The population genetics of the selfincompatibility polymorphism in papaver rhoeas. Viii. sampling effects as a possible cause of unequal allele frequencies

This paper considers the hypothesis that the unequal S-allele frequencies in the British populations of Papaver rhoeas we have examined are caused by a sampling effect over and above that due to drift. The results obtained by simulating the polymorphism on the computer show that it takes between 70 and 80 non-overlapping generations before stochastic equilibrium is achieved with respect to allele frequency in a population of size 3720 containing 31 alleles, both when the frequencies of the alleles are initially as unequal as they were in the R106 sample (Campbell and Lawrence, 1981b) and when these frequencies are exactly equal. These results show that the strength of the frequency-dependent selection that maintains the polymorphism becomes very considerably attenuated as the number of alleles in the population increases. A review of the relevant genetical and ecological evidence suggests that the size of the British populations is considerably larger than the minimum required to maintain the number of alleles they are estimated to contain, so that opportunities for sampling effects to perturb the frequencies of these alleles appear to be limited. The results obtained from an investigation of the reproductive biology of the species, however, show that the distribution of progeny size is markedly non-random. It is possible that this may be the cause of the unequal allele frequencies observed in the British populations.