Abstract
Fish production is expected to grow significantly in the coming years. The welfare of captive animals in increasingly intensive systems has historically raised ethical concerns, suggesting that such intensification in fish production will drive public debate and inform policy discussions. Traditional assessments in fish welfare have relied on physical, physiological and/or behavioural measures of biological dysfunction, offering insights into health and functional integrities, yet these approaches face notable limitations. By contrast, affective-based approaches, which use behavioural measures to assess affective state and are widely employed in welfare science, remain underexplored in fishes. Recent advances, such as judgement bias paradigms, provide reliable tools to measure affective states. These support integrative welfare assessments that combine affective evaluations with measures of health and biological functioning. A multi-level approach ensures a comprehensive and robust evaluation of fish welfare, avoiding reliance on any single type of evidence and enabling the exploration of different facets of the welfare construct. While this work synthesizes existing indicators, its main contribution lies in proposing a functional framework that prioritizes affective state evaluation while systematically integrating and validating complementary metrics. By aligning with contemporary ethical and scientific standards, this approach aims to advance the conceptualization and operationalization of fish welfare.