NATURAL SELECTION RESISTING INBREEDING DEPRESSION IN CAPTIVE WILD HOUSEMICE (MUS MUSCULUS).

Abstract

The role which heterozygote selection plays in maintaining the genetic variability of natural populations has often been debated, and the resulting controversy has contributed substantially to populationgenetic theory and experimentation (Cain and Sheppard, 1954; Dobzhansky, 1955; Dobzhansky et al., 1963; Berry, 1978). With respect to wild Mus musculus, several studies have demonstrated that natural populations do possess large amounts of genetic variability (Dunn et al., 1960; Selander and Yang, 1969; Berry, 1978). However, it is difficult to determine the role which heterozygote selection plays in maintaining this genetic variability since a variety of obstacles impede the direct measurement of heterozygote selection; for example, linkage of alleles potentially favored by selection, the environmental specificity of any selection advantage, the population specificity of any selection advantage, and other problems discussed by Berry (1978). Therefore it is worthwhile to examine heterozygote advantage using systematic inbreeding as an indirect method of detecting it (cf. Dobzhansky et al., 1963), both to examine the incidence of heterozygote selection as an evolutionary mechanism and to broaden our understanding of the inbreeding process itself. Our experiments showed that a delayed the effects of inbreeding beyond what would normally have been expected had there been no selection for heterozygotes. That is if natural selection favors heterozygotes at individual loci, then the rate of homozygote production should lag behind that predicted by F (Wright, 1965), the theoretical coeffecient of inbreeding. We will call this situation a selection-inbreeding equilibrium model since selection and inbreeding occur simultaneously and produce opposing effects (Hayman and Mather, 1953; Reeve, 1955; Li, 1967). This may be either stable (permanent) or unstable (transient): (A) A stable results if balancing selection is sufficiently intense relative to inbreeding and the size of the breeding population is sufficiently large (Hayman and Mather, 1953); as a result the breeding population never becomes genetically fixed. (B) An unstable results if balancing selection is weak relative to inbreeding in a large population, or if the breeding population is small (Li, 1967); as a result, genetic fixation eventually occurs but is delayed relative to an unselected population with the same inbreeding intensity. The small population size (approximately eight breeder pairs per generation) of our own sibmated lines makes an unstable more likely. Accordingly we expected that heterozygote selection would merely delay the eventual fixation of our sibmated lines. Thus we captured a sample from a population of wild housemice, sequestered them in our laboratory and immediately sibmated their progeny. We measured three effects as inbreeding was continued: changes in fertility, changes in homozygosity as measured by skin transplantation techniques, and homozygosity as measured by isoenzyme assay techniques. We hypothesized that inbreeding depression would occur, not at a steady rate, but interrupted by long pauses reflecting resistance by balancing selection. We also hypothesized that the amount of homozygosity produced by 20 generations of sibmating would lag behind the amount