The objective of this study was to compare the growth of rainbow trout of different colorations by Gompertz model. The experiment was carried out at Instituto de Pesca, Campos do Jordão, Brazil. The wild, blue, yellow-albino, and white-albino strais were obtained by crossing white trout with wild ones. The 3.07±0.14 g fingerlings were cultivated in triplicate in 2 m³ circular tanks in an open system with constant water flow. The minimum water temperature during the period being 13.55±1.89°C and the maximum being 15.64±1.56°C. Dissolved oxygen was 6.75±0.52 mg/L. The fish were fed three times a day ad libitum with specific extruded feed for each growth phase. At each age 1, 60, 90, 150, 210 and 270 days, 30 fish were weighed and the Gompertz model was fit to each batch. Estimates were obtained using weighted least squares due to heteroscedasticity, considering autoregressive errors.
Gompertz model was given by y = Ae(-Be-Kx). In addition, they were measured the absolute (AGR) and relative (RGR) growth rates, given by Kyln (u−1) and Kln (u−1), respectively, and u = yA−1, weight and age at inflection points (WI and AI), given by Ae−1 and (lnB)K−1, respectively These parameters were compared between strains by Tukey test at 5%.
The estimate final weight were 277.70±5.79, 287.13± 9.32, 320.19±10.92, 203.71±12.60 g to wild, blue, yellow and white, respectively. The white-albino final weight was lower than other strains (P<0.05). The age and RGR at inflection point were lower in the wild and blue strains when compared to the white one (Table 1). The higher values of weight and age at inflection and the absolute growth rates of the white and yellow strains may reflect the higher estimate of asymptotic values due to the greater variability of batch weights (Figure 1). So in this case, other models could be fit. It can be concluded that the wild, blue and yellow-albino strains, according to the Gompertz model, are more suitable for commercial cultivation in a shorter time.