AI Tracks Painted Stork Nest Fidelity in Delhi Zoo

Researchers in India have demonstrated the use of artificial intelligence to monitor nest site fidelity in painted storks at the National Zoological Park, Delhi. The study focused on a single male stork, informally named “Ringo”, tracked over four breeding seasons from 2022 to 2025. The work highlights how non-invasive digital tools can advance behavioural ecology studies without disturbing wildlife.

Study Focus and Species Behaviour

Painted storks (Mycteria leucocephala) are colonial waterbirds known for nesting in large groups. Nest site fidelity refers to the tendency of birds to return to the same nesting site across breeding seasons. Understanding this behaviour is important for conservation planning, especially in urban or semi-captive habitats like zoological parks.

Data Collection and Image Dataset

The researchers selected a male stork with a distinctive scar on its neck for individual tracking. A total of 2,349 high-resolution images of the bird were collected, capturing both sides and folded wing patterns. Additionally, 1,755 images of other storks were compiled to create a comparative dataset. These images, gathered over four years, enabled precise identification based on visual features.

Use of AI and Computer Vision

The study applied deep transfer learning (DTL), a machine learning technique that adapts pre-trained models for new tasks. Researchers also used the scale-invariant feature transform (SIFT) algorithm to extract unique image features. The scar on the stork’s neck and distinct feather patterns acted as biological markers. The AI model successfully identified the bird with 98% accuracy, confirming repeated use of the same nesting site.

Important Facts for Exams

• Painted stork (Mycteria leucocephala) is a large wading bird native to South Asia.
• Deep Transfer Learning (DTL) enables reuse of pre-trained AI models for specialised tasks.
• SIFT is a computer vision algorithm used for feature detection and image matching.
• Royal Society Open Science is a peer-reviewed open-access scientific journal.

Significance for Wildlife Monitoring

The findings demonstrate that AI-based, non-invasive techniques can reliably monitor individual animals over long periods. Such methods reduce the need for tagging or physical interference, making them suitable for sensitive species. The study underscores the growing role of pattern recognition technologies in ecological research and conservation of colonial waterbirds.