The Holocene origins of Indians nearing conclusion

At the Eurogenes blog there has been a lot of analysis of South Asian genetic history in light of ancient DNA recently. Part of this is probably due to the fact that “Euro” genes (that is, the genetic history of European peoples) are now understood to be inextricably tied to demographic pulses and shifts which are deeply rooted in Eurasian cultural revolutions over the past 4 to 10,000 years. Only a very small fraction of the ancestry of modern Europeans dates to the period before the Last Glacial Maximum ~20,000 years ago; at least according to ancient DNA. And most of the ancestry is conditional on events which occurred during the Holocene, the past 10,000 years.* To give a sense of how recent all this is, when the first cuneiform tablets were being written, the demographic-genetic revolution which was to begin the transformation of Northern Europeans into what we now know as Northern Europeans had not completed itself, and in many regions not even begun (e.g., the Swedish-Battle Axe culture begins in 2800 BCE, several hundred years after the earliest writing in Sumerian).

They say you need two hands to clap. And India is the other hand that we have now when comes to understanding this process. There’s now a lot of circumstantial evidence to tie Indians to Europeans in moderately complex ways. To the left is a figure from the supplements of Punctuated bursts in human male demography inferred from 1,244 worldwide Y-chromosome sequences. You see that the phylogeny of R1a in South Asia really is a burst. There’s just not that much genetic difference between all of us South Asian R1a males, irrespective of region and caste. We’re pretty much all on a particular branch, Z93. Rare outside of South Asia today (it is found in the Altai and in Central Asia), it has been discovered in ancient males from the Srubna culture of eastern Ukraine ~4,000 years ago. In Europe R1a is found mostly in Northern Europe, and especially Eastern Europe. And yet if you look further up in the supplements you see that for haplogroup J2 most of the males are partitioned between South Asians and Southwest Europeans. Additionally, the two large branches of J2 have both South Asians and Southwest Europeans, suggesting that divergence in J2 predates the arrival of J2 bearing males to South Asia and Southwest Europe. Finally, unlike R1a you can see visually that the phylogeny of J2 is less explosive; there are more clean sequences of mutations to differentiate the various branches of this patrilineage. J2 is common, but it did not undergo nearly the same burst as R1a. It’s prevalence is due to more continuous and gradual demographic processes.

What does this mean? Various genome bloggers have been arguing since the late aughts that there seem to be two West Eurasian admixtures into South Asia, who are clearly a compound of West Eurasians and non-West Eurasians. This was supported to some extent by the 2013 publication of Genetic Evidence for Recent Population Mixture in India, which found evidence for more than one admixture event in a subset of populations. Then, ancient DNA from the Caucasus added more evidence, Upper Palaeolithic genomes reveal deep roots of modern Eurasians. If you dig deep into the paper you see that Indian populations which are likely to be due to one(ish) admixture event are best modeled as a synthesis between the Kotias Caucasus hunter-gatherer and a non-West Eurasian population (i.e., Ancient South Indians or ASI). But some groups, such as high caste North Indians, seem to be better modeled with ancient Eurasian steppe groups as the source population (these groups themselves have ancestry from a Kotias-related/descended group).

In 2016 the ultimate judge is going to be ancient DNA. In the next year or so I think it will tell the tale that we’ve been hearing in the winds of modern contemporary genetic variation. What I think is that it will confirm part of the narrative and model pushed forward in First Farmers, Peter Bellwood’s book from the middle aughts. The Dravidian languages were brought to India from West Asia in the early-to-middle Holocene by agriculturalists descended from or related to the hunter-gatherers of the Caucasus. They mixed with indigenous hunter-gatherer populations, and gave rise to the first people we would recognize as modern South Asians genetically. Eventually they built what we term the Indus Valley civilization. The relatively evenness of this mix between West Asian descended Ancestral North Indians (ANI) and ASI across South Asia is due to the fact that much of the subcontinent was sparsely populated, and the farmers demographically overwhelmed the indigenous groups. The fact the non-Brahmin upper and middle castes are genetically somewhat different from tribal populations and Dalits in South India is probably due to the fact that the indigenous populations were absorbed at the lower levels of the nascent civilization.

The arrival of Indo-Europeans may or may not have been an “invasion” in a classical sense, but it was highly disruptive. The phylogeny of R1a in South Asia is strongly indicative of an incredible reproductive advantage for males bearing this haplogroup. In fact, R1a is more expansive than Indo-Aryan languages, and is found across language and caste, including among tribal populations. Previously this was argued as a reason for why R1a in South Asia must be old and indigenous to South Asia. Next generation sequencing of the Y chromosome has looked closely, and that is unlikely. The expansion of R1a, and the South Asian branch, is very recent. It hints at cultural processes of male domination and elite diffusion of lineages which we do not have a good theoretical grasp of at this moment.

But this is not the end of the story. I have spoken only of West Eurasians. What of the other half of the ancestral glass, the ASI? I have not explored this literature in detail, but there is now suggestive evidence I believe that ASI themselves may have been intrusive to the subcontinent, perhaps as hunter-gatherers migration out of Pleistocene Southeast Asia. The closest modern population to the “pure” ASI ghost group are the Andaman Islanders, and they arrived where they are today not from the Indian subcontinent, but from Burma. There is now a modest amount of evidence through various angles that the ancestors of the Munda people of South Asia must have arrived as part of the Austro-Asiatic agricultural migrations from what is today interior South China. They are not primal. There is no reason to think that that this was the first eruption of humans from this region into South Asia. Those with more understanding of paleoclimatology need to weigh in, but it may be that in the drier conditions of the Pleistocene South Asia had a naturally smaller population than Southeast Asia, so that the latter was always going to be a source and the former a sink, in terms of demographics.

In any case, if the model fits, eventually a preprint you must submit. I think the age of speculations will give way to the age of understanding, though I have no inside information at this point….

* To unpack this, take a conventional idea of Europe today, bounded by Gibraltar on the southwest, the Bosporus on the southeast, and in the far east the Urals and far southeast the Caucasus. The majority of Southern European ancestry seems to derive from Early European Farmers (EEF), and these almost certainly are from Anatolia. In Northern Europe the ancestry is a compound of EEF, Indo-Europeans from the east, and probably residual hunter-gatherers (some researchers believe nearly all of the hunter-gatherer ancestry in Northern Europe is actually from the Indo-Europeans and EEF, with no true relic populations by 5,000 BC). Of the original Yamnaya Indo-European ancestry, the ~25% or so that is Ancestral North Eurasia (ANE) probably is extra-European in provenance. Specifically, a migration out of central Siberia or thereabouts. Nearly half of the remaining is similar to EEF, but we now know it is likely derived from a different Near Eastern farming populations which descend in part from hunter-gatherers from the Caucasus. The balance is from indigenous European hunter-gatherer ancestry which amalgamated with ANE. This last portion is indigenous to Europe broadly construed before the Holocene. I haven’t done the detailed math, but it seems difficult then to imagine a scenario where anything by a minority of the ancestors of modern white indigenous Europeans actually lived in Europe during the late Pleistocene.