The Unz Review - Mobile
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 Gene Expression BlogTeasers
Conjectures About Southeast Asian Genetic History

Lipson, Mark, et al. “Reconstructing Austronesian population history in Island Southeast Asia.” Nature communications 5 (2014).

Austroasiatic-en.svgThere was a question below about the relationship of the Cambodians to the Vietnamese. Both these groups speak Austro-Asiatic languages. As can be inferred from the figure to the left, these languages are predominant in Indo-China. In particular, Cambodian and Vietnamese are both Austro-Asiatic. Traditionally, as noted in the comments, Cambodian is classed with the Mon language of Burma, under Mon-Khmer (as is faintly evident on the map Mon languages seem to have been spoken in what is today central Thailand before their replacement by Thai). There are also Austro-Asiatic languages on the fringes of southern China, though again the map is very illuminating, as the pattern of fragmentation is often indicative of marginalization and language replacement. Finally, there are Austro-Asiatic languages spoken in India, and among the indigenous people of central Malaysia, who are often termed Negritos (as well as the Nicobarese). As I have explored in depth elsewhere, there is now strong suggestive evidence from the genetics that the Austro-Asiatic languages are intrusive to South Asia from Southeast Asia. As the Negritos of the Phillipines also speak the language of nearby Austronesian agriculturalists, as do the Pygmy of the Congo, it seems likely that Malaysian Negritos received their language from agriculturalists.

51IZQjMbVlL._SX346_BO1,204,203,200_ Peter Bellowood fleshes out most of the details in First Farmers of how agricultural came to Southeast Asia (highly recommended, though it’s a little out of date now in some areas). At an archaeological site in northern Vietnam Bellwood describes burial grounds dating to 4,000 years in the past where two distinct groups are evident in the remains. One set of skeletons resembles modern East Asians morphologically, while the other element exhibits broad similarities to Near Oceanian peoples. He terms these “Austro-Melanesians.” Frankly, I think this is a confusing term. Though it seems likely that these groups are part of the broader range of populations which gave rise to modern Southeast Asian Negritos, like Papuans and Australian Aborigines they were in no way diminutive. So terming them “Proto-Negrito” would seem misleading. Therefore, I will term then “Ancestral Southeast Asians,” or ASA. The genetics points to the likelihood that as substantial minority of the ancestry of modern Southeast Asians derives from the ASA, in various quanta.

The best paper I know of in relation to the genetic history of Southeast Asia, maritime and mainland, is Reconstructing Austronesian population history in Island Southeast Asia. They used the PanAsian data set, which is somewhat thin on SNPs (<100,000), and also spotty in population coverage. The figure above shows one of the primary results. It seems that agriculture came to Southeast Asia in two major waves. First, with Austro-Asiatic peoples. And later, with Austronesians. The latter seem to have settled maritime Southeast Asia, where archaeological evidence of agriculture is thin to nonexistent before they arrived. But, as you can see from the figure many maritime Southeast Asian peoples also have signatures of Austro-Asiatic ancestry. The likely case then is that they picked this up en route, though there may also have been indigenous people in the islands when they arrived. But curiously, not in the east. There a Melanesian ancestral component is present, which has affinities to that contributed to modern Filipino ancestry from Negritos. The 2011 paper which posits two distinct elements before agriculture between mainland Southeast Asia and Papua would make sense of this pattern. The division probably followed Huxley’s Wallace Line.

As I said above, the PanAsian data set is spotty on population coverage. There are lots of obscure tribes, but not so much when it comes to the numerous people of mainland Southeast Asia. I have some data to probe these questions. Unfortunately not all of it is public, so I can’t release it (though some of it is from the 1000 Genomes, Estonian Biocentre, and HGDP, so you can find much of it it elsewhere).

The data set has 150,000 SNPs, with ~0 missingness (I just removed anything that had missing calls). I labeled samples from countries without ethnic provenance by those nation names. Additionally, I already did some preliminary outlier removal (e.g., removing Filipinos with non-trivial European ancestry, etc.).

