Razib Khan One-stop-shopping for all of my content

July 31, 2020

Humanity’s second “cradle” in Southeast Asia

Filed under: Flores,Hobbit,Human Evolution,naledi,Southeast Asia — Razib Khan @ 11:31 am
Pleistocene Sundaland

This week on The Insight (Apple Podcasts, Spotify, Stitcher, and Google Podcasts) Razib and Spencer talk about a topic which they have visited before, the role of Southeast Asia in understanding the origin of our species.

A generation ago our understanding of human origins was simple, clear, and elegant. The story was that 50,000 years ago a small group of East Africans expanded all across the world and replaced everyone else. Neanderthals were “dead ends.” All the other human lineages were pruned by the expansion of Homo sapiens.

Today, the reality that we know from ancient DNA and fossils is more complicated. 55,000 years ago Neanderthals mixed into the lineage that led to all people outside of Africa. 50,000 years ago a newly discovered human lineage, the Denisovans, seem to have mixed into the people who spread out across eastern Asia and Oceania, and into the New World. And in Oceania, it seems that multiple additional mixtures of Denisovans into the local peoples occurred!

While Neanderthals were a well-known part of the human family tree, spare as it was at the turn of the century, the Denisovans were entirely new to science. Initially known only through their genome, they are closer to the Neanderthals than they are to modern humans. The ancestors of the Neanderthals and Denisovans seem to have left Africa 750,000 years ago, and separated from each as the former moved north and west, while the latter occupied East and Southeast Asia. While the Neanderthals seem to have been a homogeneous population from the Atlantic to the Altai, the genetic legacy of the Denisovans indicates that they branched off into separate groups rather early.

But wait, there is more! In the 21st century three, seemingly small hominins have been discovered through fossils. Homo floresiensis, the “Hobbits”, Homo naledi, and Homo luzonensis. Two of these, the Hobbits and the Luzon humans, are flourished in Pleistocene Southeast Asia, on the islands of Flores in Indonesia, and Luzon in the Phillippines. Southeast Asia is also the site of the discovery of Java man, one of the earliest remains of our human lineage, Homo erectus.

While Africa is undoubtedly the cradle of humankind, with the most ancient precursors of our species present in that continent, and repeated migrations outward, such as that of erectus, the ancestors of Neanderthals and Denisovans, Southeast Asia seems to be the second locus of diversity and human evolution. Not only is Southeast Asia a center of hominin diversity, like Africa it is also the home of many ape species, from three kinds of orangutans, and 20 species of gibbons, the “lesser apes.”

Homo naledi

The reason that Southeast Asia is special probably has to do with its similarity to Africa, and its uniqueness in relation to other regions of the world. During the Pleistocene period, Southeast Asia was climatologically a microcosm of Africa, with vast expanses of savanna at the heart of the core of Sundaland, and rich tropical rainforests in the highlands of what are today Borneo and Sumatra. Biogeographically this promontory of Asia has the climatic and ecological regime most similar to Africa, so it should be no surprise that African species with the ability to migrate long distances rapidly colonized the region.

But Southeast Asia is also unique, as rising and falling sea levels produces a fragmented landscape of islands and maritime fringes. Even during the height of the Ice Age, while Sundaland rose in the west and the merged continent of Australia and New Guinea, Sahul, existed in the east, deep trenches separated the two ecozones. Islands such as Flores were always isolated and remained islands throughout the whole period.

A “Hobbit”

This fragmented aspect of the region is the reason that the megafauna of Southeast Asia exhibits both gigantism and dwarfism. The Komodo Dragons of Flores seem like something out of the Mesozoic, while the dwarf elephants which the hobbits once hunted were no doubt quaint to the modern humans who may have interacted with them until 12,000 years ago. Perhaps even more farcical were the “giant” rats.

A new paper suggests that the best way to find Denisovan remains in Southeast Asia is to track where the other megafauna are copious because ultimately that is what humans are, Pleistocene megafauna. The researchers looked deep in the genomes of modern humans in the region, and only found Denisovan ancestry, rather than the deeper heritage one might expect from very different hobbits and Luzon humans. If there is only the ancestry of Denisovans, where are they? It turns out that both Sulawesi and the island of Mindoro off the coast of the Phillippines have large mammals, and, have not been extensively explored. That is where they suggest researchers go digging and excavating.

But what about another possibility: could the Hobbits and Luzon humans have been “Denisovans”?

