Despite its popularity in research, there is very little scientifically validated knowledge about the best practices on zebrafish (Danio rerio) husbandry, which has led to several facilities having their own husbandry protocols. This study was performed to expand knowledge on the effects of enrichment and fish density on the welfare of zebrafish, with hopes of providing a scientific basis for future recommendations and legislations. Zebrafish were reared at three different stocking densities, (1, 3 or 6 fish/L), in tanks with or without environmental enrichment. Agonistic behavior was observed twice a week for 9 weeks directly in the housing tanks. Aspects of welfare is known to be reflected in neuroendocrine stress responses. Thus, cortisol secretion in response to lowering the water level was analyzed for each group. In addition, we assessed cortisol secretion in response to confinement and risk-taking behavior (boldness) using the novel tank diving test for individual fish. At termination of the experiment fish were subjected to stress by transfer to a novel environment and brain tissue was sampled for analysis of brain monoaminergic activity. Fish kept at the lowest density (1 fish/L) showed a significantly higher level of aggression than fish kept at 3 or 6 fish/L. Moreover, fish kept at this low density showed significantly higher cortisol secretion on a group level than fish kept at the higher stocking densities, when subjected to lowering of the water level. Keeping fish at 1 fish/L also had effects on brain monoamines, these fish showing higher brain dopamine concentrations but lower dopamine turnover than fish kept at higher densities. Neither stocking density or enrichment had any clear effects on the behavior of individual fish in the novel tank diving test. However, fish kept at high densities showed lower and more variable growth rates than fish kept at 1 fish/L. Taken together these results suggest that zebrafish should not be kept at a density of 1 fish/L. The optimal stocking density is likely to be in the range of 3-6 fish/L.
Read full abstract