Let me give you plots of PC 1 to 4 below. Click to enlarge.
Rplot06Rplot07

The first plot shows Indians and Papuans away from a cluster of Southeast and Northeast Asians. There are a lot of Southern Chinese from the 1000 Genomes, as well as Koreans. The Burmese are the first out toward Indians. The cluster that pushing itself toward Papuans are Filipinos. This makes sense in light of what we know bout Philippine Negritos. They are probably not descended from ASA, but rather a sister population, highly diverged, and with greater affinities to the peoples of Near Oceania. While the first plot shows PCs which separate both Southeast and Northeast Asians from Indians an Papuans, the second plot separates Southeast Asians among themselves. The north to south axis seems to align with a cline of Austro-Asiatics. The axis east to west runs toward Austronesians. Intriguingly there are three Indonesian samples which span the two axes, exactly in lines with the results of the paper above. Vietnamese and Dai are pulled more toward the Cambodians. Toward the top of the plot are Koreans, while the very dense cluster includes Southern Chinese, as well as assorted Southern Chinese ethnic minorities. There’s a few Malaysian samples in there. Unlike the Indonesians they are drawn much closer to the Southern Chinese cluster, but not quite in it. These may be Baba Chinese. I was surprised there weren’t more Overseas Chinese in these data. But there were some. It’s interesting that the Indians are close to the Chinese cluster, rather far from the Cambodians. I think that the Cambodian cline is probably indicative of ASA ancestry fraction.



To get more clarity a bigger plot with more labels:

Rplot08

The Lahu are a group that is in both southern China and Burma. You see the influence of geography, as they are right below the Burmese samples. Thailand is interesting, because one of the individuals is close to the Cambodians. The others sit atop the Dai and the Vietnamese. With only four individuals you can only say much. But I think we’re seeing a cline or structure within Thailand. Some regions are linguistically Thai, but have barely any genetic footprint from the Tai migrations. In other regions the impact of the newcomers might have been overwhelming. Until we get more samples from indigenous Thai we can’t say much. For all I know the three individuals clustering with the Dai and Vietnamese are of partial Chinese heritage.

Let’s look at the first plot zoomed in with some more labels:

Rplot09

You see now that one of the Indonesians seems to have a lot of Papuan-like admixture. They may be from the east. Another is very like the Burmese. And a few are clearly Overseas Chinese, and a last is shifted toward the Indians. The Burmese in these samples really seem to be two groups. I bet the ones further away from the Dai are ethnic Burmans or Mon. The others may be Shan, the descendants of Tai, at least in part.

I’m going to run TreeMix now. I will remove the Overseas Chinese, and the Malaysians. I’ll also remove the Thailand samples, as I don’t know whether they’re Chinese or not (I could run ADMIXTURE, but I’m tired). I rooted it with Papuan and did 5 migrations. Plots below.

SoutheastAsiaOut.6 SoutheastAsiaOut.7 SoutheastAsiaOut.8 SoutheastAsiaOut.9 SoutheastAsiaOut.10 SoutheastAsiaOut.1 SoutheastAsiaOut.2 SoutheastAsiaOut.4 SoutheastAsiaOut.5 SoutheastAsiaOut.3

OK, the strangest thing you’ll notice is that Burmese are placed near the Koreans, but they seem to have admixture from a population with a lot of Indian affinities. The Shan show this pattern, but much more mildly. I think this actually answers even more fully what’s going on among East Bengalis. The Burmese are a compound population, of newcomers from the north, rather deep into China, and the long resident Austro-Asiatic population. Mons. It was probably a pretty singular pulse event to get picked up so well. These later went to eastern India, and mixed with the proto-Bengalis, who probably already had some Austro-Asiatic Munda ancestry. Additionally, the connection between the Burmese and Naxi minority in China is not surprising; they speak related languages. The Miao are near the Southern Chinese. They are also quite closely related to the Hmong, who are a southern branch. The gene flow between the Filipinos and Indonesians makes sense. These are the two Austronesian groups. Finally, notice that the Dai are closest to the Chinese, then the Vietnamese, and most distant the Cambodians.

Let’s make the TreeMix plots cleaner by removing some of the groups:

SoutheastAsiaTOut.7 SoutheastAsiaTOut.8 SoutheastAsiaTOut.9 SoutheastAsiaTOut.10 SoutheastAsiaTOut.1 SoutheastAsiaTOut.2 SoutheastAsiaTOut.3 SoutheastAsiaTOut.4 SoutheastAsiaTOut.5

SoutheastAsiaTOut.6

Alright, what’s the upshot of these plots?

