Through time AND space?

Ever since the publication of Lazaridis et al. 2016, the comments section here has seen regular debates about the nature and source of steppe-related ancestry in South Asia.

According to mixture models featured in that paper, the populations that brought steppe ancestry to South Asia probably lacked early European farmer (EEF) admixture. In other words, they were more like the samples from Early to Middle Bronze Age (EMBA) cultures Yamnaya, Afanasievo, and Poltavka, than those from Middle to Late Bronze Age (MLBA) cultures Sintashta, Andronovo and Srubnaya.

This of course poses a major dilemma to those of us interested in early Indo-European expansions, because the consensus amongst historical linguists is that Indo-Iranian languages were introduced into South Asia during the Late Bronze Age from the Andronovo horizon.

So how do we reconcile ancient genomics with historical linguistics in this case? Should we assume that the linguists are way off, and posit that Indo-Iranian languages were introduced into South Asia straight from the Poltavka or even Yamnaya culture, much earlier than generally accepted?

Not necessarily.

Lazaridis et al. 2016 identified three post-Poltavka steppe individuals in their dataset that lacked EEF ancestry and were thus more similar to samples from Poltavka than Andronovo: Potapovka I0246, Potapovka I0418 and Srubnaya_outlier I0354. So where did these outliers come from and how is it that their steppe ancestors managed to stay free of EEF admixture?

One possible explanation is that most of the population on the MLBA steppe didn't carry significant levels of EEF admixture, because it was largely limited to the elites. As a result, we might be getting a skewed picture of the genetic structure of the steppe at this time, because for obvious reasons the vast majority of Bronze Age steppe samples being tested are from the best preserved burials, which are usually elite Kurgan burials.

So why would these elites harbor EEF ancestry and the commoners lack it? Perhaps because the former migrated from deep within the European part of the steppe, and imposed their culture on populations derived from, say, Afanasievo, Catacomb, Poltavka and late Yamnaya? Potential evidence of such an expansion exists in the form of chariot burials with similar horse cheek pieces found all the way from the Carpathian Basin to Central Asia (refer to the third map from Allentoft et al. 2015 here).

In any case, one way or another Poltavka-like people managed to survive on the steppe, perhaps in considerable numbers, well into the Andronovo period and probably beyond. So considering that this type of genetic structure was transmitted on the steppe across the millennia, then why not also across space into South Asia?

Interestingly, it's often claimed that some of the rituals described in the early Indo-Aryan Rig Veda hymns are very similar to the Kurgan burial rituals practiced by Potapovka people (see here). This is open to interpretation and impossible to prove, but I can test whether the above mentioned three post-Poltavka steppe outliers, including the two Potapovka individuals, show the right type of genetic structure to be potentially ancestral to modern-day South Asians.

So using the qpAdm algorithm let's test a model in which the descendants or close relatives of these three samples, labeled as Potapovka2-Srubnaya_outlier, move into the Andronovo horizon and then onto South Asia, contributing significantly to the genetic structure of modern-day South Asians.

Balochi

Brahmin

Brahui

Burusho

Gond

Gupta

Kalash

Kapu

Kshatriya

Pathan

Punjabi

These models look fine in terms of the statistical fits. In fact, much more than just fine in most cases. My prediction is that a population like Potapovka2-Srubnaya_outlier will eventually be discovered on the Late Bronze Age steppe, perhaps even at a site linked to the Andronovo horizon, and it'll fit the bill as a main player in the story of the peopling of South Asia.

However, this population might not necessarily be isolated from its EEF-rich neighbors by geography, but rather by culture and even social class. In other words, we should expect significant substructures on the steppe at this late stage of the game, after a couple of millennia of intense mobility, and in a complex way too, not simply defined by geography.

Update 01/04/2018: Andronovo pastoralists brought steppe ancestry to South Asia (Vagheesh et al. 2018 preprint)