A test of speciation via sexual selection on female preferences
Sexual selection for divergent female preferences has been proposed to stimulate speciation. We tested this basic model by selecting for divergence in the courtship repertoire of the house fly Musca domestica L. Specifically, we subjected replicate strains to artificial selection for differentiation along the first two principal components of the phenotypic intercorrelation structure of five courtship traits. Highly significant differentiation in courtship repertoire resulted, and the magnitude of the selection response was highest along the first principal component (representing the 'size', or general intensity, of courtship). Videotaped matings of the crosses between divergent lines (i.e. males of one strain mating with females from a different line) showed that the selection responses in the intensities of male performances were due to shifts in female preferences. In particular, the males were able to accommodate the demands of 'foreign' females (as well as their own) in the no-choice situation (i.e. only one male and one female per mating chamber). In contrast to this plasticity of the males, the females were consistent in their differential resistance responses, regardless of the type of male involved in the courtship. Multiple-choice mate choice tests revealed significant reproductive isolation among some lines, although the effect was asymmetrical. The patterns of nonrandom mating were largely due to females from more genetically healthy lines (i.e. with overall high mating propensity) discriminating against males from populations with more inbreeding depression. We suggest that the inability to achieve true (symmetrical) reproductive isolation could have been due to low evolutionary potential in the 'shape' of courtship, as defined by the second principal component. (C) 2002 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.