The journey to Nature Positive: Methods for Sampling and Quantifying Nature in Landscape-Scale and Dispersed Projects

This is the fourth article in our The journey to Nature Positive series, where our specialists illustrate how Nature-based Solutions (NbS) can tackle societal challenges, offer additional benefits over engineered solutions, and promote a Nature-positive world.

In this article, co-authored by our specialist Dr. Ida Bailey and Dr. Max Bodmer, Head of Nature-based Consultancy at rePLanet, we explore methods for sampling and quantifying nature and biodiversity for large and dispersed projects. These approaches are equally applicable to other types of nature restoration projects and not just those providing NbS.

NbS are increasingly central to global strategies for climate resilience, biodiversity recovery, and sustainable development. Yet, as projects scale up, from river basin restorations to networks of urban green spaces, the challenge of quantifying biodiversity outcomes becomes more complex and urgent.

Robust biodiversity metrics are essential for:

• Demonstrating ecological impact.
• Avoiding greenwashing by ensuring Nature Positive claims are real, measured, and verifiable.
• Attracting investment through biodiversity credits and carbon credits with quantified biodiversity benefits.
• Aligning with frameworks like the Taskforce on Nature-related Financial Disclosures (TNFD).

This piece explores how spatial sampling strategies, technological innovations, and cost-effective monitoring frameworks can support biodiversity quantification in both landscape-scale and highly dispersed NbS projects.

Landscape-Scale and Dispersed Projects

In large-scale projects, such as river basin restoration, landscape recovery or regional rewilding, ecological processes unfold across diverse habitats and spatial gradients. These projects are often so large, spanning 100 km² or more, that even a stratified or grid sampling-based approach would not be practical or affordable. The sheer spatial extent, combined with logistical and financial constraints, necessitates a more strategic approach.

Dispersed NbS, such as urban green infrastructure, portfolios of small marine protected areas, or forestry portfolios are made up of many small sites. Their ecological impacts are localised and heterogeneous, making aggregation difficult and it would not be practical or affordable to undertake intensive sampling at all of them.

Site Selection

Before biodiversity can be monitored, sites must be selected. Here, systematic conservation planning (SCP) and spatially explicit sampling can be useful. SCP provides a robust framework for identifying priority areas based on ecological value, threat levels, and feasibility and tools like Conservation Area Prioritization Through Artificial Intelligence (CAPTAIN) support site selection by using reinforcement learning to identify optimal conservation areas.

What to Measure & Alignment with TNFD and Carbon Standards

Biodiversity is highly varied, and it is not practical or desirable to measure everything, everywhere and instead, the leading approach is to select a representative basket of indicators/ metrics. A basket-of-metrics approach is considered especially robust for biodiversity quantification because it captures the complexity of ecosystems more effectively than single indicator approaches like BNG (a habitats only metric).

By combining multiple metrics such as species richness, habitat condition, and ecosystem function, this method provides a more comprehensive and context-sensitive picture of biodiversity change.

The Wallacea Trust’s methodology exemplifies this basket of metrics approach by tailoring metrics to specific ecoregions and using median changes over time to reflect biodiversity gains. This flexible approach can be used in any ecosystem from coral reefs to rainforests and deserts.

Basket of metrics approaches/frameworks are valuable as they provide architecture within which to build out a sampling approach that is common to all projects in a portfolio, regardless of ecosystem and geography, and offer flexibility to ensure that the approach can be adjusted to accommodate different ecosystems and project idiosyncrasies.

To address these challenges, monitoring frameworks could:

• Use project/portfolio relevant metrics that are appropriate to the ecosystem and standardised over time for consistency and replicability (i.e. A basket of five metrics approach following the Wallacea Trust Methodology such as, habitats extent x condition, breeding birds, beetles, pollinators and soil invertebrates).
• Employ low-cost, low skill technologies and methodologies like acoustic sensors, eDNA kits, mobile apps, remote sensing and pit-fall traps.
• Leverage citizen science and local stewardship for data collection.

Platforms like Mozaic Earth exemplify this approach by combining satellite data with community inputs, enabling real-time biodiversity tracking whilst empowering local stakeholders to participate in monitoring and decision-making. This democratisation of data collection reduces costs and builds local capacity, and enhances transparency, all of which are key for investor confidence and long-term project success.

