Year: 1985
Type: dissertation
Abstract: A large body of data from a single nucleus herd was used to investigate non-genetic and genetic sources of variation mainly as affecting litter size (LS) in pigs. Of particular interest were the effects of rearing treatment on guts subsequent performance. Rearing treatment had a small negative effect on subsequent LS although not significant (P>0.05) due to a large standard error. However, a significant effect was observed on days to 90kg. Yet no evidence of bias on genetic parameters for LS. from maternal (fraternal or rearing) environmental effects were detected. Thus, daughter on dam regressions (bd,D) gave good estimates of genetic parameters with small standard errors. Heritability estimates from paternal half-sib analyses agreed well with bd,D estimates. A pooled estimate for the heritability from bd,D for LS was 0.11 ± 0.01 with repeatability (pooled) of 0.14 ± 0.01 among parities. Only the genetic correlations of first with third and fourth parities for LS deviated substantially from unity. Despite the low heritability, high rates of genetic change are predicted for LS using conventional selection methods. The highest rates are predicted from selecting males and females born to gilts on an index combining litter records on the dam, her full-sibs and half-sibs as well as on the full-sibs of the sire, his half-sibs and dam. Two records would be available on each relative. Rates of just under 1/2 a piglet per year of selection are predicted assuming a simple additive genetic model. As a percent of the mean basis this is substantially higher than the expectation for much more heritable traits. The loss in selection efficiency from possible fraternal effects are small with family index. Continuous screening and breeding from 'hyper prolific females' is much less effective due to the time needed for each cycle of selection. A single cycle of the scheme can be effective in starting an improvement program for LS or a specialized dam line. Increasing the accuracy of selection through incorporating family information is preferred to increasing the number of individual records because of the reduction in genetic lag. High rates of response are also predicted from selecting females on a family index as before combined with two stage selection of males, on the same family index and then on progeny test results. The extra economic response from adding all family information for LS to a selection index including growth and carcass traits for selection of general purpose stocks is small (less than 5 percent) for UK market conditions. However, the same index information used for selection of specialised dam stocks could bring about an increase in the economic response of 5 to 19 percent, which is worthwhile and is recommended.
Author: Eduardo Avalos Raz Guzman
Language: English
Subfield: Genetics
Domain: Life Sciences
SDG: Zero hunger
Open Access status: green
OpenAlex ID: W2214168076