Evaluating forest landscape management for ecosystem integrity

Abstract

Protecting forest ecosystems is a critical action for addressing both the climate and biodiversity crises. Effective long-term management of forests requires landscape approaches, but evaluating the management actions is a key challenge. Previous research has suggested evaluation should focus on three interrelated pillars: ecosystem integrity, effective planning, and strong governance. This paper presents a framework for evaluating ecosystem integrity based on the ‘Principle, Criteria, Indicator and Verifier’ (PCIV) method. The key principle used is ecosystem autopoiesis – the ability of a system for self-generation and maintenance by creating its own parts. Four key criteria are applied, accompanied by a set of nine indicators. Verifiers for each indicator are suggested for which feasible data sources are likely available. The use of the three-pillar framework, including ecosystem integrity, is illustrated using three hypothetical cases representing different forest landscape contexts. Such evaluation can provide practical, consistent, repeatable, and comparable information for stakeholders and decision makers.

Keywords:

• Ecosystem integrity
• landscape management
• forest landscapes
• primary forests
• landscape evaluation
• landscape governance
• landscape planning

Introduction

Primary forest landscapes provide a wide range of highly valued ecosystem services (Morgan, Buckwell, et al., 2022; Taye et al., 2021). Protecting primary forests and restoring degraded forests are critical to addressing both the climate and biodiversity crises given their global significance for mitigation, adaptation and conservation (DellaSala et al., 2020; Gibson et al., 2011; Mackey et al., 2020). They provide many other ecosystem services that benefit society, including clean water, genetic resources and ecotourism (Taye et al., 2021). However, forest loss and degradation continues in all the world’s forest biomes (FAO., 2020).

It is widely acknowledged that effective long-term management of forests requires landscape approaches that seek to avoid the resource extraction focus of conventional forest management for commodity production that results in loss and degradation of many ecosystem services (Arts et al., 2017; Morgan, Cadman, & Mackey, 2021; Reed, van Vianen, Barlow, & Sunderland, 2017; Sayer et al., 2013). Such landscape approaches can also facilitate just processes and outcomes for Indigenous and local communities (Fa et al., 2020; Larson et al., 2022; Morgan et al., 2021; Zimmerman et al., 2020). These communities are often the traditional custodian owners of primary forests who have lived in harmony with these forests for millennia, obtaining sustainable livelihoods including cultural benefits that are irreplaceable, while avoiding large scale deforestation and degradation (Zimmerman et al., 2020). A landscape approach recognises the importance of traditional and local communities, and other stakeholders, their multiple values they ascribe to forests, as well as the conflicts and synergies that arise from their use. While a wide range of landscape approaches have been proposed and implemented, a key challenge is evaluating their outcomes as these are often long-term with trends that are complex to untangle or difficult to identify until changes are irreversible (Reed et al., 2017; Sayer et al., 2017). The socio-ecological complexity of landscapes means that insufficient measurements at a point in time risk missing important changes that could limit the supply of ecosystem services. Measurements need to be sufficiently comprehensive to cover all aspects of ecosystem integrity within landscapes and maintain consistency between different contexts and over time.

Ecosystem integrity is a multifaceted and complex concept (De Leo & Levin, 1997; Roche & Campagne, 2017), that provides a scientifically-based framework for evaluating the outcomes of landscape approaches from an ecological perspective (United Nations Department of Economic & Social Affairs, 2021). As used here, the term encompasses the condition of ecosystem structure, composition and function, system level characteristics including stability, resilience and adaptive capacity, and how these are linked to the stocks of ecosystem assets and flows of ecosystem services and benefits to people (De Leo & Levin, 1997; Kay, 1991).

Morgan et al. (2021) synthesised landscape approach principles to suggest three key interrelated pillars for effective forest landscape management: ecosystem integrity; effective planning and strong governance. This three pillars framework is designed to provide the basis of tools for evaluating the outcomes of forest landscape management approaches. Principle, criteria, indicator, verifiers (PCIV) evaluation frameworks for strong governance and effective planning have been developed and published (Cadman, 2012; Morgan, Osborne, et al., 2022). The objective of this paper is to explain the logic and arguments in support of a framework for evaluating forest landscape management in terms of ecosystem integrity. This framework follows the same methodology used for the planning and governance pillars by developing a set of PCIVs based on a synthesis and distillation of key points from existing approaches.

The paper discusses first the concept of ecosystem integrity and briefly summarises existing measures of ecosystem integrity for forests. We then present a framework for evaluating ecosystem integrity comprising a set of principles, criteria, indicators and verifiers based on a synthesis of ecosystem integrity definitions and metrics, and informed by findings from forest landscape projects in developed and developing countries. Finally, we consider how this new framework can be used within the three pillars framework and complement evaluations of strong governance and effective planning.

Primary forest landscapes and landscape approaches

Primary forests – importance, including people

Primary forests are defined as naturally regenerated forests of native species where there are no clearly visible indications of human activities and ecological processes are not significantly disturbed (FAO 2018). These forests are home to significant biodiversity that play a vital role in global cycles of carbon, water and nutrients. As a result, protecting these forests is essential to addressing the biodiversity and climate crises. In addition, they provide other ecosystem services that benefit people at global, regional and local scales, including water supply and filtration, climate regulation and sources of food, fibre and fuel. However, governments and communities are under pressure to allow forests to be cleared or degraded for other land uses for commercial production, including logging, mining, agriculture, and plantations (Bebbington et al., 2018; Curtis, Slay, Harris, Tyukavina, & Hansen, 2018; Kim, Sexton, & Townshend, 2015; Leblois, Damette, & Wolfersberger, 2017; Notess et al., 2018). Over a quarter (27%) of global forest loss is attributed to deforestation from permanent land use change for commodity production; with forest loss in other areas due to forest degradation largely driven by logging (26%) and shifting agriculture (24%), especially in tropical forest areas (Curtis et al., 2018).

