MONITORING EARLY-STAGE BEHAVIOR OF ATLANTIC SALMON Salmo salar USING A “PING360” SONAR: WHAT HAPPENS TO
Atlantic salmon (Salmo salar) aquaculture production in Norway focuses increasingly on better fish welfare and more sustainable practices. Understanding fish distribution in sea- cages is necessary for optimal farm management, e.g. to feed them adequately, and to better control welfare over time. The task is difficult because of the size of sea-cages on a commercial scale (60 m to 240 m circumference) and the number of individuals per cage (up to 200,000 in Norway). The Ping360 (Blue Robotics Inc) , a relatively low-cost single beam mechanical scanning imaging sonar, was successfully used in a n aquaculture context to study avoidance behavior of salmon towards intrusive objects in cages . We assumed it also had potential to study fish distribution in a commercial sea-cage if placed to scan vertically. Indeed, the Ping360 enabled us to scan of a complete section of a sea-cage, when most of the sonars and echosounders used in aquaculture nowadays have a limited coverage and are extremely costly.
We followed post-smolt salmon (123,000 individuals, 500 g on average) distribution during a sensitive period of the production: their transfer at sea from the land-based facility to a commercial sea-cage (18 500 m3) . The objective was to observe the evolution of their behavior (feeding, diurnal pattern, schooling) with time when adapting to a new environment after the stress of the transport . We collected data with the Ping360 continuously for 1.5 months. It was placed vertically in the center of the cage at 7 m depth . Single 360° scan images were extracted via Python (v 3.8.5) and edited into videos for manual observations and registration of behavioral patterns. Average scans with one hour of data compiled were also computed for easier comparison between days (fig 1).
The results are still under interpretation but having the cage outline visible on the results allows for a great overview of the vertical and horizontal distribution of the salmon in the cage . Our methodology also has great potential to assess cage deformation due to currents. The first week, salmon mainly stayed in the upper part of the cage ( 0-7 m depth) and no significant change was observed between day and night. After 3 days a diurnal pattern became clearer. After a week, evident schooling patterns appeared for some hours of the day and fish feeding organization became more structured. S almon distribution will be correlated to environmental parameters (T°C, DO, current, salinity, light).