Studies on artificial and natural selection
Frank W. Nicholas · ERA · 1974
A Very little experimental evaluation of these predictions has been reported, although James (1965) did use equation (4) to estimate half-life in several previously reported selection experiments.Agreement between observed half-lives and those expected by James was not particularly evident, with predictions being much greater than those observed, but then Robertson's (1960) half-life prediction of 1.4W generations (for additive loci) was likewise a considerable over-estimate.Such a result, however, may simply indicate that artificial selection has successfully achieved fixation of most of the desired alleles, as Robertson has pointed out.James' study was useful in that it was the first attempt to analyse the effect of the interaction between natural and artificial selection on the limit to artificial selection.In dealing with the optimum model of natural selection, however, the study did not emphasise that interactions between loci are a vital component.James' selective values took this into account, but his derivation of expressions for the metric mean at any time t, and at the limit J tended to play down the importance of these interactions, with the result that the expression obtained do not necessarily apply to the type of limit most likely to be reached with the optimum model.A more detailed discussion of this problem is given later.It would seem then, that both the homeostatic and optimum models of stabilizing.naturalselection need to be more fully explored in the context of artificial selection, before a proper understanding of the interplay between artificial and natural selection can be achieved.20.