Modeling oxygen consumption of the spotted steed larvae, Hemibarbus maculatus (Bleeker, 1871)

The oxygen consumption rate of fish is closely associated with biotic and abiotic factors. Dissolved oxygen levels may decline below the asphyxiation point into the zone of respiratory dependence. The Box–Behnken design was utilized to set up our experimental procedure in order to examine variations in oxygen consumption and asphyxiation point under the simultaneous influences of water temperature, rearing density, and body mass in the spotted steed larvae up to its juvenile stage. Results showed that water temperature, rearing density, and body mass had highly significant effects on oxygen consumption and asphyxiation point. The interaction between temperature and body mass was also significant. The models of oxygen consumption and asphyxiation point towards water temperature, rearing density, and body mass were established for the first time, with >98% of the variation in oxygen consumption and asphyxiation point accounted for by the significant model terms. The predictive R2 (coefficient of determination) values indicated that the models explained at least 89% and 94% of the variability in predicting new oxygen consumption and asphyxiation point observations, respectively. Model adequacy checking showed that the models were highly adequate, and diagnostics analyses demonstrated that several basic model assumptions were satisfied. The models developed in this study could be applied to monitor the oxygen consumption over the larviculture, contributing to refining the success rate of spotted steed seed production. •Three-factor predictive models of oxygen consumption rate and asphyxiation point were established using the Box-Cox design.•The models could be used for monitoring the variations in oxygen consumption of Hemibarbus maculatus larvae.•The models developed in this study contribute to refining the success rate of seed production of this species.