COMPARISON OF WOOD, GAINES, PARABOLIC, HAYASHI, DHANNO AND POLYNOMIAL MODELS FOR LACTATION SEASON CURVE OF SIMMENTAL COWS

Abstract

The aim of this study was to determine a suitable nonlinear model explaining the lactation season curve of Simmental cows. Monthly milk yield records representing the test days milk yield of 777 Simmental cows were used to estimate lactation curve parameters by using Wood, Gaines, Parabolic, Hayashi, Dhanno and (second degree) polynomial models and to compare the shape of lactation season curves resulted from fitting all of these models. These models were given in explanation of the parameters and according to the lactation season, the parameters a, b and c for these models at all the periods were estimated. Secondly, the formula of some criteria, such as time to peak, peak production, turning point time and turning point production, are presented in mathematical procedure and then t the criteria values are calculated for all of the examined models. Furthermore, partial derivatives of the models according to the parameters were given in mathematical procedure. For the season, winter, spring, summer, fall and without season the lactation curve graphs for all the nonlinear models were drawn, respectively. Moreover, the best nonlinear model was used to determine the adjusted coefficient determination (R(adj)(2)), mean square prediction error (MSPE) and Bayesian Information Criteria (BIC). The runs test was used for determining whether data differ systematically according theoretical curves. Generally, polynomial model gave the best nonlinearmodel to the data compared to the other models.