The Hardy–Weinberg law assumes random mating in an infinitely large population; only when population size is infinite will the gametes carry genes that perfectly represent the parental gene pool. But no real population is infinitely large, and when population size is limited, the gametes that unite to form individuals of the next generation carry a sample of alleles present in the parental gene pool. Just by chance, the composition of this sample will often deviate from that of the parental gene pool, and this deviation may cause allelic frequencies to change. The smaller the gametic sample, the greater the chance that its composition will deviate from that of the entire gene pool.

479

The role of chance in altering allelic frequencies is analogous to the flip of a coin. Each time we flip a coin, we have a 50% chance of getting a head and a 50% chance of getting a tail. If we flip a coin 1000 times, the observed ratio of heads to tails will be very close to the expected 50 : 50 ratio. If, however, we flip a coin only 10 times, there is a good chance that we will obtain not exactly five heads and five tails, but perhaps seven heads and three tails, or eight tails and two heads. This kind of deviation from an expected ratio due to limited sample size is referred to as sampling error.

Sampling error arises when gametes unite to produce progeny. Many organisms produce a large number of gametes, but when population size is small, a limited number of gametes unite to produce the individuals of the next generation. Chance influences which alleles are present in this limited sample. In this way, sampling error may lead to genetic drift, or changes in allelic frequencies. Because the deviations from the expected ratios are random, the direction of change is unpredictable. We can nevertheless predict the magnitude of the change.

THE MAGNITUDE OF GENETIC DRIFT The amount of change resulting from genetic drift is determined largely by the population size (N): genetic drift is higher when the population size is small. For ecological and demographic studies, population size is usually defined as the number of individuals in a group. The evolution of a gene pool depends, however, only on those individuals who contribute genes to the next generation. Population geneticists usually define population size as the equivalent number of breeding adults, referred to as the effective population size (N_e).

THE CAUSES OF GENETIC DRIFT All genetic drift arises from sampling error, but sampling error can arise in several different ways. First, a population can be reduced in size for a number of generations because of limitations in space, food, or some other critical resource. Genetic drift in a small population over multiple generations can significantly affect the composition of a population’s gene pool.

A second way in which sampling error can arise is through the founder effect, which occurs when a population is established by a small number of individuals. The population of bighorn sheep at the National Bison Range, discussed in the introduction to this chapter, underwent a founder effect. Although the population can increase and become quite large, the genes carried by all its members are derived from the few genes originally present in the founders (assuming no migration or mutation). Chance events affecting which genes were present in the founders have an important influence on the makeup of the entire population.

A third way in which genetic drift arises is through a genetic bottleneck, which develops when a population undergoes a drastic reduction in size. An example is seen in northern elephant seals (Figure 18.5). Before 1800, thousands of northern elephant seals were found along the California coast, but hunting between 1820 and 1880 devastated the population. By 1884, as few as 20 seals survived on a remote beach of Isla de Guadalupe west of Baja California, Mexico. Restrictions on hunting enacted by the United States and Mexico allowed the seals to recover, and there are now an estimated 100,000 seals. All the seals in the population today are genetically similar, however, because they have only those genes that were carried by the few survivors of the population bottleneck.

THE EFFECTS OF GENETIC DRIFT Genetic drift has several important effects on the genetic composition of a population. First, it produces change in allelic frequencies within a population. Because drift is random, the frequency of any allele is just as likely to increase as it is to decrease and will wander with the passage of time (hence the name genetic drift). Figure 18.6 illustrates a computer simulation of genetic drift in five populations over 30 generations, starting with q = 0.5 and maintaining a constant population size of 10 males and 10 females. These allelic frequencies change randomly from generation to generation.

480

A second effect of genetic drift is reduction of genetic variation within populations. Through random change, an allele may eventually reach a frequency of either 1 or 0, at which point all individuals in the population are homozygous for one allele. When an allele has reached a frequency of 1, we say that it has reached fixation. Other alleles are lost (reach a frequency of 0) and can be restored only by migration from another population or by mutation. Fixation, then, leads to a loss of genetic variation within a population. This loss can be seen in the northern elephant seals just described. Today, these seals have low levels of genetic variation; a study of 24 protein-encoding genes found no individual or population differences in these genes. A subsequent study of sequence variation in mitochondrial DNA also revealed low levels of genetic variation. In contrast, southern elephant seals have much higher levels of mitochondrial DNA variation. Southern elephant seals, which are found in Antarctic and sub-Antarctic waters, were also hunted, but their population size never dropped below 1000; therefore, unlike the northern elephant seals, they did not experience a genetic bottleneck.

Given enough time, all small populations will become fixed for one allele or another. Which allele becomes fixed is random but is influenced by the initial frequency of the allele. If the population begins with two alleles, each with a frequency of 0.5, both alleles have an equal probability of fixation. However, if one allele is initially more common, it is more likely to become fixed.

A third effect of genetic drift is that different populations diverge genetically with time. In Figure 18.6, all five populations begin with the same allelic frequency (q = 0.5) but, because drift is random, the frequencies in different populations do not change in the same way, and so populations gradually acquire genetic differences. Eventually, all the populations will reach fixation; some will become fixed for one allele, and others will become fixed for the alternative allele.

The effect of genetic drift on variation among populations is illustrated by a study conducted by Luca Cavalli-Sforza and his colleagues. They studied variation in blood types among villagers in the Parma Valley of Italy, where the amount of migration between villages was limited. They found that variation in allelic frequencies was greatest between small isolated villages in the upper valley, but decreased between larger villages and towns farther down the valley. This result is exactly what we expect with genetic drift: there should be more genetic drift, and thus more variation among villages, when population sizes are small.

The three results of genetic drift (allelic frequency change, loss of genetic variation within populations, and genetic divergence between populations) take place simultaneously, and all result from sampling error. The first two results take place within populations, whereas the third takes place between populations. TRY PROBLEM 32