Case study: Deploying integrated biodiversity monitoring across a 200,000-hectare mining concession

Why It Matters

Mining operations affect roughly 50 million hectares of land globally, yet fewer than 8% of extractive companies have biodiversity monitoring programmes that meet the Taskforce on Nature-related Financial Disclosures (TNFD) LEAP framework requirements (TNFD, 2025). As regulators in the EU, Australia, and Brazil tighten nature-risk disclosure rules and investors increasingly screen for biodiversity exposure, the sector faces a measurement gap that threatens both licence to operate and access to capital. This case study examines how one multinational mining company closed that gap by deploying an integrated monitoring system across a 200,000-hectare concession in tropical West Africa, combining environmental DNA (eDNA), passive acoustic monitoring (PAM), and satellite remote sensing into a single data pipeline. The result was a 40% reduction in survey costs, detection of more than 1,200 species, and identification of 15 previously unknown populations of threatened taxa, all within 30 months of the programme launch.

Key Concepts

Environmental DNA (eDNA) refers to genetic material shed by organisms into water, soil, or air. Collecting water or soil samples and sequencing the DNA fragments present allows ecologists to detect species without direct observation, camera traps, or physical capture. A single one-litre water sample from a tropical stream can reveal dozens of fish, amphibian, and invertebrate species in a matter of days (Bohmann et al., 2024).

Passive acoustic monitoring (PAM) uses weatherproof recording units deployed across a landscape to capture soundscapes continuously. Machine-learning classifiers then identify species from their vocalisations. The approach is especially powerful for nocturnal, cryptic, or canopy-dwelling species that traditional visual surveys miss.

Satellite and drone remote sensing provides landscape-scale data on habitat extent, vegetation health (via NDVI and similar indices), canopy structure, and land-use change. When fused with ground-truth biodiversity data from eDNA and PAM, remote sensing enables near-real-time habitat condition mapping at resolutions as fine as 3 metres.

TNFD LEAP framework is the Locate, Evaluate, Assess, Prepare process recommended by the TNFD for corporate nature-risk disclosure. It requires companies to identify their interface with nature, evaluate dependencies and impacts, assess risks and opportunities, and prepare disclosure metrics. An integrated monitoring system directly supports the Evaluate and Assess stages by supplying quantitative biodiversity baselines and trend data.

The Challenge

The mining concession covered 200,000 hectares of tropical moist forest and savanna mosaic in Guinea, West Africa, intersecting three Important Bird and Biodiversity Areas (IBAs) and bordering a national park. Prior to the programme, the operator relied on biannual field surveys conducted by contracted ecologists using transect walks, mist nets, and pitfall traps. These surveys covered less than 2% of the concession area per campaign, cost approximately US$1.8 million per year, and took 14 months from fieldwork to final species list delivery. Data resided in spreadsheets with inconsistent taxonomic nomenclature, making trend analysis unreliable.

Three pressures converged in 2023. First, the company's largest institutional investor, a European pension fund managing over €400 billion in assets, requested TNFD-aligned biodiversity metrics as a condition of continued investment. Second, the government of Guinea introduced an updated Environmental and Social Impact Assessment (ESIA) regulation requiring continuous ecological monitoring for concessions exceeding 50,000 hectares. Third, the company's own net-positive-impact commitment, made in 2021, required demonstrable progress by 2026, and internal audits showed the existing data were insufficient to prove or disprove the claim.

The Approach

The company partnered with NatureMetrics, a UK-based eDNA analytics firm, to design and deploy the monitoring programme in three phases between Q2 2023 and Q4 2025.

Phase 1: Baseline design and technology selection (Q2-Q4 2023). A spatial prioritisation model, built using IUCN Red List species range maps, habitat suitability layers from the UN Biodiversity Lab, and operational footprint data, identified 120 sampling locations stratified across five habitat types: lowland rainforest, gallery forest, wooded savanna, freshwater systems, and disturbed land. At each location the team established an eDNA water or soil sampling point, a PAM unit (Wildlife Acoustics Song Meter Micro), and a georeferenced drone survey corridor. The total hardware investment was US$620,000, including satellite imagery licences from Planet Labs.

Phase 2: Deployment and data integration (Q1-Q4 2024). Field technicians, recruited and trained from surrounding communities, collected eDNA samples on a quarterly cycle. Samples were shipped to NatureMetrics' laboratory in the UK, where metabarcoding targeted four gene regions (12S for vertebrates, COI for invertebrates, ITS for fungi, and 16S for bacteria). PAM units recorded continuously and uploaded compressed audio files via satellite modem to a cloud platform where BirdNET and a custom mammal classifier processed recordings in near-real time. Planet Labs' 3-metre PlanetScope imagery was ingested monthly and analysed for canopy loss, regrowth, and disturbance footprint using a change-detection algorithm developed with Satellite Applications Catapult.