Data transparency is becoming an important consideration when designing biodiversity monitoring strategies as organisations are becoming increasingly aware of the reputational damage that can be caused by accusations of greenwashing. By adopting methods and operational procedures that produce easily auditable data (preferably with a geotagged digital signature) projects maximise transparency and significantly reduce the risk of greenwashing.

Alignment and Future Proofing

Picking monitoring metrics that align with these additional data needs is therefore an important consideration. For example:

• The TNFD framework can align well with the biodiversity monitoring strategies outlined above, if the basket of metrics is selected carefully to represent biodiversity in similar terms. TNFD considers the state of nature in four broad categories:
  • Ecosystem extent (e.g. type, distinctiveness & extent of habitats).
  • Ecosystem condition (e.g. habitat condition, and landscape scale index of ecosystem connectivity or integrity or intactness could also be considered).
  • Species populations (species diversity and abundance).
  • Species extinction risk (species conservation status – in the Operation Wallacea methodology, species data are weighted by their conservation importance and data on priority species can be split out for separate examination).
• The UK’s Peatland Code and Woodland Carbon Code (WCC) have both initiated efforts to integrate biodiversity crediting into their existing carbon standards. A 2025 project led by the IUCN UK Peatland Programme, rePLANET, Soil Association Certification, and SRUC explored how biodiversity uplift could be measured, validated, and monetised within these frameworks. The Wallacea Trust’s methodology was selected as the foundation due to its scientific rigor and adaptability, with data collected by SLR contributing to this process.

Strategic Sampling for Scalable Monitoring

Once sites are selected, the focus shifts to monitoring.

Traditional biodiversity monitoring often relies on intensive fieldwork by ecological specialists. While scientifically rigorous, this approach is costly and difficult to scale, especially in regions with limited ecological capacity. To address this, recent literature advocates tiered spatial sampling strategies that balance precision with practicality.

A tiered sampling strategy, combining permanent plots, stratified random sampling, and model-based inference, offers a scalable, cost-effective solution. This hybrid approach allows for scalable biodiversity assessment with potentially only minor compromises to data integrity.

Key Components of Strategic Sampling

The following components are equally applicable to both landscape-scale and dispersed projects. We would expect all projects to have at least a proportion of permanent sample plots, with modelling supported by temporary plot data reducing costs and increasing the practicality of monitoring large and dispersed projects whilst retaining the robustness of outputs as far as practical:

• Permanent sample plots: A small number of intensively monitored sites provide high-quality, long-term data and serve as anchors for model calibration. These are essential for detecting ecological change and ensuring reproducibility in biodiversity science (Damerow et al., 2025; Sims, 2022; Lindenmayer et al., 2022).
• Temporary sample plots: Strategically deployed, in addition to permanent plots, across under-sampled or environmentally heterogeneous areas, can fill spatial gaps and improve the representativeness of the dataset. This hybrid approach enables more accurate model-based extrapolation to unsampled areas by increasing the diversity of environmental conditions captured in the training data, thereby reducing prediction bias and variance (Kissling et al., 2018). This can yield more robust estimates of species richness and community composition than uniform sampling alone (Lindenmayer & Likens, 2010). They allow for flexible, cost-effective deployment in response to emerging conservation priorities or logistical constraints, making them a valuable complement to permanent monitoring infrastructure (Yoccoz et al., 2001).
• Stratified random sampling: Ensures representative placement of sampling plots across habitat types or environmental gradients, is particularly effective in heterogeneous landscapes (Franklin et al., 2022; Latpate et al., 2021). Stratification allows for efficient allocation of sampling effort, especially when certain habitats are rare or ecologically significant. Results would likely need weighting to avoid over representation of rarer habitats.
• Grid sampling: A systematic/grid sampling design may be more cost effective than stratified sampling in many instances and reduces the risk of over representation of rarer habitats. They have become increasingly important in ecological research for biodiversity quantification, offering spatially explicit and standardised frameworks for monitoring. Recent studies, such as the LIFEPLAN project, demonstrate how grid-based designs enable integration of sensor and DNA-based data across ecosystems, enhancing comparability and scalability (Roslin et al., 2021; Bálint et al., 2023).
• Fractal and multi-scale sampling designs: These are useful when biodiversity patterns exhibit spatial autocorrelation or scale-dependent variation. These designs improve spatial inference and model performance by capturing variability at multiple spatial resolutions (Fortin et al., 2023; Fletcher & Fortin, 2019; Voss, 2025; Hou et al., 2021).
• Selecting the sampling design: The sampling design (stratified, grid or fractal) must be carefully chosen to reflect the aim of the project as well as its practical and financial constraint. Whether a proponent chooses to adopt a stratified random or grid sampling, approach will be determined by the idiosyncrasies and core objectives of the biodiversity monitoring strategy. If the primary aim is to estimate the level of biodiversity within a project boundary for the purposes of credit generation, grid sampling will provide a site-level average and help to reduce monitoring costs. However, if the project is seeking to understand how biodiversity is distributed across a project site, possibly to help design an evidence-based land management plan, a stratified random sampling approach will likely be required.
• Model-based inference: Statistical models (e.g., hierarchical Bayesian models, kriging) extrapolate biodiversity metrics from sampled to unsampled areas. This approach reduces field effort while maintaining accuracy and is particularly valuable when full coverage is infeasible. Model accuracy can be checked and refined if combined with a degree of site-based sampling, such as permanent or temporary plots.

