Expanded vision for the spatial distribution of Atlantic salmon in sea cages

Abstract

The environment in large aquaculture sea cages is never uniform. Abiotic factors such as oxygen level, water temperature, and ambient light, as well as farm operations such as feed distribution variate within a cage. Fish in cages look for the optimal conditions they can find, which results in higher stocking densities at specific areas. Understanding their spatial distribution patterns may help farmers to improve fish welfare, because unnormal distribution may imply sub-optimal living conditions, and the farmers can act for improvement. Good knowledge of fish distribution patterns in cages may also enable more efficient farm operations such as behaviour-based feeding. In the salmon farming sector, there has been a tremendous effort to understand fish behaviour by using diverse approaches. Much of the work was conducted in experimental cages which were smaller than the industrial scale sea cages commonly used nowadays, by using underwater cameras or single-beam echosounders. Every approach has its own pros and cons. Valuable knowledge about fish behaviour was brought by previous studies, but they had a disadvantage of seeing only a small proportion of fish in the entire cage. Consequently, their findings might be inapplicable to the industrial scale farming. Thus, there is a need for more representative information of fish behaviour in large cages, but we lack monitoring methods which can provide such information. The presented study aimed at filling this gap. The multi-beam sonar Wide Angle Sonar Seafloor Profiler, henceforth WASSP was used for the first time to understand fish distribution patterns in industrial aquaculture cages. The sonar had a 120° fan-shaped coverage, and it was rotated by the side wall of the cage to scan about 80% of the cage volume. The sonar data collected under three different farming conditions were used for 2D and 3D data visualization to test whether WASSP was able to capture changes in fish distribution patterns. The 3D distribution patterns captured by the sonar was validated by the optical data obtained by underwater cameras, which highlighted the suitability of the multi-beam sonar to monitor fish distribution patterns over the entire cage. Although the tested method exhibited some challenges such as noisy data near the water surface, it expanded our “vision” for the fish distribution in aquaculture cages. This innovative pilot study opened the possibility for continuous monitoring of 3D fish distribution in commercial scale cages, which will be focused in our future studies.