Gene flow can influence the adaptive course of natural selection both by determining what genetic variation is available within a deme's gene pool and by directly altering allele frequencies (Chapter 6). To see this, consider the simple model from Chapter 6 of symmetrical gene flow at rate m between two demes, 1 and 2. The change in allele frequency at an autosomal locus in deme 1 is given by Ap1 = -m(p1 - p2) where pi is the frequency of the allele in deme i(equation 6.2). Combining the effects of selection and gene flow, equation 12.2 becomes
Suppose that aA > 0 in deme 1; that is, natural selection favors an increase in the frequency of allele A. However, if p1 were initially zero, this adaptive course could never start. Suppose further that the A allele is present in population 2. Once A is introduced into deme 1 through gene flow, natural selection can now operate to increase its frequency. As we saw in Chapter 11, natural selection does not create genetic variants but only operates upon the genetic variation available in the gene pool. Gene flow can be an important source of such variation.
The sign of the component of allele frequency change induced by gene flow in equation 12.13 is determined solely by the initial difference in allele frequencies among the demes, in this case p1 - p2. The sign of the selective component of equation 12.13 is determined solely by the average excess of fitness, which in turn is a function of within-deme allele frequencies, system of mating, and genotypic deviations of fitness in the deme. As a result, there is no biological necessity for the selective and gene flow components to have the same sign. In some cases, selection and gene flow can operate in the same direction, allowing an allele to increase (or decrease) in frequency more rapidly than possible through either selection or gene flow alone; in other cases selection and gene flow will be in opposite directions and the evolutionary outcome will depend upon their balance. To illustrate these diverse outcomes, we return to the example of African Americans, a population influenced by asymmetric gene flow involving demes originally derived from Europe and West Africa (Chapter 6). In this case, the impact of asymmetric gene flow over several generations was measured by (equation 6.4)
M = p A - pw pe - pw change in allele frequency in African Americans from West Africans initial difference in allele frequency between Europeans and West Africans
Because M is standardized for the initial allele frequency difference, M should be identical for all alleles at all polymorphic loci showing any initial difference in allele frequency between Europeans and West Africans if gene flow were the only evolutionary force operating.
Adams and Ward (1973) estimated M for alleles at several loci in African Americans from Claxton, Georgia (Table 12.1), and found significant heterogeneity across alleles. Such heterogeneity implies that selection may have altered the allele frequency dynamics at some loci. A majority of the alleles yielded an estimate of M of about 0.11, and combining the estimates across all alleles as weighted by the variances of the estimators yields an overall M of 0.13. The most straightforward interpretation of these results is that these alleles are neutral or nearly neutral in this population and are reflecting the overall impact of asymmetric gene flow that has resulted in about 13% admixture. However, there are many allelic outliers from this apparently neutral background M of 0.13. For example, the A allele at the ABO blood group locus shows a slightly negative M,but not significantly different from zero. Based on this allele alone, there seems to be no admixture at all. Waterhouse and Hogben (1947) have implicated the A allele as a cause of fetal wastage (spontaneous abortions) of
Was this article helpful?