Importantly, primary forests, and especially tropical primary forests, are largely situated in developing countries and are often home to Indigenous Peoples and local communities that harness the forests for fuel, food and fibre and have strong cultural attachments to the forest (Jupiter, 2017; Larson et al., 2022; Notess et al., 2018; Schwartzman & Zimmerman, 2005). These communities are facing the consequences from those development pressures that result in forest loss and degradation, including pressure to sell or lease their land and forest resources for commercial use or face illegal exploitation of the forest (Bebbington et al., 2018; Filer, 2012; Nelson et al., 2014; Notess et al., 2018; Zimmerman et al., 2020), which creates tensions between development and forest protection. The resulting forest loss and degradation impacts ecosystem integrity and hence the type and quality of ecosystem services available into the future.

Primary forests and integrated landscape approaches

As noted by Morgan et al. (2021), taking an integrated landscape approach is useful in addressing the increasingly common situation of areas of primary forest comprising remnant patches in otherwise cleared agricultural or degraded land, rather than extensive areas such as ‘Intact Forest Landscapes’ (Potapov et al., 2017). Landscapes with significant areas of remnant primary forest still provide ecosystem services but at a reduced level due to the overall decline in natural ecosystem extent and functioning (Rogers et al., 2022). Similarly, many boundary regions of intact forest landscapes are being eroded due to incursions from competing land uses (Pinheiro et al., 2016). However, despite widespread calls, implementation and results of real-world applications of the landscape approach are limited and uncertain (Arts et al., 2017; Reed et al., 2017; Sayer et al., 2017). In large part, this is due to the socio-ecological complexity and diversity of landscapes, the mix of often conflicting land uses, and the long-term nature of management impacts and outcomes. It is not always clear what should be evaluated.

Primary forests occur in varying degrees of extent within other land uses, and face multiple, interacting drivers of landscape change from a diversity of stakeholders. Landscape approaches to forest management are needed that recognise how people can benefit from the provision of multiple ecosystem services. From this perspective, an approach to evaluating the effect of landscape management on ecosystem integrity is needed that can be applied in different contexts, consider more ecologically nuanced factors than extent, and reflect the influences of human activities.

Having the means to measure the current ecosystem integrity of a forest landscape, project the likely level resulting from management and development plans, monitor change in the level of integrity, and evaluate the outcomes of implementation of the management activities, provides critical information to support effective decision making. Ecosystem integrity evaluations therefore can provide useful information about how forested landscapes can be used and managed sustainably by recognising, valuing and accounting for multiple ecosystem services (Buckwell, Fleming, Smart, Ware, & Mackey, 2020; H. Keith, Vardon, Stein, Stein, & Lindenmayer, 2017; Morgan, Buckwell, et al., 2022; United Nations Department of Economic & Social Affairs, 2021).

The Three Pillars Framework (Morgan et al., 2021) provides a framework for forest landscape management evaluation based on (1) ecosystem integrity, (2) effective planning and (3) strong governance. Evaluation frameworks for governance (Cadman, 2012) and planning (Morgan, Osborne, et al., 2022) have already been developed and are in the process of being tested on-the-ground (Morgan, Zambo, et al., 2022; Shrestha et al., 2022). Here we develop the third pillar of ecosystem integrity to enable a full evaluation of integrated landscape management.

The PCIV approach

The evaluation framework is organised following the ‘principle, criteria, indicator and verifier ‘(PCIV) approach that is commonly used for forest landscape management, planning and governance (CIFOR, 1999; Cadman, 2012, FAO, 2015; Lammerts van Bueren & Blom, 1997; Morgan, Osborne, et al., 2022). Three of those terms (principles, criteria, indicators) are commonly used in different settings and combinations and have therefore acquired various definitions. For example, the Global Biodiversity Framework uses the terminology of goals, targets and indicators where indicators are the measured factors (Biosafety Unit & CBD, 2023). Here we have adopted the definitions of principles, criteria and indicators commonly used for creating standards (see for example, Business and Biodiversity Offsets Program (BBOP) (2012) and Lopez-Casero, Cadman, and Maraseni (2016)) and as explained in detail by Cadman (2012). While principles are perhaps more commonly used as normative statements, we use them here to describe desired outcomes that represent higher order system states amenable to objective measures. We then understand criteria as parameters operating at a level below principles that demonstrate compliance with specific aspects of the biophysical components and properties defined by the principles. Criteria therefore are categories of conditions of entities or processes that need to be met in order for a principle to be achieved and are designed to facilitate assessment of principles. We define indicators as variables that describe the state of the system and degree of compliance with the related criteria. They are therefore measurable states that allow the assessment of whether or not a particular criterion has been met which can indicate a negative or positive trend or difference from a baseline condition or be absolute measures. For each indicator, there is one or more verifier which are attributes that can be measured qualitatively or quantitatively, usually using a single dataset, though multiple verifiers might be used for a given indicator.

Ecosystem integrity

Ecosystem integrity is a long-used concept in ecological science and natural resource management (Karr, 1993), and is referenced in international agreements on climate change and biodiversity, including the Paris Agreement (United Nations, 2015) and the Global Framework for Biodiversity (Convention on Biological Diversity, 2022). Various definitions and operational frameworks have been proposed and tested and a range of related terms are in use (Roche & Campagne, 2017). The term as used here encompasses the concept of ecological integrity which in turn incorporates direct measures of biological condition sensu Karr et al. (Karr, 1993; Karr, Larson, & Chu, 2022).

Ecological integrity has been defined as a measure of the composition, structure, and function of an ecosystem in relation to the system’s natural range of variation. It integrates different characteristics of an ecosystem that collectively describe its ability to achieve and maintain its optimum operating state, given the prevailing environmental drivers and perturbations, and continue its self-organisation and regeneration. This high order system state has been referred to as ‘autopoiesis’ (Kay, 1991).

