ON THE POSSIBLE ADAPTIVE SIGNIFICANCE OF MALE STERILITY IN PREDOMINANTLY INBREEDING POPULATIONS

Abstract

N a theoretical study of the possible adaptive significance of gynodioecy as a Ibreeding system, LEWIS (1941) had postulated that the cytoplasmic male sterility was favoured under natural selection due to an increased hybridization in the population. Later, DARLINGTON and MATHER (1949) argued that each species maintains a characteristic hybridity optimum in order to meet its oppos-ing requirements of immediate fitness and long-term flexibility. This issue is particularly important in the case of predominantly self -fertilizing species where occasional bursts of outcrossing were assumed to originate the necessary hybridity (STEBBINS 1957). Recently, an analysis of changes in the genotypic frequencies at certain marker loci was carried out in a bulk-hybrid population of barley (JAIN and ALLARD 1960). It was found that a striking heterozygote advantage and one to two percent outcrossing together were able to retain substantial amounts of heterozygosity after 18 generations of selection and under as much as 98 to 99 percent selfing. In this paper it is suggested that naturally occurring male steriles in such a population might be adaptively favoured for their produc-ing hybrid progeny and thus effectively contributing to an enhanced outcrossing. Such individuals, although reproductively less fit themselves, would therefore have a selective advantage due to their higher social fitness, or stated otherwise, they possess a higher average survival value in the population than the one ex-pected on the basis of their individual effects (FISHER 1941). In order to study the survival of male sterility factors with the help of a theo-retical model, consider for simplicity the case of a recessive male sterile mutant ms, with genotypes M s M s and M s ms fully male fertile. Let P,, Qn, R, denote the relative frequencies of M s Ms, msms and M s m s respectively in the nth generation, and assume the proportions of selfing (s) and random mating (t) to be .95 and .05, respectively. The selection coefficients for the genotypes M s Ms, M s ms and ms ms may be taken as 1 -z, 1, and 1 -y which implies heterozygote advantage for 0 < z, y < 1. Next, an inbreeding population in nature would not