AI turns millions of bat calls into a map for woodland recovery

Conservation teams can now see how bats use England’s woodlands at a scale that manual listening could never reach. The National Woodland Bat Survey collected 9,734 nights of sound from 310 sites, then used artificial intelligence to identify more than 98 million calls.

The response combines machines with people. Open-source tools sorted roughly 90,000 hours of recordings, while 19 trained verifiers checked a sample. The Bat Conservation Trust says every British bat species or species group except the hard-to-identify serotine appeared; common pipistrelles were detected at every site.

It is a baseline, rather than proof that populations are recovering. Recorders also detect activity, not exact animal numbers, and rare calls still need expert judgement. Even so, the survey reveals where threatened species are active and where records were previously sparse. That gives woodland managers a practical national evidence map for protecting habitat and measuring whether nature recovery works over time.

Sources

• Forest Research — AI helps reveal Britain’s bats in largest-ever national woodland bat survey
• Bat Conservation Trust — AI reveals 98 million bat calls in Britain's largest woodland bat survey