Building upon the Kay definition, Rogers et al. (2022) argued that key foundational elements of ecosystem integrity for forests include: (1) dissipative structures – especially forest structural complexity because of its importance for stored emergy as manifested in biomass and carbon stocks; (2) ecosystem processes – particularly those related to productivity (e.g. nutrient cycling) and regenerative capacity; and (3) stability sensu Grimm and Wissel (1997) which encompasses ecosystem resistance/constancy, resilience and persistence. A fourth key element is the natural adaptive capacity potential of an ecosystem in the face of environmental change. These foundational elements of ecosystem integrity – dissipative structures, ecosystem processes, stability and adaptive capacity – are all derivatives of the underlying biodiversity of a forest ecosystem. Rogers and colleagues argued that these functional roles of biodiversity at all levels (genetic, species and community) exist because natural selection yields the characteristic biodiversity and phenotypic plasticity best suited to prevailing environmental conditions, including fluctuating resource inputs, extreme events, periods of stress, and natural disturbances. Many interdependencies exist among composition, structure, and function. However, the variables that can be measured, their spatial and temporal scale, and the characteristics that they describe are limited by available data and resources. We also currently lack the knowledge to fully account for the role of biodiversity in ecosystem function and thus how to link biodiversity changes to ecosystem services (Weiskopf et al., 2022).

Kay further argued that to be useful, the concept of ecosystem integrity must enable consideration of how changes in the ecosystem are valued by people and considered desirable or not. From an anthropocentric perspective, forest landscapes provide a range of ecosystem services that can be harnessed through appropriate forest management including conservation management and customary practices which provide benefits to local communities, various stakeholders and society (de Groot, Wilson, & Boumans, 2002; Vallés-Planells, Galiana, & Van Eetvelde, 2014).

The quality and quantity of these services varies with the type, extent and ecological condition of the ecosystem (Rogers et al., 2022) and standardised systems are available to account for them and the benefits they provide (Czúcz et al., 2021; United Nations Department of Economic & Social Affairs, 2021). The terminology and structure of SEEA EA is somewhat different to the PCIV approach but has similar outputs and its application demonstrates how these frameworks can be used to inform decision-making. For example, Keith et al. (2017) produced spatial accounts of the ecosystem services generated from the tall, wet forest of the Central Highlands of Victoria, Australia, using the U,N. System of Environmental Economic Accounting - Ecosystem Accounts (SEEA-EA). The accounts revealed how impact of land use which reduces forest age results in decreases in the ecosystem condition of the forest for water yield and carbon storage, as well as biodiversity and recreational services.

Land use and related activities, such as infrastructure development and industrial-scale extractive industries, can impact on a forest landscapes ecosystem integrity, degrading the quantity and quality of the ecosystem services and the benefits they provide people. A change in the mix of ecosystem types, with different levels of ecosystem integrity, will also alter the type of ecosystem services available. Planning determines the choice of land uses and activities, based both on how ecosystem services are valued by stakeholders, but also by the range of available ecosystem services, as determined by ecosystem integrity. Governance is important as it determines how, when and where land uses and management activities are carried out, and how benefits are shared.

A PCIV framework for assessing ecosystem integrity

Framework structure

A range of sets of indicators for evaluating forest ecosystem integrity have been proposed (Table 1). Most indicators focus on elements of ecosystem structure, function and composition. Using the Tierney, Faber-Langendoen, Mitchell, Shriver, and Gibbs (2009) definition, an assessment for a specific ecosystem’s integrity can be based on long-term monitoring data that populate metrics of status and trend in structure, composition, and function of forests impacted by multiple agents of change. Some approaches recognise the different scales that need to be considered for examining landscape structure and landscape connectivity (Reza & Abdullah, 2011), while others consider land-use or human impact indicators, reflecting an emphasis on ecosystem integrity as a measure of naturalness (Theobald, 2013). The approach of Becker (2005) highlights resilience, collaboration and auto-sufficiency.

Table 1. Examples of proposed indicators of ecosystem integrity.

Criteria/indicators	Reference
Self-organisation, naturally self-regenerating, stability/resilience, adaptive capacity, framework species and characteristic biodiversity	(Rogers et al., 2022)
Land-use intensity, structural diversity, rareness, endangerment, species, structures	(Müller, Hoffmann-Kroll, & Wiggering, 2000)
Resilience, collaboration and auto-sufficiency	(Becker, 2005)
Structure, landscape structure, composition, function	(Tierney et al., 2009)
Fragmentation, representativeness, ecosystem sensitivity, landscape connectivity	(Reza & Abdullah, 2011)
Use intensity, structural diversity, rarity/threat	(Walz, 2011)
Surface temperature, stand structure, plant life history traits	(Norris et al., 2012)
Level of human modification	(Theobald, 2013)
Exergy capture, entropy production, storage capacity, cycling & nutrient loss reduction, biotic water flows, metabolic efficiency, heterogeneity, biotic diversity	(Kandziora et al., 2013)
Micro-climate, patch heterogeneity, biodiversity, territorialisation, seston (for riparian forested areas)	(Michez et al., 2013)
Composition, structure, function	(Wurtzebach & Schultz, 2016)
Ecosystem structure, composition, function, species composition, diversity, and functional organisation	(Brown & Williams, 2016)
Physical and chemical state characteristics, biotic composition, structure and function, landscape level characteristics	(United Nations Department of Economic & Social Affairs, 2021)
Rates of decline in ecosystem distribution; restricted distributions with continuing declines or threats; rates of environmental (abiotic) degradation; rates of disruption to biotic processes; and risk of ecosystem collapse	(D. A. Keith et al., 2013)

Our focus here is on providing guidance that facilitates consideration of ecosystem integrity in the context of forest landscape planning and management in ways that are compatible with the existing frameworks for assessing effective planning and strong governance. This requires prioritising verifiers for which data are more likely to be available. It also necessitates translating complex and strictly defined ecological terms into generalised language more familiar to real world decision-makers.

The framework for evaluating a forest landscape’s ecosystem integrity is based on a reference condition of the primary forest ecosystem which represents the highest level of ecosystem integrity, following the System of Environmental Economic Accounting Ecosystem Accounts (SEEA-EA; (Czúcz et al., 2021; H. Keith et al., 2020) and the IUCN Ecosystem Red List assessment criteria (D. A. Keith et al., 2013) and (Rogers et al., 2022) (Figure 1).

