Experimental arena size alters larval zebrafish photolocomotor behaviors and influences bioactivity responses to a model neurostimulant

Zebrafish behavior is increasingly common in biomedical and environmental studies of chemical bioactivity. Multiple experimental arena sizes have been used to measure photolocomotion in zebrafish depending on age, endpoints observed, and instrumentation, among other factors. However, the extent to which methodological parameters may influence naïve behavioral performance and detection of behavioral changes is poorly understood. Here we measured photolocomotion and behavioral profiles of naïve larval zebrafish across arena sizes. We then performed concentration response studies with the model neurostimulant caffeine, again across various arena dimensions. We found total swimming distance of unexposed fish to increase logarithmically with arena size, which as related to circumference, area, and volume. Photomotor response during light/dark transitions also increased with arena size. Following caffeine exposure, total distance travelled was significantly (p < 0.001) affected by well size, caffeine treatment (p < 0.001), and the interaction of these two experimental factors (p < 0.001). In addition, behavioral response profiles showed differences between 96 well plates and larger well sizes. Biphasic response, with stimulation at lower concentrations and refraction at the highest concentration, was observed in dark conditions for the 96 well size only, though almost no effects were identified in the light. However, swimming behavior was significantly (p < 0.1) altered in the highest studied caffeine treatment level in larger well sizes during both light and dark periods. Our results indicate zebrafish swim more in larger arenas and arena size influences behavioral response profiles to caffeine, though differences were mostly observed between very small and large arenas. Further, careful consideration should be given when choosing arena size, because small wells may lead to restriction, while larger wells may differentially reflect biologically relevant effects. These findings can improve comparability among experimental designs and demonstrates the importance of understanding confounding methodological variables.