R. J. Berry

Heterogeneous heterozygosities in Mus musculus populations

Both chance and adaptation have effects in determining the genetical constitution of local populations of any organism, but opinions differ widely over their relative importance. This study describes the frequencies of electrophoretically detected alleles at 22 loci in 1538 house mice (Mus musculus L.) from 27 population samples collected from the Faroe, Shetland and Orkney archipelagoes; the mainland of Great Britain plus three small off-shore islands; and a sub-Antarctic island (Macquarie) lying between Australia and the Antarctic Continent. Neither the average heterozygosities nor the distribution of allele frequencies in the different populations showed any discernible pattern, but at least three loci (Hbb, Es-2, Dip-1) underwent seasonal changes in frequency which could only be due to selection. Moreover the overall variances of allele frequencies were significantly heterogeneous (P ≈ 0.001), suggesting that different factors affect different loci. The key to understanding this apparent randomness of frequencies is recognizing that selection pressures are non-constant, and particular traits may affect fitness only spasmodically. The occurrence and frequency of an allele in any population may reflect only chance historical factors, but the trait(s) affected by it is potentially subject to selection at any time. For example, the proportion of heterozygotes at the Hbb locus in one population (Skokholm) living entirely independent of man, increased in five summers out of six. Winter death in mice is colddependent, and the exceptional summer followed a particularly mild winter when the usual decrease in heterozygotes did not take place. An r-selected species like the house mouse is more likely to reveal the interplay of genetical chance and purpose than a numerically less volatile one in which short-term genetical adjustment will be relatively uncommon.