Genetic Selection Using Pooled Semen

Abstract The use of pooled semen (PS), standard procedure at the commercial level of the swine industry, limits genetic progress, since the offspring parentage is unknown. Literature suggests that some boars have a greater number of progeny than others in the same litter when PS is used. The objective of this study was to evaluate different pedigree-relationship matrices for selection purposes when PS is used. Data on a trait with heritability of 0.4 were simulated 1,000 times for 12 scenarios: combination of sires per pool (2 or 3), number of phenotyped progeny (1 or 12), and three boar dominance levels: no dominance (equal probability of parentage), medium dominance, and complete dominance (all progeny from one boar). Ten pools were created for each scenario based on 5 sires and used for 5 dams each. Breeding values (BV) of the progeny were estimated (EBV) using three relationship matrices: known parents (A), using equal probabilities of parentage (E), and probabilities based on known boar dominance (D). Results for each relationship matrix were compared using the average of the true BV (TBV) of the 10% best animals selected based on EBV. Results are presented as percentage TBV of the selected animals compared to using A (Table). In general, D resulted in better results than E as boar dominance increased. Similar results between E and D were obtained when 12 progenies were phenotyped. When one was phenotyped, D was superior than E in the presence of boar dominance. Knowing the probability of each sire contributing to the progeny increases response to selection when pooled semen is used.