and Conditions for Validation

Integrating agroforestry into the Global Scientific Annex should constitute an identifiable methodological workstream within the overall UDTR Global Forest Atlas programme. It should address:

• Harmonizing definitions and agroforestry families.
• Ensuring compatibility between land-use classification and ecological characterization.
• Validating an operational global nomenclature and its regional correspondences.
• Methods for detection, mapping, and verification.
• Identifying appropriate ecological reference conditions.
• Selecting robust indicators for composition, function, management, and trajectory.
• Recognizing biocultural systems and rights.
• Quantifying uncertainty and misclassification risk.
• Tracing conversions and ecological gains and losses.
• Procedures for scientific revision, contestation, and version control.

The programme should involve expertise in agroforestry, agronomy, forest ecology, landscape ecology, botany, soil science, hydrology, biodiversity, remote sensing, environmental history, social sciences, and land law. Local and Indigenous knowledge-holders should be involved in ways that respect their authority and rights. The scientific protocol must undergo critical review of the literature, regional comparisons, and methodological testing.

This section does not constitute a nomenclature validated by those actors. It proposes a framework for their expert scrutiny and possible revision. The participation of a researcher, citation of their work, or incorporation of their comments must never be represented as their endorsement of the overall doctrine or of the Universal Declaration of Tree Rights.

The scientific programme must explicitly address classification errors, limits of satellite detection, independent field-validation samples, regionally appropriate ecological references, the comparability of land-use definitions, and transparent procedures for challenging and revising classifications. Researchers' participation, citation, or contribution of comments must not be represented as endorsement of the whole doctrine, the proposed eight-family nomenclature, or the Universal Declaration of Tree Rights.

15. Synthesis: Proposed Doctrinal Principle

Agroforestry must neither be made invisible in global classifications of tree-covered landscapes nor indiscriminately equated with natural forest ecosystems. It comprises diverse socioecological systems in which trees support agricultural and pastoral production, biological interactions, and human practices potentially of substantial food-related, economic, ecological, and biocultural value.

Such value should be documented through evidence-based criteria without erasing ecosystem origins, historical losses, differences in integrity, or the conditions of future development. Protecting natural forest ecosystems and improving tree-based agricultural systems are not competing objectives. They entail complementary responsibilities, but require separate assessment criteria and ecological accounts.

Agroforestry does not replace natural forest. When adapted to its territory and appropriately managed, it can contribute to greater structural complexity, functionality, and resilience in agricultural landscapes. The UDTR Global Forest Atlas should be able to identify its forms, document performance, disclose limitations, and follow change without sacrificing scientific precision to an undifferentiated representation of tree cover.

The aim remains that of the wider forest doctrine: accurately characterize living realities, establish what the evidence allows us to affirm, recognize ecological differences, and inform corresponding human responsibilities.

This framework therefore assigns agroforestry a distinct role without allowing it to stand in for natural forest conservation. It recognizes agricultural, food, economic, ecological, and biocultural value only to the extent supported by appropriate evidence, while making earlier forest conversion, losses of integrity, and unresolved uncertainty visible. The Atlas should document both the potential contributions and the limits of each system, rather than merging all wooded land into an undifferentiated canopy-cover statistic.

VI. The Planet’s Major Forest Biomes

Degree of naturalness alone does not describe the world’s forest diversity. The classification should be cross-referenced with the major climatic domains. FAO reports the 2025 global forest area across tropical, boreal, temperate, and subtropical domains. That broad framework should then be refined by precipitation, seasonality, elevation, soils, and biogeography. Consult the Global Forest Resources Assessment 2025.

This bioclimatic classification should remain compatible with more detailed frameworks, including the IUCN Global Ecosystem Typology, which distinguishes ecosystems by their functions, biotic assemblages, and ecological drivers. The doctrine advanced under the Declaration is not intended to replace those typologies, but to add an analysis of naturalness, protection, and trajectory. Consult the IUCN Global Ecosystem Typology.

VII. Forests in Relation to Water, Soils, and Landforms

A forest never exists independently of its substrate. The same assemblage of species does not have the same ecological meaning on alluvial soil, peat, a dune, a volcanic slope, karst, or permafrost. Water, soil, landform, and physical disturbance should therefore form a separate axis of classification.

1. Major Physical Settings

• lowland and plateau forests;
• slope, ravine, and scree forests;
• montane, subalpine, and cloud forests;
• alluvial and riverine forests;
• riparian and gallery forests;
• swamp and seasonally flooded forests;
• forests on peatlands;
• mangroves and coastal forests;
• dune forests;
• karst and limestone forests;
• volcanic forests;
• forests on permafrost;
• dry and xeric forests or forests subject to pronounced seasonal water limitation.

2. Disturbance Regimes Are Part of the Ecosystem

Fire, wind, floods, avalanches, insects, windthrow, and natural mortality are not always anomalies external to a forest. In many systems they create gaps, recycle nutrients, trigger regeneration, or maintain an age mosaic. The doctrine distinguishes disturbances that are inherent to the ecosystem from direct anthropogenic disturbances, climate-amplified disturbances, and chronic stresses that human activity has intensified, displaced, or rendered destructive.

Protection therefore cannot mean the universal suppression of all fire, flooding, or mortality. It should preserve natural ecological regimes while reducing human-caused risk, ecological threshold crossings, and dangerous consequences for human communities.

A disturbance should be assessed as part of a regime rather than recorded as an isolated event. Relevant attributes include origin, intensity, frequency, duration, spatial extent, recurrence, interactions with other pressures, retained biological legacies, post-disturbance succession and seral state, and—where a demonstrated loss relative to an appropriate reference makes the term applicable—recovery. The analysis should distinguish natural disturbance, direct anthropogenic disturbance, and natural processes whose frequency or severity has been amplified by anthropogenic climate change.

Severity alone does not establish degradation. A severe natural disturbance may remain within the characteristic regime of an ecosystem, initiate biodiversity-rich early-seral stages, and retain the biological legacies required for succession, habitat continuity, and continued ecological functioning. Conversely, degradation risk increases where disturbance intervals become unnaturally short, disturbances become chronic or interact cumulatively, or climate-amplified drought, fire, or insect activity combines with logging, roads, mining, grazing, drainage, fragmentation, or other anthropogenic pressures. In such cases the relevant question is whether the disturbance regime and interacting pressures are pushing the ecosystem beyond its characteristic range of variation toward persistent loss of integrity or ecological type conversion. DellaSala et al. (2025); DellaSala et al. (2014).

3. Natural Post-Disturbance Succession Is Not Presumptively “Recovery”

Recovery is a reference-dependent descriptor; it should not be the default name for every post-disturbance forest trajectory. Where a fire, windthrow, flood, insect outbreak, or other disturbance remains compatible with the ecosystem’s characteristic regime and ecological continuity, the forest may be undergoing natural post-disturbance succession, ecological reorganization, or a seral transition rather than recovering from degradation. The fact that the post-disturbance forest looks radically different from the pre-disturbance forest does not, by itself, establish ecological loss.

This distinction is especially important for complex early-seral forests. After stand-replacing natural disturbance, retained live trees, standing dead trees, downed wood, seed banks, resprouting tissues, fungi, shrubs, regenerating trees, and other biological legacies can generate structurally heterogeneous and species-rich habitat. In such systems, early-seral forest is not merely an incomplete version of mature green forest; it is a functional stage in a connected successional system. DellaSala et al. (2014) document this explicitly for the Sierra Nevada, while DellaSala (2020) emphasizes the multi-scale role of fire-mediated biological legacies and the need to distinguish natural pulse disturbances from chronic land-use disturbances. DellaSala et al. (2014); DellaSala (2020).

Renewal, maturation, senescence, mortality and regeneration are linked and recurring phases of forest dynamics, rather than a linear progression from an ecologically inferior state toward a superior one.

The characteristic seral states of an ecosystem, when expressed through its natural dynamics, each possess ecological value in their own right. They should not be ranked according solely to their resemblance to mature or old-growth forest; their importance must be understood through their ecological functions, biological legacies, role within disturbance and succession regimes, and landscape context.

4. Fire Vulnerability Is Multifactorial and Biogeographically Specific

Wildfire vulnerability and burn severity result from interactions among fire weather, moisture availability, topography, vegetation composition, fuel loads, stand age, forest structure, landscape continuity, and management history. These relationships are non-linear and cannot be reduced to a universal rule according to which older forests always burn more or less severely.

In the sectors analysed after the 2023 Quebec megafires, burn severity was associated most strongly with topographic wetness, fire weather, topographic position, forest age, and vegetation type. Severity was comparatively elevated in forests approximately twenty to forty years old within that study area, followed by a moderate decline and stabilization in older age classes. This regional result should not be universalized, but it demonstrates that uniform landscape rejuvenation cannot be presumed to reduce fire severity.

Fire-resilient landscapes should retain hydrological refuges, natural age diversity, mature and old forest classes, structural heterogeneity, and ecologically appropriate vegetation mosaics: Mackey et al. (2025), Quebec megafire severity.

VIII. Forest Structure and Ecological Integrity

Two areas with the same canopy cover may have radically different structures. A global description should therefore integrate the horizontal, vertical, temporal, and functional dimensions of a stand.

Complementary Ecological and Biocultural Dimensions of Forest Assessment

Fauna, microbial communities, and the interactions that drive regeneration require explicit indicators. Ecological-integrity assessment should not infer biological integrity solely from vegetation structure. Vertebrates, invertebrates, fungi, bacteria, and other microorganisms perform essential functions in dispersal, pollination, herbivory, predation, decomposition, nutrient cycling, and soil processes. Their presence, abundance, functional roles, and community composition should therefore be included where appropriate. Where interactions essential to regeneration are severely depleted—for example dispersal or pollination networks—the persistence of canopy or biomass may conceal a substantial loss of regenerative resilience. No single indicator can determine whether a forest has lost resilience or whether it will recover without assistance; that conclusion requires converging evidence concerning composition, ecological interactions, regeneration, structure, soils, disturbance history, connectivity, and trajectory. See Velásquez-C et al. (2024).

Faunal integrity must be distinguished from the forest’s structural appearance. A continuous canopy or high biomass may coexist with severe defaunation, particularly where hunting, access expansion, or other pressures have reduced large vertebrates, specialist forest birds, certain reptiles, insects, or other functional groups. Research on illegal forest activities shows, in particular, that opening and exploiting large forest areas can increase access and hunting of large seed-dispersing vertebrates and alter community composition, with possible consequences for regeneration and ecosystem functioning. Biological indicators must not, however, be universalized: relevant species or assemblages should be defined according to biome, ecoregion, forest type, the pressure being assessed, sampling methods, and detection limits. See Pacheco et al. (2016) on the multiple and intertwined impacts of illegal forest activities.

High-value timber species may also provide complementary evidence. Their depletion, local disappearance, or altered population structure may indicate a history of selective logging that does not necessarily result in immediate canopy loss. Such signals must be interpreted relative to the expected composition of the forest type, harvesting history, available inventories, and other ecological evidence; the absence or low abundance of a single timber species is never sufficient, by itself, to diagnose degradation.

Below-Ground Biodiversity, Fungal Communities, and Mycorrhizal Evidence

Below-ground biological evidence can materially enrich forest assessment, but its present scientific operability is strongly scale-dependent. In specialist scientific comments provided on 21 September 2026, Professor Leho Tedersoo, Professor of Mycorrhizal Studies at the University of Tartu, advised that mycorrhizal and soil-fungal information can contribute to assessment of ecological condition, integrity, continuity, or trajectory, but only to a limited extent under current knowledge. He identified, among potentially meaningful evidence, the diversity of mycorrhizal associations per area and the presence of protected fungal species in soil or wood, while stressing substantial detection difficulties for the latter. This caution is consistent with the broader scientific literature on the ecological significance and heterogeneity of mycorrhizal types. Tedersoo and Bahram (2019); Tedersoo, Bahram, and Zobel (2020).

Fungal-community variables should therefore not presently be treated as universally standardized forest-condition metrics. Their strongest interpretive value is generally biome-, region-, landscape-, or site-specific. Carefully defined aggregate measures, including average or weighted alpha and gamma diversity, may support broader comparison where datasets are sufficiently representative, but a global value must not be presented as though it captured the full local community or its ecological meaning.

Molecular, metabarcoding, and global fungal datasets can contribute to this evidence layer, but their sampling design, detection limits, taxonomic and functional scope, spatial representativeness, and capacity to capture local variation must be made explicit. Non-detection in a sample or database is not evidence of ecological absence. Global databases of plant mycorrhizal associations are valuable reference resources, but they do not eliminate the need for local and regional validation. Soudzilovskaia et al. (2020), FungalRoot.

Mycorrhizal strategy can nevertheless provide important functional context. Ectomycorrhizal versus arbuscular-mycorrhizal status is a major below-ground trait of trees, but natural systems commonly form a continuum of AM-to-EcM dominance rather than two discrete ecological classes. Some ectomycorrhizal tree lineages also possess physiological traits that do not conform neatly to a binary functional division. The Atlas should therefore record mycorrhizal association or dominance as contextual ecological evidence where relevant, not use AM and EcM as universal mutually exclusive forest classes.

Causal attribution requires a higher evidentiary threshold than description. A difference in fungal-community composition must not, by itself, be interpreted as proof of anthropogenic degradation, ecological recovery, or loss of continuity. Natural biogeographical variation, characteristic disturbance, drought and other microclimatic variation, soil conditions, and management may generate overlapping signals. Trait-based ecology and phylogenetic groupings may improve interpretation, but the present evidence base does not justify a universal causal diagnostic. As emphasized by Tedersoo’s 2026 specialist review, the central scientific error to avoid is the overinterpretation of correlative relationships as demonstrated mechanisms.

