Atlantic salmon (Salmo salar) distribution patterns related to the environment in commercial sea cages

Abstract

Studies have shown that Atlantic salmon in captivity will alter their vertical distribution in sea cages based on thermal gradients, though how they respond to small changes in temperature and how their horizontal distribution is characterised is less well known. Here, a study was performed at three marine cage farms using a computer vision AI algorithm that automatically analyses fish group swimming behaviour in real-time. Multiple video cameras were deployed within one cage in each farm (5 in Farms A and B, and 4 in Farm C), for high spatial coverage of fish behaviour and distribution throughout the cage. The algorithm processed video footage from these cameras and outputted behavioural data termed ‘activity’, which encompasses fish abundance, speed, and shoal cohesion. This dataset includes this activity data along with the date/time, depth of the camera, and temperature at the camera's location. Additionally, environmental data such as photoperiod, weather, and current speed are also included for context.