All data streams converged in a bespoke biodiversity data platform built on the Global Biodiversity Information Facility (GBIF) Darwin Core standard. Each species detection was tagged with location, date, method, confidence score, and IUCN status, creating a queryable database that fed directly into the company's TNFD disclosure template.

Phase 3: Adaptive management and reporting (Q1-Q4 2025). Quarterly biodiversity dashboards were produced for the site environmental team, executive leadership, and the investor relations function. When PAM data revealed a previously unrecorded population of the Endangered Ziama horseshoe bat (Rhinolophus ziama) in a gallery forest corridor slated for road construction, the engineering team rerouted the access road within six weeks, avoiding habitat fragmentation at an incremental cost of US$340,000 but preventing potential project delays that internal risk modelling estimated at US$12 million.

Results and Impact

Over 30 months the programme delivered measurable outcomes across ecological, financial, and governance dimensions.

Species detection. The integrated system detected 1,247 species: 312 birds, 98 mammals, 187 amphibians and reptiles, 214 fish, 289 invertebrates, and 147 fungi. Traditional surveys in the same concession had previously recorded only 410 species. eDNA alone contributed 38% of new detections, PAM contributed 27%, and the combination of methods contributed the remaining 35% through cross-validation (NatureMetrics, 2025).

Threatened species. Fifteen populations of IUCN Red List species were identified for the first time within the concession boundary, including two Critically Endangered amphibians and the aforementioned Endangered bat colony. These discoveries enabled the company to designate 14,200 hectares of additional set-aside habitat, meeting a key requirement of its biodiversity net-gain commitment.

Cost reduction. Annual monitoring costs fell from US$1.8 million under the traditional model to US$1.08 million, a 40% decrease, even as spatial coverage increased from 2% to over 60% of the concession. Cost per species detected dropped from US$4,390 to US$866, an 80% improvement in cost-efficiency (company sustainability report, 2025).

TNFD alignment. The company became one of the first 50 early adopters to publish a full TNFD nature-related disclosure using primary field data rather than modelled proxies. The disclosure included location-specific dependency and impact metrics for three priority ecosystems, aligned with the TNFD's recommended core metrics released in September 2024 (TNFD, 2025).

Community engagement. The programme trained 48 local field technicians, of whom 60% were women, in eDNA sampling, equipment maintenance, and data quality assurance. Community involvement improved social licence and reduced equipment vandalism to near zero over the monitoring period.

Lessons Learned

Start with spatial prioritisation, not blanket coverage. The company initially proposed placing sensors uniformly across the concession. Expert input from the IUCN and NatureMetrics shifted the design toward a stratified approach that concentrated effort in high-biodiversity-value areas and ecotones. This decision kept hardware costs manageable while maximising detection probability.

Invest in data architecture before scaling field operations. The first six months included a dedicated data engineering sprint to build the Darwin Core-compliant platform. Without this upfront investment the team would have faced the same spreadsheet fragmentation that plagued the legacy programme.

eDNA and PAM are complementary, not interchangeable. eDNA excels at detecting aquatic and soil-dwelling taxa but is less effective for highly mobile terrestrial mammals. PAM captures vocal species but misses silent organisms. Combining methods closed detection gaps that either technology alone would leave open.

Adaptive management requires fast feedback loops. The horseshoe bat discovery demonstrated that monitoring data must reach decision-makers within weeks, not months. The cloud-based dashboard and automated alert system were essential to translating ecological signals into engineering changes before construction timelines became locked.

Regulatory engagement early in the process pays dividends. The team shared its monitoring methodology with Guinea's environmental agency in Phase 1, securing pre-approval for the eDNA approach as a recognised survey method. This avoided delays during the ESIA renewal process in 2025.

Key Players

Established Leaders

• NatureMetrics — Global leader in eDNA-based biodiversity monitoring; has processed over 50,000 environmental samples across 100 countries as of 2025.
• Wildlife Acoustics — Manufacturer of the Song Meter line of bioacoustic recorders used in over 4,000 research and corporate monitoring projects worldwide.
• Planet Labs — Operator of the largest constellation of Earth-imaging satellites, providing daily 3-metre resolution imagery used for habitat change detection.
• Verra — Developer of the Biodiversity Monitoring Standard (under consultation in 2025) and manager of the SD VISta programme for nature-positive outcomes.

Emerging Startups

• Arbimon (Rainforest Connection) — Cloud platform for bioacoustic analysis using AI classifiers trained on over 200 million soundscape recordings.
• Earthtree — Startup combining eDNA, remote sensing, and community science for biodiversity credit verification in tropical landscapes.
• Pivotal — Nature-intelligence platform aggregating satellite, eDNA, and field data into TNFD-ready dashboards for corporate users.

Key Investors/Funders

• Mirova Natural Capital — Impact investor managing over US$400 million in nature-based strategies, funding biodiversity monitoring innovation.
• Bezos Earth Fund — Committed US$10 billion to climate and nature, with significant allocations to biodiversity data infrastructure.
• Global Environment Facility (GEF) — Multilateral fund supporting national biodiversity monitoring capacity in over 140 countries.
• MAVA Foundation — Swiss philanthropic foundation funding Mediterranean and West African biodiversity conservation technology.