No dimension should be treated as a universal substitute for another. High biomass cannot automatically compensate for the loss of fauna, soils, connectivity, or regenerative capacity. Biocultural continuity, tenure, and governance must be documented as distinct and complementary dimensions, without being incorporated as automatic substitutes for or sub-scores of ecological integrity. Depending on the objective, certain criteria may require minimum thresholds or rules of non-compensability.

Ecological integrity results from coherence among composition, structure, function, and process. It does not mean the absolute absence of change. An intact forest evolves, may burn, undergo mortality, migrate, or change in composition. It remains ecologically integral when those transformations are compatible with its fundamental processes, diversity, continuity, and capacity for renewal. This multidimensional approach is consistent with Kormos et al. (2017), Rogers et al. (2022), and Mackey, Morgan, and Keith (2024).

Landscape-scale integrity must also be assessed. The Forest Landscape Integrity Index, for example, combines observed pressures, inferred pressures, and loss of connectivity to show that tree cover alone does not describe the condition of remaining forests. Consult the foundational scientific paper on the Forest Landscape Integrity Index.

Forest degradation may occur without deforestation. A landscape can remain statistically forested while losing old-age classes, large interior cores, connectivity, biological legacies, and functional habitat. Roads, repeated harvesting, edge effects, and the cumulative spatial arrangement of disturbances must therefore be assessed across the landscape and over time. See Mackey et al. (2024) on cumulative impacts in boreal landscapes

IX. Ecological Condition, Integrity, Stability, Trajectory, Origin, Risk, and Vulnerability

These concepts are related, but they answer different questions and should not be merged into a single undifferentiated score. Condition and integrity primarily describe present ecological attributes; stability and trajectory are explicitly temporal; origin and causal attribution explain how observed change arose; risk and vulnerability concern future loss and the system’s susceptibility to it. Their separation is necessary for interpreting change, identifying restoration priorities, and avoiding false conclusions from a single satellite image or inventory.

A stable canopy is not necessarily an integral ecosystem: a simplified plantation or degraded forest may remain spectrally stable for years. Conversely, an unstable or abruptly opened canopy may reflect a normal post-disturbance seral transition or natural succession rather than ecological degradation. Stability indicators therefore require interpretation through ecological history, origin and causal attribution, disturbance regime, biological legacies, vulnerability, and field evidence.

Recovery should therefore be treated as conditional vocabulary. It is appropriate where an ecosystem is being evaluated relative to a demonstrated loss, degradation state, restoration target, or other explicit reference-dependent deficit. It should not be used automatically to imply that every naturally disturbed ecosystem is moving back toward a single preferred prior state. Natural succession may be directional, cyclically connected across seral stages, or otherwise dynamic while remaining compatible with ecological integrity. DellaSala et al. (2025); DellaSala et al. (2014).

This separation is informed particularly by the ecosystem-integrity framework and the boreal canopy-stability analysis: Rogers et al. (2022); Mackey et al. (2024), Canopy Stability Index.

Ecological thresholds and tipping points should be treated with comparable caution. In many forest systems an exact universal threshold cannot be established in advance, even when cumulative evidence shows movement toward a fundamentally altered state. The absence of a precise threshold is not evidence that degradation is absent. Assessment should therefore disclose the indicators of approaching regime change, the interacting pressures involved, and the associated uncertainty rather than manufacture a false numerical certainty. DellaSala et al. (2025).

X. The Global State of Forests in 2025

FAO’s Global Forest Resources Assessment 2025 provides the most comprehensive international statistical baseline. It draws on national reports, common terms, and a coordinated process covering the world’s countries and territories. Its figures should be used rigorously while recognizing differences in methods, inventory capacity, and national categories. Consult the Global Forest Resources Assessment 2025.

These data do not permit the area of “real forests” to be calculated by simple subtraction. Deducting the 312 million hectares of planted forests from the official total of 4.14 billion hectares would be methodologically incorrect. Some planted forests develop into complex ecosystems. Conversely, some naturally regenerating forests are highly degraded, severely fragmented, or depleted of fauna.

The conclusion is therefore not to replace existing statistics, but to supplement them with an accounting of naturalness and integrity.

XI. Primary Forests: Definition, Extent, Distribution, and Threats

1. A Definition Based on Predominant Natural Processes, Not the Absence of People

A primary forest is a naturally regenerated forest ecosystem dominated by native species, in which natural ecological and evolutionary processes, biological legacies, characteristic disturbance regimes, and ecological continuity remain predominant, and whose composition, structure, functions, ecological relationships, and trajectory have not been fundamentally reorganized by industrial intervention or other intensive human intervention. Primary status does not require the absence of human presence. Indigenous Peoples and local communities may live in, govern, and use such forests where their traditional, cultural, or subsistence practices maintain ecological continuity and integrity. The case of the Mẽbêngôkre-Kayapó territories documents the coexistence of long-standing human occupation with extensive areas of largely intact primary forest and underscores that the ecological effects of uses must be assessed empirically rather than inferred from human presence alone. Kormos et al. (2017); Zimmerman et al. (2020).

The relevant distinction is therefore not between a forest “without people” and one “with people.” It distinguishes relationships that maintain the continuity and integrity of the living system from interventions that fundamentally reorganize its composition, structure, functions, ecological relationships, or trajectory. Primary status may coexist with compatible Indigenous, traditional, cultural, or subsistence practices; it may be lost where industrial intervention or other intensive human intervention produces a substantial break in ecological continuity.

Natural disturbances—including fire, storms, flooding, insect outbreaks, or windthrow—and the young or intermediate seral stages that result from them do not, by themselves, terminate primary status where they remain compatible with the ecosystem’s characteristic disturbance regime and ecological continuity. Primary status cannot be excluded solely on the basis of forest age, canopy openness, fragment size, absence of a legal protection status, the existence of a management plan, or historical low-intensity human use.

Qualification must be evidentiary and multidimensional. It must rest on a converging body of evidence concerning, in particular, natural regeneration, ecological continuity, biological legacies, disturbance regimes, the history of interventions and land use, present ecological attributes, and landscape context. No single threshold, canopy appearance, age class, or administrative listing is conclusive by itself. The absence of a forest from an inventory, statistical category, or national or international map does not constitute evidence that it is not primary.

Where credible evidence indicates possible primary status but the available evidence remains insufficient to reach a conclusion with an adequate level of confidence, the classification should remain explicitly provisional or not yet determined. For any decision capable of causing conversion, extraction, or irreversible harm, that uncertainty should not be interpreted against the forest: precautionary protection should apply until an independent scientific assessment has been completed.

2. Resilience, Regeneration, and the Boundary of Primary Continuity

There is no single ecological indicator that can determine whether a forest has lost resilience, whether it can continue through natural succession without assistance or—where degradation has occurred—recover toward an appropriate reference condition, or whether an intervention has crossed the boundary from compatible use into substantial degradation or rupture of primary continuity. The assessment must consider the intensity, frequency, duration, spatial extent, and cumulative effects of interventions together with composition, structure, soils, biological communities, ecological legacies, disturbance regimes, regeneration, connectivity, self-organization, and long-term trajectory. This process-based approach is consistent with Ghazoul and Chazdon (2017), who frame degradation and recovery through changes in regenerative capacity, ecological interactions and feedbacks, resilience, and recovery trajectories rather than through a single universal threshold. Ghazoul and Chazdon (2017).

Regeneration is central to this assessment, but it must be understood as a process involving an ecological community rather than tree recruitment alone. Seed dispersal, pollination, herbivory, predation, recruitment, soil biota, and other interactions help determine whether the forest can continue to renew itself. Their severe loss may leave an area that remains visibly forested while becoming increasingly dependent on human intervention for recovery.

“Recovering forest” is a reference-dependent trajectory descriptor, not a sufficient ecological origin or the default label for a post-disturbance forest. The assessment must first ask whether recovery is even the appropriate concept and, if it is, recovering from what, relative to which reference, and after which disturbance? A primary forest following a natural fire, storm, insect outbreak, or other characteristic disturbance may instead be undergoing post-disturbance succession or occupying a complex early-seral state. Where biological legacies, native processes, natural regenerative pathways, and ecological continuity persist, the system need not be interpreted as degraded merely because it differs from the green forest that preceded the disturbance. By contrast, recovery after logging, road development, conversion, drainage, or another substantial anthropogenic disturbance must retain that causal history in the classification—for example as logged natural forest, degraded natural forest, second-growth forest, or a restoration trajectory, as the evidence warrants. Whenever the term recovery is used, present progress must never erase the causal history and reference condition from which that recovery is being assessed. DellaSala et al. (2014); DellaSala (2020).

The existence of human use does not itself establish a rupture of primary continuity. Archaeological, ecological, and Indigenous evidence shows that areas now regarded as primary may have long histories of stewardship, harvesting, or fire use. The relevant scientific question is therefore the ecological effect and cumulative trajectory of the use, not the mere fact of human presence. Indigenous or community status likewise does not, by itself, constitute evidence of integrity or primary continuity: those properties must be established independently through ecological evidence. This scientific qualification question remains distinct from the strict protection rule adopted by this doctrine, under which commercial or industrial timber or biomass exploitation is excluded from primary forests as a precautionary and non-substitution requirement. Zimmerman et al. (2020).

3. Distinguishing Primary Forest, Old Forest, Mature Forest, Intact Landscape, and High Integrity

• Primary forest refers to natural regeneration, predominance of natural ecological and evolutionary processes, biological legacies, characteristic disturbance regimes, and ecological continuity, without fundamental reorganization of composition, structure, functions, ecological relationships, or trajectory by industrial intervention or other intensive human intervention.
• Old-growth or old forest refers to advanced structural development, including old trees, large diameters, cavities, and deadwood.
• Mature forest refers to a developmental stage and does not, by itself, establish primary continuity.
• Intact forest landscape refers to a large, relatively unfragmented landscape-scale condition and should not erase the value of smaller primary remnants.
• High-integrity forest refers to composition, structure, ecological processes, connectivity, stability, regenerative capacity, and other ecological attributes close to an appropriate ecological reference. Biocultural continuity, tenure, and governance are documented separately.

These concepts should also be compared with operational monitoring frameworks. Global Forest Watch / Global Nature Watch, for example, treats primary forests through ecological-integrity and human-disturbance considerations while recognizing that Indigenous presence is compatible with primary-forest status; intact forest landscapes operate at a larger spatial scale and use explicit size and fragmentation criteria. These approaches are complementary rather than interchangeable. Consult the Global Forest Watch discussion of primary forests.

4. An Uneven Global Distribution That Is Still Imperfectly Documented

Major primary-forest continuities are concentrated in particular within the tropical basins of the Amazon and Congo, the boreal regions of North America and Eurasia, parts of New Guinea and Southeast Asia, and temperate, montane, or island massifs that remain relatively intact. That distribution should not obscure small relict primary forests, which may be essential to an ecoregion and far more vulnerable to a single infrastructure project. Even small primary-forest fragments can be disproportionately important as habitats, sources of seeds and other propagules, refuges for dispersers and pollinators, and ecological anchors for regeneration in surrounding degraded or deforested landscapes. Their small area is therefore not evidence of low conservation value.

Mapping should combine several levels of analysis:

• continent, State, and territory;
• biome and ecoregion;
• river basin;
• size and continuity of the forest complex;
• protection status;
• public, community, Indigenous, private, or shared governance;
• degree of fragmentation;
• current pressures and authorized projects;
• quality and date of the data.

At landscape scale, fragmentation should be treated as a process capable of changing ecological function even where forest cover remains. The intact-forest-landscape framework provides one operational example of explicit scale and fragmentation criteria and of temporal monitoring of intactness. See WRI on fragmentation and intact forest landscapes.

5. Direct and Indirect Threats

Primary forests are threatened by agricultural conversion, livestock expansion, mining, oil and gas development, roads, dams, energy infrastructure, logging, human-caused fires, drainage, pollution, invasive species, unsustainable hunting, land conflict, and land grabbing.

Climate change acts as a threat multiplier. It alters rainfall regimes, increases drought and heat waves, shifts climatic zones, intensifies some fires, and may exceed species’ capacity to migrate or adapt. A primary forest that is legally protected but isolated, subject to drought, hydrologically altered, or surrounded by infrastructure may therefore progressively lose its integrity.

6. Degradation That May Be Invisible from Above

An apparently continuous canopy does not always reveal selective extraction, the disappearance of large animals, the loss of deadwood, hunting, understory alteration, drainage, or soil damage. Remote sensing is indispensable, but it should be complemented by field inventories, Indigenous and local knowledge, and indicators of fauna, soils, structure, and functioning.

The detectability of selective logging is conditional. Specialized methods using Landsat and analyses of canopy structure or opening have detected and mapped moderate- or high-intensity selective logging across very large areas of the Amazon. But that capacity depends on logging intensity, spatial and temporal resolution, analytical method, observation date, and canopy response. Very low-intensity or highly selective extraction may remain undetected. Asner et al. (2006); Matricardi et al. (2020).

Road networks illustrate this requirement directly. In a study of tropical Asia-Pacific, Engert et al. mapped a total road network 3.0 to 6.6 times more extensive than that represented in the principal global datasets used for comparison, and found road density to be the strongest spatial correlate of forest loss among the variables examined. Unmapped roads, logging tracks, and other access routes should therefore be sought as pressures, accessibility signals, and risk factors, without automatically converting them into an ecological diagnosis: the presence of a road does not, by itself, establish the intensity or nature of degradation, and its absence from a database does not prove its absence on the ground. Engert et al. (2024).

