ABSTRACT
A Monte-Carlo simulation program is described for a polygenic mating model introduced in the first paper in this series (Sved 1981). The program is used to simulate the situation in laboratory experiments in which two strains are allowed to mass-mate, hybrids are artificially eliminated and the establishment of mating isolation is studied. It is shown that, if mating choice is sufficiently precise, a combination of chance fluctuation and selection can lead to divergence in mating behavior. However, for small population sizes, the variability would usually be considerably reduced by the time some divergence is established, leading to low eventual levels of isolation. For larger population sizes, on the other hand, it may take many generations for any divergence to be established. —A dissection of the selective forces involved in the divergence shows that the major force potentially responsible for initial selective response is the tendency for divergent females and males to reject mates from the wrong strain. However, this is nullified in mixed-strain matings by the tendency of such individuals equally to reject mates from the correct strain. To overcome this problem, it is suggested that the usual mixed-strain mating procedure be replaced by procedures specifically designed to select for rejection of interstrain matings. Two procedures are suggested for this, and computer simulation shows that one or other of the procedures will work under the assumptions of the mating model. Other possible outcomes of selection, including asymmetrical divergence, are discussed for cases in which the assumptions of the mating model are invalid.
Some estimates on the space scales of vortex pairs emitted from river mouths