The evaluation framework by necessity must be hierarchical and flexible in its application given that data gaps will be the norm in most forest landscapes, especially in developing countries. It follows that the framework must allow for assessments that are qualitative and based on the best available information including expert opinion and local knowledge, where quantitative data are limited or absent. Note that we provide example verifiers here as an aid to explaining, and demonstrating the benefits of, the framework. Other verifiers may be applicable to the different indicators, given different contexts or data availability. Alternative verifiers will need to align with the indicators, however. The verifiers presented here provide examples to aid understanding and are likely available in, and applicable to, a wide-range of contexts.

Explanation of components

Ecosystem autopoiesis

We posit ecosystem autopoiesis as the key principle for evaluation of ecosystem integrity. As noted above, autopoiesis is defined as the ability of a system for self-generation and maintenance by creating its own parts (Maturana & Varela, 1991). In a biological context, the term describes the capacity of living entities to construct their own metabolism, and use it to maintain themselves, grow and reproduce (Nurse, D., 2020). The criteria listed under this principle identify key components of an ecosystem that contribute to its autopoiesis, and which are amenable to measurement, namely: ecosystem dissipative structures; ecosystem processes; ecological community interactions; and landscape characteristics.

Structural exergy

The concept of structural exergy warrants some background discussion. It is founded in ecological thermodynamic system theory as proposed by Jørgensen (1992) Ecosystems are a special category of complex dissipative systems as their autopoietic self-generation, organisation and maintenance is due to a host of natural biological, ecological and evolutionary processes. Consequently, they construct their own metabolism, the products of which are used to maintain themselves, grow and reproduce, including components that store and transmit information across generations (i.e. the genomes of the constituent species populations). Ecosystems are considered to be ‘dissipative systems’ (Nicolis, 1986). because they lose heat energy to their external environment, but also have a capacity for taking in energy, conserving and storing it to support metabolic processes along with the production of biomass and the growth and maintenance of the supporting structures and functioning parts.

As ecosystem mature, energy and matter is accumulated in the growing (and expanding) living and dead ecosystem biomass (Kandziora, Burkhard, & Müller, 2013). The solar energy that is captured by an ecosystem through photosynthesis and stored as chemical energy in the bonds of the carbohydrates of biomass and that constitutes the energy available for metabolic work, is called exergy (Jørgensen, 1992). The main store of exergy in a forest is the woody biomass of trees and it is big old trees that is the largest stock of stored energy and carbon (Stephenson et al., 2014).

While this structural exergy has been suggested as an appropriate measure of ecological integrity (Jørgensen, 1992), it is difficult to calculate. From a practical perspective, a more feasible indicator is based on measurements of forest biomass including stand and landscape level estimates of: living and dead biomass; age class distribution of trees; complexity of vertical vegetation structure; coarse woody debris.

The vegetation structure verifiers require measurements from both field-based observations, typically from a network of representative sample plots, plus spatial analyses using remotely sensed data from sensors mounted on satellites and/or drones. Field-based data can be taken by appropriately trained citizen scientists including community ranges (e.g. (NAILSMA, 2023) but requires expertise in planning, quality control and data analysis, and remotely sensed data with a high spatial and temporal resolution are now freely available globally and being regularly applied to map changes in forest structure (DellaSala et al., 2022; Shestakova et al., 2022).

Ecosystem processes

Here we highlight two critical ecosystem processes that are fundamental for ecosystem autopoiesis. Nutrient cycling and nutrient loss reduction ensures that forests retain a sufficient stock of essential macro- and micro-nutrients to support plant metabolism, growth and reproduction, and in turn the entire food chain. This process involves complex suites of interactions and feedbacks between plants, fungi, invertebrates and micro-organisations (Moreau, Bardgett, Finlay, Jones, & Philippot, 2019). Moreover, in forests ecosystems, the spatial and temporal measurement of soil nutrients and carbon pools and flows are complex and expensive to measure (Cao, Domke, Russell, & Walters, 2019). Regional studies provide research data describing nutrient and carbon cycles at temporal and spatial scales (for example, (Fahey et al., 2005; H. Keith et al., 2009; R. H. Whittaker, Likens, Bormann, Easton, & Siccama, 1979), but data at continental scales are limited, with a notable exception of the conterminous USA (Cao et al., 2019). In practice, there are few forest landscapes where adequate data are available and in a form suitable to populate verifiers for ongoing monitoring that would support forest management decision-making.

The second process noted here is referred to as hydro-ecology which refers to the functional roles that the vegetation plays in regulating surface and subsurface hydrological flows, as well as atmospheric moisture, at local and regional scales, and the importance of water availability to ecosystems and animal habitat (Soulé et al., 2004). As with soils, this process is critically important for ecosystem integrity, yet reliable data requires complex and expensive monitoring at multiple spatial scales from stand to watershed. Therefore, there are few forested catchments with the necessary in situ instrumentation required.

It follows that for this criterion the verifiers will usually require expert judgement and local knowledge to provide qualitative and relative assessments ratings. However, citizen science is developing and has been shown to be capable of producing reliable observational data on water quality in data poor catchments (Quinlivan, Chapman, & Sullivan, 2020).

Ecological composition

Characteristic native species are a subset of all native species found in a forest ecosystem that distinguishes an ecosystem from other types; there are often many other species that are common with other ecosystems (D. A. Keith et al., 2013). We identify two kinds of characteristic species here. Framework species play a dominant function role and the most obvious and important in a forest are the canopy tree species. Focal species are those of interest in a given context because they are threatened, charismatic (and therefore popular) or have cultural significance such customary totem species (Raven, Robinson, & Hunter, 2021). Interactive species play major role in ecosystem process, such as the creation of structures such as cavities, burrows, and dams, and interactions such as predation, pollination, decomposition, competition and bioturbation (Soulé et al., 2004). These species also need to be present at a population size that enables them to be functionally effective (Soulé, Estes, Berger, & Del Rio, 2003).