This distinction is all the more necessary because degradation may persist for long periods without conversion to non-forest land use. Across the Brazilian Amazon, Matricardi et al. estimated that from 1992 to 2014 the cumulative area of degraded forest exceeded the area deforested, and showed that logging, understory fire, edge effects, and fragment isolation are distinct forms of disturbance. The Atlas should therefore document deforestation, degradation, canopy persistence, and ecological trajectory separately. Matricardi et al. (2020).

XII. Why Primary Forests Are Irreplaceable

A primary forest is not merely a group of older trees. It is the accumulation of an ecological history that cannot be reproduced within the time frame of a human project. Its soils, trees, microorganisms, dispersal networks, predator–prey relationships, cavities, deadwood, genetic variation, and evolutionary continuities have developed over periods that often exceed many human generations. The concept of ecological irreplaceability requires attention to attributes, places, and processes for which replacement or offset claims are scientifically inadequate or subject to major temporal and ecological limits. See Maron et al. (2026).

1. Biological and Genetic Diversity

Primary forests support specialized species that depend on old trees, large diameters, cavities, microhabitats, long-decomposing deadwood, undisturbed soils, or complex ecological relationships. Many cannot persist within short rotations, uniform stands, or fragmented landscapes. A broad tropical meta-analysis found primary forests to be irreplaceable for sustaining tropical biodiversity relative to human-modified forest systems. Gibson et al. (2011).

2. Soils and Underground Networks

A decisive share of the ecosystem is invisible: soil carbon, roots, fungi, bacteria, pore structures, organic horizons, and relationships among plants. Planting young trees does not immediately reconstruct these components. Some may partly recover; others depend on long continuity that is difficult or impossible to recreate in identical form.

The ecological importance of these below-ground components does not mean that present fungal data can serve as a universal proxy for forest continuity or integrity. Mycorrhizal and soil-fungal evidence must be interpreted against appropriate biome- and region-specific references, with sampling representativeness, detection limits, and natural biogeographical variation explicitly considered. Tedersoo’s 2026 specialist review of this framework specifically cautions against inferring degradation or recovery from fungal-community correlations alone.

3. Carbon: Protect Existing Stocks Before Promising Future Flows

Primary forests contain major carbon stocks in biomass and soils. Their destruction causes immediate or progressive emissions, whereas a new plantation absorbs carbon only over time. Comparing only future growth flows may therefore conceal the loss of an already accumulated stock and uncertainty regarding long-term storage.

4. Water, Regional Climates, and Resilience

Large forests influence evapotranspiration, rainfall, runoff, water quality, temperatures, and soil stability. Simplification or fragmentation may reduce those functions even where some tree cover remains.

5. Cultures, Knowledge, and Relationships to Territory

Many primary forests are territories of life, memory, spirituality, food, and knowledge. Their value cannot be reduced to a carbon price, timber volume, or species count. Protecting the ecosystem therefore entails recognizing the peoples and communities connected to it.

6. Temporal Asymmetry: Loss Can Be Rapid, Full Recovery Can Take Centuries

Forest loss and the destruction of ecological legacies can occur over very short periods, whereas complete recovery after that loss may require centuries or, for some compositional and ecological attributes, even longer. This asymmetry is fundamental to conservation decisions: a future possibility of recovery is not temporally or ecologically equivalent to maintaining an existing primary ecosystem.

Recovery rates differ strongly among ecosystem attributes. Canopy closure, structural development, or aboveground biomass may recover substantially before native species composition, rare or poorly dispersed species, old-growth microhabitats, large trees, faunal communities, and ecological interactions have converged toward an old-growth reference. Structural recovery or aboveground biomass should therefore never be used as a universal proxy for recovery of native biodiversity. Ghazoul and Chazdon (2017) emphasize that recovery rates for timber volume, biomass, hydrological services, biodiversity, and other ecosystem functions can differ widely, and that the interpretation of degradation and recovery depends on the attributes and temporal scales assessed. Ghazoul and Chazdon (2017). Recent studies further reinforce this multi-metric and long-term understanding: Metz et al. (2026); Macdonald et al. (2026); Brancalion et al. (2025).

7. Recovery Debt: Ecological Loss Accumulates During the Time to Recovery

Ghazoul and Chazdon (2017) use the concept of recovery debt to connect degradation and recovery through time. Recovery debt represents the cumulative loss of selected ecosystem functions, services, biodiversity, or other attributes relative to a reference or target condition during the period of degradation and subsequent recovery. A system that eventually recovers may therefore still have incurred substantial ecological losses throughout the interval preceding that recovery; where recovery is arrested, the debt may remain open-ended. Ghazoul and Chazdon (2017).

For the purposes of this doctrine, recovery debt strengthens the temporal dimension of non-substitution but does not create a conversion factor by which future restoration can authorize present destruction. Its proper role is to make visible the magnitude and duration of ecological losses once degradation has occurred, to compare recovery trajectories, and to reinforce the priority of avoiding preventable damage to ecosystems whose existing continuity can still be conserved.

Recovery debt must not be assigned merely because a naturally disturbed forest differs from its pre-disturbance appearance. The concept presupposes a demonstrated loss relative to an explicit reference or target condition. A characteristic natural fire that produces a biologically rich early-seral stage does not, by that fact alone, create a recovery debt; the relevant assessment is whether ecological integrity, continuity, or other reference-dependent attributes have actually been lost.

Recovery-Debt Principle

Recovery debt applies to demonstrated ecological loss, not to the mere passage of a forest through a characteristic natural seral stage. Where recovery is genuinely occurring, it can reduce future ecological loss; it does not erase losses accumulated during degradation and recovery, and it cannot be invoked to justify the avoidable destruction of an existing primary forest.

Principle of Irreplaceability

What has developed over centuries or millennia cannot be deemed replaced by planting, by a restoration promise, or by the expectation that ecological recovery may occur in the future.

XIII. The Principle of Non-Substitution: Existing Natural Forests, Plantations, and Future Recovery

The principle of non-substitution provides that the destruction or degradation of a primary forest—and, more broadly, of a high-integrity natural forest—cannot be treated as ecologically offset or justified by establishing a plantation, restoring another site, invoking assisted natural regeneration, or anticipating future ecological recovery, even where the future area is equal or larger.

One hectare of recent plantation, regenerating forest, or restoration project does not replace one hectare of existing primary forest because it does not reproduce its historical continuity, soils, biological legacies, specialized species, diversity of ages, ecological interactions, carbon stocks, or evolutionary continuity. Restoration and regeneration can generate major ecological gains; those gains must be recognized without converting them into a retrospective equivalence with the ecosystem that was lost.

Non-substitution applies not only to land categories, but also to carbon reservoirs, temporal processes, ecological structures, and landscape functions. Fossil carbon emissions are not physically neutralized by temporary terrestrial storage; an immediate loss is not equivalent to a future and uncertain recovery; the return of canopy cover is not equivalent to the recovery of integrity; and deadwood retained within an ecosystem is not ecologically equivalent to material removed and burned as an energy feedstock. Mackey et al. (2013); Mackey et al. (2020); Mackey et al. (2025) on biomass.

Relationship to the international restoration Standards. This doctrinal rule finds important methodological support in, but is not identical to, the third edition of the Society for Ecological Restoration Standards. Those Standards state as an underpinning assumption that the potential for ecological restoration should never be invoked as justification for destroying, damaging, or unsustainably utilizing existing native ecosystems. They nevertheless recognize compensatory restoration and offsets within some legal frameworks and recommend offsets only after the earlier steps of the mitigation hierarchy have been thoroughly examined and shown to be ineffective and technically unachievable. The principle of non-substitution advanced in this article is therefore a distinct requirement of the Declaration and should not be presented as a rule adopted by SER. Gann et al. (2026).

Proper use of recovery-potential evidence. Expected natural regeneration, assisted natural regeneration, and restoration potential are relevant to identifying suitable recovery areas, choosing interventions, estimating time horizons, and designing monitoring. They should not be used to reduce the conservation requirement applying to an existing primary forest or to justify its destruction or conversion. This distinction is especially important where primary remnants themselves provide the seed sources, habitat, dispersers, connectivity, or ecological legacies on which surrounding regeneration depends.

1. The Mitigation Hierarchy and the Doctrine’s Non-Substitution Rule

The sequence below corresponds to the widely used mitigation hierarchy. The doctrine applies it across multiple ecological dimensions and adds an explicit rule of non-substitution: mitigation or compensation in one dimension cannot automatically erase losses in another, and residual measures cannot be presented as ecological equivalence. See Chazdon (2020). The following hierarchy should apply:

1. avoid destruction or degradation;
2. minimize impacts that cannot be avoided;
3. restore on site the functions that have been altered;
4. remedy residual harm without presenting that remedy as ecological equivalence;
5. create additional gains elsewhere only where those gains are additional to protection of the threatened ecosystem and are not used to authorize its destruction.

2. The Difference Between Accounting Compensation and Ecological Restoration

An offset may balance an administrative, financial, or carbon unit. It does not necessarily recreate the ecosystem that was destroyed. Restoration seeks to recover composition, structure, function, and trajectory. Even when successful, it requires time and does not guarantee the return of every lost attribute.

3. The Disclosure Rule

Any project causing the loss of a natural forest should separately disclose:

• the area and ecosystem type lost;
• the degree of naturalness and integrity;
• the species, soils, stocks, and functions affected;
• permanent and temporary impacts;
• avoidance and minimization measures;
• on-site restoration measures;
• plantations or actions carried out elsewhere, without aggregating them into the balance of the natural forest.

XIV. Second-Growth (Secondary) Forests and Restoration Trajectories

The higher conservation priority accorded to primary-forest protection should not lead to the undervaluation of second-growth (secondary) forests. In many territories, they constitute much of the remaining forest, provide major habitat, protect water, store carbon, and offer a principal opportunity to recover ecological integrity and high naturalness. Young regrowth forests are themselves conservation priorities because their persistence strongly affects future recovery trajectories. See Chazdon (2025).

1. Recognizing Their Value Without Erasing Their History

A mature second-growth (secondary) forest may become highly complex and functional. It may regain a diversified canopy, substantial biomass, decomposition cycles, rich fauna, and self-regeneration. It does not thereby become historically primary, nor should a restored or naturally regenerating ecosystem be assumed to be ecologically equivalent to an existing or previous primary ecosystem merely because some attributes have recovered. Different components recover at different rates, and species composition, rare or dispersal-limited taxa, old-growth structures, and ecological interactions may require far longer than canopy or biomass. Maintaining the distinction preserves the truth of the trajectory without diminishing the ecological value acquired.

2. Favoring Natural Regeneration First

Where soils, hydrology, seed sources, landscape connectivity, ecological interactions, and human pressure permit, natural regeneration—including assisted natural regeneration where necessary—often provides the closest fit to local conditions. It mobilizes existing genetic diversity, allows locally adapted recruitment, and avoids some of the impacts associated with intensive site preparation. Remaining native and primary vegetation can be decisive: even small fragments may provide seeds, habitat, dispersers, pollinators, microclimatic refuges, and other ecological legacies that support recovery of surrounding areas.

Natural regeneration may be assisted by protection from grazing or cutting, targeted control of invasive species, hydrological restoration, limited enrichment planting, creation of ecotones, reconnection of fragments, or—where key ecological interactions have been lost—carefully designed measures intended to restore the processes required for regeneration. Assisted natural regeneration should remove or reduce barriers to recovery rather than substitute permanent management for ecosystem self-organization.

3. Where Recovery Is the Appropriate Descriptor, It Is Multidimensional: Structure and Biomass Are Not Proxies for Biodiversity

Where a documented loss, degradation state, restoration objective, or other explicit reference makes recovery the appropriate descriptor, recovery must be assessed through multiple, non-interchangeable dimensions. Canopy cover, basal area, height, or aboveground biomass can recover relatively rapidly while taxonomic composition, functional composition, old-growth-associated species, large trees, faunal communities, and ecological interactions remain substantially different from reference conditions. A structural metric can therefore demonstrate structural recovery without demonstrating full biological or ecological recovery.

Recent evidence illustrates this divergence. In a tropical rainforest, recovery across sixteen taxonomic groups showed strong recovery of abundance and diversity while community composition converged more slowly toward old-growth conditions; boreal evidence likewise shows that biodiversity recovery after clear-cut harvest can extend over many decades and, for some groups and forest types, beyond conventional harvest rotations. Biodiversity-centred restoration research further emphasizes that successful restoration must monitor taxa and ecological outcomes beyond tree establishment alone. Metz et al. (2026); Macdonald et al. (2026); Brancalion et al. (2025).

4. Intervening Without Locking the Ecosystem into Permanent Dependence

Restoration may require substantial initial work. Its purpose should nevertheless be to restore to the ecosystem an increasing capacity to function without continual intervention. A project that indefinitely requires planting, inputs, artificial drainage, or intensive control may improve some functions, but it has not yet recovered advanced naturalness. Where scientifically and practically feasible, treated areas should be evaluated against appropriate reference conditions and against comparable areas undergoing unassisted recovery over time, so that the effect of intervention can be distinguished from recovery that would have occurred without it.

5. Measuring Long Time Horizons

Results should be monitored across several horizons:

• approximately fifteen years: establishment, survival, hydrology, erosion, early regeneration, and immediate pressures;
• approximately twenty-five years: structural development, recruitment, soil and nutrient trajectories, fauna, and connectivity;
• approximately fifty years: diversification, self-regeneration, functional and compositional trajectories, mature structural attributes, large-tree development, and landscape integration, without presuming convergence with old-growth reference conditions;
• one century: maturity, old trees, deadwood, soils, and functional continuities;
• several centuries: very high ecological naturalness, advanced succession, and intergenerational continuity.

