THE GENETICS OF VIABILITY IN DROSOPHILA MELANOGASTER: EFFECTS OF INBREEDING AND ARTIFICIAL SELECTION

Inbreeding and artificial selection experiments were conducted to investigate the genetic properties of egg-to-pupa viability in a population of Drosophila melanogaster. The effect of different levels of inbreeding (F = 0, 0.25, 0.50, and 0.73) was studied. Up to F = 0.50, a linear depression of the mean viability was observed, accompanied by a significant increase of both within-line additive variance and between-line variance. At F = 0.73, no further changes were detected. This can be attributed to natural selection opposing high levels of homozygosity. In parallel, artificial selection to increase viability was performed for 27 generations in (1) a single undivided population (U) and (2) two populations with cycles of subdivision and between-line selection, followed by reconstitution of selected lines (S_O and S_I ). During the first cycle (generations 0-4), most of the final total response was achieved under all selection regimes. An advantage of the S_O and S_I strategies was observed after the completion of the first cycle. However, the same limit was reached in all cases because of a delayed response experienced by line U. Reverse selection for viability resulted in positive correlated responses for fecundity and mating success. Both inbreeding and selection results are compatible with the genetic variance of viability in the base population being generated by segregation at a few loci with substantial additive effects and several deleterious recessives at low initial frequencies. Possible reasons for the maintenance of that variance in natural populations are discussed.