The identification of the target characteristic, focal and interactive species is highly context-related and requires deep expert and local knowledge of a forest’s plant and animal biota. While survey and monitoring of wildlife is increasing in both high-, middle- and low-income countries, including a substantive proportion by citizen science, monitoring in most taxonomic groups remains sparse and uncoordinated, and most of the data generated are elusive and unlikely to feed into wider biodiversity conservation processes (Moussy et al., 2022; Oliveira et al., 2017). Integration of data sources is progressing, for example including from citizen science and local knowledge in the Atlas of Living Australia, and Traditional Knowledge as a source of information on wildlife presence and abundance (Ziembicki, Woinarski, & Mackey, 2013). Open access to global biodiversity repositories and analytical platforms is improving, for example the Botanical Information and Ecology Network (BIEN, 2023).

Species introduced to new regions through human activities are termed alien species. Invasive alien species represent animals, plants, and other organisms – known to have established and spread with negative impacts on biodiversity, local ecosystems and species. Invasive alien species are recognised as one of the five major direct drivers of change in nature globally, alongside land- and sea-use change, direct exploitation of organisms, climate change, and pollution (IPBES, 2023). Frameworks have been developed to facilitate the systematic collating and reporting on invasive species at the landscape level (Shackleton et al., 2020) along with global and national invasive data repositories are under active development (CABI, 2023; Ziller et al., 2020).

Ecosystem emergent and system properties

A critical property of ecosystem integrity that can be considered an emergent property is ecosystem stability. Following Grimm and Wissel (1997) we recognise three types of ecosystem stability that are relevant in different contexts: (1) resistance (or constancy) – the ability of an ecosystem to resist an external stressor, i.e. not be significantly impacted by it; (2) resilience – an ecosystem’s ability to recover from an external stressor back to something like its pre-disturbance state; and (3) persistence over time, which collectively represent an ecosystem’s ability to resist or be resilient to change at both short and long time scales. Resilience is now a commonly used term in many contexts including disaster risk reduction (Tiernan et al., 2019) and climate change adaptation (Choko et al., 2019). However, here we are using the term in an ecological context (Rogers et al., 2022).

Tropical wet primary forests, for example, have closed canopies and are highly resistant to wildfire. However, human land use can reduce forest cover which induces canopy desiccation and increases wildfire risk (Briant, Gond, & Laurance, 2010). Australian native forests with canopy species dominated by Eucalyptus species are highly resilient to wildfire because of many adaptive traits that enable them to survive and quickly recover from wildfires (Pausas & Keeley, 2014). Wildfire is also a common natural disturbance in boreal forests which can result in mosaic of forest patches some of which are in early seral stages of re-growth (DellaSala, 2011). From this system’s perspective, forest landscapes can follow a number of different trajectories in response to human and natural disturbances. The impacts could be so severe as to trigger ecosystem collapse and/or conversion to another ecosystem type (D. A. Keith et al., 2015; Lindenmayer, Hobbs, Likens, Krebs, & Banks, 2011). Alternatively, the impacts could be ongoing, leading to changes in the condition of the ecosystem such as the vegetation structure that occurs under a logging regime that maintains the canopy trees in a regrowth phase and promotes those species of commercial interest (Pérez, Carmona, Fariña, & Armesto, 2009).

Ecosystems can be understood to have a natural adaptive capacity which is derived from their component biodiversity, which includes: the available pool of species; the genetic diversity among the populations that constitute these species; the phenotypic expression of that genetic diversity (i.e. the observable characteristics or traits of an organism); the functional roles in ecosystem processes played by species; and the many and complex interactions between plant, animal, fungal and microorganism species (including food webs and the role of animals in plant reproductive cycles). All these dimensions of biodiversity provide potential options for responding to changing environmental conditions in ways that enables an ecosystem’s composition, structure and processes to persist and avoid ecosystem collapse (Lindenmayer et al., 2011; Tebbett, Morais, Goatley, & Bellwood, 2021).

An ecosystem’s adaptive capacity is greater when the diversity of native species is higher than lower, all other factors being equal; noting that species diversity in terms of the number of species in a given area (i.e. richness), can be considered at local, landscape and macro-scales (R. J. Whittaker, Willis, & Field, 2001). A greater diversity of species affords more options for adaptation to new conditions that will maintain provision of ecological functions such as seed dispersal and pollination. The genetic diversity within and among the populations of species also provides the potential for new phenotypic traits that may allow adaptation in the face of environmental change. There is also a link here with the criterion of the presence of focal species and their effective population size as these will possess more genetic variation, thereby facilitating trait evolution (Kelly, 2019). Adaptive capacity is also enhanced when populations of focal native species are maintained across their natural ranges as this helps ensure genetic diversity and related adaptive responses such as phenotypic plasticity which is the ability of organisms to produce distinct phenotypes (i.e. observable characteristics like body size) in response to environmental variation (Gardner et al., 2014). Approaches have been developed to assess this verifier such as the forest stability index (FSI) which can be used to measure decline in tree species populations across their range (Stanke, Finley, Domke, Weed, & MacFarlane, 2021).

Landscape characteristics

Deforestation has catastrophic impacts on ecosystem integrity and the severity of the impact scale with the proportion of a forest ecosystem that has been cleared relative to its natural distribution (Rogers et al., 2022; Seymour & Harris, 2019). Applying this verifier requires information on the current spatial extent of a given forest ecosystem and the fraction this constitutes of its pre-disturbed extent. The second verifier here is the proportion of primary forest to secondary growth forest. In this context primary forests are defined as forests that are largely undisturbed by industrial-scale land uses (such as logging, mining, human-caused fires, dam, and road construction) and are the result of ecological and evolutionary processes including the full range of successional stages (Kormos et al. 2018). Land use impacts that convert primary forest to secondary regrowth forest is a type of forest degradation, as are the impacts from logging, mining and infrastructure development on ecosystem structure, composition and functioning and associated ecosystem services including carbon storage and clean water supply (Qin et al., 2021).

A major category of degradation is forest fragmentation from roading which breaks up large intact forest landscapes into smaller fragments, increasing edge effects, reducing the area of interior forest micro-climates and increasing forest desiccation and wildfire risk (Briant et al., 2010; Ibisch et al., 2016). The spatial configuration of forest in heavily disturbed landscapes is therefore also an important consideration and a range of patch statistics are now routinely applied for such assessments (Hesselbarth, Nowosad, Signer, & Graham, 2021). The data needed for monitoring changes in the landscape characteristics of spatial extent and spatial configuration ecosystem can now be readily obtained from the satellite-based sensor data used to map forest vegetation structure noted above (Fischer et al., 2021; Galiatsatos et al., 2020).

Another important landscape configuration verifier is ecological connectivity which refers to the natural ecological flows between ecosystems at landscape and macro-scales. These flows include the movement of plants (through seed dispersal) and animal migrations including so-called meta-population dynamics. This describes movement of animals in partially isolated local populations that are subject to extinction, recolonisation, fluctuations in size over time and limited gene flow (Mitchell-Olds, Willis, & Goldstein, 2007). Maintaining ecological connectivity at landscape and macro-scales is therefore an important conservation priority (Mackey et al., 2023). GIS based tools are available for mapping patch connections and when they are paramaterised for a specific species habitat requirements they provide information to help identify wildlife corridors for ecological connectivity (Norman & Mackey, 2023).

Applying ecosystem integrity into a landscape evaluation

The ecosystem integrity PCIV developed here is the third pillar of a framework for evaluating forest landscape management approaches, alongside the pillars of effective planning and strong governance (Morgan et al., 2021). Ecosystem integrity provides the basis for assessing the key biophysical properties of the forest landscape. A landscape level ecosystem integrity evaluation could highlight areas of high ecosystem integrity that need to be protected, such as areas of primary forest, or areas where there is scope to improve ecosystem integrity, such as by connecting remnant areas of primary forest or areas of rapidly recovering secondary forest. It could also highlight key threats to ecosystem integrity, such as from invasive species or increasing anthropogenic forest degradation. This understanding about the level of a forest landscape’s ecosystem integrity can underpin decisions about the use of the landscape and the impact on ecosystem services, including which ones are maintained, optimised or lost (Taye et al., 2021). In this way, ecosystem integrity provides the ‘why’ of landscape management. We use the concept of ‘ecosystem services’ to provide a cross-cutting bridge between the three framework pillars. The quality and quantity of ecosystem services scales with the level of ecosystem integrity as this underpins ecosystem composition, structure and function.

The integrity of the planning and governance (the ‘what’ and ‘how’ of landscape management) are indicative of whether these ecosystems services provide benefits and to which people, and if they are maintained over time, improved or degraded (Kandziora et al., 2013; Morgan, Buckwell, et al., 2022). Planning evaluation assesses ‘what’ decisions are being made about the landscape. Effective planning can be defined as a high integrity process of choosing future land uses and activities. Creating integrity in planning requires shared learning to bring stakeholders together, integration as holistically as possible to ensure the landscape is managed as a whole, and processes to ensure the planning process and outcomes are considered through a justice lens. Note that this definition includes informal and community planning as well as Indigenous decision-making about land uses and activities (Morgan, Osborne, et al., 2022). Evaluating the integrity of the planning will indicate whether future decisions about land and resources in the landscape are likely to be effective. Note that this may be formal planning or informal planning being undertaken by communities (Healey & Hillier, 2010; Thorpe, 2017), which is more common in many primary forest landscape contexts. A planning evaluation focuses on assessing shared learning about the landscape, integration around activities, and whether benefits of the landscape are being equitably shared to provide a situated justice (Morgan, Osborne, et al., 2022). The planning evaluation can, if appropriate, provide the basis for more formal planning, leading to the development of a landscape plan.

Evaluation of governance integrity assesses ‘how’ decisions are being made. Strong governance can be defined as decision-making with a high degree of legitimacy and integrity (Cadman, 2012). Legitimacy and integrity in governance requires participation of all stakeholders and a robust process for deliberation. The governance evaluation is based on assessing participation to be meaningful, and deliberation to be productive (Cadman, 2012). Importantly, the governance evaluation is undertaken by the stakeholders, including local communities. This participatory governance assessment helps ensure decision-making processes, including planning processes, are legitimate. Ensuring legitimacy of governance strengthens the likelihood of actions being taken, as well as providing confidence that decisions are agreed by the stakeholders. The evaluation can also result in the development of a governance standard, produced by the stakeholders (Lopez-Casero et al., 2016).

Note that although the terminology of the evaluation implies a technocratic, and therefore ‘top down’ approach that might ignore communities, in fact the combined framework provides a bottom-up evaluation. The PCIV approach specifically allows for the stakeholders to assess the indicators (see e.g. Morgan (2023)) and even choose the verifiers (Lopez-Casero et al., 2016). This participatory approach essentially translates what communities are doing into the technocratic language. Although local communities may not be thinking or talking about governance and planning, they have governance and are doing planning. This is largely unrecognised, resulting in top down approaches in order to fit people to national or international conventions or standards. Similarly, although central (government-led) planning and governance may be poor, local governance and more informal/local ‘planning’ (often not called planning (see Morgan, Buckwell, et al., 2022) may actually be quite strong in some places. This approach enables assessment of what communities are doing and that links back to the technocratic language of theory and Western practice.

The combined evaluation from considering the three pillars provides structured and consistent information about a forest landscape’s integrity that can be used to identify trends, understand the impacts and outcomes of different approaches, values, prioritisations and socio-economic contexts. An example using hypothetical evaluations of three theoretical forests landscapes is described in Table 2 and Figure 2 to illustrate how evaluating all three pillars provides insights into pressures, weaknesses and strengths. Landscape 1 represents an intact forest landscape (i.e. an extensive area of largely primary forests sensu (Potapov et al., 2017)) in an economically developing country context, with recognised customary land rights and strong governance. However, resources and capacity for formal planning may be limited and increasing population and industrial economic pressures present a growing risk to maintain high ecosystem integrity levels. Landscape 2 represents an economically developed country context where there is highly formalised and centralised planning and governance in place, but the priority given to urban development and industrial extractive land uses results in low ecosystem integrity due to ongoing impacts. Here the solution space requires a focus on protecting remnant primary forest areas and ecological restoration, with a focus on increasing public awareness of ecological impacts and then improving participation of a broad range of stakeholders and societal values. Landscape 3 represents another high forest cover, developing country context but where there is a mix of extensive primary forest areas and industrial logging. There are strong customary land rights in place but spread among a large number of traditional owners who are under pervasive outside influences pressuring them to open up the forest to extractive industries, and there is a lack of central governance support to prevent illegal logging and mining. Here the solution space directs attention to strengthening existing governance and seeking forest protection opportunities through building planning capacity.

