On the relative roles of faster-X evolution and dominance in the establishment of intrinsic postzygotic isolating barriers.

The modern theory of speciation assigns a prominent role to the recessivity of genetic incompatibilities in the two rules of speciation, namely Haldane's rule and the 'large X effect', and considers that the contribution of faster evolution of the X versus the autosomes to those patterns is generally of relatively minor importance. By extending Turelli and Orr's previous analysis of the model of two-locus Dobzhansky-Muller incompatibilities, I first show that when the X and the autosomes evolve at the same rate, the two dominance parameters involved in that model are not equally important for the declaration of a large X effect, but that the degree of recessivity of homozygous-homozygous incompatibilities is the major determinant for such a declaration. When the X evolves faster than the autosomes, the model obviously predicts that the importance of both dominance parameters will progressively vanish. It is then of importance to obtain estimates of the relative evolutionary rate of X-linked incompatibility loci. Several different procedures to obtain such estimates from the perspective of the large X effect are suggested. The application of the appropriate test to the only suitable data from Drosophila hybridizations so far available leads to the conclusion that the X actually evolves at least 2.5 times faster than the autosomes, as far as hybrid male sterility determinants are concerned, thus making dominance considerations absolutely irrelevant. Notwithstanding the necessity of further tests, the relative roles currently assigned to faster-X evolution and dominance in the theory of speciation should be revised, giving due prominence to faster-X evolution, at least for hybrid male sterility in the genus Drosophila.