A comparison of the light/dark and novel tank tests in zebrafish

The recent introduction of tasks to assess the behavior of zebrafish in novel and/or aversive environments has spurred great interest, prompting attempts to determine which constructs are modeled by these tasks (e.g., fear, anxiety, or some other construct). A review of the pharmacological and behavioral experiments indicates that not all behavioral testing models are equivalent. A more precise understanding of the parameters that influence task performance affords a wider selection of experimental procedures for investigating a particular construct, and also provides tools for differentiating the various constructs that may ultimately be of interest. In this review we will more closely examine two behavioral assays commonly used to measure the construct of 'anxiety' in adult zebrafish, with the conclusion that they do not both appear to be measuring a single underlying state.