1) We see that the Cambodians are a hybrid of a population like the Dai of South China, and something somewhat Indian-like, but not totally.

ORDER IT NOW

2) The Vietnamese have a very faint gene flow from the Cambodians. Some of the samples which were Vietnamese look to me like they were ethnic minorities. I left them in because that is part of Vietnam’s genetics. They may simply be assimilated minorities or Khmer.

3) The gene flow into the Burmese is between Indian-like and Cambodian-like. It’s hard to parse out the distinctions, but there’s probably recent gene flow into Burma (I removed a few outliers as it is). Additionally, the ASA ancestry in the Cambodians may be higher than in the Burmese, because the latter have had more dilutions. But remember that I think there’s Indian ancestry in the Cambodians too.

4) The Negrito ancestry among Filipinos is pretty obvious in the gene flow.

 
• Category: Science • Tags: Southeast Asia 
Email This Page to Someone

 Remember My Information



=>
Commenters to Ignore...to FollowEndorsed Only
[]
  1. From the data you used, approximately what percentage of Filipinos had ‘non-trivial European ancestry?’ That has always been a burning question, because while on one hand, many Filipinos claim to be part-Spanish (the Philippines was a Spanish colony for 300 years), one study showed that only 3.6% of Filipino men have a Y chromosome from Europe (which of course would not account for, say, a man with European ancestry from his mother’s side). There apparently is a study by a Spanish academic that will reportedly try to gauge the amount of Spanish ancestry in Filipinos, but it has not come out yet.

    Read More
    ReplyAgree/Disagree/Etc.
    AgreeDisagreeLOLTroll
    These buttons register your public Agreement, Disagreement, Troll, or LOL with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used once per hour.
    Sharing Comment via Twitter
    http://www.unz.com/gnxp/conjectures-about-southeast-asian-genetic-history/#comment-1047632
    More... This Commenter Display All Comments
  2. I have a question here. Could there have been a gene flow from India to Burma in the ancient/early medieval period akin to the one in Cambodia ?

    I think there is some possibility of gene flow of Burmese like people into Bengal. Burma gets its name from the Burmans. This surname is found among Bengalis as well. However I do not know of other such connections.

    Read More
  3. Between these papers, looks like there been a lot more sampling of Southeast Asian insular minority populations in the past year with high SNP counts:

    Early modern human dispersal from Africa: genomic evidence for multiple waves of migration – Tassi – http://biorxiv.org/content/early/2015/07/20/022889.full-text.pdf+html

    Unravelling the Genetic History of Negritos and Indigenous Populations of Southeast Asia – Farhang Aghakhanian – http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4453060/ (Open access)

    Denisovan Ancestry in East Eurasian and Native American Populations. – Qin – http://biorxiv.org/content/early/2015/04/03/017475

    Maybe this will all be put together into something more “complete” for Southeast Asia in the future.

    One of the interesting things re: Southeast Asian and Australo-Melanesian affinities is that in mainland SE Asia the cranial form is in some ways like this and in some ways not –
    http://www.unsworks.unsw.edu.au/primo_library/libweb/action/dlDisplay.do?vid=UNSWORKS&docId=unsworks_1471 – Cranial variation of contemporary East Asians in a global context –

    “East Asians are distinguishable from non-Asians on the basis of their tall, round, vault, shorter cranial length, tall faces that are flattened in the upper and mid-facial regions, short malars (anteroposterior length), narrow interorbital breadth and orthognathism. A north-south East Asian cline was also detected, with the northern samples exhibiting tall, orthognathic faces, and a long low vault. This long, low vault shape is in contradiction to the purported shape of cold-climate adapted populations. Southern East Asians possess a tall, rounded vault and a short, projecting (prognathic) face. Island Southeast Asians inhabiting the Andaman and Nicobar Islands exhibit a ‘mixed’ morphology, possessing the southern East Asian facial form, but the long, low vault seen in northern East Asian samples. The long, low vault also characterises non-Asian samples from Australia, Africa and Melanesia.

