Improvement of bioenergetics model predictions for fish undergoing compensatory growth

A previous evaluation of a bioenergetics model applied to juvenile hybrid sunfish (F₁ hybrid of female green sunfish Lepomis cyanellus X male bluegill L. macrochirus) undergoing compensatory growth (CG) indicated that the model substantially overestimated growth and underestimated cumulative consumption. This result suggested that fish bioenergetics models might not adequately account for physiological shifts that occur during CG. However, we demonstrate that application of a recently developed procedure for correcting consumption- and growth-rate-dependent systematic errors common among bioenergetics models negates much of the predictive error that had been attributed to the physiological complexities of CG. Correction equations for estimating the model-relative growth rate error (predicted less observed; g·g^-1·d^-1) from the observed mean daily consumption rate (g·g^-1·d^-1) and the consumption rate error (predicted less observed; g·g ^-1·d^-1) from the observed relative growth rate (g ·g^-1·d^-1) were derived by applying linear regression analysis to data from individual hybrid sunfish not undergoing CG. These independently generated correction equations significantly improved model predictions of growth and cumulative consumption for three groups of fish undergoing CG at one temperature near their growth optimum. The findings indicate that the high consumption and growth rates characteristic of fish undergoing CG merely amplify the consumption- and growth-rate-dependent errors inherent in bioenergetics models and that model predictions for fish undergoing CG can be significantly improved through application of the correction procedure.