Effects of cis-trans viability selection on some two-locus models (Q1897496)

Theoretical analyses of two-locus viability models have, with few exceptions, centered on models in which the interactions between the two loci have been additive or multiplicative, or that follow the ``symmetric viability'' scheme. In most of these studies, the viabilities of the coupling \textit{cis} and repulsion \textit{trans} double heterozygotes have been assumed to be equal. The motivation for most two-locus studies has been to examine the consequences of strong interlocus interactions, such as are known, for example, in self-incompatibility or mimicry systems, to study the more general effects of weaker interactions, such as multiplicative, or to examine the evolutionary function of recombination. In recent years, considerable attention has focussed on questions of nucleotide diversity and the maintenance of genetic variability within a locus. A number of surveys of nucleotide diversity around known loci, particularly in Drosophila, have shown that strong non-random associations may exist between nucleotide polymorhisms separated by up to 10 kb or more (corresponding to a recombination rate of around 0.02\% in Drosophila). Commonly, the patterns of nucleotide diversity and the associations between nucleotides at different sites have been attributed to population bottlenecks and to selective sweeps, or hitchhiking. However, there is ample opportunity for selectively generated disequilibrium between polymorphic sites within such short spans of DNA, and because the recombination distances are so small, the selective forces need not be large. Further, as we shall show, interactions that increase the viability in cis, but reduce the viability in trans, are capable of generating very strong linkage disequilibrium, much stronger than that usually predicted for interlocus interactions in which there are no such differences between the two kinds of double heterozygote.