Effectiveness of genetic selection for prolificacy in pigs

This paper attempts to summarize and discuss the new evidence on the effectiveness of selection for prolificacy. In recent years selection between lines, and the adoption of formal cross-breeding programmes, have led to considerable improvements at a commercial level. Within-line selection has not been tackled seriously except in a few experiments. The mainly negative results from these should not lead to the conclusion that progress is impossible, but rather that it needs larger resources, applied consistently for several generations, to overcome the inherent problems. There are now many estimates of genetic parameters, based on large data sets, and these have helped to clear up some of the confusion resulting from previous inadequate material. A heritability of around 0.1, repeatability of 0.15, and fairly high genetic correlations between successive records seem justified. The negative environmental effect of being reared in a large litter is frequently present, but there is no good reason why this should seriously affect attempts to improve the trait by selection. Information from a candidate's relatives other than its dam can undoubtedly increase the accuracy with which its breeding value is estimated, but the routine use of such data in practical programmes is not straightforward. Prolificacy is only one group of traits contributing to net economic worth. The need to strike a balance between all relevant traits is explored in the particular context of a scheme based on hyperprolific sows. Finally, selection can only have lasting and cumulative effectiveness if it is applied at the top of a pyramid of improvement. Both culling and selection of replacement gilts within the commercial sector are largely irrelevant.