TY - CONF

T1 - Lactation curves in Valle del Belice sheep using random regression models

AU - Maizon, Do

AU - Portolano, Baldassare

AU - Riggio, Valentina

AU - Tolone, Marco

PY - 2007

Y1 - 2007

N2 - In recent years, several mathematical functions have been proposed to fit lactation curves. The goal of this study was to find the function that best fits the test-day records collected by the University of Palermo in Valle del Belice sheep, in order to implement a genetic evaluation. The lactation curve was modelled with different cubic spline (7, 9, 11, 13 knots) and Legendre polynomial (of degree 3, 4, 5 and 6) functions using ASReml. The dataset used comprised test-day records of milk production traits of Valle del Belice sheep recorded in accordance with the A4 method. A total of 30,428 test-day records from 4,424 lactations belonging to 2,136 ewes, offspring of 140 sires, were used. The average number of test-days per ewe per lactation was equal to nine. A sire model was used considering in the construction of the pedigree sires and maternal grandsires. In the model, the days in milk and the permanent environmental effect were modelled with cubic splines and the Legendre polynomials. Based on minimising the Akaike crterion, a Legendre polynomial model of degree four was the most adequate model. However, there was not a great deal of difference among the models compared.

AB - In recent years, several mathematical functions have been proposed to fit lactation curves. The goal of this study was to find the function that best fits the test-day records collected by the University of Palermo in Valle del Belice sheep, in order to implement a genetic evaluation. The lactation curve was modelled with different cubic spline (7, 9, 11, 13 knots) and Legendre polynomial (of degree 3, 4, 5 and 6) functions using ASReml. The dataset used comprised test-day records of milk production traits of Valle del Belice sheep recorded in accordance with the A4 method. A total of 30,428 test-day records from 4,424 lactations belonging to 2,136 ewes, offspring of 140 sires, were used. The average number of test-days per ewe per lactation was equal to nine. A sire model was used considering in the construction of the pedigree sires and maternal grandsires. In the model, the days in milk and the permanent environmental effect were modelled with cubic splines and the Legendre polynomials. Based on minimising the Akaike crterion, a Legendre polynomial model of degree four was the most adequate model. However, there was not a great deal of difference among the models compared.

KW - lactation curve, random regression model, dairy sheep

UR - http://hdl.handle.net/10447/76958

M3 - Paper

ER -