## N

size of N = 10 for a range of allele frequencies. The standard error of the allele frequency decreases as the allele frequency approaches fixation or loss. In the same way, genetic drift is less effective at spreading out the distribution of allele frequencies as alleles approach fixation or loss. The standard deviation is 0 when the allele frequencies are 0 or 1 since there is no genetic variation and any size sample will faithfully reproduce the allele frequencies in the source population.

when one allele is very nearly fixed except for one copy of the alternate allele (p = 1 - -

2N 2N

drift has a reasonable chance of only several sampling errors, such as no, one, two, or three copies of the low-frequency allele. Sampling error that causes fixation of the high-frequency allele is quite likely. However, sampling error that results in greatly increased frequencies of the low-frequency allele in one generation would be very, very unlikely.

There is a graphical metaphor to summarize the consequences of initial allele frequency for the range of outcomes in allele frequency under genetic drift. Figure 3.8 shows the range of possible allele frequencies (0-1) in a population and indicates the effects of genetic drift by the width of the arrows and the vertical scale. The range of outcomes probable under drift depends on the allele frequency in a population. When both alleles are equally frequent (pq = 0.25, its maximum), the sampling error is the largest so that the rate of genetic drift in changing allele frequencies is greatest. When the allele frequency is closer to fixation or loss (pq < 0.25), the sampling error is smaller and the rate of genetic

## Post a comment