Can mating preferences explain changes in mtDNA haplotype frequencies?

SINGH and HALE ( 1 990) propose that differences in mating preference and mating speed of Drosophila pseudoobscura females from Bogota and California may explain the frequency changes of mitochondrial DNA (mtDNA) haplotypes that we (MACRAE and ANDERSON 1988) observed in population cages. In the course of considering various aspects of mating behavior as possible causes for the frequency changes in mtDNA variants, we cited PRAKASH (1972) as having reported no mating preferences between D. pseudoobscura strains from Bogota and North America. We failed to cite SINGH (1983), who reported a strong mating preference of Bogota females for North America males, as well as an advantage in mating speed for the Bogota females, and we apologize for this oversight. We do not believe that mating preferences and mating speed can account for our results, however, and we outline our reasons below. 1. We chose mtDNA variants from Bogota and California for our experiments in order to compare conspecific haplotypes that were as different as possible. SINGH and HALE (1990) treat the Bogota and North America subspecies as if they were distinct species, but we feel they are properly described as conspecific. The genetic differences between them justify their description as subspecies (AYALA and DOBZHANSKY 1974; SINGH 1983), but they are separated by a partial reproductive isolation only. They can mate and produce fertile offspring. 2. It is worth recounting some puzzling aspects of the sterility of F1 hybrid males from crosses of Bogota females to North American males. The Bogota strain we employed, BogER, was one of several brought back from Bogota in 1979 by JEFFREY POWELL and one of us (W.A.). It is the same strain used by ORR (1989) in his genetic analysis of sterility in Bogota/ North America hybrids, but it is different from the strain SINGH (1983) used in his study of mating preference. POWELL ( 1 982) reported that crosses between females from these new Bogota strains and males from North America did yield quite a few fertile F1 male offspring when the crosses were made only two generations after the new strains were obtained. This F1 male fertility was increased sixfold when the food medium contained tetracycline, leading POWELL (1982) to suggest that a microbial symbiont might be involved in the F1 male sterility. F1 male sterility had increased considerably after four or five generations of laboratory propagation but was still not complete, even in crosses with no tetracycline treatment. Five years after the new strains were brought back from Bogota, all F1 males from crosses of the Bogota fe-