ESTIMATING THE DEGREE OF POLYANDRY IN NATURAL POPULATIONS.

Abstract

A mating system in which one dominant male sires all the offspring implies that a deme practicing it has a small effective size. A group of isolated demes with this mating system would become more genetically divergent in time due to genetic drift alone. When one encounters a group of large demes that are close together yet genetically well differentiated from another, he could suppose that drift due to this mating system (plus isolation) explains the genetic differentiation, rather than different selection regimes. This hypothesis would be supported (though not proven) by finding that the typical female mates with only one male. hypothesis would be falsified by finding that the typical female mates with several males. Birdsall and Nash (1973) pointedly express their interest in mating system this way: The possibility that populations of small mammals are subdivided into highly isolated demes of small size has been a point of contention in the literature . . . this paper is a report of direct evidence of frequent multiple male parentage within single litters in natural populations of deer mice . . . thus, in these populations the social units (if, in fact, they exist) must be large enough to include two, or probably more, breeding males. This is quite unlike the social structure . . . in Apodemus populations and . . . in Mus populations in which one dominant male is thought to do most, if not all, of the breeding. This area of inquiry has stimulated several investigators to develop indirect methods of estimating the number of times a female has mated to produce a brood of offspring. An indirect method is desirable because of the general difficulty of directly observing matings in a natural population and verifying that the matings have issue. Murray (1964) studied multiple mating in the hermaphroditic, obligate-outcrossing land snail, Cepaea nemoralis. He examined a population segregating for the shell color alleles dominant pink and recessive yellow. In addition to ascertaining the proportions of pinks and yellows in the colony, he scored the numbers of pink and yellow offspring within each brood of 35 homozygous yellow mothers. broad idea of his analysis was this: the more times a yellow snail mates, the more probable that at least one of the brood's sires will be pink and, therefore, the more probable that at least one of the offspring in the brood will be pink. Thus, the greater the proportion of broods of yellow mothers containing pink offspring, the greater the average number of matings. assumptions he invoked to make the idea quantitative were: (a) the population was in binomial genotype proportions; (b) the number of offspring per mating were large enough that all heterozygous pink sires produce pink offspring; (c) the mates of the yellow snails are drawn at random from the colony as surveyed; i.e., selection, migration or other systematic effects have not rendered the 'male' mating pool different from the colony; and (d) all the yellow mothers have mated the same number of times. result of the appli1 Dr. Wilson passed away on November 18, 1980, while this manuscript was in review. We would like to thank the editors of Evolution for their help and comments. Dr. Timothy Prout provided welcome assistance. Finally, we wish to thank Dr. Outi Muona for her help in the final revision. Joel Weintraub and Lon McClanahan, California State University, Fullerton.