Effects of change in amount of information from relatives of candidate on expected genetic gain in selection for increased litter size in a closed swine herd

Family selection index and computer simulation were used to investigate the effects of change in amount of information from relatives of candidate on expected genetic gain in the selection for increased litter size in a closed swine herd. Selection accuracy (rIg) and expected genetic gain (E[Delta-g]) were computed using information on the candidate's dam, sire and their full and half sibs found in F-1 generation grand parents and their full and half sibs found in generation F-2 and all ancestral information in generation F-1 dating back to generation F-4. The candidate for selection was found in the current generation F-0. Each female animal except those in current generation was assumed to have one record for litter size. Mating ratios (number of dams per sire) of 5 and 10 were assumed. Alternatively E[Delta-g] was estimated from the data generated using Monte Carlo computer simulation by either using dam's phenotype of candidate for selection or animal-model BLUP including all information. For simulation 100 or200 boars were mated to 1000 gilts randomly in each generation. Pedigree information comprised generations F-4 to F-0 whereas phenotypic information for litter size was recorded for generation F-4 to F-1. The numbers of full sibs and sisters per litter for the breeding stock were varied from 4 to 6 and from 2 to 4 respectively. A heritability of 0.1 for litter size was used for generating data, estimating breeding values and E[Delta-g]. Animal-model BLUP was used for estimating breeding values and E[Delta-g] was computed using true breeding values. The rIg calculated using information from all relatives was twice as large as that computed using only dam's phenotypic information of candidate for the selection. The change in E[Delta-g] computed using only the candidate's dam information was small when the number of full sibs per litter for breeding stock was held constant but the number of sisters per litter was varied