Comparison some nonlinear functions for describing of growth curve in Makooei sheep

For doing this study 6841 body weight records at different ages of 1877 Makooei sheep which were collected at Makooei sheep breeding station in Maku city of West Azarbaijan province was used. Parameters of growth curve based on body weight at different ages and different initial amounts using NLIN procedure of SAS software were estimated. Four non-linear models including Brody, Von-Bertalanffy, logistic and Gompertz, were fitted on data. After setting the most appropriate describing non-linear growth curve, using by Gauss-Newton repeat method fitted in way that growth curve parameters obtained separately for all individuals. Using GLM procedure of SAS software, effective fixed effects on these parameters were identified. The Brody model with having the highest accuracy (R2 = 0.955432) and lowest error (MSE = 37.78) was recognized as the best model for describing growth pattern of Makooei sheep. The year and season of birth, lamb sex and birth type had significant effect on mature weight and growth rate (P < 0.01) but lamb sex on mature rate was not significant (P < 0.01). The results showed that, compared with the other evaluated models, for fitting growth curve of Makooei sheep was more suitable.