Europäisches Journal für Experimentalbiologie Offener Zugang

Abstrakt

Use of the random regression model for the genetic parameters estimation of milk yield in a Holstein cows herd

Shahabodin Gharahveysi*, Mehrdad Irani, Ruhollah Abdullahpour and Zabihollah Kalashi

To estimate genetic parameters for daily milk yield, test day records of first lactation Holstein cows were used. The data included 9782 test day records of 1140 Mahdasht herd cows recorded from 2001 to 2009. Analysis of test day records was performed using the REML procedure in DFREML software. Legendre polynomials on days in milk of order 5 were used to fit the population mean as fixed regression and of order 3 for both additive genetic and permanent environmental effect as random regression. Residual variances for the model considered to be constant for all days in milk and estimated as 305. The minimum value for heritability estimate (0.039) occurred in early lactation but had increase increased toward the lactation and in the ninth month of lactation reached the maximum value (0.280). Phenotypic variance of daily milk yield was higher in early and late lactation. The maximum (18.750) and minimum (2.360) values of genetic variance were in month 10 and early lactation respectively. The maximum values for genetic and phenotypic correlations were between records of adjacent days in milk and as the distance between the days increased the correlations decreased. Using test day records, increased accuracy for genetic parameters will be achieved.