| || |
| Variation Within and Among Human Populations |
At the DNA level, humans are both very similar to and very different from one another. On average, pairs of individual humans share 99.9 percent DNA sequence identity. Due to the sheer size of our genomes, however, we possess numerous differences from one another. The human genome consists of just over three billion nucleotides; that 0.1 percent of difference represents three million variants between the average pair. The vast majority of these variants have no functional significance. However, even if one in a thousand did, that would still mean that we would each differ at thousands of functionally important sites.
How does this variation compare with that of other species? Humans actually have less genetic variation than do their closest relatives. For instance, the average difference between two randomly selected chimpanzees is roughly four times greater than between two humans. This is, at first glance, surprising. Based on population genetic theory, levels of genetic variation within species should correlate positively with population size. This predicted correlation comes about because the strength of random genetic drift - which results in the loss of genetic variation - increases at lower population sizes. Yet, the human population numbers in the billions and the population sizes of chimpanzees and gorillas is fewer than a hundred thousand.
What could explain that discrepancy? The strength of genetic drift is dependent not on the current census population size but on the historical population sizes. The relatively low levels of genetic variation in humans can be explained by a severe, but short-lasting, population bottleneck, where the population of our species was likely reduced to a few thousand. It could also be explained by a more moderate, sustained bottleneck. During this bottleneck, the population was possibly in the tens to hundreds of thousands for a more considerable time. Alternately, natural selection could also either increase or decrease the extent of variation in one of the species. Yet, because it is unlikely that natural selection would act in the same way on multiple regions of the genome, the difference in the extent of genetic variation between humans and chimpanzees is more likely a consequence of historical demography.
How is this variation partitioned according to known racial groups? During the 1970s the then state-of-the-art technique of electrophoresis of protein variants showed that around eighty to ninety percent of human genetic variation was within ethnic populations, five to ten percent was among ethnic populations within the major racial groups, and only about five to ten percent was among the major racial groups. In other words, "if everybody on earth became extinct except for the Kikiyu of East Africa, about eighty-five percent of all human variation would still be present in the reconstituted species"2. More recent analyses of DNA sequence data strongly confirm the results of earlier protein electrophoresis studies. In both the protein electrophoresis and the DNA sequence studies, the differences between racial groups are generally ones of frequencies and not kind. The situation of "fixed differences" - in which all individuals in one group have variant A and all individuals of another group have variant B - is extremely rare in humans. Instead, groups vary by having different frequencies of genetic variants. There are cases of "private alleles," however, where genetic variants are found in low to intermediate frequencies in some populations, but are virtually absent from others.