Modelling growth of brown trout, Salmo trutta, in terms of weight and energy units

1. The chief objectives were: (i) to compare two growth models, one based o n weight and the other on energy, using the same data set for the analyses; (ii) to discover if weight and energy units can be simply interchanged for growth assessment. The data set was for 183 brown trout, Salmo trutta (liv e weight 1-300 g), fed to satiation on shrimps, Gammarus pulex, and grown i ndividually over 42 days at constant temperatures (range 3.8-20.4 degreesC) . 2. Rates of change in weight or energy content, and final weight or energy content at the end of 42 days growth, were estimated from the models and we re excellent fits to the experimental data (P < 0.001). The shape of the te mperature relationship for rates of change or final values was triangular f or the weight model and curvilinear for the energetics model. Optimum tempe ratures for growth according to the weight and energetics models were 13.1 and 13.9 degreesC, respectively, for rates of change and 13.1 and 13.5 degr eesC, respectively, for final values. When the growth period was extended t o 100 and then 300 days, the triangular relationship and optimum temperatur e remained the same for the weight model, but the curvilinear relationship became more triangular for the energetics model and the optimum temperature identical to that in the weight model. The relationship between gross effi ciency and temperature also differed in shape between the two models but ma ximum efficiencies occurred at a similar value of 9 +/- 0.1 degreesC (18 an d 32% for weight and energetics models). As fish weight increased, gross ef ficiency remained constant in terms of energy units, but decreased markedly in terms of weight. 3. These comparisons showed that different conclusions can be drawn from th e two models, even if the same data set was analysed. There was a close rel ationship between initial wet weight and energy content for stock trout use d in the experiments, but the relationship was not so close at the end of t he experiments, and interchangeability of units could no longer be assumed. A variable error, often as high as 10-12%, would occur if the relationship for initial values was used to predict one unit from the other. Therefore, weight and energy units cannot be simply interchanged for growth assessmen t, especially in comparisons for trout of different sizes.