98 million bat calls could reveal which English woods need closer protection

For someone deciding how to manage a woodland, knowing that a vulnerable bat uses it could change what needs protecting. An animal missed by earlier surveys may still depend on that habitat. That is the practical promise of an enormous listening exercise across England’s woods: to reveal places whose importance has been overlooked, not simply to collect more wildlife sounds.

The National Woodland Bat Survey has amassed more than 98 million bat calls from 310 woodland sites, with artificial intelligence helping researchers detect and identify species in 217 terabytes of recordings. Forest Research and University College London announced those totals on October 30, 2026. The scale is impressive; the conservation test is what researchers can reliably establish from it. ^{10 9}

BBC News reports that barbastelles have been detected in areas where they had not previously been recorded, especially in south-west England. The Bat Conservation Trust classifies the barbastelle, Barbastella barbastellus, as Vulnerable in both Great Britain and England. A confirmed record of this rare species could give conservationists a reason to look more closely at a woodland. ^{1 18}

But the individual woodlands behind those reported discoveries have not been identified in the available evidence. The primary project information establishes the survey’s overall scale without independently confirming the concentration of newly recorded barbastelles in the south-west. That geographical finding should therefore remain attributed to the BBC, rather than presented as a detailed, published map of the species’ whereabouts. ^{1 10}

The distinction between discovering a bat and demonstrating a change in its fortunes matters. A first record tells researchers that the animal was detected there. It does not tell them that it has only recently arrived. More effective listening can uncover wildlife that previous surveys missed.

Nor does a new detection alone establish breeding, an expanding range or a growing population. Those are different questions requiring additional evidence. The survey’s reported discoveries are promising leads about habitat use, not proof that barbastelles are recovering.

The headline call total needs similar care. An individual bat can call repeatedly, so more than 98 million calls cannot be translated into more than 98 million animals. Acoustic records help researchers investigate activity and identify species; they are not, by themselves, a census.

The precise project totals also fit within the BBC’s initial description of more than 90 million calls from more than 300 woodlands. Those were lower bounds, not conflicting counts. And although the project’s name and some announcements use national or British framing, these 310 sites are a sample of English woodland—not equivalent coverage across the United Kingdom, and not every woodland in England. ^{1 10 14}

The BBC describes devices recording over three years. The exact beginning and end of that period have not been established. Separately, the project’s data catalogue says the partnership has operated since 2021; that date should not be mistaken for a confirmed start to the three-year recording period. ^{12 14}

Forest Research and the Bat Conservation Trust lead the project, funded through the Department for Environment, Food and Rural Affairs’ Natural Capital and Ecosystem Assessment programme. University College London, known as UCL, is also involved. At this volume, analysing the audio entirely by hand would be extremely laborious. Automated processing makes the archive more practical to use. ^{3 9 12}

The recordings are processed through the Bat Conservation Trust’s Sound Classification System. AI detects bat sounds and helps classify them by species using patterns in spectrograms—pictures showing sound across frequency and time. The available documentation does not establish the precise underlying classifier or software version, so there is no basis here for attaching a particular model name or an accuracy percentage to the results. ^{8 1}

The consequential step is verification: separating a machine-generated label from a checked biological record. For a rare species, that is not a technical footnote. Whether a woodland appears to support a vulnerable bat may depend on whether researchers can confirm that the recorded call really belongs to it.

The project’s 2024 example dataset explicitly contains unvalidated automated bat echolocation-call classifications. That description applies to the example; it does not establish that every finding from the wider survey is unvalidated. A separate dataset is catalogued as Verified Bat Activity. Equally, the existence of those checked records does not mean that all 98 million calls have been individually verified. The survey-wide share verified remains unclear. ^{3 14}

This is where the enormous total becomes less important than the quality of particular records. A clearly documented detection tied to a particular woodland can inform a habitat assessment. A vast quantity of uncertain labels cannot do the same job merely by being vast.

Forest Research says all UK bats depend on woodland for feeding, roosting or travelling between places. The survey therefore has potential value beyond rare-species discoveries: a better picture of which bats use different woods could help identify important habitat and guide future management. ¹⁶

What has not yet been demonstrated is a specific management change, a new protected-site designation or a measured improvement in bat populations resulting from this survey. Its immediate achievement is a substantial increase in the capacity to listen. Turning that into conservation requires a clear connection between verified species records, the woods those animals use and decisions about those woods.

The next useful details are consequently concrete: which woodlands contain the newly recorded barbastelles, how those detections were checked and what managers will do differently. Those answers will matter more to the bats than another impressive total.