Effects of population size and selection intensity on responses to disruptive selection in Drosophila melangaster.

Disruptive selection for sternopleural bristle number with opportunity for random mating was done in the four treatment combinations of two population sizes (40 pairs and 8 pairs of selected parents) and two selection intensities (1 in 40 and 1 in 2). In each generation, matings among selected parents were observed in a mating chamber, and progeny collected separately from each female parent. In the high number, high selection intensity treatment, divergence between the high and low parts ceased about generation 11. The isolation index increased rapidly to generation 3, but then fluctuated to termination of the population at generation 17. The overall isolation index was significant, indicating a real tendency to assortative mating. The failure of the isolation index to increase after generation 3 was attributed to lower average mating fitness of high males (due to inbreeding) and reduced receptivity of low females (due to a homozygous lethal gene with a large effect on sternopleural bristle number in heterozygotes). In the two low number treatments, isolation indices fluctuated from generation to generation with no obvious trends, and none of the overall isolation indices were significantly different from zero. The high number, low selection intensity treatment showed very little divergence, and one of the replicates showed, in contrast with expectation and the high number, high selection intensity treatment, a significant tendency to disassortative mating. Intense disruptive selection may lead to assortative mating.