Reproducibility and replicability remain a challenge in behavioural neuroscience, including zebrafish research. One source of this issue may stem from differences in how we raise and keep these fish in the laboratory. While there are efforts to standardize husbandry conditions, one potentially important source of variability is often ignored: human-animal interaction. In this opinion piece, we argue that routine husbandry procedures such as netting, feeding, tank cleaning and water changes can constitute repeated acute stressors that can lead to chronic alteration of zebrafish physiology and behaviour. These human-induced stress effects are known to interact with experimental treatments in other species, and thus are expected to confound results obtained with zebrafish too. Furthermore, timing of maintenance tasks, including human handling, and other temporal factors can introduce systematic facility-, study- or handler-specific effects. We suggest that these unreported human-animal interactions and unstandardized temporal maintenance patterns represent hidden variables that can undermine reproducibility of data, and argue that recognition and documentation of these factors are necessary for improving the reliability of zebrafish research.
Tsang et al. (Wed,) studied this question.