The objective of this study was comparing non-linear models to describe the growth of quails males and females from birth to 42 days old. We used 24,679 records of body weight of quails from the female line LF1, belonging to UFVJM´s Improvement Program of Quail in Diamantina, MG. The three successive generations of quails were weighed every seven days (birth, 7, 14, 21, 35 and 42 days old). The models Brody, von Bertalanffy, Richards, Logistic and Gompertz were fitted to the data set. The model parameters were estimated for males and females separately, the Gauss Newton algorithm using the NLIN procedure of SAS The criteria used to choose the model best fit the curve of growth were the coefficient of determination (R²), the asymptotic standard deviation (ASD), the mean absolute deviation of the waste (DMA) and the asymptotic index (AI). Of the five models used only the Richards model showed no convergence, both for males and females. The lower asymptotic value for the index (AI) was observed for the Gompertz model in males (26.57) for females and Logistics (25.02), indicating that these models that best fit the data.
Coturnix coturnix coturnix; growth curve; non-linear models