These horizons are monitoring milestones, not predicted dates of ecological equivalence. Recovery may be rapid for some attributes and extremely slow for others. Published evidence therefore supports comparing multiple structural, taxonomic, functional, and compositional metrics over decades rather than inferring recovery from a single indicator. Ghazoul and Chazdon (2017) show that recovery rates differ among biomass, biodiversity, hydrological services, and other ecosystem attributes; more recent studies likewise document divergent and often prolonged recovery trajectories across biological and structural dimensions. Ghazoul and Chazdon (2017); Metz et al. (2026); Macdonald et al. (2026).

International restoration standards reinforce this trajectory-based approach: ecological restoration should be guided by an appropriate reference model of a native ecosystem, measurable attributes, transparent monitoring, and adaptive management. The 2026 third edition also places ecological restoration within a broader continuum of restorative activities and aligns it with the United Nations Decade on Ecosystem Restoration. Consult the International Principles and Standards for the Practice of Ecological Restoration.

XV. The Global Program for the Recovery of Forest Naturalness

The Declaration could establish an international program for areas that are no longer primary but may recover very high ecological naturalness over multiple generations.

In this programmatic title, recovery is used for areas whose documented anthropogenic alteration, degradation, conversion history, or loss of ecological continuity provides a reference-dependent basis for seeking greater naturalness. It is not intended to describe a primary forest merely passing through a characteristic natural disturbance cycle or seral stage.

1. Eligible Areas

• mature second-growth (secondary) forests;
• long-continuity forests that have been exploited but retain major ecological continuities;
• degraded but still connected natural forest complexes;
• former plantations capable of developing toward autonomous forest structure;
• buffer zones around primary forests;
• corridors between natural forest complexes;
• strategic watersheds;
• restorable alluvial forests and forested peatlands;
• former agricultural land undergoing natural or assisted natural regeneration;
• territories in which a community seeks to commit to very long-term succession.

2. Conditions for Registration

Registration should not be granted on the basis of a promise alone. It would require:

• an independent initial assessment;
• public mapping and stable boundaries;
• identification of ownership, tenure, land-use rights, customary rights, and governance arrangements;
• an ecological baseline and reference condition;
• a stated trajectory and measurable indicators;
• a prohibition on conversion;
• a financial regime ensuring continuity of the undertaking;
• a scientific and community-based monitoring system;
• review, alert, and accountability mechanisms.

Admission and verification require an institutional architecture. Eligibility should be determined through a transparent protocol by an independent or pluralistic assessment body whose composition, expertise, conflicts of interest, procedures, evidence standards, and review mechanisms are public. The program should specify who conducts the initial assessment, who validates registration, how contested rights or data are handled, and how periodic re-assessment may confirm, revise, suspend, or withdraw status.

Long-term protection also requires durable incentives and finance. Possible mechanisms may include public conservation finance, payments for verified ecosystem stewardship, community and Indigenous direct-access funding, conservation easements or covenants where domestic law permits, endowments, trusts, fiscal incentives, and other arrangements that reward protection without converting the registered area into an offset authorizing ecological loss elsewhere.

3. Long-Term Regime

Depending on the needs of each site, registered areas would be subject to a regime including:

• no industrial exploitation;
• priority for natural or assisted natural regeneration where ecologically appropriate;
• initial restoration of soils and hydrology;
• progressive removal of unnecessary infrastructure;
• targeted control of invasive species where indispensable;
• reconnection with neighboring ecosystems;
• retention of old trees, deadwood, and microhabitats;
• local, Indigenous, community, public, private, or shared governance consistent with recognized ownership, tenure, customary, and land-use rights;
• protection against land speculation and changes of use;
• indefinite duration or a legally secured commitment extending across several centuries.

4. A Transparent International Register

For every site, the program should publish its status, initial condition, owners or custodians, use rights, funding, indicators, interventions, results, and any failures. Registration must not become a decorative label or an opaque carbon asset.

XVI. Sanctuaries for Multi-Century Forest Succession

Within the Global Program, certain areas could receive the enhanced status of a “regenerating forest sanctuary” or, where more appropriate, a “forest under protected natural succession.” These terms may include regenerating degraded forests as well as other eligible recovery areas. “Sanctuary” does not imply exclusion of human communities. It denotes protection of the ecological trajectory against conversion, production rotations, and incompatible changes in land use.

A forest of recovered naturalness may perform many of the ecological functions associated with a primary forest. It may become long-established, mature, rich in deadwood, and highly diverse. It does not, however, recreate exactly the history that was lost. Preserving that distinction prevents restoration from becoming an argument for destroying the last areas of primary-forest continuity.

Legally Securing Duration Across Generations

The principal challenge is institutional. Political mandates, budgets, and contracts are short; forest development unfolds over centuries. Regenerating forest sanctuaries or areas under protected natural succession therefore require instruments capable of maintaining continuity beyond individual owners, governments, and generations: conservation easements or covenants adapted to domestic law, trusts, foundations, protected areas, legally recognized Indigenous territories, very long-term agreements, perpetual endowment funds, and public oversight mechanisms.

XVII. Global Objectives for 2030, 2050, 2100, and Beyond

A single global target for forest cover would be scientifically dangerous. It could lead to tree planting in savannas, grasslands, steppes, heathlands, open peatlands, or tundra whose ecological value depends precisely on their non-forest character.

Objectives should therefore be defined by biome, ecoregion, watershed, potential natural vegetation, species requirements, connectivity, climate projections, land and territorial rights, and the need to protect other natural ecosystems from inappropriate afforestation.

1. Proposed Baseline for 2030

The restoration objective corresponds to Kunming–Montreal Target 2, which calls for at least 30 percent of areas of degraded terrestrial, inland water, coastal, and marine ecosystems to be under effective restoration by 2030. The conservation objective corresponds to Kunming–Montreal Target 3, which calls for ecologically representative, well-connected, and equitably governed systems, while recognizing Indigenous and traditional territories.

The 30 percent threshold constitutes a minimum floor for representative conservation, not a presumed level of ecological sufficiency. Depending on the ecological functioning, history, degree of fragmentation, hydrological requirements, species, connectivity, and vulnerability specific to each biome or ecoregion, the ecologically necessary proportion may be substantially higher—40 percent, 60 percent, 80 percent, or more where scientific assessment so justifies.

This minimum does not in any way limit the protection of all remaining primary forests or the higher levels of conservation required by the ecological characteristics of a territory.

Relationship to Existing Global Forest Frameworks

The proposed objectives should be read in relation to, rather than in isolation from, existing forest commitments. The New York Declaration on Forests (2014) is a voluntary, multi-stakeholder political declaration launched at the 2014 UN Climate Summit; it is not a treaty or a resolution of the United Nations General Assembly. Its goals include ending natural forest loss, restoring degraded landscapes and forestlands, improving governance and finance, and reducing emissions from deforestation and degradation. Consult the New York Declaration on Forests.

The United Nations Forest Instrument, renamed in 2015 and adopted by the General Assembly through resolution A/RES/70/199, provides a globally agreed, non-legally binding framework for sustainable forest management. The United Nations Strategic Plan for Forests 2017–2030, adopted by the General Assembly through resolution A/RES/71/285, establishes six voluntary and universal Global Forest Goals. Consult the United Nations Strategic Plan for Forests 2017–2030.

The contribution proposed by this doctrine is more specific and ecological: it separates natural forest loss from plantation gain; distinguishes ecological origin, condition, management, and trajectory; applies a principle of non-substitution to primary and high-integrity natural forests; requires reporting of ecological quality as well as area; and extends restoration and natural-succession commitments across intergenerational and multi-century horizons. These propositions are intended to complement existing instruments, not to replace or reinterpret their legal status.

2. Objectives for 2050

The Global Biodiversity Framework calls for ecosystem integrity, connectivity, and resilience to be maintained, enhanced, or restored, and for the area of natural ecosystems to increase substantially by 2050. The proposed forest doctrine translates that direction into the following objectives:

• a substantial increase in the area of natural forests;
• no gross or net loss of natural forests;
• restoration of continuity among major forest complexes;
• re-establishment of essential hydrological regimes;
• a major reduction in fragmentation;
• restoration of all severely depleted forest ecoregions;
• a representative network of forests placed under very long-term natural-process management in every biome;
• transition of the most vulnerable production plantations toward more diversified landscape mosaics;
• legal security for the rights, financing, and governance required for conservation.

Consult the 2050 goals of the Global Biodiversity Framework.

3. The Horizon of 2100 and Beyond

By 2100, States and territories should be able to demonstrate:

• that an ecologically sufficient share of every forest ecoregion is placed under indefinite natural succession;
• that continental and transboundary corridors function in practice;
• that primary forests are surrounded by buffer zones and ecologically compatible landscapes whose land uses maintain or restore connectivity, hydrological function, fire-regime compatibility, low edge pressure, permeability to species movement, and respect for recognized land and territorial rights;
• that mature forests are being rebuilt across several human generations;
• that the area of forests of very high ecological naturalness is measurably increasing;
• that legal commitments are carried forward across generations.

“Compatible landscape” is not a residual label. It means a surrounding landscape whose uses and management do not undermine the ecological continuity of the protected forest and, where possible, actively support connectivity, hydrology, disturbance-regime coherence, climate adaptation, community rights, and reduced conversion pressure. Compatibility must be assessed against the needs of the particular biome, ecoregion, species assemblages, and rights-holders rather than presumed from a generic land-use category.

4. Why No Single Global Percentage Should Be Declared at This Stage

The ecological need for forest varies by region. A historically forested and heavily deforested ecoregion may require extensive restoration. A natural steppe or savanna may, by contrast, be degraded by afforestation. Percentages should therefore remain indicators derived from ecological analysis, not abstract objectives imposed uniformly.

XVIII. Protection of the Rights of Indigenous Peoples and Local Communities

A global doctrine of forest protection would be legally and morally unacceptable if it turned peoples who live in forests into obstacles to conservation. The territorial, cultural, political, and economic rights of Indigenous Peoples and local communities must be integrated into the very definition of effective protection. This recognition does not, however, remove the fundamental scientific requirement to distinguish rights, tenure, governance, uses, and the ecological condition of the forest. Consult the United Nations Declaration on the Rights of Indigenous Peoples. Consult Target 22 of the Kunming-Montreal Global Biodiversity Framework.

1. Do Not Presume Either Degradation or Integrity from the Identity of Populations

Many primary or high-integrity forests have been maintained in territories inhabited and governed over long periods. Traditional practices may contribute to regeneration, fire stewardship, habitat diversity, seed dispersal, and protection against industrial conversion. But no favorable ecological relationship should be presumed solely from the Indigenous, community, or traditional status of a territory.

Indigenous or community presence, traditional occupation of a territory, customary rights, or local governance do not, by themselves, constitute either evidence of degradation or evidence of ecological integrity. The effects of a use must be characterized from its history, intensity, duration, extent, and actually observed consequences for composition, structure, functions, regeneration, soils, hydrology, fauna, connectivity, and ecosystem continuity.

The documented experience of the Mẽbêngôkre-Kayapó illustrates this requirement. Zimmerman et al. (2020) show that extensive Indigenous territories can retain largely intact primary forest where territorial control, institutions, economic alternatives, and conservation partnerships support ecological continuity; they also show that other sectors belonging to the same people and subject to different extractive pressures may follow trajectories of degradation. Garcia et al. (2024) formalize this distinction by assessing ecosystem integrity, governance, and planning separately. Zimmerman et al. (2020); Garcia et al. (2024).

2. Distinguish Recognized Rights from Effective Territorial Control

Legal recognition of a territory, ownership right, tenure, or land-use rights does not, by itself, demonstrate that those rights can be effectively exercised or that the forest is protected in practice. Assessment must distinguish legal status from actual governance and territorial-control capacities: representative institutions, participation, decision-making capacity, surveillance, access control, human and financial resources, enforcement capacity, continuity of funding, the presence of violations or illegal extractive activities, and the degree of exposure to external pressures.

The Kayapó case is particularly instructive: the existence of the same broad territorial and cultural setting has not produced the same ecological outcomes everywhere. Available studies identify a strong association between institutional and territorial capacity supported by Kayapó organizations and their partners, on the one hand, and the maintenance of an extensive area of primary forest, on the other. This relationship must nevertheless be interpreted cautiously: even a strong spatial association supported by plausible mechanisms must not automatically be converted into a universal causal relationship applicable to all peoples and territories. Garcia et al. (2024).

3. Document Economic and Institutional Pressures Without Confusing Them with Ecological Condition

Forest trajectories may be influenced by economic, technological, extractive, or institutional pressures: logging, mining, roads, agricultural frontiers, monetary economies, access to illegal markets, weak law enforcement, corruption, tenure insecurity, dependence on particular income sources, or the absence of economic alternatives compatible with conservation. These factors may help explain a trajectory, but they are not themselves measures of ecological integrity.

Conversely, mechanisms such as income from non-timber forest products, territorial surveillance, community institutions, long-term partnerships, or conservation financing may contribute to forest maintenance without their mere existence establishing good ecological condition. Ecological outcomes must remain independently verified.

