Journal or Book Title
Journal of Animal Science
Four procedures, one using genetic relationships, were evaluated to rank centrally tested boars from records on 9,093 boars tested in central stations. Breeds with larger numbers of tested boars were found to have more total relationship ties than were breeds with smaller numbers of boars tested. Of the breeds with greater tested numbers, Durocs and Yorkshires Were genetically more tied than were the Hampshire or Spotted breeds. Among breeds with fewer observations, Berkshires had the most relationship ties, while Poland Chinas had the fewest. When the number of ties between boars in each season-year with boars in the most recent season-year were evaluated, only the last six or seven-season-year groups were needed to provide sufficient relationship ties with pigs currently tested. The four evaluation procedures were performance value (PV), deviation of the performances value from the station-season-year subclass mean (SSYD), mixed model without relationships (MM) and mixed model including relationships (MMR). The MM procedure produced much lower prediction error variances than did the PV or SSYD procedures; whereas the MMR procedure produced the lowest prediction error variances. Although the MM and MMR procedures gave far more precise estimates than did PV and SSYD, they do involve more complex computing procedures. When the cost of these morecomplex computations is reasonable, the MMR procedure should be considered to evaluate centrally tested boars because it allows a fair comparison of boars across stations in the same season. The value of across station evaluation is discussed.
American Society of Animal Science
Carlson, John P.; Christian, L. L.; Rothschild, Max F.; and Willham, Richard L., "An Evaluation of Four Procedures to Rank Centrally Tested Boars" (1984). Animal Science Publications. 315.