Rapid evolution in Tibet?

[Heraclitus]

http://www.nytimes.com/2010/07/02/sc...ibet.html?_r=1

Tibetans live at altitudes of 13,000 feet, breathing air that has 40 percent less oxygen than is available at sea level, yet suffer very little mountain sickness. The reason, according to a team of biologists in China, is human evolution, in what may be the most recent and fastest instance detected so far.
RSS Feed
RSS Get Science News From The New York Times »

Comparing the genomes of Tibetans and Han Chinese, the majority ethnic group in China, the biologists found that at least 30 genes had undergone evolutionary change in the Tibetans as they adapted to life on the high plateau. Tibetans and Han Chinese split apart as recently as 3,000 years ago, say the biologists, a group at the Beijing Genomics Institute led by Xin Yi and Jian Wang. The report appears in Friday’s issue of Science.

....

The Beijing team analyzed the 3 percent of the human genome in which known genes lie in 50 Tibetans from two villages at an altitude of 14,000 feet and in 40 Han Chinese from Beijing, which is 160 feet above sea level. Many genes exist in a population in alternative versions. The biologists found about 30 genes in which a version rare among the Han had become common among the Tibetans. The most striking instance was a version of a gene possessed by 9 percent of Han but 87 percent of Tibetans.

Such an enormous difference indicates that the version typical among Tibetans is being strongly favored by natural selection. In other words, its owners are evidently leaving more children than those with different versions of the gene.

For those that bother to read the entire article. Archeologists are disagreeing about the dates since they claim 3000 years ago is way to late and that the plains have been settled long ago.

Archaeologists say they believe that the Tibetan plateau has been inhabited for at least 7,000 years and maybe for as long as 21,000 years.

“The separation of Tibetans and Hans at 3,000 years ago is simply not tenable by anything we know from the historical, archaeological or linguistic record,” said Mark Aldenderfer, a Tibetan expert at the University of California, Merced.

Dr. Aldenderfer said that there had probably been many migrations onto the Tibetan plateau, and that there was indirect evidence that pastoralists had entered the plateau from the north-northeast around 6,000 years ago. Earlier genetic studies have found that Tibetans are more similar to northern Han than to those from southern China, and have some admixture of genes from Central Asia, he said.

Rasmus Nielsen, a Danish researcher at the University of California, Berkeley, did the statistical calculations for the Beijing study. “We feel fairly confident that something on the order of 3,000 years is correct,” he said. But in a later e-mail message, Dr. Nielsen said, “I cannot with confidence rule out that the divergence time is 6,000 instead of 3,000.”

I don't know about this. My gut feeling is that isn't a new mutation because it has no long, linked haplotype around it that also differs in frequency in the two populations.

However from another source:

none of the models we fitted could explain the data with a divergence time much larger than 3000 years. If you look at the figure in the paper, you can see that there is an extremely strong correlation between the allele frequencies in Hans and in Tibetans. This is very difficult to explain with a long divergence time of genetically separated populations. To maintain such a strong correlation for a large amount of time, the Tibetan population (and the Han population) would have to be enormously large - and this is incompatible with the observed levels of variation in the population. We could not find a model that fit the data and which included a large divergence time no matter what we did. But there are of course many factors going into these estimates - including a calibration of number of mutations with the chimp, a number of demographic assumptions, and assumptions regarding generation times. If we are making errors on these assumptions - the estimates could change in one way or another. For that reason I feel it is most conservative to avoid arguing that our analysis definitely rejects that the divergence time could be 6000 years. The main objective of the paper was after all to investigate the evolution of altitude adaptation. The demographic analysis was there mostly to allow us to do the coalescence simulations - but we also used them to make the argument that this selection has occurred quite recently - and not say 10k or 20k years ago. It is quite clear from the data that such long divergence times cannot be supported by the data

This being said, we of course want to know if this short genetic divergence time is compatible with other evidence. I would argue that it is. There has been several migrations into Tibet. It is entirely compatible with the archaeological record that individuals living in Tibet today genetically mostly are descendants of migrants arriving around 3000 years ago even though the first migrants appeared much earlier. In terms of the selection - and when it has been acting - we want to determine when selection acted to increase the frequency on EPAS1 mutations in the ancestry of the individuals living in Tibet today. If they are genetically descendants of individuals migrating into Tibet just a few thousand years ago - then this is the relevant data for describing when selection has been acting on the EPAS1 mutations. As an aside I should also say that this has nothing to do with when the mutation(s) arose. Selection has in this case most likely been acting on standing variation.

This being said, we of course want to know if this short genetic divergence time is compatible with other evidence. I would argue that it is. There has been several migrations into Tibet. It is entirely compatible with the archaeological record that individuals living in Tibet today genetically mostly are descendants of migrants arriving around 3000 years ago even though the first migrants appeared much earlier. In terms of the selection - and when it has been acting - we want to determine when selection acted to increase the frequency on EPAS1 mutations in the ancestry of the individuals living in Tibet today. If they are genetically descendants of individuals migrating into Tibet just a few thousand years ago - then this is the relevant data for describing when selection has been acting on the EPAS1 mutations. As an aside I should also say that this has nothing to do with when the mutation(s) arose. Selection has in this case most likely been acting on standing variation.

Echos of the 10 000 year explosion? Jhon Hawkes (professor of anthropology at Wisconson) writes:

So it seems to me that the data must reflect the high incidence of recent selection within mainland China. This is exactly what we expect based on the real demography of massive population growth across the same interval and adaptation to post-agricultural ecologies. Although the headline of the paper is about high altitude adaptation in Tibet, the real story is the massive selection in China of other genes.

Did the Archaic Han picku up the gene from admixture with locals much like Eurasian Homo Sapiens picked up the 4% of Neanderthal genes because they where usefull, while selection worked against most of the other ones and rapid population growth of the Archaic Han meant any new sources of neutral admixture was vonureable to dissapearing?

The adaptations of the Tibetans are probably older in origin even if most of their ancestors did arrive just 2750 years ago. I base this on a comparison with the Andeans who's adaptations to high altitudes seems a more brute force solution. And the current model of history means they couldn't have been in the Andes longer than ~12k years.

[Barnabas]

Andean Amerindians have much less genetic diversity than Tibetans, all amerindians descend from a small number (perhaps less than 2000) siberian hunters.

[Heraclitus]

Andean Amerindians have much less genetic diversity than Tibetans, all amerindians descend from a small number (perhaps less than 2000) siberian hunters.

Goog catch, higher genetic diversity in Eurasia compared to American Natives could be the reason.

http://www.ncbi.nlm.nih.gov/pubmed/18039031

Abstract
We examined genetic diversity and population structure in the American landmass using 678 autosomal microsatellite markers genotyped in 422 individuals representing 24 Native American populations sampled from North, Central, and South America. These data were analyzed jointly with similar data available in 54 other indigenous populations worldwide, including an additional five Native American groups. The Native American populations have lower genetic diversity and greater differentiation than populations from other continental regions. We observe gradients both of decreasing genetic diversity as a function of geographic distance from the Bering Strait and of decreasing genetic similarity to Siberians--signals of the southward dispersal of human populations from the northwestern tip of the Americas. We also observe evidence of: (1) a higher level of diversity and lower level of population structure in western South America compared to eastern South America, (2) a relative lack of differentiation between Mesoamerican and Andean populations, (3) a scenario in which coastal routes were easier for migrating peoples to traverse in comparison with inland routes, and (4) a partial agreement on a local scale between genetic similarity and the linguistic classification of populations. These findings offer new insights into the process of population dispersal and differentiation during the peopling of the Americas.