4. Minimum Safeguards

• recognition of lands, territories, and resources;
• respect for representative institutions;
• full, effective, and equitable participation in decision-making;
• free, prior and informed consent where required by applicable standards;
• protection of customary uses compatible with conservation;
• direct access to financing where relevant;
• protection of environmental and human-rights defenders;
• respect for ownership, collective authority, and confidentiality of knowledge;
• fair benefit-sharing;
• access to justice and effective remedies.

These requirements are consistent with the United Nations Declaration on the Rights of Indigenous Peoples and with the participation, information, justice, and rights safeguards reflected in Target 22 of the Global Biodiversity Framework. Practical rights-based forest governance should also draw on the experience and documentation of organizations working directly on Indigenous Peoples’ and forest peoples’ rights, including land and territorial rights, self-determination, free, prior and informed consent, and access to justice. Consult the United Nations Declaration on the Rights of Indigenous Peoples Consult Target 22 of the Global Biodiversity Framework. See also Forest Peoples Programme.

5. Reject Exclusionary Conservation

The creation of a protected area must not result in displacement, criminalization, or loss of livelihoods for populations whose rights must be respected and whose practices may have contributed to maintaining the ecosystem. Protection must be equitably governed, recognize Indigenous territories, and integrate communities into monitoring, decision-making, and resource allocation.

6. Indigenous and Local Knowledge: Provenance, Authority, Consent, and Limits on Use

Indigenous and local knowledge may provide irreplaceable information about territorial history, species, uses, disturbances, seasonal cycles, stewardship practices, and ecological transformations. It must not, however, be treated as a freely extractable database. Its use within the Atlas must respect provenance, legitimate knowledge-holders and authorities, applicable consent requirements, disclosure restrictions, confidentiality of sensitive information, and any collective conditions governing access or reuse.

The evidentiary value of Indigenous or local knowledge depends on the question being examined, its context, its transmission, whether it can be documented without violating the rights of knowledge-holders, and its relationship with other available sources. Recognition must neither assimilate it to conventional scientific data nor exclude it because it belongs to a different knowledge system.

XIX. Global Indicators of Ecological Naturalness, Integrity, Management, and Connectivity

The Declaration should propose a global dashboard complementary to existing inventories. Indicators should be measurable, comparable, and sufficiently disaggregated to prevent gains in plantation area from concealing losses of natural forest.

From Doctrine to Verification: Principle, Criteria, Indicators, and Verifiers

A doctrinal principle becomes operational only when it is connected to criteria, indicators, and verifiers. This chain of evidence should be adapted to the biome, ecoregion, natural disturbance regime, spatial scale, decision context, and quality of available data.

This architecture is reinforced by the multi-scale degradation framework developed by DellaSala et al. (2025), which applies a Principle–Criteria–Indicators–Verifiers approach to departures from high-integrity reference conditions. Its illustrative verifier families include age and diameter structure, large old trees, snags and coarse woody debris, carbon stocks, native and non-native species, threatened species, hydrological condition, soil compaction and productivity, mycorrhizal functionality, climate buffering, patch size, road density, roadless area, connectivity, and cumulative disturbance. These examples should not become universal fixed thresholds: the specific indicators and verifiers used by the Atlas must remain calibrated to forest type, spatial scale, available evidence, and the decision being examined. For mycorrhizal functionality and fungal-community evidence in particular, specialist review by Professor Leho Tedersoo (2026) requires an additional safeguard: such evidence may be scientifically informative at site, landscape, regional, or biome scales, but should not presently be treated as a globally standardized standalone indicator of integrity, degradation, continuity, or recovery; interpretation must disclose sampling representativeness, detection limits, biogeographical context, and the distinction between correlation and demonstrated mechanism.

This architecture follows the methodological framework developed for forest-landscape integrity: Mackey, Morgan, and Keith (2024).

The Central Accounting Principle

National and international reports should separately disclose:

• losses of primary forest;
• losses and degradation of natural forest;
• managed natural forests, with management objective and intensity reported separately;
• gains through natural expansion or regeneration;
• areas under ecological restoration;
• planted establishment used within ecological restoration or rehabilitation pathways, identified separately from the ecological condition ultimately attained;
• production-oriented tree plantations;
• offset or carbon plantations;
• agroforests and other agroforestry or silvopastoral systems;
• other agricultural, urban, and linear tree cover.

Losses should be disclosed on a gross basis before any net balance is presented. A gain in tree cover, plantation area, or recovering forest elsewhere should not cancel the record of primary-forest loss or degradation of a high-integrity natural forest. Gross accounting preserves the ecological identity of what was lost and what was gained and prevents a positive aggregate balance from concealing irreversible or long-duration losses. DellaSala et al. (2025).

XX. Complementary Ecological Accounting: Forest Ecosystems, Plantations, and Other Tree-Based Systems

Complementary ecological accounting does not abolish FAO categories and does not mean that forests are divided into only two ecological classes. It proposes two reporting layers: the existing international statistical account and a multi-category ecological account that disaggregates ecological origin, condition, management regime, trajectory, and type of tree-based system. A State could therefore continue reporting forest area under international criteria while separately publishing the ecological nature and management of the areas concerned.

Carbon-Stock Clarification

Consistent with Brancalion and Chazdon (2017), pre-existing or residual carbon stocks retained in natural forest ecosystems must be distinguished, both quantitatively and qualitatively, from carbon newly accumulated through natural regeneration, ecological restoration, or tree planting. Existing stocks and future removals are not ecologically or temporally interchangeable. The clearing of an old-growth or high-integrity forest causes immediate carbon emissions and losses of biodiversity, structure, soils, and ecosystem functions, whereas carbon accumulation through regeneration or restoration is gradual, less predictable, and may require decades or longer. Restoration reporting should therefore disclose separately the conservation or loss of pre-existing stocks, avoided emissions, newly accumulated carbon, the period required for its accumulation, permanence risks, and the ecological nature of the system in which that carbon is stored. Brancalion and Chazdon (2017).

An Example of Reporting That Should No Longer Be Acceptable

A country that loses 100,000 hectares of natural forest and establishes 120,000 hectares of plantations should not report only a “net gain” of 20,000 hectares. Its report should simultaneously disclose:

• 100,000 hectares of natural forest lost;
• the naturalness and integrity of the forests destroyed;
• 120,000 hectares of plantations established;
• their composition, purpose, location, and management regime;
• the statistical balance, presented as such and not as an ecological gain.

XXI. Forest Carbon, Carbon-Carrying Capacity, and Biomass Energy

Forest carbon policy must distinguish existing stocks, potential ecological capacity, gross emissions, gross removals, avoided emissions, foregone sequestration, and permanence risks. A single net balance can conceal the loss of old stocks and the degradation of the living system that stores them.

1. Current Carbon Stock and Ecological Carbon-Carrying Capacity

Current Carbon Stock (CCS) is the carbon presently stored in an ecosystem, reflecting both environmental conditions and the accumulated effects of land use and forest management. Carbon-Carrying Capacity (CCC) is the stock that an ecosystem type could maintain under its environmental conditions and characteristic natural disturbance regime in the absence of direct anthropogenic depletion. The difference between CCC and CCS can reveal historical depletion, foregone storage, and recovery potential.

European primary-forest data show the importance of using an ecological reference rather than treating historically depleted stocks as normal: Keith et al. (2024).

2. Gross Emissions, Gross Removals, and Foregone Sequestration

Gross emissions and gross removals should be disclosed separately. A territorial forest sector may remain a net sink while harvesting, clearing, or combustion still generates substantial emissions. Accounting should also include foregone sequestration: the carbon that a forest would probably have continued to accumulate if it had been protected and allowed to mature.

The Tongass rainforest provides a concrete illustration of this temporal asymmetry. Drawing on prior carbon simulations, DellaSala and Furnish (2020) report that only a no-logging scenario maintained carbon stores over time and that, after clearcutting an old forest, regenerating young forest may remain a net CO₂ emitter for approximately five to fifty years depending on site productivity. This regional case should not be generalized as a universal recovery period, but it demonstrates why future regrowth cannot be booked as an immediate climatic equivalent to the loss of an existing old-growth carbon stock. DellaSala and Furnish (2020).

3. Terrestrial Carbon Does Not Physically Offset Fossil Carbon

Terrestrial carbon storage is finite and reversible, whereas fossil carbon introduces additional carbon into the active atmosphere–ocean–land system from geological reservoirs. Forest protection and restoration are indispensable complements to rapid reductions in fossil-fuel emissions, not substitutes for them.

The distinction among reservoirs, stocks, flows, and time horizons is established in: Mackey et al. (2013).

4. Forest Biomass Energy and the Absence of Presumed Carbon Neutrality

The biological renewability of wood does not create a presumption of immediate climate neutrality. Combustion releases carbon at the time it occurs, whereas any subsequent removal through regrowth is delayed, uncertain, and dependent on future management, climate, disturbance, and land-use permanence.

Forest residues are ecological components contributing to soils, nutrient cycles, moisture regulation, habitat, decomposition, and carbon storage. They should not automatically be classified as waste. Any assessment of forest biomass energy should compare the full bioenergy pathway with a counterfactual protection or restoration pathway and account for gross emissions, foregone sequestration, processing and transport, ecosystem integrity, biodiversity adaptation, and the time required for any carbon recovery.

The scientific basis for this precaution is developed in: Mackey, Lindenmayer, Keith, and de Bie (2025).

XXII. What This Doctrine Changes for Silviculture, Climate, and Biodiversity

1. For Silviculture

The doctrine does not abolish silviculture. It clarifies its respective domains. Timber production should be organized in places and through methods compatible with soils, water, biodiversity, resilience, and social needs. It must not extend into primary forests, which are subject to strict and permanent protection excluding all commercial or industrial timber or biomass harvesting, including harvesting described as selective, sanitary, salvage, or post-disturbance. A natural disturbance must never be treated as an automatic opening to exploitation.

Post-disturbance logging requires particular caution because it can transform an ecologically functional disturbance into a second, anthropogenic disturbance. Removal of surviving trees, snags, deadwood, and other biological legacies after fire, windstorm, or insect outbreak may alter nutrient cycling, soils, succession, habitat, hydrology, and the ecosystem’s capacity to continue through natural post-disturbance pathways. In complex early-seral forests, those legacies are not merely debris awaiting removal; they are structural and biological components of the post-disturbance ecosystem. The fact that trees may subsequently be planted does not by itself reverse those losses. DellaSala et al. (2025); DellaSala et al. (2014); DellaSala (2020).

It calls for:

• placing primary forests under strict and permanent protection, outside all commercial or industrial timber or biomass exploitation, while maintaining Indigenous, traditional, cultural, or subsistence uses compatible with their ecological continuity and integrity;
• conserving high-integrity forests and ancient structures;
• diversifying production plantations at stand and landscape scales;
• extending rotations where doing so improves soils, carbon stocks, and habitats;
• reducing large, continuous clear-cuts;
• maintaining ecotones, riparian forests, old trees, corridors, and senescence islands;
• clearly separating production, restoration, and conservation objectives;
• providing remuneration for ecological functions that do not immediately produce timber.

The Tongass case also illustrates a broader planning principle. Where timber production remains economically and socially necessary, transitions can be designed to reduce or end reliance on remaining old-growth forests by shifting supply, where ecologically appropriate, toward previously logged young-growth stands or other already transformed production areas. DellaSala and Furnish (2020) concluded that a faster Tongass transition could be supported by young growth already regenerating from earlier clearcuts, on a substantially smaller timber land base than the transition then proposed, while retaining greater ecological and climate benefits in old growth. This is an illustration, not a universal prescription: second-growth forests can themselves have high conservation value, and any production allocation must remain subject to site-specific ecological classification, landscape needs, rights, and cumulative-impact assessment. DellaSala and Furnish (2020).

2. For Climate Policy

The doctrine calls for a clear distinction between existing carbon stocks and future removals. Protecting an existing stock avoids emissions and maintains an ecosystem. Planting may create a future sink, but that sink is subject to time, fire, drought, harvesting, and permanence of land use.

Climate policies should therefore report separately:

• emissions avoided through protection of natural forests;
• removals attributed to natural regeneration;
• removals from ecological restoration;
• temporary stocks held in production plantations;
• emissions associated with harvesting, soils, inputs, and land-use change;
• non-permanence risks.

Climate claims should also disclose the ecological category in which carbon is stored, the difference between current stock and ecological capacity, gross emissions from harvesting or combustion, foregone sequestration, the expected recovery period, and risks of non-permanence. Forest biomass energy should receive no general presumption of carbon neutrality.

3. For Biodiversity

The doctrine shifts evaluation from a simple count of hectares to ecosystem representativeness, habitat quality, and continuity of processes. It calls for the protection not only of iconic large forest complexes, but also of underrepresented relict, alluvial, dry, island, montane, coastal, and other forest ecosystems.

4. For Finance, Certification, and Corporate Commitments

Public and private actors should separately report impacts on primary forests, other natural forests, degraded forests, and plantations. A “zero-deforestation” commitment should not authorize degradation. A “nature-positive” claim should not aggregate plantations and natural ecosystems. A carbon obligation should not override biodiversity and rights criteria.

5. For Land-Use Planning

Decisions should be made at the scale of watersheds, ecoregions, biological continuities, and fire regimes. Administrative and property boundaries correspond neither to water flows, species movement, nor local climate.

XXIII. What the Declaration Does Not Say

The doctrinal scope of this article requires explicit limits.

The Declaration and Legal Personality

The Universal Declaration of Tree Rights does not confer legal personality upon trees; legal personality is not provided for in its text. The Declaration does not, by itself, create a cause of action, confer legal capacity to hold or administer assets, or reallocate public powers or jurisdiction. Legal systems may adopt mechanisms of representation, protection, or rights of nature, but those mechanisms should not be attributed to the Declaration when they do not appear in its text.

