Bioacoustics & passive ecological monitoring

01Overview and value chain

Markers: [EC: Kunming-Montreal Global Biodiversity Framework + EU Nature Restoration Law | OECD: Sustainable land use, Agricultural biotechnology | Regulator: EPA (USA), ADEME (France), MOA (China)]

Bioacoustics & passive ecological monitoring measures biodiversity without trapping or observing individual animals — autonomous recorders capture soundscapes, AI identifies species from audio, and environmental DNA (eDNA) reads the genetic traces organisms leave in water, soil and air. Driven by the Kunming-Montreal Global Biodiversity Framework and the EU Nature Restoration Law, the field turns cheap field data into verifiable biodiversity metrics for conservation, real-time enforcement against illegal logging (a $50–150B/yr crime), and corporate nature-risk disclosure under the TNFD. Rainforest Connection alone has catalogued 160M+ audio files and identified 4,208 species; Cornell Lab’s BirdNET AI recognises 6,522 species and Merlin Bird ID has 10M active users.

The key directions of bioacoustics & passive ecological monitoring are:

1. Autonomous bioacoustic recorders & PAM (Passive Acoustic Monitoring): weatherproof, long-deployment recorders — Wildlife Acoustics’ Song Meter SM5BAT runs 4 channels (2 ultrasonic + 2 acoustic) for simultaneous bat and bird surveys.
2. Acoustic AI & citizen science (Acoustic AI): machine-learning species ID — BirdNET classifies 6,522 species; RFCx’s edge AI auto-recognises 955 species and alerts rangers to chainsaws and gunshots in real time.
3. eDNA metabarcoding (eDNA Metabarcoding): genetic traces in water/soil/air detect all taxa from microbes to mammals — Spygen’s VigiDNA® covers freshwater vertebrates at €290/kit; NatureMetrics serves Unilever and Anglo American.
4. Biodiversity data platforms & TNFD disclosure (Biodiversity Platforms): subscription services convert field data into decision-ready nature-risk metrics — NatureMetrics’ Portfolio Assessment AI maps nature risk across site portfolios.

Sectoral value chain

[sensor deployment] ──> [data capture] ──> [AI species ID] ──> [biodiversity analytics]
                              │
                  (acoustic / eDNA / satellite)
                              │
                              ▼
[conservation action] <─── [TNFD disclosure] <─────┘

Value chain levels

Cross-cutting technologies of the sector:

• Edge + cloud acoustic AI (Edge + Cloud Acoustic AI): on-device detection (RFCx satellite alerts) plus cloud-scale analysis (Arbimon, BirdNET).
• eDNA metabarcoding (eDNA Metabarcoding): multi-primer PCR + high-throughput sequencing identifies whole communities from one water/soil sample.
• Handheld field genetics (Handheld Field Genetics): Conservation X Labs’ NABIT runs point-of-contact DNA barcode tests by non-technical users.

02US

The US leads bioacoustic hardware, open acoustic AI and field-portable eDNA, spanning research, conservation enforcement and citizen science.

acoustic-recorder incumbency, edge AI illegal-logging alerts, open acoustic AI + field eDNA

• Wildlife Acoustics (Maynard, MA, founded 2003): the global bioacoustic hardware leader; its Song Meter SM5BAT runs 4 channels (2 ultrasonic bat + 2 acoustic bird) with IP67 weatherproofing, 4 TB storage, onboard GPS/GLONASS (30 µs time-sync) and a 2027 cellular-access expansion port.
• Rainforest Connection / RFCx (San Francisco, founded 2014 by Topher White): solar-powered Guardian devices stream or edge-analyse forest soundscapes — 587 devices across 37 countries monitor 736,200 ha, having catalogued 160M+ audio files and 4,208 species (408 threatened), with edge AI auto-recognising 955 species and alerting rangers to chainsaws/gunshots.
• Cornell Lab of Ornithology (Ithaca, NY): open acoustic-AI anchor — BirdNET identifies 6,522 species and underpins community tools (BirdNET-Go adds Google Perch v2’s 14,795 species); the Merlin Bird ID app has 10M active users since 2014, and Macaulay Library + eBird feed global science.
• Conservation X Labs (Washington DC, founded 2015): NABIT (Nucleic Acid Barcode Identification Tool) — a handheld, battery-powered device running point-of-contact DNA tests by non-technical users, paired with the MDK molecular development kit (Capgemini Gene Genius partnership).

03CN

China’s passive ecological monitoring is research-led rather than vendor-led — the Chinese Academy of Sciences sets the eDNA methodological frontier while commercial supply remains thin.