    This is all somewhat odd in terms of the idea of a “Two Layer” Australo-Melanesian+East Asian synthesis in mainland Southeast Asia (in which “preceramic hunter-gatherers of Southeast Asia have morphology similar to Australian Aborigines and Melanesians, namely relatively long, narrow cranial vaults and robust faces and jaws”http://tinyurl.com/pufmq6l), if the most “East Asian” cranial vault shape seems to be found in those populations.

    Read More
  4. I have always wondered how thoroughly the Vietnamese exterminated the Chams when they conquered Champa. Is anything known about this and which ethnic group so the Austronesian speaking Chams most resemble genetically? I read a book suggestion the Vietnamese who colonised Champa absorbed elements of their culture. That would seem unlikelyif they wiped them out.

    Read More
    • Replies: @22pp22
    And now without the spelling mistakes. I got called away to the phone and missed the chance to edit.

    I have always wondered how thoroughly the Vietnamese exterminated the Chams when they conquered Champa. Is anything known about this? Which ethnic group do the Austronesian speaking Chams most resemble genetically? I read a book suggesting the Vietnamese who colonised Champa absorbed elements of their culture. That would seem unlikely if they wiped them out.
  5. I’m very interested by the genetic impact implied in the highlander conquest of Lowland Mainland SEA. Are the Irrawaddy basin Burmese basically Lolo-Burman speaking Mons, or are they already half Lolo-Burman by blood? The same question for the Bangkok Thai wrt Mons and Khmers.

    Read More
  6. This is of personal interest to me, as 23andme are telling me that I am of M33c ancestry, from a Szechuanese branch. Just like all those other Ashkenazim who have that mDNA…

    M33c is not a natively Han signature. It is a minority even in Hunan and Guangxi; 1.8%, is one number I’ve read. I’m looking at the other ethnic groups in south China – Burmo-Tibetans, Zhuang-Thai, and Hmong – and I’m seeing them as autochthonous to the region, or near-enough. If anything they are intrusive to south Asia: Burma, Thailand, and the north Vietnamese hills respectively.

    M33 happens to be endemic to eastern India and western Burma. Given its M33c offspring’s low numbers in China, this looks to me like a population of immigrants that never achieved mastery over the region. My fourth choice of language-families is Mon-Khmer-Munda.

    I should like to see a mDNA study of Bolyu-speaking villagers in Guangxi.

    Read More
  7. @22pp22
    I have always wondered how thoroughly the Vietnamese exterminated the Chams when they conquered Champa. Is anything known about this and which ethnic group so the Austronesian speaking Chams most resemble genetically? I read a book suggestion the Vietnamese who colonised Champa absorbed elements of their culture. That would seem unlikelyif they wiped them out.

    And now without the spelling mistakes. I got called away to the phone and missed the chance to edit.

    I have always wondered how thoroughly the Vietnamese exterminated the Chams when they conquered Champa. Is anything known about this? Which ethnic group do the Austronesian speaking Chams most resemble genetically? I read a book suggesting the Vietnamese who colonised Champa absorbed elements of their culture. That would seem unlikely if they wiped them out.

    Read More
    • Replies: @Razib Khan
    the vietnamese don't look very austronesian. the chams were closely related to the malays. i don't know if they wiped them out, or if the chams were not numerous. there are vietnamese samples that look khmer. but none that look filipino.
    , @terryt
    I don't think it's necessary to postulate extermination of Chams. It is most probable that Vietnam was already heavily populated by the time of the Austronesian expansion and so they were limited in where they could gain a foothold. It appears they were mainly confined to the delta region of the Mekong, which makes sense because the Austronesians were primarily a boating people.
  8. @22pp22
    And now without the spelling mistakes. I got called away to the phone and missed the chance to edit.

    I have always wondered how thoroughly the Vietnamese exterminated the Chams when they conquered Champa. Is anything known about this? Which ethnic group do the Austronesian speaking Chams most resemble genetically? I read a book suggesting the Vietnamese who colonised Champa absorbed elements of their culture. That would seem unlikely if they wiped them out.

    the vietnamese don’t look very austronesian. the chams were closely related to the malays. i don’t know if they wiped them out, or if the chams were not numerous. there are vietnamese samples that look khmer. but none that look filipino.