Its tree-specific scope should likewise not be read as denying or competing with broader rights-of-nature frameworks or proposals for a general bill of rights for nature. The Declaration develops a specific architecture around the Tree and the ecological systems in which trees participate. Broader legal frameworks may recognize rights of ecosystems, species, rivers, landscapes, or nature as a whole; those approaches can be compared, combined, or independently adopted by competent legal systems, but they remain conceptually and legally distinct and should not be attributed to the Declaration unless they are expressly incorporated through a separate instrument.

The Declaration and Production

The distinction between forest and plantation is not intended to disqualify all economic activity. It is intended to prevent economic use from erasing the nature of the system being used. A production plantation may be acknowledged, improved, and integrated within a landscape. It should not become the replacement metric for a natural forest.

XXIV. Conclusion — Passing On Forests, Not Merely Planted Areas

The global forest crisis cannot be understood from the number of hectares covered by trees alone. The world may gain plantations while simultaneously losing natural forests, ancient soils, specialized species, ecological continuities, and cultures connected to forest territories.

The doctrine advanced under the Universal Declaration of Tree Rights rests on a simple requirement: the realities involved must be named before they are counted. A primary forest, a second-growth (secondary) forest, a degraded forest, a restoration area, a production plantation, an agroforest, a hedgerow landscape, and an urban park may all contain trees. They do not share the same history, autonomy, function, or replacement value.

Within the doctrine advanced here, protecting primary forests constitutes a paramount requirement of strict and permanent conservation, incompatible with commercial or industrial timber or biomass exploitation; restoring degraded forests constitutes a requirement of ecological repair. Enabling new forests to recover high naturalness is a responsibility toward the future. Producing timber within transparent, resilient, and appropriately located systems is an economic necessity that should no longer be confused with the conservation of natural ecosystems.

The true ambition is therefore not to maximize a single global figure for forest cover. It is to guarantee, in every biome and ecoregion, a sufficient area of natural forests that are ecologically integral, connected, diverse, and capable of regeneration, while rebuilding over centuries the continuities that human societies have interrupted.

Article III of the Universal Declaration of Tree Rights

“Human beings, endowed with reason and conscience, must act with the Tree in a spirit of fraternity and solidarity.”

To act with the Tree, at the global level, means protecting what cannot be replaced, restoring what can still be restored, producing without concealing impacts, and accepting responsibility for trajectories whose full realization will belong to future generations.

We must pass on not merely trees, but forests.

XXV. Legal and Methodological Note

The classification set out in this article constitutes a doctrinal proposal advanced under the Universal Declaration of Tree Rights. It serves ecological, interpretive, and programmatic functions. It may provide a basis for voluntary commitments, public policies, charters, scientific work, monitoring systems, or future standards. Consult the Official Framework of the Universal Declaration of Tree Rights.

It does not constitute:

• an amendment to FAO terms and definitions;
• a new binding category of international law;
• an automatic amendment of forest codes, property rights, or protected-area regimes;
• a site-specific expert assessment in the absence of field data;
• a presumption that every planted forest is devoid of ecological functions;
• a presumption that every naturally regenerating forest is intact.

The proposed classifications should be applied on the basis of verifiable data, ecological assessment, historical analysis, local and Indigenous knowledge, mapping, indicators, and a transparent and procedurally fair review in which affected rights-holders and stakeholders can be heard.

XXVI. Global Scientific Annex and the Global Forest Atlas

The doctrine advanced in this article proposes the progressive establishment of a Global Scientific Annex and a Global Forest Atlas. The Atlas is conceived as an independent international scientific instrument. It may be recognized and used for the purposes of the Universal Declaration of Tree Rights, the International Convention on Tree Rights, or other instruments without being established, administered, or governed by their bodies. Their purpose is not to replace existing international forest inventories, national forest classifications, scientific ecosystem typologies, or remote-sensing platforms. Their purpose is to connect those sources, where scientifically and legally appropriate, through the multidimensional ecological framework developed in this article.

The Global Forest Atlas should therefore not be conceived as another map of tree cover. It should operate as an evolving, spatially explicit, temporal, and evidentiary system capable of distinguishing what kind of forest or tree-based system is present, what ecological condition it is in, how that condition has changed, what pressures affect it, what rights and governance arrangements apply, how certain the available evidence is, and what conservation, restoration, monitoring, or management needs follow from that evidence.

Its central purpose is to make visible a distinction that conventional area statistics cannot establish by themselves: where trees occur is not the same question as what kind of living forest ecosystem exists there, what has happened to it, or what it is becoming.

The Atlas should be developed progressively. It should not claim global completeness before sufficient evidence exists. Areas or variables for which reliable evidence is unavailable should be identified as such rather than assigned an artificial classification.

1. Institutional and Scientific Purpose

The Global Scientific Annex should provide the methodological framework through which the doctrine may progressively be translated into operational assessment. The Global Forest Atlas, with autonomous scientific, methodological, and institutional governance, should provide the corresponding spatial and temporal expression of that framework. This methodological relationship should neither subordinate the Atlas nor confer upon institutions associated with the Declaration, bodies of the Convention, or States any authority over its classifications, assessments, data governance, or publication decisions.

The two instruments have distinct but complementary functions:

• the Global Scientific Annex should define concepts, criteria, indicators, verifiers, evidentiary standards, methodological safeguards, uncertainty rules, validation procedures, and revision procedures;
• the Global Forest Atlas should organize and display geographically explicit evidence and classifications produced through those methods.

Neither instrument should create a presumption that a mapped classification is legally binding on a State, landowner, Indigenous People, community, public authority, or other rights-holder. Their function is ecological, scientific, interpretive, informational, and programmatic unless a competent legal authority independently decides otherwise.

Institutional stewardship of the project should remain distinct from scientific validation. The independent organization responsible for developing and maintaining the Atlas should guarantee methodological transparency, while scientific classifications and revisions should be subject to pluralistic expertise, documented evidence, declared conflicts of interest, and review procedures appropriate to the consequences of the classification concerned.

2. Data to Be Established for Each Biome, Ecoregion, and Relevant Forest Unit

Assessment should operate across several nested spatial scales. The ecoregion provides a major ecological reference, but it should not constitute the only spatial unit. Depending on the issue being examined, assessment may also concern a biome, river basin, landscape, forest complex, forest ecosystem, stand, patch, restoration site, or other scientifically relevant unit.

For each ecoregion and, where appropriate, for finer-scale forest units, the Atlas should progressively document:

• historical, reconstructed, or ecologically potential distribution of forest ecosystems, with the limitations of such reconstruction explicitly stated;
• current area meeting international statistical forest definitions;
• primary-forest area;
• forests of very high ecological naturalness;
• second-growth or secondary forests, including relevant maturity, secondary-succession, or recovery stages;
• degraded natural forests;
• managed natural forests, with management objective and intensity reported separately;
• forests and other eligible areas undergoing ecological restoration;
• areas under protected natural succession;
• planted establishment used within restoration or rehabilitation pathways;
• production-oriented tree plantations;
• agroforests, agroforestry systems, and silvopastoral systems;
• other relevant forms of tree cover;
• connectivity, fragmentation, interior habitat, ecological corridors, and landscape permeability;
• hydrological condition and major soil constraints;
• below-ground biological evidence, including fungal and mycorrhizal data where sufficiently sampled, with taxonomic or functional scope, sampling design, detection limits, representativeness, and biome- or region-specific reference context explicitly recorded;
• characteristic disturbance regimes;
• post-disturbance seral or developmental state, retained biological legacies, and evidence distinguishing natural succession from reference-dependent recovery;
• anthropogenic disturbances and chronic pressures;
• climate exposure, vulnerability, adaptive capacity, and risk of ecological transition;
• current carbon stocks, ecological carbon-carrying capacity where scientifically assessable, gross emissions, gross removals, foregone sequestration, and permanence risks;
• ownership, tenure, land-use rights, customary rights, Indigenous and community territories, and governance arrangements where such information may lawfully and ethically be disclosed;
• protection status and evidence of effective protection;
• essential human needs and compatible uses;
• non-forest ecosystems that require protection from inappropriate afforestation;
• priorities for conservation, restoration, protected succession, field investigation, risk reduction, or compatible production;
• the quality, age, spatial resolution, and limitations of the evidence supporting each assessment.

The absence of data must never be interpreted as evidence of ecological absence, low ecological value, absence of rights, or absence of degradation.

3. A Multidimensional Atlas Rather Than a Single Forest Score

The Atlas should not reduce forest condition to a universal composite score.

Naturalness, integrity, stability, carbon, biodiversity, connectivity, governance, vulnerability, and trajectory answer different questions. A high value in one dimension should not automatically compensate for serious degradation in another.

For example, high biomass cannot by itself compensate for the disappearance of fauna, hydrological disruption, loss of connectivity, impaired regeneration, or violation of legitimate territorial rights. Likewise, a legally protected status cannot by itself establish effective ecological protection, and stable canopy cover cannot establish ecological integrity.

The Atlas should therefore preserve the principal dimensions of assessment separately and allow synthetic indicators only where their construction, weighting, limits, and non-compensability rules are scientifically justified and publicly documented.

4. Evidentiary Architecture of the Global Forest Atlas

The Atlas should operate as a temporal system of evidence, not as a static representation of tree cover.

Its minimum architecture should distinguish the following ten analytically distinct layers:

These layers should be interoperable but should remain analytically distinguishable.

5. Provenance, Metadata, and Chain of Evidence

Every mapped value or classification should be traceable to the evidence from which it is derived.

Where relevant, the Atlas should disclose at least:

• date or period of observation;
• geographical extent;
• spatial resolution;
• data producer;
• original source;
• version of the source data;
• method of acquisition;
• sampling design, sampling intensity, detection limits, and representativeness where the evidence derives from biological sampling;
• analytical method;
• ecological reference used;
• principal assumptions;
• known limitations;
• uncertainty;
• confidence level or evidentiary status;
• date of Atlas assessment;
• applicable version of the scientific protocol;
• reviewing or validating body where applicable.

A user should therefore be able to distinguish an observation directly measured in the field from a satellite-derived indicator, a modelled estimate, a historical reconstruction, an expert interpretation, a community-defined indicator, or an inference based on several converging sources.

Remote sensing is indispensable for global screening, repeated observation, disturbance detection, and temporal monitoring. It cannot, by itself, establish every component of ecological integrity, historical continuity, land and territorial rights, biological interactions, or the condition of soils and microbial communities. Field assessment, historical evidence, ecological expertise, and Indigenous and local knowledge remain indispensable where the question being asked requires them.

6. Uncertainty, Confidence, and Insufficient Evidence

Scientific uncertainty must be visible rather than concealed.

The Atlas should distinguish between the ecological classification itself and the degree of confidence that can reasonably be placed in that classification. The Global Forest Atlas Scientific Protocol, in its current Version 0.1.6 dated 25 October 2026, defines a standardized evidentiary scale capable of distinguishing, at minimum:

• classifications supported by strong and convergent evidence;
• classifications supported by substantial but incomplete evidence;
• provisional classifications requiring further verification;
• contested classifications;
• areas or variables for which evidence is insufficient to support a classification.

The category “insufficient evidence / not yet determined” is therefore an essential component of the Atlas.

Uncertainty should never be converted automatically into a lower ecological classification. A poorly documented forest is not thereby a degraded forest, a secondary forest, a plantation, or an ecologically unimportant area.

7. Reference Conditions and Regional Calibration

No global methodology can treat all forests as though the same structural characteristics, species diversity, fire regime, carbon density, regeneration dynamics, or hydrological conditions were universally normal.

Indicators and thresholds must therefore be calibrated to appropriate ecological references and spatial scales.

For degradation assessment, primary or old-growth forests should be used as reference conditions where suitable examples remain. Where they do not, the reference may draw on near-natural or mature forests, historical reconstruction, naturally regenerating structurally complex forests, and other evidence appropriate to the ecosystem concerned. The Atlas should record the basis of the reference, its spatial and temporal relevance, and the confidence attached to it rather than treating a depleted contemporary baseline as ecologically normal. DellaSala et al. (2025).

A naturally monodominant tropical forest should not be penalized for low tree-species diversity merely because another tropical forest type is species-rich. An open boreal woodland should not be assessed against the canopy closure of a temperate rainforest. A naturally fire-dependent ecosystem should not be evaluated according to a universal objective of fire exclusion.

The Atlas should establish common global principles while permitting scientifically justified biome-, ecoregion-, ecosystem-, and indicator-specific calibration.

8. Below-Ground Biodiversity, Fungal Communities, and Mycorrhizal Evidence

The Atlas should recognize below-ground biodiversity as an ecologically important but methodologically constrained evidence domain. Specialist review by Professor Leho Tedersoo, University of Tartu, on 21 September 2026 indicates that current knowledge does not support treating mycorrhizal or soil-fungal information as a universally standardized diagnostic of forest condition, integrity, continuity, or trajectory. Its evidentiary role should therefore be explicitly tiered by spatial scale, sampling quality, and interpretive purpose.

At site and landscape scales, the Atlas may record directly sampled fungal-community composition, diversity of mycorrhizal associations, protected fungal species, functional groups, molecular or metabarcoding observations, and relevant soil or wood-associated fungal evidence. At regional or biome scales, such evidence may be compared only against ecologically appropriate references and sampling frameworks. At global scale, carefully harmonized aggregate measures may be informative, but should not imply universal ecological equivalence or local completeness.

