Selección del modelo matemático que caracterice la curva de lactancia en ganado vacuno en el sistema doble propósito
Botero Arango, Luz Mercedes | 2020-01
Objective. To select a mathematical model for the lactation curve of milk production in bovine females. Materials and methods. During 11 months, the production of milk was studied in 500 bovine females double purpose Bos taurus x Bos indicus, at the Savannas of the Colombian Low Tropic. Production was quantified in kilograms. Data of milk production were included in time (dry-rainy) and number of lactations (first; second and third and, more than three). Data make part of the Cattle raising XB file, located in Bolivar’s Savanna, Colombia, gathered from the year 1990 up to the 20000. The lineal simple, quadratic, lineal logarithmic, quadratic logarithmic, incomplete gamma, lineal hyperbolic and inverse polynomial models were applied to these data. nursings (first; second and third and, more than three). The data make part of the file of the Cattle raising XB, located in Bolívar’s Savannas, Colombia, gathered from the year 1990 up to the 2000. The parameters for the incomplete gamma and inverse polynomial models were estimated starting from the “Gauss-Newton” method to non lineal regression; the other ones were adjusted by lineal regression of the productions in function of the months in lactation by the minimum squares method. Results. It was observed that the inverse polynomial model is the one that better characterizes the lactation curve, presenting the biggest values for the Durbin-Watson statistic and the coefficient of determination (R2) and about this model necessary information exists to obtain practical parameters from the lactation curve equation. Conclusion. The mathematical polynomial model inverse is an excellent tool to approach the administration and desicions management of cattle double purpose.