Ecoacoustics in Nature Tech: Making Sound Decisions
Session Description
This session covered the use of ecoacoustics as a nature tech monitoring method, opening with a soil invertebrate case study comparing acoustic activity in good-quality versus degraded woodland habitat, then moving into an extended Q&A on hardware, recording methodology, and data standards, before a live hands-on demonstration using BirdNET-based species identification software on audio recorded from the room.
Speakers
Dr Carlos Abrahams, Director, Nature Sound Ltd
Natasha Fell, Nature Sound Ltd
Watch the Session Recording
Key Takeaways
Good-quality woodland habitat showed more detectable soil biological sound activity and greater invertebrate diversity than degraded habitat in a Sheffield Woodlands study.
Passive acoustic monitoring captures continuous raw environmental data that can be archived and compared year on year.
Choice between eDNA and acoustics depends on the question: eDNA reflects presence over weeks or months, acoustics pinpoints activity to a specific time and place.
Microphone type and recording setup directly affect detection radius and data volume; detection range varies from tens of metres for small birds to 100+ miles for whales in water.
Soil bioacoustics has a detection radius of roughly 50cm due to sound attenuation in soil.
Citizen science models, such as BTO's recorder-lending libraries, have generated large-scale acoustic datasets across whole UK regions.
Large repositories like the Macaulay Library and Xeno-Canto exist, but well-labelled "gold standard" datasets remain scarce.
A live demo used BirdNET-based software to analyse recorded audio and generate species identification confidence scores.
The session did not reach its planned focus on standardising and scaling acoustics for use in formal MRV systems.
