Acoustic Monitoring

Listening to the birds of University of Agricultural Sciences Veterinary campus

Over five days, an autonomous audio recorder captured the soundscape of the University of Agricultural Sciences Veterinary campus. Using a custom acoustic classifier that identifies birds from their calls, those recordings were turned into a picture of which species are present and when they are most vocal.

Site: University of Agricultural Sciences Veterinary campus · Period: 31 Jul – 05 Aug 2026 · Method: continuous passive acoustic recording, analysed in 3-second windows

The survey at a glance

The recorder gathered 3,675 one-minute clips — around 61 hours of audio. The classifier scanned every clip and returned 16,052 individual call detections spanning 45 species.

61Hours analysed

16,052Call detections

45Species

3,675Recordings

Which birds were heard

The campus hosts a mix of common urban residents and a few quieter specialists. Crows, parakeets, barbets and magpie-robins dominate the daytime chorus, while herons, kingfishers and waterhens work the nearby water bodies. Owls and nightjars add a nocturnal layer that visual surveys usually miss.

List of bird species detected at the site — Every species identified from the recordings, with scientific names.

When they call through the day

Grouping detections by the hour they were recorded reveals each species' daily rhythm. Most birds cluster around the morning chorus and taper through the afternoon — but the pattern also exposes the night shift: the Indian Scops-Owl lights up the pre-dawn and late-evening hours when almost everything else is silent, a signature only acoustic monitoring can capture.

Heatmap of bird activity by hour of day — Diel activity heatmap — brighter cells mean more detections in that hour. Rows are ordered by overall detection frequency.

Note: detection counts reflect vocal activity, not the number of individual birds. A single vocal species can produce many detections, and automated identifications are occasionally imperfect — but the broad patterns of presence and timing are robust.