Mycorrhizal association should be treated as a functional trait rather than a rigid global forest class. Ectomycorrhizal and arbuscular-mycorrhizal dominance frequently form a continuum, and the ecological consequences of mycorrhizal status interact with host identity, soils, climate, nutrient availability, disturbance, and other traits. Tedersoo and Bahram (2019); Tedersoo, Bahram, and Zobel (2020).

Every below-ground biological record used interpretively should disclose, where applicable, sampling location and extent, sampling date, number and distribution of samples, laboratory or molecular method, taxonomic or functional resolution, detection limits, database coverage, representativeness, relevant soil and microclimatic conditions, and the ecological reference against which interpretation is made. Existing global datasets such as FungalRoot may support trait and association context, but cannot substitute for site-level evidence when a site-level conclusion is required. Soudzilovskaia et al. (2020).

Observation, association, and causal attribution must remain separate evidentiary levels. A fungal-community pattern may be described and compared without being interpreted as proof that a particular disturbance, management action, degradation process, or recovery mechanism caused it. Where several causal explanations remain scientifically plausible, the Atlas should preserve those alternatives and classify causal attribution as unresolved rather than manufacture mechanistic certainty.

9. Scientific Governance, Review, and Conflicts of Interest

The Global Scientific Annex should be developed through pluralistic scientific governance. The Atlas should have its own pluralistic and autonomous scientific, methodological, and institutional governance. Recognition of the Atlas or use of its outputs should not confer upon any institution associated with the Declaration, any body of the Convention, or any State authority to direct, alter, delay, or suppress its methods, classifications, assessments, or scientific publications.

Relevant expertise should include, according to the questions examined:

• forest ecology;
• botany;
• zoology;
• mycology and microbial ecology;
• soil science;
• hydrology;
• biogeography;
• climate science;
• disturbance ecology and fire science;
• remote sensing and geospatial science;
• restoration ecology;
• forestry and silviculture;
• conservation biology;
• landscape ecology;
• carbon science;
• environmental and land law;
• social sciences;
• Indigenous knowledge and governance;
• community-based forest management.

Scientific participation should not be interpreted as endorsement of the Universal Declaration of Tree Rights, of every doctrinal proposition contained in this article, or of every classification subsequently produced by the Atlas.

Membership, mandates, methodological roles, conflicts of interest, funding relationships, declarations of interest, and review procedures should be transparent.

10. Indigenous Peoples, Local Communities, Sensitive Knowledge, and Data Governance

The Atlas must not become a mechanism through which ecological knowledge is extracted from peoples or communities and made public without legitimate authority.

Indigenous Peoples and local communities should participate in the governance and interpretation of information concerning their territories where appropriate. Their knowledge should not be treated merely as an auxiliary data source to be absorbed into an external classification system.

The Atlas should recognize that some information must not be openly displayed. This may include, depending on context and the wishes or rights of those concerned:

• sacred or culturally sensitive sites;
• confidential traditional knowledge;
• precise locations of threatened species vulnerable to collection or persecution;
• information capable of facilitating illegal exploitation;
• data subject to legal, ethical, contractual, customary, or community-defined restrictions.

The Global Forest Atlas Scientific Protocol, in its current Version 0.1.6 dated 25 October 2026, therefore distinguishes between publicly accessible data, restricted scientific data, and confidential or non-disclosable information.

11. Contestability, Correction, and Scientific Versioning

No classification should be treated as permanently immune from challenge.

The Atlas should establish a transparent procedure through which researchers, public institutions, Indigenous Peoples, local communities, rights-holders, civil-society organizations, and other persons with relevant evidence may request review of a classification.

A review request should be capable of relying on, among other evidence:

• peer-reviewed scientific studies;
• field inventories;
• ecological assessments;
• historical documentation;
• remote-sensing analyses;
• land and territorial records;
• verified local observations;
• Indigenous and local knowledge supplied under appropriate conditions;
• evidence of disturbance, restoration, or land-use change occurring after the previous assessment.

Changes should be versioned rather than silently overwritten.

Where practicable, the Atlas should retain a record of:

Scientific Versioning

previous classification → evidence submitted → review → decision → revised classification

together with the applicable dates and protocol versions.

Scientific disagreement should be disclosed when it cannot reasonably be resolved by the evidence presently available.

12. Monitoring Protocols and the Global Forest Atlas Scientific Protocol

This article defines the principal questions the Atlas should address. It does not yet prescribe a complete universal methodology for measuring every variable.

A dedicated Global Forest Atlas Scientific Protocol has therefore been established as the methodological framework for identification, classification, evidence, validation, uncertainty, scientific review, and temporal monitoring within the Atlas. Its current Version 0.1.6, dated 25 October 2026, incorporates the successive methodological safeguards developed since the foundational version.

The protocol defines, at minimum:

• spatial units and rules for changing scale;
• criteria and indicators for each assessment dimension;
• verifiers and acceptable data sources;
• minimum metadata;
• sampling design;
• remote-sensing calibration and validation;
• field protocols;
• ecological reference models;
• treatment of historical evidence;
• procedures for Indigenous and local knowledge;
• treatment of missing data;
• uncertainty estimation;
• confidence and evidentiary-status rules;
• quality assurance and quality control;
• non-compensability rules among ecological dimensions where justified;
• conditions for automated or model-assisted classification;
• conditions requiring expert review;
• review and appeal procedures;
• data-governance rules;
• update frequency;
• version control;
• rules for revising or withdrawing classifications.

Monitoring frequency should depend on the variable concerned. Rapid disturbances may require continuous or near-real-time alert systems where technically possible. Tree cover, fragmentation, fire, drought, and infrastructure pressures may be assessed annually or at other appropriate intervals. Structure, soils, hydrology, fauna, fungal and mycorrhizal communities, other microorganisms, regeneration, and management effects may require periodic field reassessment or appropriately designed molecular sampling, with intervals and sampling intensity calibrated to the variable and ecological context. Very long ecological trajectories should also be reviewed across the multi-decadal and multi-century horizons described elsewhere in this doctrine.

The intervals identified in this article—approximately fifteen, twenty-five, fifty, one hundred, and five hundred years—are long-term monitoring horizons for the recovery of ecological naturalness. They are not universal measurement frequencies for every indicator.

13. Progressive Implementation and Pilot Regions

The Atlas should be developed progressively rather than through an immediate claim of complete global classification.

Its first operational phase should test the scientific protocol across ecologically contrasted forest systems. Pilot regions should be selected so that the methodology is confronted with substantially different ecological histories, disturbance regimes, management contexts, data availability, governance systems, and restoration challenges.

A pilot phase should therefore include, where possible, contrasting examples such as:

• tropical humid forest;
• boreal forest;
• temperate forest;
• Mediterranean or seasonally dry forest;
• flooded, peat-forming, coastal, or mangrove forest systems.

Pilot classifications should be explicitly identified as such and used to test, revise, and strengthen the methodology before broader deployment.

Global coverage should subsequently distinguish between areas that are:

• methodologically assessed;
• undergoing assessment;
• provisionally classified;
• contested;
• supported only by partial evidence;
• not yet evaluated.

Scientific completeness should never be simulated by filling evidentiary gaps with unjustified certainty.

14. Interoperability with Existing Scientific and Institutional Systems

The Global Forest Atlas should be designed to complement rather than reproduce or supersede existing systems.

Where legally, technically, and scientifically possible, it should remain interoperable with established international statistical definitions, ecosystem typologies, national forest inventories, biodiversity databases, remote-sensing programs, protected-area datasets, climate and carbon datasets, restoration monitoring systems, and relevant territorial or governance information.

The original category, terminology, resolution, methodology, and limitations of imported or linked data should remain identifiable.

The Atlas should never relabel a statistical category as an ecological conclusion without the additional evidence required by this doctrine.

15. No Automatic Equivalence Between Mapping and Legal Status

Inclusion in the Atlas should not, by itself:

• alter ownership or tenure;
• create or extinguish land-use rights;
• establish a protected area;
• confer legal personality;
• create a cause of action;
• prohibit an activity otherwise lawful under the applicable legal system;
• authorize an activity otherwise unlawful;
• determine compensation;
• replace environmental assessment;
• substitute for site-specific ecological expertise;
• resolve a territorial or jurisdictional dispute.

Conversely, the absence of a forest from the Atlas should not be interpreted as evidence that the forest does not exist, lacks ecological importance, or is available for conversion.

The Atlas is an instrument of ecological knowledge, transparency, comparison, monitoring, and decision support. Any binding legal consequence must arise from the competent legal authority and the applicable legal framework.

16. From Observation to Responsibility

The ultimate purpose of the Global Forest Atlas is not cartographic completeness.

It is to make ecological differences, trajectories, natural successions, reference-dependent recoveries, losses, uncertainties, and responsibilities visible at the scales at which decisions are actually made.

A forest should therefore be capable of being followed not only as a location, but as a documented ecological history:

Temporal Chain of Evidence

what it was → what happened to it → what remains → what seral, successional, recovery, or decline trajectory it is following → what risks it faces → what protection, non-intervention, or restoration is required → what evidence supports that conclusion

Such an Atlas would permit forest policy to move beyond the question of how many hectares are covered by trees and toward the more demanding questions posed throughout this doctrine: what kind of forest exists, what degree of integrity and continuity it retains, what trajectory it is following, and what must be done to ensure that a living forest ecosystem can still be passed to future generations.

XXVII. Glossary of Key Terms

This glossary clarifies the principal scientific, ecological, statistical, legal, governance, and doctrinal terms used throughout this article. Its purpose is terminological precision: several expressions are established scientific or institutional terms, while others are definitions or interpretive formulations developed specifically for the ecological doctrine advanced under the Universal Declaration of Tree Rights.

A

B

C

D

E

F

G

H

I

L

M

N

O

P

R

S

T

Agroforestry — Additional Terminology for Section V bis

Agroforestry — Land-use systems intentionally combining perennial woody plants with crops, livestock, or pastoral activities in spatial arrangements or temporal sequences, involving ecological or socioeconomic interactions.

Agroforest — An agroforestry system with a potentially multilayered tree structure; the label does not itself establish natural, primary, or high-integrity forest status.

Agrisilviculture — Association of trees and agricultural crops.

Silvopastoralism — Association of trees, pasture, and/or animals in a managed system.

Agrosilvopastoralism — Combination of agricultural crops, trees, and livestock.

Sequential agroforestry — A managed temporal arrangement alternating or sequencing agricultural and woody components.

Principal References

The revised article distinguishes the core scientific corpus that directly informed the consolidation from institutional and methodological references, complementary doctrinal and contextual materials, and documented external scientific review. Inclusion in this bibliography, or incorporation of scientific comments, does not imply endorsement of the article, its classifications, or its programmatic proposals by the authors, reviewers, correspondents, or institutions cited.

