Factor analysis for sire evaluation in poultry
U.C. JAISWAL AND R.P. SINGH
Department of Animal Breeding C.C.S.
Haryana Agricultural University, Hisar – 125 004
The genetic merit of a sire for a particular trait is judged by the average performance of its daughters for that trait. Various methods of sire evaluation in livestock and poultry breeding have been proposed and compared by researchers from time to time. In the present paper a new method of sire evaluation through factor analysis has been proposed and illustrated with live data from poultry. The data on 481 pullets, progeny of 19 sires for four economically important traits, body weight at 18 weeks (gm) age at first egg (days), egg number up to 40 weeks and egg weight at 40 weeks (gm) were collected for the present investigation. Correlation matrix of four traits was subjected to principal component analysis for the extraction of factors. The results revealed that 86.56 percent of the variance among all the variables is accounted for by three common factors. First factor having high loadings on body weight at 18 weeks (gm) and age at first egg (days) was labeled as ‘maturity factor’. Second and third factors having high positive and negative loadings on egg weight at 40 weeks and egg number up to 40 weeks were designated as ‘production factors’. For estimation of scores on each sire, regression coefficients of factors on variables were estimated by using multivariate multiple regression analysis. Average performance of the daughters of each sire for different traits after standardization of their values and the estimated regression coefficients were used in obtaining scores on each sire in respect of the ‘maturity factor’ and ‘production factors’. Sires were ranked on the basis of their scores with highest as first and so on.