Darker Means More Dangerous? Understanding Fish Mentality Through Color: Visualizing Stress in Amazon's Star Tambaqui with AI

A research team from São Paulo State University (UNESP) and EMBRAPA has developed an AI tool to determine the stress levels of tambaqui (Colossoma macropomum), a major aquaculture fish native to the Amazon, based on the "shades of body color." The tool extracts the color of the lower body (counter-shading) from photographs and quantifies stress using a threshold based on the black/white pixel ratio. It was trained on image data from 3,780 fish, showing that stress resistance has a "medium to high degree of heritability." Experiments involving α-MSH hormone stimulation and confinement in small tanks demonstrated that body color significantly darkens under stress. In 2022, Brazil's tambaqui production reached 110,000 tons, and the AI tool is likely to contribute to welfare improvement and selective breeding (selection of stress-resistant strains). Since its release, the tool has gained attention in scientific news media and aquaculture communities, with expectations for "continuous monitoring using cameras and AI." However, challenges such as standardizing light conditions and photographic environments for implementation have been noted.