Figure 1. Comparative evaluation framework for forest landscape ecosystem integrity.

Figure 2. Hypothetical evaluation of different theoretical landscapes, as described in Table 2. Landscape 1 represents an intact forest landscape (i.e. extensive area >50,000 ha of primary forest) set in a developing tropical country where the forest retains a high level of ecosystem integrity and is managed under strong local, participatory governance and planning regimes underpinned by legalised customary land rights. Landscape 2 is set in a temperate forest in a developed country subject to a long history of legal industrial logging, and while this is subject to significant formal planning and governance, these are not participatory and the major stakeholder’s focus is on resource extraction. Landscape 3 represents a developing country with strong customary land rights but weaker centralised governance and more illegal logging.

Table 2. Illustrative assessment of three hypothetical forest landscape contexts using the three pillar framework.

Landscape	Context	Ecosystem Integrity	Effective Planning	Strong Governance	Priorities for solutions to address integrity shortfalls
1	Economically developing country Dominantly intact forest landscape (i.e. extensive areas of primary forest Traditional village-based communities Legally recognised customary land rights and well-organised local community organisations Growing populations, changing demographics, strong influence from outside Western communities, growing climate change impacts	High	Low	Medium-high	Developing planning to respond to future changes and identify key priorities for forest protection Formalise customary governance through application of a governance standard Recognise the benefits from primary ecosystem services Establish sustainable income generation through payments for ecosystems services
2	Economically developed country Some areas of remnant primary forest, but historical and ongoing logging Highly centralised government-led decision-making, with land use policies influenced by extractive industry vested interests Formal planning processes in place Illegal land use activities largely prevented	Low	Medium-high	Medium-low	Improving wider participation in governance and planning Advocating for planning to include ecosystem integrity considerations recognising values and benefits associated with primary forest protection and restoration
3	Economically developing country Extensive areas of primary forest plus large areas of heavy and ongoing logging Customary land rights and decision-making legally recognises Development pressure from external extractive economic interests Limited capacity of centralised government to prevent illegal logging and mining	Medium	Low	Medium	Building capacity of customary owners for participatory planning Strengthen governance to prevent illegal land uses Recognise the benefits from primary ecosystem services Establish sustainable income generation through payments for ecosystems services

Importantly, the information from evaluation using any one of the three pillars helps improve the other two (Figure 3). For example, data gathered from the ecosystem integrity assessment can inform the planning process. This will improve the knowledge about the condition of the forest landscape and the flows of ecosystem services, as well as guide analysis of potential impacts. Ensuring legitimate governance will result in improved participatory planning processes, as well as supporting the inclusion of local knowledge in the ecosystem integrity evaluation. Evaluations therefore should extend beyond being a ‘tick-box’ exercise, instead producing information and processes that support ongoing improvement of forest landscape management.

Figure 3. Examples of flows of information between the three pillars can be used to improve forest landscape management.

The three pillar assessment framework therefore can be used to improve forest landscape management over time (Figure 3) which in turn provides a more robust and reliable basis for the creation of sustainable income generation activities, such as payments for ecosystem services (PES) and eco-certification schemes (Morgan, Buckwell, et al., 2022). Fundamental to the successful operation of these schemes is confidence that it is genuine. Paying a premium for eco-tourism or eco-certified non-timber forest products requires the consumer to have confidence that the forest ecosystem values are being protected now and into the future. Similarly, there is a growing need for PES-type schemes to have transparent integrity, as well as comparability across different landscapes and contexts, as highlighted by recent concerns over the integrity of carbon credits, which emphasises the need for improved information (Blaufelder, Levy, Mannion, & Pinner, 2021; Greenfield, 2023; West, Börner, Sills, & Kontoleon, 2020).

Discussion

The ecosystem integrity evaluation framework developed here provides clear and transparent information about the ecological condition of ecosystems and the quantity and quality of the services they provide. A key advantage of the PCIV approach is its flexibility that allows for different context-specific verifiers to be chosen depending on the type of forest ecosystem (e.g. tropical, temperate, boreal), the landscape context (e.g. intact forest landscape, remnant primary forest, border region), socio-economic-cultural factors, and data availability. The lack of data for verifiers is probably the single biggest barrier to implementation of assessment frameworks. This lack of data can translate into selective use of data without adequate criteria for their selection and interpretation. Monitoring of forest ecosystem integrity can be expensive and time-consuming depending on the verifiers chosen and there are only a few landscapes anywhere in the world that have sufficient data currently available. However, from a management perspective, progress is being made to provide cost-effective data through a combination of remotely sensed sources, available national level data bases, synthesis of research studies, citizen science and local knowledge. While the actual verifiers and their data sources will vary between forest landscapes in different contexts, comparisons of their indication of relative ecosystem integrity can still be made so long as a shared set of indicators and criteria are used.

