A MODEL FOR THE EVOLUTION OF ASYMMETRICAL MALE HYBRID STERILITY AND ITS IMPLICATIONS FOR SPECIATION.

Chromosomal analysis of several cases of asymmetrical male hybrid sterility in Drosophila has implicated the X- or the Y-chromosome and one or more autosomes. Here, I develop a model for the evolution of this phenomenon. An autosomal locus is assumed to affect viability and to interact with a Y-linked or an X-linked locus to determine male fertility. In a new environment, selection for viability favors a new allele at the autosomal locus, but incompatibility of this new allele with the sex-chromosome-linked gene generates male sterility. The incompatibility can be resolved if a new allele at the sex-linked locus invades the population. This results in nonreciprocal male hybrid sterility, the direction of the nonreciprocity being determined by the dominance or recessiveness of the new autosomal gene in its effect on fertility. It is shown that stable polymorphism for the autosomal locus is possible and that, if the equilibrium frequency of the new allele is above a critical value, the population will be constantly at the verge of speciation, "waiting" for the sex-linked mutation to occur. The appearance of this mutation causes a runaway process leading to rapid fixation of the new autosomal and sex-linked alleles. If the equilibrium frequency of the new autosomal allele is less than the critical value, deterministic speciation is impossible, but random drift may increase the frequency above the critical value and predispose the population to the invasion of the new sex-linked allele. Thus, both deterministic and stochastic modes of speciation are possible. Because deterministic speciation requires large selection coefficients, which impose a severe genetic load on the population, and because stochastic speciation requires repeated population bottlenecks, it is concluded that relative to the number of successful speciation events there will be many more events of deme extinction. © 1986 The Society for the Study of Evolution.