A. Core Scientific Corpus

1. Mackey, B., Prentice, I. C., Steffen, W., House, J. I., Lindenmayer, D. B., Keith, H., and Berry, S. (2013). “Untangling the Confusion Around Land Carbon Science and Climate Change Mitigation Policy.” Nature Climate Change, 3, 552–557.
2. DellaSala, D. A., Bond, M. L., Hanson, C. T., Hutto, R. L., and Odion, D. C. (2014). “Complex Early Seral Forests of the Sierra Nevada: What Are They and How Can They Be Managed for Ecological Integrity?” Natural Areas Journal, 34(3), 310–324.
3. Mackey, B., DellaSala, D. A., Kormos, C., Lindenmayer, D., Kumpel, N., Zimmerman, B., Hugh, S., Young, V., Foley, S., Arsenis, K., and Watson, J. E. M. (2015). “Policy Options for the World’s Primary Forests in Multilateral Environmental Agreements.” Conservation Letters, 8(2), 139–147.
4. Chazdon, R. L., Brancalion, P. H. S., Laestadius, L., Bennett-Curry, A., Buckingham, K., Kumar, C., Moll-Rocek, J., Vieira, I. C. G., and Wilson, S. J. (2016). “When Is a Forest a Forest? Forest Concepts and Definitions in the Era of Forest and Landscape Restoration.” Ambio, 45, 538–550.
5. Kormos, C. F., Mackey, B., DellaSala, D. A., Kumpel, N., Jaeger, T., Mittermeier, R. A., and Filardi, C. (2017). “Primary Forests: Definition, Status and Future Prospects for Global Conservation.” Encyclopedia of the Anthropocene.
6. Brancalion, P. H. S., and Chazdon, R. L. (2017). “Beyond Hectares: Four Principles to Guide Reforestation in the Context of Tropical Forest and Landscape Restoration.” Restoration Ecology, 25(4), 491–496.
7. Ghazoul, J., and Chazdon, R. L. (2017). “Degradation and Recovery in Changing Forest Landscapes: A Multiscale Conceptual Framework.” Annual Review of Environment and Resources, 42, 161–188.
8. Tedersoo, L., and Bahram, M. (2019). “Mycorrhizal Types Differ in Ecophysiology and Alter Plant Nutrition and Soil Processes.” Biological Reviews, 94(5), 1857–1880.
9. Tedersoo, L., Bahram, M., and Zobel, M. (2020). “How Mycorrhizal Associations Drive Plant Population and Community Biology.” Science, 367(6480), eaba1223.
10. Soudzilovskaia, N. A., Vaessen, S., Barcelo, M., He, J., Rahimlou, S., Abarenkov, K., Brundrett, M. C., Gomes, S. I. F., Merckx, V., and Tedersoo, L. (2020). “FungalRoot: Global Online Database of Plant Mycorrhizal Associations.” New Phytologist, 227(3), 955–966.
11. Mackey, B., Kormos, C. F., Keith, H., Moomaw, W. R., Houghton, R. A., Mittermeier, R. A., Hole, D., and Hugh, S. (2020). “Understanding the Importance of Primary Tropical Forest Protection as a Mitigation Strategy.” Mitigation and Adaptation Strategies for Global Change, 25, 763–787.
12. DellaSala, D. A. (2020). “Forest Biome: Trees of Life.” In M. I. Goldstein and D. A. DellaSala (eds.), Encyclopedia of the World’s Biomes, Volume 3, 1–15. Elsevier.
13. DellaSala, D. A. (2020). “Fire-mediated Biological Legacies in Dry Forested Ecosystems of the Pacific Northwest, USA.” In E. A. Beever, S. Prange, and D. A. DellaSala (eds.), Disturbance Ecology and Biological Diversity: Scale, Context, and Nature. CRC Press.
14. DellaSala, D. A., and Furnish, J. (2020). “Can Young-Growth Forests Save the Tongass Rainforest in Southeast Alaska?” Encyclopedia of the World’s Biomes, Volume 3, 218–225.
15. Rogers, B. M., Mackey, B., Shestakova, T. A., Keith, H., Young, V., Kormos, C. F., DellaSala, D. A., Dean, J., Birdsey, R., Bush, G., Houghton, R. A., and Moomaw, W. R. (2022). “Using Ecosystem Integrity to Maximize Climate Mitigation and Minimize Risk in International Forest Policy.” Frontiers in Forests and Global Change, 5, 929281.
16. Mackey, B., Morgan, E., and Keith, H. (2024). “Evaluating Forest Landscape Management for Ecosystem Integrity.” Landscape Research, 49(2), 246–267.
17. Keith, H., et al. (2024). “Carbon Carrying Capacity in Primary Forests Shows Potential for Mitigation Achieving the European Green Deal 2030 Target.” Communications Earth & Environment, 5, 256.
18. Mackey, B., Campbell, C., Norman, P., Hugh, S., DellaSala, D. A., Malcolm, J. R., Desrochers, M., and Drapeau, P. (2024). “Assessing the Cumulative Impacts of Forest Management on Forest Age Structure Development and Woodland Caribou Habitat in Boreal Landscapes: A Case Study from Two Canadian Provinces.” Land, 13, 6.
19. Mackey, B., Hugh, S., Norman, P., Rogers, B. M., and DellaSala, D. (2024). “Insights into Boreal Forest Disturbance from Canopy Stability Index.” Land, 13, 1644.
20. DellaSala, D. A., Mackey, B., Kormos, C. F., Young, V., Boan, J. J., Skene, J. L., Lindenmayer, D. B., Kun, Z., Selva, N., Malcolm, J. R., and Laurance, W. F. (2025). “Measuring Forest Degradation via Ecological-Integrity Indicators at Multiple Spatial Scales.” Biological Conservation, 302, 110939.
21. Mackey, B. G., Lindenmayer, D. B., Keith, H., and de Bie, J. (2025). “Burning Forest Biomass Is Not an Effective Climate Mitigation Response and Conflicts With Biodiversity Adaptation.” Climate Resilience and Sustainability, 4, e70015.
22. Mackey, B., Bhatt, T. R., Norman, P., DellaSala, D., Rogers, B. M., and Hugh, S. (2025). “Exploring the Determinants of the 2023 Quebec, Canada, Mega-Wildfire Burn Severity.” International Journal of Wildland Fire, 34, WF24175.
23. Gibson, L., Lee, T. M., Koh, L. P., Brook, B. W., Gardner, T. A., Barlow, J., Peres, C. A., Bradshaw, C. J. A., Laurance, W. F., Lovejoy, T. E., and Sodhi, N. S. (2011). “Primary Forests Are Irreplaceable for Sustaining Tropical Biodiversity.” Nature, 478, 378–381.
24. Chazdon, R. L. (2020). “Creating a Culture of Caretaking through Restoring Ecosystems and Landscapes.” One Earth, 3(6), 653–656.
25. Velásquez-C, K. L., Pérez-Maqueo, O., Guevara, R., Verde Arregoitia, L. D., and Munguía-Carrara, M. (2024). “A Systematic Review of the Role of Terrestrial Vertebrates in Ecological Integrity Assessment.” Environmental and Sustainability Indicators, 23, 100426.
26. Chazdon, R. L. (2025). “Young Regrowth Forests Are Worth Saving.” Nature Ecology & Evolution, 9, 1090–1091.
27. Metz, T., et al. (2026). “Biodiversity Resilience in a Tropical Rainforest.” Nature, 652, 1232–1239.
28. Macdonald, S. E., et al. (2026). “Biodiversity Recovery Is Slow Following Clear-Cut Harvest of Boreal Forests.” Nature Sustainability.
29. Brancalion, P. H. S., Hua, F., Joyce, F. H., Antonelli, A., and Holl, K. D. (2025). “Moving Biodiversity from an Afterthought to a Key Outcome of Forest Restoration.” Nature Reviews Biodiversity, 1, 248–261.
30. Maron, M., Ward, M., Simmonds, J. S., Wintle, B. A., Possingham, H. P., Venegas, R., Butler, D., Macintosh, A., Reside, A. E., Sonter, L., Dunn, D. C., Kerswell, A., Dielenberg, J., and Watson, J. E. M. (2026). “Ecological Irreplaceability in the Era of Nature Positive.” Conservation Letters, 19(4), e70073.
31. Gann, G. D., McDonald, T., Walder, B., Aronson, J., Nelson, C. R., Hallett, J. G., Guariguata, M. R., Gonzales, E. K. G., Hua, F., Echeverría, C., Eisenberg, C., Liu, J., Decleer, K., Barr, Z. E., Bartholomew, D. C. B., Best, M., Chazdon, R., Cliquet, A., Cortina-Segarra, J., Kalirai, H., MacCallum, K., Mosyaftiani, A., Pedrini, S., Young, R., and Dixon, K. W. (2026). “International Principles and Standards for the Practice of Ecological Restoration. Third Edition.” Restoration Ecology, 34, e70441.
32. Zimmerman, B., Schwartzman, S., Jerozolimski, A., Esllei, J., Santini, E., and Hugh, S. (2020). “Large Scale Forest Conservation With an Indigenous People in the Highly Threatened Southeastern Amazon of Brazil: The Kayapo.” In M. I. Goldstein and D. A. DellaSala (eds.), Encyclopedia of the World’s Biomes, volume 3, 27–34. Elsevier.
33. Garcia, B., Morgan, E. A., Aruch, M., Ferreira, I. R., Jerozolimski, A., Mackey, B., and Hugh, S. (2024). “Large-scale forest protection: the successful case of the Kayapo people in the Brazilian Amazon.” Regional Environmental Change, 24, 148.
34. Asner, G. P., Broadbent, E. N., Oliveira, P. J. C., Keller, M., Knapp, D. E., and Silva, J. N. M. (2006). “Condition and fate of logged forests in the Brazilian Amazon.” Proceedings of the National Academy of Sciences, 103(34), 12947–12950.
35. Matricardi, E. A. T., Skole, D. L., Costa, O. B., Pedlowski, M. A., Samek, J. H., and Miguel, E. P. (2020). “Long-term forest degradation surpasses deforestation in the Brazilian Amazon.” Science, 369(6509), 1378–1382.
36. Engert, J. E., Campbell, M. J., Cinner, J. E., Ishida, Y., Sloan, S., Supriatna, J., Alamgir, M., Cislowski, J., and Laurance, W. F. (2024). “Ghost roads and the destruction of Asia-Pacific tropical forests.” Nature.
37. Pacheco, P., Cerutti, P. O., Edwards, D. P., Lescuyer, G., Mejía, E., Navarro, G., Obidzinski, K., Pokorny, B., and Sist, P. (2016). “Multiple and Intertwined Impacts of Illegal Forest Activities.” Chapter 6, synthesis of the social, economic, political, and environmental impacts of illegal forest activities.

B. Institutional and Methodological References

1. FAO — Global Forest Resources Assessment 2025
2. FAO — Global Forest Resources Assessment portal
3. FAO — FRA 2025, Terms and Definitions
4. FAO — FRA global data platform
5. Convention on Biological Diversity — Target 2
6. Convention on Biological Diversity — Target 3
7. Convention on Biological Diversity — Global goals for 2050
8. Convention on Biological Diversity — Target 22
9. IPBES–IPCC — Co-sponsored workshop report on biodiversity and climate change
10. IUCN — Global Ecosystem Typology
11. United Nations — United Nations Declaration on the Rights of Indigenous Peoples
12. Society for Ecological Restoration — International Principles and Standards for the Practice of Ecological Restoration
13. Grantham et al. — Forest Landscape Integrity Index, Nature Communications
14. Official Framework of the Universal Declaration of Tree Rights
15. United Nations Decade on Ecosystem Restoration — Standards of Practice to Guide Ecosystem Restoration
16. New York Declaration on Forests (2014; refreshed 2021)
17. United Nations — United Nations Forest Instrument, A/RES/70/199
18. United Nations Forum on Forests — United Nations Strategic Plan for Forests 2017–2030
19. Global Forest Watch / Global Nature Watch — Primary Forests: Definition and Protection
20. World Resources Institute — Fragmentation and Fires Are Threatening the World’s Last Intact Forests
21. Forest Peoples Programme — Rights, Territorial Governance, and Forest Peoples

C. Complementary Doctrinal and Contextual Materials

These materials informed ethical or contextual interpretation. They are not treated as substitutes for peer-reviewed empirical evidence; where a source is a public scientific briefing rather than a peer-reviewed publication, that status is stated explicitly.

1. DellaSala, D. A. (2020). “Has Anthropocentrism Replaced Ecocentrism in Conservation?” In H. Kopnina and H. Washington (eds.), Conservation: Integrating Social and Ecological Justice, 91–103. Springer.
2. DellaSala, D. A., and Moomaw, W. R. (2020). “Scientists Announce Importance of the World’s Primary Forests and Large, Old Trees in Climate Regulation and Biodiversity Conservation.” Geos Institute / Forest Legacies, public scientific briefing (30 March 2020; not a peer-reviewed journal article).

D. External Scientific Review and Correspondence

The following correspondence is cited only for the specific methodological or scientific comments it contains. It is not a peer-reviewed publication and does not imply endorsement of the article, the Universal Declaration of Tree Rights, or the Global Forest Atlas by the correspondents or their affiliated institutions.

1. Tedersoo, L. (2026). “Specialist scientific comments on the mycorrhizal and below-ground dimensions of a global forest classification and the UDTR Global Forest Atlas.” Written scientific correspondence to Ricardo Rey, 21 September 2026. Professor of Mycorrhizal Studies, Institute of Ecology and Earth Sciences, University of Tartu.
2. Zimmerman, B. L. (2026). Full review of the doctrinal article and written observations concerning the role of Indigenous Peoples in forest protection, post-contact transformations, territorial, cultural and economic autonomy, and the experience of the Kayapo Project; transmission of two scientific publications for examination. Scientific correspondence addressed to Ricardo Rey, 25 October 2026. Senior Advisor, Kayapo Project. This correspondence does not constitute endorsement of the doctrine or validation of the Global Forest Atlas; Barbara Zimmerman indicates that she does not propose an additional general revision beyond the contributions of Brendan Mackey and Dominick A. DellaSala, while inviting examination of the publications transmitted to identify potentially relevant elements concerning Indigenous Peoples.
3. Zimmerman, B. L. (2026). Additional scientific observations addressed to Ricardo Rey on 26 October 2026 concerning animal species as indicators of forest integrity in the Amazon, the sensitivity of certain taxa to hunting, the potential value of regional bird indicators, the informative value of high-value timber species, and the limits of detecting logging through GIS and remote sensing; transmission of four additional publications on selective logging, forest degradation, unmapped roads, and illegal forest activities. Written scientific correspondence, Senior Advisor, Kayapo Project. These observations constitute neither endorsement of the doctrine nor validation of the Global Forest Atlas.

Additional Scientific and Institutional References — Agroforestry (Section V bis)

1. FAO — Agroforestry: Overview
2. FAO — Agroforestry: Frequently Asked Questions
3. FAO (2026) — Monitoring global agroforestry extent: An analysis of methodological approaches and data sources — Part A
4. Torquebiau, E., and Balaguer, F. (2022) — For a universal classification of agroforestry
5. Ortolan, E., Maciel, E. A., and Martins, V. F. (2025) — Biodiversity in agroforestry systems implemented in tropical ecoregions: a systematic review
6. Le, T. H., Bonari, G., Sauerwein, M., Plieninger, T., and Zerbe, S. (2025) — Traditional agroforestry systems in Europe revisited: a systematic review
7. Osewold, J., et al. (2022) — Support trees in vanilla agroforests of Madagascar: diversity, composition and origin
8. Osen, K., et al. (2021) — Land-use history determines stand structure and tree diversity in vanilla agroforests of northeastern Madagascar
9. Pappo, E., et al. (2025) — Carbon payment strategies in coffee agroforests shape climate and biodiversity outcomes
10. Boffa, J.-M. / FAO (1999) — Agroforestry parklands in sub-Saharan Africa
11. FAO (2020) — Global Forest Resources Assessment: Terms and Definitions

By Ricardo Rey
Author and Founder of the Universal Declaration of Tree Rights
President of La Compagnie des Papillons Bleus

Prepared and published within the framework of the work of La Compagnie des Papillons Bleus, the official steward organization of the Universal Declaration of Tree Rights.

Official website: www.declarationuniverselledesdroitsdelarbre.org.