As defined here, ecosystem integrity encompasses assessment frameworks and metrics for ‘ecological integrity’ and ‘biological integrity’ which incorporate direct measures of biological value, particularly losses due to human impacts, to diagnose human and non-human causes of ecological degradation, and propose preventive or restorative measures (Karr et al., 2022). These frameworks and metrics have also been used to examine the interrelations between ecosystem integrity, ecosystem services and human well-being criteria (Kandziora et al., 2013). As the quantity, quality and range of ecosystem services varies with the level of ecosystem integrity, they enable an assessment of ecosystem condition (Roche & Campagne, 2017) which can be used to value the economic contributions of the benefits to society from ecosystem services. For example, H. Keith et al. (2020) in a case study of the tall, wet forests of the Central Highlands of Victoria, Australia, used by the U.N. System of Environmental Economic Accounting – Ecosystem Account (UNSEEA) to compare how different management regimes impacted the forest’s ecosystem integrity and ecosystem services. The level of ecosystem integrity is also indicative of a forest landscape’s risk profiles including their exposure and vulnerability to wildfire, drought, biodiversity loss, depletion of carbon stocks, and various climate change impacts (Rogers et al., 2022). Another potential application of the framework is to use the verifier values in calculating a multi-criteria distance function which generate a single normalised index of the aggregate sustainability with which an ecosystem is being managed (Diaz-Balteiro et al., 2018). Further research is therefore warranted into the different ways in which verifier values can be applied to support decision making, including examining the correlation between the indicators and the extent to which an increase in one can compensate for a decrease in another.

All the above applications of assessments of ecosystem integrity provide relevant information for stakeholders and decision makers concerned with forest landscapes management. While the benefits for forest landscape management from improving planning and governance are recognised, to date in practice these are largely or at least not explicitly coupled to the ecosystem integrity of those landscapes. The result therefore can be landscapes that are effectively planned with strong governance, including through formal regulatory mechanisms, but with forests that nonetheless end up with low ecosystem integrity leading to a significant reduction in ecosystem service benefits and increased risk profiles, including the potential for ecosystem collapse (DellaSala et al., 2021; Wilson et al. 2022).

Landscape planning and management often entail conflicting land use values and priorities, particularly when biodiversity and heritage conservation assets, freshwater supply or Indigenous customary livelihoods, which depend on ecosystem services from forests with a high level of ecosystem integrity, are at risk from the impacts of extractive commercial industries such as logging and mining which degrade the forest’s ecosystem integrity (Schwartzman & Zimmerman, 2005; Taye et al., 2021). Applying our ecosystem integrity framework in such conflicted situations would provide information that can assist stakeholders and decision makers identify the risks from different land use options and the gains from identifying appropriate mixes of complementary land uses (Rogers et al., 2022).

While some of the theoretical underpinnings of ecosystem integrity will be obscure for experts from other disciplines, practitioners and stakeholders to readily grasp, their value rests on providing a solid scientific foundation. The concept of autopoiesis – when explained as the capacity of a system for self-organisation – is relatable and provides a transparent justification for use of the term ‘integrity’ which is necessary given that it is more typically considered to reflect a normative social value rather than, as used here, a system capacity that is amenable to objective measurement.

The significance of the term ‘structural exergy’ is also obtuse, but the indicator of vegetation structure is relatable to a common sense understanding of a forest’s ecological condition. The age, size and complexity of the vegetation structure plays a key role in buffering ecosystems from external stressors through regulating interior micro-climatic conditions and maintaining a flow of energy, water and nutrients, and the necessary habitat resources needed for wildlife shelter, food and reproduction (Norris, Hobson, & Ibisch, 2012). Human land uses and impacts – particularly industrial scale logging, mining, roading, and capital works – can destroy or degrade this buffering capacity, deplete the stocks and flows of energy, water and nutrients, as well as habitat resources. These impacts all reduce the level of ecosystem integrity and flow and quality of ecosystem services.

Conclusion

We present a Principle, Criteria, Indicator and Verifier (PCIV) framework for evaluating the ecosystem integrity of forest landscapes. The framework provides the basis for evaluating from an ecological science perspective the outcomes of landscape management as part of a three-pillar framework for evaluating landscape management inclusive of governance and planning considerations.

The ecosystem integrity framework comprises the core principle of ecosystem autopoiesis, five key criteria (structural exergy, ecosystem processes, ecological composition, ecosystem emergent and system properties, landscape characteristics) and a set of nine indicators. Verifiers for each indicator are proposed for which feasible data sources are likely available though at differing levels for economically developed and developing countries. This framework is based on translating well-established scientific ecological concepts into a hierarchical, logical framework for evaluation purposes.

Collectively, the three pillars of ecosystem integrity, effective planning and strong governance provide for a comprehensive assessment of forest landscape management, which can provide consistent and repeatable information for stakeholders and decision makers regarding the impacts and consequences of current or proposed actions. Further research is needed to implement, test and refine the ecosystem integrity framework presented here, as well as its application as part of a comprehensive assessment process, using case studies across different forest landscapes contexts.

Acknowledgements

This research was funded by a charitable organization which neither seeks nor permits publicity for its efforts. The trust has had no influence on the design, analysis, interpretation and documentation of this research.

Disclosure statement

The authors declare no conflicts of interest.

Additional information

Notes on contributors

Brendan Mackey

Prof. Brendan Mackey is Director of the Griffith Climate Action Beacon at Griffith University, Queensland. He has a PhD in plant ecology from The Australian National University. He has over 300 academic publications in the fields of climate change research and biodiversity as well as related topics in environmental science and policy. Brendan was Coordinating Lead Author for the 2022 Intergovernmental Panel on Climate Change IPCC 6th Assessment Report, Working Group II - impacts, vulnerability & adaptation.

Edward Morgan

Dr Edward A. Morgan is a transdisciplinary research fellow at the Policy Innovation Hub and Climate Action Beacon, Griffith University. His research focuses on developing, implementing and evaluating policy, planning and governance for landscape and natural resource management, sustainable livelihoods, ecosystem-based climate change adaptation and environmental protection. His projects include transdisciplinary participatory action research to develop and implement landscape planning and governance for biodiversity conservation and sustainable community livelihoods in the Papua New Guinea, the Amazon, and the Democratic Republic of the Congo. He is interested in how to turn knowledge into action to respond to climate change and biodiversity loss and create more equitable and sustainable livelihoods.

Heather Keith

Dr Heather Keith’s research is aimed at understanding the functioning of ecosystems, particularly forests, to improve their management for conservation and climate change mitigation. Her research integrates field and experimental data with spatial data to scale up information about ecosystems across landscapes. Her expertise includes reporting of data in ecosystem accounts that provide an internationally standardised format used to inform ecosystem management and climate change policy.