Let's talk: editor@tmv.in
IIT Mandi Develops AI Model To Predict Wildlife Population Changes

IIT Mandi Develops AI Model To Predict Wildlife Population Changes

Saikiran Y
November 22, 2026

Researchers at the Indian Institute of Technology (IIT) Mandi have developed an artificial intelligence (AI)-based mathematical modelling framework to predict changes in wildlife populations, potentially providing scientists with a new tool to understand predator-prey relationships and forecast ecological trends up to five years in advance. The model combines physics-informed neural networks and transfer learning to analyse historical population data and estimate how different species influence each other's survival and growth, even when environmental information is limited or uncertain.

The research focuses on wolves and moose in Isle Royale National Park, a 544-square-kilometre island in Lake Superior, Michigan, USA. Wolves are the primary natural predators of moose, which account for nearly 90 per cent of their winter diet. Their populations have been monitored annually since 1959, creating one of the world's longest-running predator-prey records.

The study was conducted by Dr Nitu Kumari, Professor at IIT Mandi's School of Mathematical and Statistical Sciences, and her PhD student Anurag Singh. Their model estimates moose reproduction rates, natural changes in wolf populations and the impact of predation. It also identifies unusual population declines in historical records.

The researchers tested two approaches to understand wolf hunting patterns and found that the model examining the relationship between predator and prey numbers performed better. For 2020-2024, it predicted increasing wolf numbers and declining moose populations, broadly matching subsequently observed trends.

Physics-informed neural networks incorporate scientific equations into AI learning, while transfer learning enables knowledge from existing datasets to improve predictions. Their combination helps researchers model ecological changes despite gaps and uncertainties in historical observations.

Isle Royale has experienced significant wildlife fluctuations over decades. Wolf numbers fell sharply in the early 1980s following a canine parvovirus outbreak, while moose populations declined during the mid-1990s. Climate change has also reduced ice coverage in Lake Superior, restricting natural movement between the island and mainland and increasing concerns about genetic isolation. The US National Park Service relocated 19 wolves in 2018 and 2019 to strengthen the island's population.

The research builds on the team's 2025 study published in Scientific Reports, which used Sparse Identification of Nonlinear Dynamics (SINDy) to examine wolf-moose interactions. The latest framework advances this work using self-adaptive AI modelling.

According to Professor Kumari, this is the first study, to her knowledge, combining physics-informed neural networks and transfer learning in a self-adaptive framework to investigate predator-prey population fluctuations. The findings could eventually support wildlife monitoring and conservation planning, although further validation across different ecosystems is necessary. IIT Mandi said Kumari recently discussed the research with US-based news outlet The Primary.

IIT Mandi Develops AI Model To Predict Wildlife Population Changes - The Morning Voice