ADVANTAGE BLUP FOR PREDICTING BREEDING VA LUES IN POULTRY BREEDING PROGRAMS WITH MULTIPLE TRAITS - A MONTE-CARLO STUDY

The objective of the study was to compare BLUP and traditional index s election over 40 generations for a typical egg-laying poultry breeding program, A multi-trait model was simulated assuming only additive gen etics effects. three sex-limited traits and non-overlapping generation s. Populations of 570, 1140 and 2280 were investigated using 150, 100 and 50 replicates. respectively. The male-female ratio was 1:3.75. To avoid inbreeding the simulated population was divided in 1. 2, 4, 8, 1 6, 32 and 64 sire families (pens) in which 64, 32, 16 8. 4. 2 and I si re was used per generation, respectively. Criterium for comparing the methods was the cumulative response to selection and the average inbre eding coeffcient after 40 generations. The selection was based on an i ndex containing performance and full- and half-sib performances was si mulated by application of BLUP by using only data and pedigree of the last generation, i.e, sires and darns were assumed as known but unrela ted. BLUP had a clear advantage over index selection after 40 generati ons for all population sizes and number of pens greater than 2.. Depen ding on the number of pens the estimated advantage was 1.3 to 4.9% for N=570. 1.2 to 4.3% for N=1140 and 1.1 to 4.0% for N=2280. Without a r estriction of inbreeding the cumulative selection response was not sig nificantly improved using BLUP. This study shows that the structure of the breeding program and the mating strategy affect the advantage of BLUP over selection index.