AI-assisted survey maps rare bats across English woodlands
The story at a glance
The National Woodland Bat Survey used automated classification and expert checks to analyse acoustic recordings from woodland sites across England. Forest Research reports more than 98 million calls recorded between 2021 and 2024. 12
The survey recorded rare and threatened bats, including barbastelle and grey long-eared bat. Its findings offer a new monitoring baseline, but do not show that bat populations have recovered. 13
What you need to know
- Forest Research reports that the survey covered 310 English woodland sites, producing more than 98 million calls and 217 terabytes of audio. An earlier technical document gives different totals, so the figures should be read alongside that unresolved discrepancy. 14
- The recordings included the UK Red List species barbastelle and grey long-eared bat, as well as near-threatened Nathusius’ pipistrelle and Leisler’s bat. The serotine was not detected, although the release says it is particularly difficult to identify acoustically. 1
- The survey reported barbastelle activity hotspots across southern England, with detections in parts of Cornwall, Devon and Dorset where records had been sparse. Common and soprano pipistrelle activity was generally highest in western regions. 1
- Automated BatDetect2 classifications were supplemented by manual review of species presence at each NFI square. A sample of calls was checked further to adjust activity indices for classification error, and the government data record advises users to consult supporting limitations documentation. 2
- Forest Research says the survey provides national-scale monitoring away from known roosts for seven listed bat species. It is also the only active national-scale monitoring in England for barbastelle, Leisler’s bat and Nathusius’ pipistrelle. 1
Why this matters
Acoustic monitoring at this scale adds information about bat presence and activity beyond known roosts, including for species with sparse previous records. That can strengthen the evidence available for future monitoring, but the survey is a baseline rather than a measure of population recovery. 13
What happens next
Repeated survey cycles will be needed to establish whether the baseline can reveal trends over time. Users also need to consult the limitations documentation, particularly when interpreting non-detections and acoustic classifications. 23
What we know
- The survey covered English woodland from 2021 to 2024 and recorded more than 98 million calls, according to the latest Forest Research release. 1
- The dataset combines automated call classifications with manual review and further checks on a sample of calls. 2
- The published findings include records of rare and threatened species and describe regional patterns in bat activity. 1
What we don't know
- The reviewed sources do not reconcile the latest report of 310 sites and more than 98 million calls with an earlier document reporting 319 unique NFI squares and more than 93 million calls. 14
- It is not yet clear how many repeated survey cycles will be needed before the baseline can establish population trends. 3
- The sources do not establish how non-detections should be interpreted for each species, particularly where acoustic identification is difficult. 12
Read the reporting
This WildNews briefing synthesises and summarises reporting from the sources below. WildNews did not produce the original reporting. Learn more about how we use AI
Was this story useful?
Your anonymous feedback helps us make future WildNews briefings more useful.