A novel automated method for the simultaneous detection of breathing frequency and amplitude in zebrafish (Danio rerio) embryos and larvae

Stress responses of fish to disruption of oxygen homeostasis include adjusted oxygen consumption rate (MO₂) as well as the hyperventilation consisting of changes in breathing frequency (f_v) and amplitude (f_ampl). However, studying the HVR in very small organisms such as zebrafish (Danio rerio) embryos and larvae is challenging, and breathing movements (i.e., f_v) are usually manually counted, which is time- and human resource-intense, error-prone and does not provide information on the amplitude of breathing movements of the response, the breathing amplitude (f_ampl). Hence, in the present study, a new automated method was developed to simultaneously measure f_v and f_ampl in small zebrafish embryos and larvae with the computer software DanioScope™. To compare HVR strategies at different life-stages of zebrafish and the physiologically linked MO₂, hatched 4 d old embryos and early gill-breathing 12 d old larvae were treated with the HVR-inducing neurotoxic compound lindane (γ-hexachlorocyclohexane; γ-HCH) as a model substance. Comparison of manually counted f_v with f_v data measured by DanioScope™ at both life-stages showed high to moderate agreement between the two methods with respect to f_v in control fish and in fish treated with lower lindane concentrations (3 - 18% deviation at 25 µg/L γ-HCH). With increasing lindane concentrations (100 and 400 µg/L γ-HCH), however, manual counts showed an average underestimation of f_v by up to 30%, mainly due to very fast, rapidly successive, and indistinct movements of the fish, which cannot be properly detected by manual counts. Automated measurement thus proved significantly more sensitive, although several pre- and post-processing steps are needed. The improved automated detection of f_v and the first reliable estimation of f_ampl in small fish embryos and larvae, as well as the inclusion of MO₂, may provide new insights into different respiratory strategies and may, thus, represent a tool to lower the detection limit for reactions of different life-stages of fish to environmental stressors. In the present study, this became evident, as early gill-breathing 12 d old zebrafish larvae showed symptoms of respiratory failure (i.e., increase in f_v, f_ampl and MO₂, followed by subsequent lethargy) after exposure to lindane, whereas skin-breathing in 4 d old embryos proved mainly insensitive to the paralytic effects of lindane.