Modelling of Growth Curve Models According to Sex in Akkaraman Lambs with Different Methods: Logistik and Gompertz Modeling Example

In this study, the live weight data of Akkaraman lambs reared in Nigde Omer Halisdemir University Ayhan Sahenk Agricultural Research Application and Research Center were analyzed. Among the growth curve models, Logistic and Gompertz growth models were estimated using Least Squares Estimation (LSE), Maximum Likelihood Estimation and Bayesian methods. LSE and Maximum Likelihood methods, which are from the frequency methods, were compared with other AIC, AICC and BIC values and the results suggested that the best model for both genders was LSE Gompertz model. For the Bayesian method, LSE estimation information used as a priori. The Gompertz model was selected for the two genders by comparing the models estimated with the Bayesian method according to their DIC values. At the end of the study, parameter point estimates of Gompertz Model for male Akkaraman lambs were beta(0) = 86.296, beta(1) = 2.778, beta(2) = 0.014; while parameter point values for female Akkaraman lambs were beta(0) = 54.784 beta(1) = 2.453, beta(2) = 0.014. In addition, the distributions of the model parameters with the Bayesian approach were estimated by the Markov Chain Monte Carlo (MCMC) method, and it was decided that the expected adult body weight of male Akkaraman lambs was higher than that of female Akkaraman lambs.