    Read More
  9. Most Muslims in Cambodia (a very visible minority in terms of dress and diet) identify as Cham, they can still speak Cham and consider themselves the descendent of Champa. Interestingly, Malaysia was a main port of call for Cham refugees during the Khmer Rouge and Cambodia-Vietnam war times; subsequently a lot of Chams married into Malay families.

    Read More
  10. Chams were surrounded by AA Bahnaric-speaking peoples though. Many Cham languages are phonologically quite Bahnarized, suggesting intensive cultural, and why not genetic exchange.

    Read More
  11. @22pp22
    And now without the spelling mistakes. I got called away to the phone and missed the chance to edit.

    I have always wondered how thoroughly the Vietnamese exterminated the Chams when they conquered Champa. Is anything known about this? Which ethnic group do the Austronesian speaking Chams most resemble genetically? I read a book suggesting the Vietnamese who colonised Champa absorbed elements of their culture. That would seem unlikely if they wiped them out.

    I don’t think it’s necessary to postulate extermination of Chams. It is most probable that Vietnam was already heavily populated by the time of the Austronesian expansion and so they were limited in where they could gain a foothold. It appears they were mainly confined to the delta region of the Mekong, which makes sense because the Austronesians were primarily a boating people.

    Read More
    • Replies: @22pp22
    Historic Champa was in central Vietnam. The Mekong Delta was Cambodians speaking until the late sixteenth- early seventeenth centiry.
  12. Hi Razib,
    Given that the Austronesians expanded through the Philippines on their way to Indonesia and Malaysia, is there any indication of a “Philippine Negrito-like” component in the ancestry of the Indonesians, Malaysians and the Chams? A interesting related question IMHO is, when did significant admixture start to occur between the Austronesian migrants and Austro-Melanesians in the Philippines? There has been no report of any such admixture showing up outside of the Philippines from the posts and papers I have been able to read but I might have just missed it. It might be that expansion through the Philippines had been a rapid process with admixture occuring somewhat later.

    Read More
    • Replies: @terryt
    @ 22pp22:

    Thanks for that information. However it does seem strange that a language that was well past its major expansion in the region (Austronesian) should replace a language that was actively expanding at the time (Cambodian). You may not have seen this post of Maju's from some time back:

    http://forwhattheywereweare.blogspot.co.nz/2012/05/cham-people-and-why-genetic-structure.html

    " It might be that expansion through the Philippines had been a rapid process with admixture occuring somewhat later".

    If the Austronesia expansion was anything like it was into and beyond New Guinea they would have occupied uninhabited islands as a first choice. The prehistoric discovery of uninhabited islands has been described as the prehistoric equivalent of winning the lottery.

  13. @terryt
    I don't think it's necessary to postulate extermination of Chams. It is most probable that Vietnam was already heavily populated by the time of the Austronesian expansion and so they were limited in where they could gain a foothold. It appears they were mainly confined to the delta region of the Mekong, which makes sense because the Austronesians were primarily a boating people.

    Historic Champa was in central Vietnam. The Mekong Delta was Cambodians speaking until the late sixteenth- early seventeenth centiry.

    Read More
  14. @roelm
    Hi Razib,
    Given that the Austronesians expanded through the Philippines on their way to Indonesia and Malaysia, is there any indication of a "Philippine Negrito-like" component in the ancestry of the Indonesians, Malaysians and the Chams? A interesting related question IMHO is, when did significant admixture start to occur between the Austronesian migrants and Austro-Melanesians in the Philippines? There has been no report of any such admixture showing up outside of the Philippines from the posts and papers I have been able to read but I might have just missed it. It might be that expansion through the Philippines had been a rapid process with admixture occuring somewhat later.

    @ 22pp22:

    Thanks for that information. However it does seem strange that a language that was well past its major expansion in the region (Austronesian) should replace a language that was actively expanding at the time (Cambodian). You may not have seen this post of Maju’s from some time back:

    http://forwhattheywereweare.blogspot.co.nz/2012/05/cham-people-and-why-genetic-structure.html

    ” It might be that expansion through the Philippines had been a rapid process with admixture occuring somewhat later”.

    If the Austronesia expansion was anything like it was into and beyond New Guinea they would have occupied uninhabited islands as a first choice. The prehistoric discovery of uninhabited islands has been described as the prehistoric equivalent of winning the lottery.

    Read More

Comments are closed.