The Effects of Social vs. Individual Housing of Zebrafish on Whole-Body Cortisol and Behavior in Two Tests of Anxiety

Two of the most used models of anxiety in zebrafish research, the novel tank-diving test (NTDT) and the black-white preference test (BWPT), are modifications of assays used in rodent research (open field test and light/dark test). There has been a thorough validation of these tests in rodents, but a similar level of knowledge is still missing in zebrafish. Adult zebrafish naturally live in shoals with conspecifics, and group housing is therefore assumed to be the optimal housing condition for zebrafish, as it allows for shoaling behavior. This study investigated how housing in social isolation affected whole-body cortisol and the behavioral responses in the NTDT and BWPT. We also examined the correlation between the behavioral responses in the two behavioral tests. We found that zebrafish housed in groups had significantly higher whole-body cortisol than individually housed zebrafish (F-1,F- 85 = 25.51, P < 0.0001). Regardless of treatment, all groups had a general preference for the lower compartment in the NTDT and the black compartment in the BWPT. Individually housed zebrafish had a higher total number of entries to the white compartment in BWPT compared to group housed zebrafish when their first test was BWPT (F-1,F- 48 = 5.79, P = 0.0201), but not when BWPT was their second behavioral test. Fish that had higher whole-body cortisol had a tendency toward fewer entries into the white compartment the first 3 min of the BWPT (F-1,F- 48 = 3.90, P = 0.0540). There was no effect of housing on the behaviors registered in the NTDT. There was a positive correlation (correlation coefficient 0.40; p = 0.003) between transitions from black to white compartment in BWPT and transitions from lower to upper compartment in NTDT, but we did not find any association between duration in white compartment in BWPT and upper compartment in NTDT. Considering this, we suggest that further model validation is needed.