Despite the physical differences between hobbits and modern humans, it is not impossible that the exigencies of island life and the splendid isolation of Flores resulted in rapid and striking adaptations in these humans. If the hobbits and Luzon humans were Denisovans, they may have been very physically difference from the Denisovans of Siberia, who seem to have been larger and more physically robust, judging by their skulls and teeth. But then again, the genetics makes it clear that Denisovan lineages were separating even before the emergence of anatomically modern humans! Why should physical diversity be surprising?

Southeast Asia, with its forests, mountains, and even desert islands, is a laboratory of human evolution. Unlike Africa, the landscape is shattered by oceans and separated by deep trenches. The waxing and waning of sea levels during the Pleistocene also had a much larger impact on Southeast Asia, as Sundaland rose and fell several times. What we see in the region is a vast evolutionary experiment.

The next few years will see many discoveries in the area of human evolution. It seems entirely likely most of those discoveries will occur in Africa, and, Southeast Asia.


Humanity’s second “cradle” in Southeast Asia was originally published in Insitome on Medium, where people are continuing the conversation by highlighting and responding to this story.

January 25, 2020

The Indian admixture into Southeast Asia is not just a function of distance

Filed under: Human Population Genetics,Southeast Asia — Razib Khan @ 10:41 pm

In the comments to the post below about Indian ancestry in Thailand, some observed that this should not be surprising due to reciprocal gene flow and proximity. Implicitly, I think what is being suggested here is that there is isolation by distance and continuous gene flow. Obviously some of this is true, but there details here which suggest that it is simply not just geography at work.

The reason I was curious about the Dusun people in coastal Borneo is that while Malays all seem to have Indian ancestry, many tribal Austronesian groups in maritime Southeast Asia do not. The Indian admixture into the Malays is not just recent. Some of it seems quite a bit older than the colonial period.

In the context of Southeast Asia, it seems that some of the more ancient Austro-Asiatic people, in particular, the Mon and Khmer, have Indian ancestry, and groups which mixed with Austro-Asiatic substrates, such as Burmans and Thai, also have this.

Additionally, some groups in the northeastern states of India have less “Indian” admixture than the Thai and Khmer. To show this, see this PCA:

The Garo and Naga live in India (some Garo are in Bangladesh). The “East Indian” samples seem to be mostly Mizo. Of course, some of these groups are intrusive to the northeast. But still. Here are admixture and TreeMix:

The issue in Southeast Asia is that ethnolinguistic groups are the product of several syntheses and migrations. Most of what is today “Thailand” was the domain of Mon-Khmer people in 500 AD. Most of the ancestry seems to date to that period, though there was an overlay from Tai people to the north. In Burma the population is a synthesis of Burman elements with connections to northern East Asians, and Austro-Asiatic people such as the Mon in the south. Additionally, a later movement of Tai people also occurred in Burma (the upland Shan). In Vietnam, the Kinh moved south and seem to have replaced the indigenous Chams and Khmer (there is very little Indian-like ancestry in any Vietnamese samples).

When looking at the map the plausible route of gene flow is clearly from the northeastern part of the subcontinent overland. But several people have pointed out to me that this is very difficult terrain. Recently, I have been convinced that a maritime intrusion of Munda languages into Odisha is plausible. One of the potential points of departure for the proto-Munda is the Tanintharyi region of today’s far southern Burma, adjacent to southern Thailand. I propose that the Tanintharyi region served as a cultural and demographic valve, initially mediating overseas expansion by a group of Southeast Asian rice farmers, who eventually established connections across the Bay of Bengal between South and Southeast Asia.

The absorption of lowland Munda domains in Odish by Indo-Aryan speakers did not entail the disruption of the flow of goods and people in the preestablished trade network. Rather, these routes which were preexistent were co-opted by the Kalinga state, and later on by various southern Indian polities facing east on the Bay of Bengal. Inspection of the Y and mtDNA haplogroup profile suggests these were elite Indian males, with few females. This is very different from a folk expansion through Arakan, which would involve both and women.

January 21, 2020

Indian Y chromosomes in Thailand

Filed under: Southeast Asia — Razib Khan @ 11:58 pm

The region of modern Thailand has gone through a major cultural shift over the last 1,000 years. Today the zone of Austro-Asiatic speech in mainland Southeast Asia is fragmented. To the east, there are the Khmer people of Cambodia, as well as various “hill-tribes” in Thailand and Laos who also speak Austro-Asiatic dialects. To the west, there are the Mon people of Burma.

But around 1000 A.D. the whole zone from the India ocean out toward the Mekong was dominated by Austro-Asiatic peoples. Modern-day Thailand was dominated by the Dvaravati polity, of which little is known, but possible Mon associations are assumed.

