Modelling growth of Colossoma macropomum (Cuvier): comparison of an empirical and an explanatory model

Accumulated data from growth experiments on the nutrition of Colossoma macropomum (Cuvier) were analysed using an empirical (linear regression) and an explanatory model called the fish growth simulator (FGS). Both methods gave satisfactory results. Differences between predicted and observed values fluctuated less with the FGS. However, the linear regression method is simple and results are easily translated to practical recommendations. The FGS covered a wider range of experimental conditions than the linear regression method. Both linear regression and the FGS stressed the importance of dietary protein in growth of C, macropomum. Dietary protein concentration is the main explaining variable of ad libitum feed uptake, while protein ration is the main variable determining growth. Dietary protein concentration also affects body protein and body lipid content. Protein utilization efficiency data indicate that substitution of fishmeal by soya meal increases dietary protein quality. Although C, macropomum can thrive under less than optimal conditions, a high-quality, high-protein diet is required for maximal growth and optimal feed utilization.