Unlocking global fish ecological trait knowledge through Artificial Intelligence

This project will use artificial intelligence to read information about how fish live directly from their images. By training AI models on global databases of fish images, body-shape information and climate data, we will predict important ecological traits such as where species live (rivers, lakes), what they eat, how far they move, and how they care for their offspring. Having this information is essential to understand how they might respond to threats such as climate change, habitat loss and overfishing, and how we can better protect them. The results will be an AI-based model that can predict the ecological traits of fish, likewise, global maps and open code that make this trait information easier to use for studying biodiversity and supporting conservation and fisheries management. For Dalhousie University, the project will provide new trait data and methods that strengthen its research on fish biodiversity and environmental change. For the Industrial University of Santander, it will build local capacity to apply AI in biodiversity research, create datasets, and deepen a long-term collaboration with Macroecology & Biodiversity Change Lab in Canada.

Faculty Supervisor:

Juan Carvajal-Quintero

Student:

Partner:

Universidad Industrial de Santander

Discipline:

Life Sciences

Sector:

Education

University:

Dalhousie University

Program: