A UNIQUE APPROACH TO THE EVOLUTION OF SEX.

Abstract

Over the last ten years books by Ghiselin (1974), Williams (1975), Maynard Smith (1978), and Bell (1982) have generated a considerable amount of discussion concerning the evolution and maintenance of sexual reproduction, but no true consensus has emerged as to whether the general phenomenon of sex evolved or is maintained by a common selective mechanism. A second, and to some degree overlapping, body of literature has concerned itself with the evolution and genetic consequences of various mating systems within sexual populations. A third body of literature has sought to explain the evolution, or adaptive significance, and genetic consequences of various dispersal strategies displayed by plants and animals, particularly in the context of patterns of habitat stability. Shields's book attempts to synthesize aspects of the literature in these three areas in such a way as to present a rather novel view of the adaptive significance of sexual reproduction. The book will no doubt stir up some controversy among evolutionary biologists, in that its central hypothesis states that mating systems that exhibit moderately high levels of inbreeding should be selected for in many species. Shields's thesis is that epistatic fitness interactions among loci will result in some combinations of alleles particularly well suited to a given local environment. Any mating system that will act to preserve these coadapted genotypes will be selected for. One mechanism that would limit the disruption of the gene complexes would be inbreeding. Shields seems to define inbreeding rather broadly as nearly any deviation from species-wide panmixia or, alternately, any probability much greater than zero that an individual will carry two alleles that are identical by descent. Individuals migrating into a local population will carry gene complexes adapted to the local environment from which they originated and which would not necessarily interact well with genes derived from individuals native to the new locality. Under this scenario, selection should act to limit dispersal distances, keeping the individual within the environment to which it is adapted and increasing the probability that paired alleles are identical by descent. It should also promote xenophobic mate se-