Have we seen the face of Rama?

One of the problems with looking up pictures of the Kalash people of Pakistan is that photographers have a bias toward highlighting the most European-looking villagers. Let’s call this “Rudyard Kipling Lost White Races” syndrome. Therefore for your edification, I post the YouTube above which is probably more representative of what the Kalash look like.

The reason I post a link to what the Kalash look like is that it is germane to the answer to the question: what did the Indo-Aryans look like? The past tense is key since “Indo-Aryans” today means a lot of people in South Asia, in a literal sense.

In the post below Zach L. made a passing comment:

(1.) The AASI’s, which are sort of co-equivalent to the Negritos and Anadamese Islanders (one of the first coastal waves out of Africa that somehow also ended up in the Amazon). It’s interesting that they are substrate to every South Asian population (I think there are trace amounts in Central Asia, Afghanistan and even Iran).

(2.) the “Dravidian” farmers out of Iran. They are probably related to the J1/J2 types and might be an olive skinned population. Prominent in Sindh and Southern Pakistan through to South India (high % in Gujarat – must have been a locus of some sort).

(3.) our beloved Aryans who are especially prevalent among Brahmins, the Punjab and Haryana (though arguably the Haryanvis and East Punjab descend from Scythians to some extent). These look “European” but it’s a very different look to #2.

The Aryans are conventional European (light eyes, light hair, white skin) the ancient Dravidians would have (probably) looked like Middle Easterners (olive skin, dark hair dark eyes) and the AASI, ” looks like Papua New Guineans.

I can’t see any disagreement with point number two.

As for the AASI (“Ancient Ancestral South Indians”), we need to be careful here. They diverged from the ancestors of the people of Papua New Guinea ~40-50 thousand years ago. The divergence from the Andamanese, who probably migrated from mainland Southeast Asia, was not too much later. Aside from being very dark-skinned, the various extant “Australasian” people can be quite distinctive in appearance. The people of Papua, and native Australians, are quite robust. A substantial minority have blonde hair color due to a mutation common among Oceanians. The “Negrito” people of Southeast Asia and India all seem to be have adapted to a narrow relic niche, and may not be representative of their ancestors.

That being said, there is a particular non-West Eurasian look that many South Asians have which we can presume is the heritage of the AASI.

The comment about Aryans looking like Europeans raised my eyebrows a bit. This is a touchy subject, and to be honest my initial reaction was to be skeptical. But the more I read the primary literature to check up on Zach, the more reasonable this seemed to be. The dominant steppe signal into South Asia does resemble the people who were pushing into Central and Western Europe 1,000 years earlier than the Indo-Aryans, who were moving southward probably ~3,500 years ago. This is clear in rather simple statistical genetic analyses-populations such as the Kalash and Pathans for example show strong evidence of “European-like” gene flow.

Current work out of David Reich’s lab suggests that the Kalash are the best modern proxies we have for the “Ancestral North Indians,” the ANI. This population is modeled as:

– ~30% “steppe”, which is very similar to the ancestry which expaned westward into Europe between 3000 and 2500 BCE
– ~70% “Indus Periphery”, which seems the likely ancestral contribution of the people of the IVC, and is a heterogenous mix of Iranian-farmer and AASI

The mid-range estimate for the emergence of the Kalash mix is ~2,500 years before the present, but these usually have some downward bias, so it is reasonable that it would be greater than ~3,000 years. The samples from the Swat Valley dating to this period show gradual increase of “steppe” ancestry over time.

So one reason to be skeptical that the Indo-Aryans were “European-like” in appearance is that by the time they were flourishing in the lands previous inhabited by the IVC they may already have been more than 50% genetically like the people of the IVC. In which case, a minority would be very European-looking, but most would look vaguely West Asia, with some looking more stereotypically South Asian. If you look at the video above I think you do see the Kalash look this way.

One reason I’ve always been skeptical of the idea that the Indo-Aryans looked European, or, that their demographic impact was large, is that it seemed unlike both could be true. The expression of blue eyes among Indians was too low of a percentage.

Here is the frequency at a major SNP which predicts a lot of the blue vs. brown eye color.

What you see here is that the Kalash have the derived (“light”) variant at 25-30%. Notice that some Northern European populations are >75%.

Here are the frequencies from the 1000 Genomes:

I was a little surprise of the lack of variation from Punjabis (PJL), to Gujaratis (GIH), and Bangladeshis (BEB). Using the above logic the ~10% result would imply that a bit more than 10% European-like Indo-Aryan ancestry. This is reasonable.

But there are more SNPs than that that impact pigmentation. SLC24A5 is derived and fixed in Europeans, but pretty high frequency in South Asians (I have two homozygote derived copies and I’m rather brown). But some SNPs in SNP SLC45A2 are much more European specific in derived allele frequencies. So the 1000 Genomes surprised me somewhat:

Here you notice that the derived variant is nearly fixed in Northern Europe.  But in South Asian populations it’s not as high as you would expect. The frequency of OCA2 derived variant is higher than SLC45A2 in South Asia, while in Northern Europe it’s the opposite.

One explanation could be in situ selection in Northern Europe or in South Asia (or Central Asia). So these two markers suggest to me we can’t draw a straight line between physical affinity and total genetic ancestry/affinity.