Heart rate responsiveness as a predictor of weight change in farmed Atlantic salmon. Advancements in welfare monitoring.

Abstract

The aquaculture industry is rapidly expanding with an increasing demand of fish for human consumption, and this comes with challenges. Early and accurate predictions of stress in farmed fish is critical for improving welfare and economic return. Further advancements are needed in the sector to provide sufficient welfare monitoring, especially in large scale systems. This study was conducted on Atlantic salmon (Salmo salar), a promising species for recirculating aquaculture systems in Sweden, and examined the individual and combined effects of operational welfare indicators (both pre-existent and a manually inflicted fin clip) and simulated stress events common in aquaculture practices (i.e., crowding, fasting and anthropogenic disturbance) on physiological and behavioural stress indicators. The data was collected over three trials, each 14 days long, on a total of 30 individuals with surgically implanted biologgers measuring heart rate and swimming activity via tri-axial accelerometry. Heart rate was found to be an effective measurement to predict weight change; individuals who lost the most weight responded with a significantly lower heart rate compared to fish with a larger appetite. This indicates a need for stimulation in order to differentiate them. While accelerometry measurements showed less obvious effects, similar trends were observed, suggesting potential for future welfare monitoring purposes. These indicators were also correlated to behavioural observations of feeding motivation and activity from video recordings. Lastly, heart rate and accelerometry did not manage to show differences in fish based on fin clipping. Reliably connecting physiological cues to camera footage could form the basis for automatic, AI-based systems for future welfare assessment in fish, ultimately removing the need for manual evaluation.