Zebrafish are becoming more common as subjects in neuroscience studies of behavior and in screens for new treatments of behavioral disorders such as addiction, attention deficits and autism spectrum disorders. The fish are small and reproduce prolifically, making it easy and relatively affordable to house large numbers of them in the laboratory. In addition, they are amenable to genetic manipulation to study the effects of specific genes on behavior.

A behavior test commonly used to assess stress levels in zebrafish is the tank diving test. A fish placed in a new tank typically dives to the bottom of the tank and hovers there, stationary, for some time before rising. The amount of time the fish spends at the bottom is considered to vary in proportion with its stress level: a longer interval reflects greater stress. Such tests may be used to analyze the efficacy of potential new treatments for anxiety disorders.

In other lab animal species such as rodents, housing and handling procedures are known to affect the results of behavioral experiments, but few studies have examined how husbandry practices may affect results of zebrafish experiments. Caroline H. Brennan and colleagues at Queen Mary University of London (UK) predicted that zebrafish housing environments would affect their stress levels and, therefore, the results of tank diving assays. Overall, they found that individually housed fish exhibited different behaviors on the tank diving test than did group-housed fish (PLoS ONE 7, e34992; 2012). They spent less time on the bottom of the tank than did group-housed fish. Individually housed fish also responded to treatment with ethanol, which has anti-anxiety effects, whereas group-housed fish did not. Finally, levels of the stress hormone cortisol were lower in individually housed fish.

The group concludes that a zebrafish's tank diving response is related specifically to the degree of change from the environment from which the fish had come. This means that housing is an important factor in obtaining reliable data from behavioral tests in zebrafish. Brennan stated in a press release, “Not only will publication of our results improve the reliability of zebrafish behavioural analysis, but, by demonstrating that by refining housing one can increase the power of our analysis and reduce the number of animals used, we contribute to the 3Rs aim of UK and international science policy—to reduce, refine and replace animals in research.”