CAS eDNA research leadership, national biomonitoring networks, research-led (not vendor-led)

• Research leadership: the Chinese Academy of Sciences Kunming Institute of Zoology Biodiversity Genomics Center runs a P2 eDNA/paleoDNA platform and, in a 2025 Gaoligongshan study, compared three eDNA sampling strategies — detecting 235 vertebrate species (water eDNA 173, leech-derived iDNA 102, air eDNA 56) across a hyperdiverse mountain ecosystem.
• National infrastructure: China’s national ecological-monitoring networks and biodiversity-survey programs under the Global Biodiversity Framework, with BGI-scale sequencing infrastructure supporting metabarcoding — but no dominant commercial bioacoustic/eDNA vendor has yet emerged, leaving the niche to CAS institutes and universities.

04EU

Europe leads eDNA biodiversity services and corporate nature-risk disclosure, backing the TNFD and CSRD reporting wave.

eDNA service leadership + TNFD, precision rare-species eDNA, corporate nature-risk SaaS

• NatureMetrics (UK, founded 2014 by Dr Kat Bruce): the leading nature-intelligence company; raised a $25M Series B (Just Climate, EDF Pulse Ventures, Monaco ReOcean Fund, 2150, BNP Paribas), using eDNA + earth observation + AI for clients including Unilever, Anglo American, WWF and Fauna & Flora; its Portfolio Assessment AI maps nature risk across corporate site portfolios, with labs in the UK, Canada and an Indonesia partner.
• Spygen (France): pioneer of precision eDNA — its VigiDNA® technology (6 international patents, science dating to 2005) specialises in rare and elusive species, with a freshwater Vertebrate Screening Kit at €290 and metabarcoding inventories built on one of the industry’s largest private reference databases.

05Leading companies and research institutes

06Tech stack and innovations

The stack rests on three platforms: autonomous bioacoustic recorders, acoustic-AI species ID, and eDNA metabarcoding / field genetics.

1. Autonomous bioacoustic recorders (PAM Recorders):
   • weatherproof, multi-month recorders capture ultrasonic (bats) and acoustic (birds, amphibians, mammals) soundscapes with GPS time-sync for localisation.
   • case: Wildlife Acoustics’ Song Meter SM5BAT (4 channels, IP67, 4 TB, 30 µs GPS sync); RFCx’s solar Guardians (24/7 canopy deployment, 50–1,500 m mic range).
2. Acoustic-AI species identification (Acoustic AI):
   • deep-learning models classify species from audio, running at the edge (real-time alerts) or in the cloud (bulk archive analysis).
   • case: Cornell Lab’s BirdNET (6,522 species), BirdNET-Go’s multi-model consensus (BirdNET + Google Perch v2’s 14,795 species + BattyBirdNET bats); RFCx’s 955 auto-recognised species.
3. eDNA metabarcoding & handheld field genetics (eDNA & Field Genetics):
   • environmental DNA in water/soil/air is PCR-amplified and sequenced to inventory whole communities; handheld devices push barcode tests into the field.
   • case: NatureMetrics’ eDNA service (microbes to mammals, Portfolio Assessment AI), Spygen’s VigiDNA® rare-species precision (6 patents), and Conservation X Labs’ NABIT handheld point-of-contact testing.

07Value chains and production pipelines

Industrial pipeline of a passive ecological-monitoring operation (TNFD / GBF)

┌───────────────────────────┐      ┌───────────────────────────┐
│ 1. Sensor deployment      │ ───> │ 2. Data capture & xmit    │
└───────────────────────────┘      └───────────────────────────┘
                                                  │
                                                  ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 4. Biodiversity analytics │ <─── │ 3. AI / eDNA species ID   │
└───────────────────────────┘      └───────────────────────────┘
              │
              ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 5. Disclosure & reporting │ ───> │ 6. Conservation action    │
└───────────────────────────┘      └───────────────────────────┘

Stage 1: Sensor deployment

Acoustic recorders (Wildlife Acoustics, RFCx Guardians) are deployed in canopies/wetlands, and eDNA samples (water/soil/air, Spygen/NatureMetrics kits) are collected across the site grid.

Stage 2: Data capture & transmission

Recorders stream via GSM/Wi-Fi or run satellite edge inference (RFCx SWARM); eDNA samples ship to central labs (NatureMetrics UK/Canada) for sequencing, or run on-site via NABIT handhelds.

Stage 3: AI / eDNA species ID

Acoustic AI (BirdNET 6,522 species, Google Perch v2 14,795) classifies audio; eDNA metabarcoding matches sequences against reference libraries (Spygen’s private database) to produce species detections.

Stage 4: Biodiversity analytics & MRV

Detections become biodiversity metrics — species richness, occupancy, baselines and trends — validated against eBird/eDNA priors (the FINCH framework fuses audio with eBird spatiotemporal data).

Stage 5: Disclosure & reporting

Metrics flow into TNFD/CSRD/GRI nature reports and corporate dashboards — NatureMetrics’ Portfolio Assessment AI aggregates site-level nature risk across a whole corporate portfolio.

Stage 6: Conservation action

Insights trigger action — RFCx real-time alerts stop illegal logging and poaching; biodiversity baselines inform restoration targeting under the EU Nature Restoration Law and China’s national ecological-monitoring programs.