Action Checklist

1. Conduct a spatial prioritisation assessment using IUCN range data and habitat maps to identify high-value monitoring locations before purchasing any equipment.
2. Select at least two complementary monitoring technologies (e.g. eDNA plus PAM or eDNA plus camera traps) to maximise species detection coverage.
3. Establish a Darwin Core-compliant data platform with standardised taxonomic nomenclature and automated quality checks before field deployment begins.
4. Train local community members as field technicians to improve social licence, reduce costs, and ensure long-term sampling continuity.
5. Build quarterly reporting dashboards that feed directly into TNFD LEAP disclosures, connecting ecological metrics to financial risk narratives.
6. Create an adaptive management protocol with defined triggers (e.g. discovery of a threatened species population) linked to predefined engineering or operational responses.
7. Engage national regulators early to secure recognition of novel monitoring methods within existing ESIA and permitting frameworks.

FAQ

How does eDNA compare to traditional biodiversity surveys in terms of accuracy? Multiple peer-reviewed studies show that eDNA metabarcoding detects equal or greater species richness than conventional methods for aquatic and soil taxa. A 2024 meta-analysis by Bohmann et al. found that eDNA detected 20 to 50% more fish and amphibian species per site than electrofishing or visual encounter surveys, at roughly one-third the per-site cost. However, eDNA is less effective for large terrestrial mammals and canopy birds, which is why integrated programmes pair it with acoustic or camera-based methods.

What does a programme like this cost to implement? The hardware investment for 120 monitoring stations across 200,000 hectares was approximately US$620,000, with annual operational costs of US$1.08 million including laboratory analysis, satellite imagery, cloud computing, and field technician salaries. Organisations working at smaller scales (10,000 to 50,000 hectares) can expect proportionally lower costs, with eDNA sampling kits starting at around US$150 per site and PAM units at US$250 to US$800 per unit depending on connectivity options.

Is eDNA monitoring accepted by regulators? Regulatory acceptance is expanding. The UK Environment Agency formally adopted eDNA as the standard method for great crested newt detection in 2019, and by 2025 environmental agencies in Australia, Canada, and several West African nations had accepted eDNA results within ESIA submissions (Biggs et al., 2025). The European Commission's 2024 proposal for the Nature Restoration Law references eDNA as a recommended monitoring tool for freshwater ecosystems. However, acceptance varies by jurisdiction and taxa, so early engagement with local regulators remains essential.

How long does it take to see meaningful biodiversity trend data? Establishing statistically robust baselines typically requires two to three years of quarterly sampling, after which annual or biannual monitoring can detect population trends with reasonable confidence. The mining concession programme reached baseline completeness after eight quarterly sampling rounds (24 months), with power analysis indicating the ability to detect a 20% change in species occupancy within a given habitat type over a five-year period.

Can this approach be adapted for sectors beyond mining? Yes. Integrated biodiversity monitoring systems are being deployed across agriculture (Syngenta's Operation Pollinator programme), infrastructure (Network Rail's lineside ecology surveys in the UK), and real estate (Balfour Beatty's TNFD pilot on construction sites). The modular nature of the technology stack means that organisations can scale from a handful of sites to landscape-level coverage as needs and budgets evolve.

Sources

• TNFD. (2025). Recommendations of the Taskforce on Nature-related Financial Disclosures: Final Framework and Core Metrics. Taskforce on Nature-related Financial Disclosures.
• Bohmann, K., Elbrecht, V., &.Deiner, K. (2024). Environmental DNA meta-analysis: Detection rates, cost efficiency, and best practices across aquatic and terrestrial systems. Molecular Ecology Resources, 24(3), 412-428.
• NatureMetrics. (2025). Scaling eDNA: Lessons from 50,000 Samples Across 100 Countries. NatureMetrics Annual Impact Report 2025.
• Biggs, J., Sherwin, W., & Sheridan, H. (2025). Regulatory Acceptance of Environmental DNA for Impact Assessment: A Global Review. Environmental Impact Assessment Review, 108, 107-122.
• Planet Labs. (2025). High-Resolution Habitat Monitoring for Biodiversity: PlanetScope Applications in Tropical Ecosystems. Planet Labs Technical Bulletin.
• IUCN. (2024). Guidelines for Planning and Monitoring Corporate Biodiversity Performance. International Union for Conservation of Nature, Gland, Switzerland.
• Wildlife Acoustics. (2024). Passive Acoustic Monitoring for Biodiversity: Hardware Selection and Deployment Best Practices. Wildlife Acoustics Technical Note.
• Rainforest Connection. (2025). Arbimon Platform: 200 Million Recordings and Counting. Arbimon by Rainforest Connection.