Biological criteria for submergence of physostome (Atlantic salmon) and physoclist (Atlantic cod) fish in sea-cages

Abstract

This thesis aims to describe behavioural responses and welfare parameters for the physostome Atlantic salmon and physoclist Atlantic cod when out of neutral buoyancy at different degrees and periods of time. The background for this approach is to improve the culture conditions by fully submergence of the farming installations, as the water below 10 m depth often is more stable in terms of environmental factors such as temperature and current in addition to the lack of waves, which again opens for alternative sites in more exposed oceanic areas. Salmon were submerged in large-scale fully submersible cages at depths between 4 to 15 m for 22 days (Paper I) and 10 to 25 m for 42 days (Paper II) under different light conditions and at various times of year. The behaviour of individual salmon in a school under submerged conditions was studied to reveal whether a range of coping abilities among individuals exists during the new and more challenging conditions (Paper III). Atlantic cod in an experimental submersible cage were raised from five different starting depths (between 30 and 8 m) and lowered from surface position to 10 m, 20 m and 30 m to test a protocol for safe lifting and lowering steps. Based on the behavioural responses, safe acclimation times before the next vertical step at high and low sea temperatures were identified (Paper IV). In Papers I and II, the general patterns of swimming depth and schooling density were studied at group level using echo-sounders in addition to swimming speed and swimming angle based on instantaneous observations with underwater cameras. Welfare parameters were defined as weight gain, feed intake, feed utilisation and fin and vertebral condition during the experimental period. The behaviour of individual salmon was studied by monitoring the swimming depth and body temperature using data storage tags implanted in randomly selected fish. Whether the development of diel vertical migration (DVM) activity during feeding was linked with individual growth rates was analysed to study individual coping styles. The immediate response and recovery time after lifting and lowering of Atlantic cod were estimated through measurements of swimming speed, swimming tail beat rate and swimming angle based on underwater camera recordings, and additionally by echo-sounder data obtained from resting fish on the net-floor (Paper IV). Atlantic salmon submerged without the opportunity to refill their swim bladder lost gas steadily over time, and the bladder was empty after about three weeks (Papers I and II). Swimming speeds were elevated on the first day after submergence and schooling became more structured (more constant speed and a greater distance to neighbouring fish). The diurnal vertical swimming pattern for the salmon in spring was broadly similar to that observed prior to submergence, as the artificial underwater lighting allowed the salmon to keep high swimming speed and lift during night (Paper III). In contrast, about 90% of the salmon submerged below 10 m during the dark winter reversed their diurnal pattern from swimming at shallow depths at night and deeper during the day to swimming deep at night and shallower during the day (reversed DVM) or to swimming with a normal or reversed diurnal pattern on different days (irregular). A separation of faster- and slower-growing salmon also gradually appeared, where the faster-growing individuals occupied the deeper part of the water column. This situation occurred also at night at the end of the experiment among the salmon given continuous light (Paper III). Growth was clearly reduced in the salmon that were submerged during the winter, with more injuries recorded on their fins and snout. Slightly compressed vertebrae in the tail region were also observed, probably due to their tilted head-up tail-down swimming angle during the dark nights (Paper II). Lifting farmed cod from five different start depths, equivalent to a 40% pressure reduction, resulted in strong downwards swimming movements dependent on the water temperature (Paper IV). The depth before cage lifting affected the immediate response, as the fish became more active after lifting events from shallow compared to deeper depths. Appetite decreased after lifting, but loss of behavioural control was never observed. During the subsequent 8-10 hr recovery periods, swimming activity gradually decreased to the same level as before lifting. The overall recovery time did not depend on start depth or temperature. Independent of final depth or temperature, rapid lowering of cod only resulted in a moderate short-term increase in upwards swimming movements, while appetite was less affected than after lifting. A compressed swim bladder after descents from the surface to 10-30 m leads to negative buoyancy, which required 18-90 h to re-fill by gas secretion, which is a temperature-related process (Paper IV). In conclusion, this thesis demonstrates that air gulping is a behavioural need for Atlantic salmon, and that long term denial of surface access will reduce their welfare Atlantic cod cope well in fully submerged cages, but 40% pressure reduction is near the upper limit for lifts of healthy farmed cod. Secondary lifts should not be done until at least 10 h after the first lift. Cage lowering should be done slowly to avoid potentially stressful crowding of negatively buoyant fish on the cage bottom, especially at low temperatures.