Date of this Version
The objectives were 1) to estimate responses from 28 generations of selection for increased pigs born alive (BA) with additional selection for increased 180 d weight (WT) and longissimus muscle area (LMA) and decreased backfat (BF) in the last eight generations, 2) to estimate genetic parameters for an array of traits, and 3) to investigate whether a plateau in response for BA has occurred. All lines were derived from the same Large White/Landrace composite population. Index selection for ovulation rate and embryo survival (G0-11) was initiated in L2 in 1981; L1 was selected randomly control line for L2 (G0-23). L2 was subsequently selection for fully formed pigs or BA during G12-28, and WT, BF and LMA (G20-28). At G8, L4 was derived from L2; L5 and L6 were derived from L1. L4 and L5 were in two-stage selection for ovulation rate and litter size (G9-16) and for BA (G17-19); L6 was selected randomly. L4 and L5 were crossed in G20 to form L45 which was selected for BA, WT, LMA and BF thereafter. Line sizes each generation were 40-60 litters by 15-20 sires. MTDFREML was used to obtain estimates of variance components, EBVs and responses for 17 traits (n = 1,883 ~ 54,174). Responses and cumulative selection differentials (CSD) were estimated by linear and quadratic regression of mean EBVs and phenotypes on generation for each line. Genetic trend and CSDs for BA were linear or quadratic in all selection lines (P < 0.05). Direct and indirect selection on BA was effective. Genetic correlations of reproduction with growth and carcass traits were small. BA, WT, and LMA increased and BF decreased in all selection lines (P < 0.05). For BA, genetic variance in selection lines during G20-28 still exhibit, and changes in CSDs and mean EBVs were linear (P < 0.01), but not quadratic. There was no evidence for a plateau in BA after 28 generations of selection.
Advisor: Rodger K. Johnson