Transparent Classification of Project by Data Integrity

The more heavily a project relies on modelled data and temporary plots rather than intensive permanent plots, the harder it is to have the same degree of confidence in the results. Even if confidence is still high for modelled data, this may make it less attractive in some circumstances depending on the level of integrity that the project requires to justify any claims that it is making. A solution to this could be a transparent, tiered approach to project classification with different users choosing different tiers depending on their ultimate aim. For example:

Tier 1

High-intensity data collection on the ground that provides the highest integrity data but also generates a large cost. Small projects wanting to generate natural capital credits that are perceived as having high value may make the investment in high-intensity sampling as it will maximise the price at which they can sell their credits. When projects get much larger than approx. 10,000 hectares (ha), this approach becomes less feasible because of rising costs and increased human resource demands.

Tier 2

Tools that are developed to allow non-experts to collect data in the field which can be analysed and evaluated by ecologists, e.g. photographs and audio recordings; this approach reduces costs significantly and has only minor impacts on data accuracy/integrity as species identifications are still made by taxonomic experts. This approach might work for small projects that are not necessarily seeking to generate credits but want to have a biodiversity assessment that allows them to track their progress towards Nature Positivity over time. This may present a cost-effective solution for organisations that need to assess biodiversity across a large portfolio of dispersed sites, such as global car manufacturers or pharmaceutical companies that have factories and offices located around the world. If the tier 2 approach is implemented across the portfolio it would allow organisations to assess the relative performance of different sites under their jurisdiction. It also reduces costs by harnessing the power of the non-expert to gather data in the field and it is therefore likely most feasible across portfolios that have sites that are not larger than 1,000 ha.

Tier 3

Combining geospatial analysis with ecological/predictive modelling (i.e. the Map of Life) verified by ground truth control points collected on the ground such as spot checks to verify the accuracy of the model. A best practice approach would involve assigning a series of permanent ground truth control plots that are assessed regularly throughout the project period alongside temporary plots that are rotated between sites within the project area where the data indicates higher resolution information is needed. When using this approach, data insights rely on inferences from geospatial analysis and ecological modelling, so they are less accurate than more traditional ‘boots on the ground’ approaches, but the significant cost reductions make it much more feasible for large-scale projects. This approach will likely work for large scale projects between 10,000 and 1 million ha.

Tier 4

A geospatial only approach should be considered as a last resort, for projects that are too large to monitor on the ground and where achieving a meaningful density of ground-truth control points would be prohibitively expensive. In such cases – such as jurisdictional REDD Projects, even a lower density of control points may add limited value. This approach should only be used for extremely large project areas, upwards of 1 million ha and should draw on field/primary data as far as practical.

