Effects of stocking density on the welfare of farmed rainbow trout (Oncorhynchus mykiss)

Abstract

The welfare of farmed fish is a subject of growing public, commercial and governmental interest. The Farm Animal Welfare Council’s report on the welfare of farmed fish (Anon., 1996a) highlighted stocking densities used in intensive fish production as a major welfare concern. This thesis investigates links between stocking density and welfare of rainbow trout (Oncorhynchus mykiss, Walbaum, 1972). The effects of stocking density and water quality deterioration on the welfare of rainbow trout were assessed in two controlled experiments. Fish welfare was assessed by measuring a range of morphometric (growth, fin condition), physiological (haematocrit, plasma cortisol and glucose) and immune response (lysozyme activity) indicators. Principal Components Analysis (PCA) was used to generate welfare indices based on coherence that existed between the individual welfare indicators. The first experiment stocked different numbers of fish into the tanks at the same inflow rate (60 1 min-1) to achieve stocking densities of 10, 40, and 80 kg m-3. Results suggested that, provided good water quality was maintained, stocking densities around 80 kg m-3 did not produce consistent negative effects on growth rate, stress response or immunological indicators of welfare. Fin erosion increased with increasing density, although the exact cause of the erosion remains unclear. Increased size variation and elevated cortisol levels in the 10 kg m-3 treatment, possibly linked to dominance hierarchies, indicated that low, as well as high, stocking densities may have a detrimental effect on trout welfare. The second experiment assessed the effect of water quality deterioration by adjusting inflow rate (20, 40 or 60 1 min-1) in tanks containing the same initial stocking density of fish (16 kg m-3). Inflow rate affected growth during the 3 month summer period when water temperatures were highest (≈14°C), with significantly better growth observed in the 60 1 min-1 treatment. There was no significant effect of inflow rate on cortisol, haematocrit or fin erosion, but higher mortality and poorer body condition were observed in the 20 1 min-1 treatment. A questionnaire survey of stocking density practices on UK trout farms found marked differences in the stocking practices and the perception of a high SD when comparing farms producing trout exclusively for the table market with fisheries and restocking farms. A lack of accurate flow rate data from respondents highlighted the difficulties of trying to apply alternative methods of quantifying stocking density rather than the conventional unit of kg m-3. On-farm work successfully applied the system of welfare assessment in a range of selected commercial systems and confirmed some of the findings from the tank based experiments and questionnaire survey. In summary, increased stocking density resulted in increased fin erosion although there was also the suggestion that there are welfare implications at low as well as high SD. Systems applying high SD or loading rates may run an increased risk of mass mortality in the event of system failure, necessitating the need for increased supervision and appropriate back-up systems. A universally applied SD limit will not guarantee good fish welfare and it is suggested that defining limits of key water quality parameters may prove to be more effective. Further work is required to establish thresholds of water quality parameters to safeguard fish welfare.