Britain’s largest woodland bat survey recorded 98 million calls over three years, a dataset that scientists state would have been impossible to process without artificial intelligence. Researchers deployed acoustic recording devices on poles across 310 woodlands throughout Britain, capturing 90,000 hours of audio data.
AI Processing Unlocks 98 Million Bat Calls Across Britain
Manual analysis of the massive acoustic dataset would have taken more than 30 years to complete. Open-source artificial intelligence tools allowed researchers to identify millions of individual bat calls across thousands of nights of recordings, flag specific species at a national scale, and map behavioral patterns.
Kate Jones from University College London stated that artificial intelligence has fundamentally changed the scale of wildlife monitoring. According to reporting by BBC News, Jones described the initiative as a notable example of technology utilized for conservation benefits, emphasizing that analyzing such a high volume of acoustic data manually was previously unfeasible.
Rare Barbastelle Bats Discovered in South-West England
The survey successfully tracked some of Britain’s rarest and most threatened bat species, including the red-listed Barbastelle. Hotspots of Barbastelle activity were heavily concentrated across the south-west of England, with significant populations detected in Somerset, Devon, Cornwall, and Dorset.
The Barbastelle typically roams ancient forests, but its elusive habits make traditional tracking methods difficult. Data from the Bat Conservation Trust indicates that Somerset serves as a strong regional stronghold for multiple native bat species.
Did you know?
The NCEA National Woodland Bat Survey deployed acoustic sensors across up to 210 one-hectare National Forest Inventory woodland sites in England each year starting in 2021, with automated classifications receiving expert manual verification to ensure data accuracy.
Collaboration Behind the National Woodland Bat Survey
The large-scale monitoring project operates as a joint initiative between Forest Research and the Bat Conservation Trust. Lia Gilmour of the Bat Conservation Trust characterized the survey as unprecedented in scope, noting that the sheer volume of 90,000 hours of recordings precluded any manual review approach.
Forest Research published a validated, multi-year dataset of accepted species records alongside an interactive data explorer. This publicly accessible platform allows researchers and the public to examine bat activity and distribution patterns across designated Special Areas of Conservation.
Frequently Asked Questions About the Woodland Bat Survey
How were the bat calls recorded in the woodlands?
Researchers placed automated acoustic recording devices on poles situated among trees across 310 designated woodlands throughout Britain over a three-year monitoring period.
Why is the Barbastelle bat difficult to track?
The Barbastelle roams ancient forests and possesses an elusive nature, which historically restricted monitoring efforts to confirming mere presence rather than establishing accurate population sizes or broader trends.
What role did automated systems play in verifying the data?
While open-source artificial intelligence tools processed millions of calls and flagged species across thousands of recording nights, automated classifications underwent expert manual verification to ensure overall accuracy.
Related reading