An experimental check on quantitative genetical theory II. The long-term effects of selection

Summary

• 1. The results of continued selection for abdominal bristles in a large population ofDrosophila melanogaster are presented and discussed.

• 2. Response had slowed down considerably in many lines after twenty generations, although in some it continued until the 30th generation. In many of the lines, the cessation of response was abrupt and lack of response did not mean exhaustion of genetic variability.

• 3. In three of the high lines, the high variability was apparently due to continued selection of heterozygotes for a lethal gene. In two high lines, such genes were present on both 2nd and 3rd chromosomes.

• 4. In the low lines, a striking phenomenon was a sudden increase of variation in females followed by a rapid response in that sex. This appeared in all lines with different times of onset. The ratio of male to female score in all low lines was greater than unity, compared to 0.8 in the initial population.

• 5. Genetic variation was maintained in many of the low lines after response had ceased. The situation appeared to be complex in that lethal genes, infertility of extreme females, and heterozygosity for inversions all played some part.

• 6. In such situations, the classical heritability approach appeared to break down completely. In some cases, the reason for the breakdown could be given in terms of the other phenomena observed, but in some of the low lines the full explanation was not clear.

• 7. The relevance of these results to practical problems of animal breeding is discussed.