Modeling the lactation curve in a herd of registered Jerseys in the dry tropics of Mexico.
Source: PubMed, NCBI / U.S. National Library of Medicine
The production performance of livestock depends strongly on the environment and management conditions, especially the capability of exotic breeds in different climatic regions. There are concerns about adaptation issues and decreased productivity of dairy breeds when raised in a distinct climate. Therefore, this research aimed to understand the productive performance of Jerseys raised with local resources under the dry tropical conditions in southern Mexico and develop model parameters for milk production throughout the lactation period. To this end, in a registered Jersey herd in Chiapas, México, over a two-year period with 42 cows and 2227 weekly records, the lactation curve for milk, fat, and protein was modeled and characterized using nonlinear mathematical models. The Brody, modified Brody, Sikka, Wood, Wilmink, and Nelder mathematical models were employed. The adjusted multiple coefficient of determination (R²), the residual standard deviation (RSD), the Akaike information criterion (AIC), the Bayesian information criterion (BIC), the Durbin Watson test (DW), and the distribution of residuals were used as comparison criteria to select the best-fitting models, also Friedman non-parametric test was used to rank models. The model that best predicted the lactation curve was Nelder's, for fat was Sikka's, and for protein it was Wilmink's, with R²= 0.995 and RSD = 0.269, R²= 0.994 and RSD = 0.068, R²= 0.983 and RSD =
Abstract
The production performance of livestock depends strongly on the environment and management conditions, especially the capability of exotic breeds in different climatic regions. There are concerns about adaptation issues and decreased productivity of dairy breeds when raised in a distinct climate. Therefore, this research aimed to understand the productive performance of Jerseys raised with local resources under the dry tropical conditions in southern Mexico and develop model parameters for milk production throughout the lactation period. To this end, in a registered Jersey herd in Chiapas, México, over a two-year period with 42 cows and 2227 weekly records, the lactation curve for milk, fat, and protein was modeled and characterized using nonlinear mathematical models. The Brody, modified Brody, Sikka, Wood, Wilmink, and Nelder mathematical models were employed. The adjusted multiple coefficient of determination (R²), the residual standard deviation (RSD), the Akaike information criterion (AIC), the Bayesian information criterion (BIC), the Durbin Watson test (DW), and the distribution of residuals were used as comparison criteria to select the best-fitting models, also Friedman non-parametric test was used to rank models. The model that best predicted the lactation curve was Nelder's, for fat was Sikka's, and for protein it was Wilmink's, with R²= 0.995 and RSD = 0.269, R²= 0.994 and RSD = 0.068, R²= 0.983 and RSD = 0.013, respectively. The average milk production per cow was 3,387 kg per lactation, with 5.1% fat and 3.40% protein. The results of this research suggest that these models can be used to predict with high level of accuracy daily milk, fat and protein production of Jerseys under dry tropical conditions, as well as to help improve production, peak, and lactation persistence.