How we can help

SLR is uniquely positioned to support clients in implementing biodiversity quantification strategies for both landscape-scale and dispersed NbS projects. Through our collaboration with rePLANET, the team behind The Wallacea Trust’s methodology, we bring technical expertise and practical experience in applying this framework.

Our industry leading specialists offer comprehensive aquatic environmental, ecology. Our team supports clients globally with corporate nature reporting, TNFD-aligned disclosures, and nature-positive strategy development. Our integrated approach combines ecological science, geospatial analysis, and stakeholder engagement to deliver credible, practical, decision-useful project outcomes.

To speak to one of our specialists, contact us today.

References

Bálint, M., Schmidt, P.A., Sharma, N., et al. (2023). A global DNA-based biodiversity monitoring system for the Anthropocene. Nature Ecology & Evolution, 7, 105–115. https://doi.org/10.1038/s41559-022-01991-0

CAPTAIN Project. (2025). Conservation Area Prioritization Through Artificial Intelligence. Available at: https://www.captain-project.eu [Accessed 1 Aug. 2025].

Damerow, J. E., et al. (2025). Metadata standards and sample tracking for long-term biodiversity monitoring. Nature Scientific Data.

Fletcher, R. J., & Fortin, M.-J. (2019). Spatial Ecology and Conservation Modeling: Applications with R. Springer.

Forest Stewardship Council (FSC). (2025). Ecosystem Services Procedure: Impact Demonstration and Market Tools. Available at: https://open.fsc.org/items/51569dac-d956-438d-9722-9f7e338c1840 [Accessed 1 Aug. 2025].

Fortin, M.-J., et al. (2023). Comparing sampling designs for detecting spatial autocorrelation in plant communities. Journal of Ecology.

Franklin, J., et al. (2022). Two-stage stratified random sampling for vegetation monitoring in arid ecosystems. Ecological Indicators.

Hou, D., et al. (2021). Multi-scale fractal analysis of spatial heterogeneity in natural systems. Natural Resources Research.

Kissling, W.D., Ahumada, J.A., Bowser, A., Fernandez, M., Fernández, N., García, E.A., Guralnick, R.P., Isaac, N.J.B., Kelling, S., Los, W. and McRae, L. (2018). Building essential biodiversity variables (EBVs) of species distribution and abundance at a global scale. Biological Reviews, 93(1), pp.600–625.

Latpate, R. V., et al. (2021). Advanced Sampling Methods for Environmental Monitoring. In: Handbook of Environmental Statistics.

Lindenmayer, D. B., et al. (2022). The role of long-term ecological monitoring in biodiversity conservation. Biological Conservation.

Lindenmayer, D.B. and Likens, G.E. (2010). The science and application of ecological monitoring. Biological Conservation, 143(6), pp.1317–1328.

Mozaic Earth. (2024). Ground-truth your nature data, at scale. Available at: https://www.mozaic.earth [Accessed 1 Aug. 2025].

Roslin, T., Antão, L.H., Bálint, M., et al. (2021). The LIFEPLAN approach for global monitoring of biodiversity. Nature Ecology & Evolution, 5, 256–267. https://doi.org/10.1038/s41559-020-01368-5

Sims, A. (2022). Statistical Design of Sample Plots in National Forest Inventories. In: Forest Inventory Methodologies. Springer.

Taskforce on Nature-related Financial Disclosures (TNFD). (2025). Recommendations of the Taskforce on Nature-related Financial Disclosures. Available at: https://tnfd.global/publication/recommendations-of-the-taskforce-on-nature-related-financial-disclosures/ [Accessed 1 Aug. 2025].

Voss, R. F. (2025). Fractals in Nature: Applications in Ecology and Earth Systems. Springer.

Wallacea Trust. (2023). Biodiversity Credit Methodology V3. Available at: https://wallaceatrust.org/wp-content/uploads/2022/12/Biodiversity-credit-methodology-V3.pdf [Accessed 1 Aug. 2025].

Yoccoz, N.G., Nichols, J.D. and Boulinier, T. (2001). Monitoring of biological diversity in space and time. Trends in Ecology & Evolution, 16(8), pp.446–453.