Modeling the lactation curves for milk, fat and protein yield of Holstein cattle in Antioquia, Colombia

Objective. Modeling lactation curves for milk yield and for fat (%) and protein (%) and to determine the main factors affecting the lactation curve. Materials and methods. Official milk control data of 1532 Holstein dairy cows from 19 farms from the department of Antioquia, Colombia were used. Non-linear mathematical models proposed by Wood, Brody, Papajcsik and Bodero and Wilmink were calculated. Regression coefficients of the best fit non linear model selected was used for the effect on parity, time and year of birth was. Results. Lactation curves for milk yield and both fat and protein percentages estimated by Wood's model showed the best values for the comparison criteria used. Parity and year of parity effects were significant for the estimated parameters in the Wood's model. Conclusions. Wood's model showed the best fit in lactation curves for milk yield and for fat (%) and protein (%). The lactation curve for protein (%) was characterized by a very consistent production and no major fluctuations throughout lactation.