I have posted several times about the reality that it seems the whole zone between Burma and Cambodia seems to be impacted by a non-trivial proportion of South Asian (Indian) ancestry. A new preprint has a lot of Y chromosomes from various groups in Thailand. Below are frequencies I pulled out of two ethnic groups with large sample sizes (from table 3 in the supplements):

R1a+RLJ2HSample Size
Mon15%2%5%2%105
Central Thai13%0%3%5%129

These lineages are clearly more evidence of Indian males settling in this region.

January 9, 2020

Indian Ancestry In Thailand During the Iron Age

Filed under: History,Southeast Asia — Razib Khan @ 11:30 pm


A follow-up to my previous post, one of the “Iron Age” samples from Thailand seems a definite outlier in comparison to the other Iron Age and Bronze Age samples. There is suggestive evidence again of Indian ancestry, as one sees in the plot above. One of the samples from Thailand overlaps with the Cambodians and Burmese, who do seem to have South Asian shift, while the other samples from Thailand do not. Today most Thais seem to show some Indian ancestry as well, at low levels.

Unfortunately, much of Southeast Asian history before 1000 A.D. is pretty much a cipher. Perhaps the best survey I’ve seen is Strange Parallels: Volume 1, Integration on the Mainland: Southeast Asia in Global Context, c.800–1830, though even there it’s rather thin before the arrival of the Tai and the shocks that entailed for the earlier Indic societies of Southeast Asia.

January 5, 2020

Indian ancestry in Cambodia was present ~2,000 years ago

Filed under: Mainland Southeast Asia,Southeast Asia — Razib Khan @ 10:55 pm

When STRUCTURE-style bar plots first emerged using the HGDP Cambodian samples, there were often strange residual components with affinities to South Asians. When Treemix was developed there were strange edges between South Asians and Cambodians. In discussions with Joe Pickrell, the author of Treemix, we both adduced this must be due to deep affinities to “Ancestral South Indians” (ASI). Though Cambodia had “Indic” cultural affinities, the standard model is that this was due to cultural diffusion, not gene flow. Then Spencer Wells told me that The Genographic Project had detected that many Cambodian males seem to carry the R1a1a lineage. Looking at the literature, several Southeast Asian groups carry West/South Eurasian haplogroups which are likey Indian-mediated (R1a, R2, and J2, to name three). The enrichment is notable in groups like the Thai and Khmer which are located at some distance from South Asia.

Out of curiosity, I decided to look at the “Cambodian Iron Age” sample from a recent ancient DNA paper. This sample dates to 100 to 300 A.D., the period of ancient Funan, which we know mostly though not exclusively through Chinese sources:

According to modern scholars drawing primarily on Chinese literary sources, a foreigner named “Huntian” [pinyin: Hùntián] established the Kingdom of Funan around the 1st century CE in the Mekong Delta of southern Vietnam. Archeological evidence shows that extensive human settlement in the region may go back as far as the 4th century BCE. Though treated by Chinese historians as a single unified empire, according to some modern scholars Funan may have been a collection of city-states that sometimes warred with one another and at other times constituted a political unity.

Look at the Iron Age sample it does seem it is notably “Indian-shifted” even compared to modern Cambodians. This could just be an artifact of ancient DNA, but when I looked at a few dozen ancient Vietnamese samples, only one exhibited this same pattern of being Indian-shifted. Reducing the dataset to the 55,000 SNPs that came back on this ancient sample, you see the result above (many of the modern samples don’t have the full complement of these SNPs).

Something on the order of ~5-10% of the ancestry of many Southeast Asian groups seems to be of Indian origin. Looking at the Malays in the Singapore Genome Project, some of them have clear recent Indian ancestry, but even removing all of those you see notable Indian-shift, just as you see with the Cambodians. In contrast, Vietnamese and Dayaks from Borneo don’t show any evidence of such admixture. Neither do samples from the Phillippines.

The question is when this admixture occurred then. A large number of Indians migrated to Southeast Asia during the colonial period to Malaysia and Burma. But some preliminary analysis suggests to me that this doesn’t account for all of the Indian ancestry there. And, it can’t account for Cambodia and Thailand at all (though there aren’t too many genome-wide samples from Thais, the Y chromosomes show the same pattern as the Khmer).

Over time the genetic data is going to coalesce and converge on the details, though I think we see where it’s pointing. At that point, it’s up to archaeologists and historians to make sense of it. This includes scholars of South as well as Southeast Asia. The genetic imprint of South Asians in Iran and Central Asia is rather modest compared to what one sees in Southeast Asia, so it’s an interesting contrast as to why.

Older Posts »

Powered by WordPress

Do NOT follow this link or you will be banned from the site!