search this blog


Wednesday, November 9, 2011

The Erzya and Moksha of the Middle Volga

Slowly, but surely, the the diversity of publicly available genome-wide SNP samples is increasing. One of the most interesting and valuable sets, released not long ago thanks to the Yunusbayev et al. study, comes from the Mordovian Republic in the Russian Federation. It includes the results for 15 people from the closely related Erzya and Moksha ethnic groups (10 and 4 samples respectively, and one mixed individual).

So why are these Mordovian Republic samples so important? Because the Erzya and Moksha have a long history in the Middle Volga region of Russia, which is a key zone for those interested in the Eurasian steppe and forest steppe. Their presence in the area as a distinct Volgaic Finnic entity probably dates back to around 500 BC, but various groups who contributed to their ethnogenesis were probably here much earlier. We can infer this from the Erzya and Moksha languages. They're classified as Finno-Ugric, and thus Uralic, but show ancient influences from Indo-Iranian, Baltic, and possibly even proto-Tocharian. What this suggests is that Erzya and Moksha genomes should provide strong clues to the various population movements through this region, possibly all the way from the metal ages to historic times.

Based on a variety of analyses I've done with these samples to date, this certainly appears to be the case. But the results give a strong impression of mostly one way traffic across Mordovia - from the west to east.

Below is an MDS featuring the Erzya and Moksha from Yunusbayev et al., plus a single Erzya from my Eurogenes Ancestry Project (Erzya1), run against samples from the North Caucasus, Chuvashia and North and East Europe. Interestingly, not only do they cluster tightly with the North and East Europeans (PL = Poland, RU = Russia, SE = Sweden, UA = Ukraine), but they're clearly separated from the North Caucasus samples and even from their immediate neighbors on the Volga, the Chuvashs.

These results are obviously totally out of whack with geography. Indeed, it appears that the Swedes are a little closer to the North Caucasus than the Erzya and Moksha across the more significant dimension 1. This might seem bizarre at first, but not if we consider that Eastern Europe wasn't affected by any significant post-Ice Age migrations directly from the south (ie. the Mediterranean Basin/Near East), unlike Western Europe. It's not just my own work showing this, but also the aforementioned Yunusbayev et al. study, which described the presence of a significant genetic gap between populations of Eastern Europe and the Caucasus. Moreover, archeologists have been saying something similar for years.

From the Eneolithic [Copper] period to the Early Iron Age there were no invasions of population from the Near East or Iran in the steppes of Eurasia. All migrations, though numerous, occurred within the steppe and forest-steppe zones.

Elena E. Kuz'mina, James P. Mallory (Editor), The Origin of the Indo-Iranians, Vol. 3, Barnes & Noble, 2007

Hence, the cline of Mediterranean/Near Eastern influence across the European continent generally runs like this: Southern/Southeastern Europe > Western Europe > North/Central Europe > Eastern Europe.

However, the Erzya and Moksha do carry direct influence from the east of the Volga. This shows up on the MDS plot along dimension 2, with the samples from Mordovia placed slightly closer to the Chuvash than most of the other North and Eastern Europeans. My ADMIXTURE analyses also indicate about 6% of North/East Asian ancestry, which is 5-6% higher than in most North/Central/Eastern European groups (see here).

This isn't surprising, considering the number of invasions from the east that have passed through or near the Middle Volga on their way to Europe, including various Turkic and Altaic onslaughts during the middle ages. What is startling, however, is how much of the older, western-like genetic ancestry the Erzya and Moksha have manged to retain. This didn't happen across the border in what is now Chuvashia, where language replacement to Turkic seems to have been accompanied by significant gene flow from Siberia. Thus, modern Chuvashs carry around 20-25% North/East Asian admixture.

Another interesting fact revealed by the Yunusbayev et al. dataset was the high level of light pigmentation amongst the Erzya and Moksha. I discovered this when plotting allele frequencies from various SNPs linked to fair hair and light eyes in Europeans (see here). The combined samples from Mordovia came out as one of the fairest groups. The result was confirmed by a reader, who sent me data from a pigmentation survey of Northeastern Europe, showing the Erzya as fairer than ethnic Russians, and the Moksha at about the same level as the Russians. A quick search via Google revealed the following two random representatives of the Erzya nation. I don't know the original source for the first image, while the second comes from Wikipedia.

I'm not sure whether these women look like the average Erzya, or even if they're representative of a significant minority of the population. The latter is more likely. In any case, it's interesting that two random hits off Google matched the genetic results of the reference samples from Mordovia. The relatively light pigmentation of such individuals, from near the Samara Bend, certainly fits with the notion of an expansion of a predominantly light eyed and fair haired population from this area during the Bronze and Iron Ages. Of course, I'm referring here to the Andronovo tribes.

According to C-14 [carbon] dating, the Andronovo and Timber-grave cultures formed at the shift of the 3rd and 2nd millennium BC on the common base of Eastern European cultures. The main components were the steppe cultures of Poltavka, late Catacomb (to far lesser degree), and the Abashevo culture of the forest steppe zone. The latter culture belonged to the Corded Ware group of cultures of Central Europe.


The subsequent development of the Timber-grave culture in the Pontic Caspian steppes was determined by the large-scale participation of the Abashevo culture in the Don region and the post-Catacomb Multi-roller Ware culture in the Ukraine.


The close genetic affinity of the Andronovo and the Timber-grave culture tribes became even closer thanks to regular contacts, especially because of the flow of tin from Kazakhstan and of copper ( as well as of finished products) from the Urals. The contact zone lay between the Volga and the Urals. The Andronovan influence reached the Dnieper.

Elena E. Kuz'mina, James P. Mallory (Editor), The Origin of the Indo-Iranians, Vol. 3, Barnes & Noble, 2007

Our autosomal, Y-chromosomal and mitochondrial DNA analyses reveal that whereas few specimens seem to be related matrilineally or patrilineally, nearly all subjects belong to haplogroup R1a1-M17 which is thought to mark the eastward migration of the early Indo-Europeans. Our results also confirm that at the Bronze and Iron Ages, south Siberia was a region of overwhelmingly predominant European settlement, suggesting an eastward migration of Kurgan people across the Russo-Kazakh steppe. Finally, our data indicate that at the Bronze and Iron Age timeframe, south Siberians were blue (or green)-eyed, fair-skinned and light-haired people and that they might have played a role in the early development of the Tarim Basin civilization.

Christine Keyser et al.,
Ancient DNA provides new insights into the history of south Siberian Kurgan people, Human Genetics, Saturday, May 16, 2009, doi: 10.1007/s00439-009-0683-0

The publically available datasets I used in this blog entry are found at the links below. My analysis can be reproduced with the PLINK software, found here.

HGDP-CEPH Human Genome Diversity Cell Line Panel

Yunusbayev et al. dataset


mikej2 said...

How do you explain that they speak Finnic language if they came from west?

Davidski said...

Most of their genes came from the west.

Their present language came from the Southern Urals.

mikej2 said...

Have you any idea how this happened? Did Uralics conquered local western people there? I just ponder how the shift between languages happened, because adapting new language needs to have reasons. Is it possible that the reason for their R1a is genetic drift and not any migration and conquest?

Davidski said...

Well, look at the Chuvash next door. They were Volgar Bolgars within historic times, and now they're Turkic Chuvashs. Voila, that easy.

The problem with the Erzya and Moksha is that their language shift happened in pre-historic times, so we don't know how it went down exactly. But it most likely did happen, because they're genetically not very different from Poles, or even Germans to be honest, who never spoke Uralic.

By the way, it's likely that the Erzya and Moksha had much higher R1a1a1 frequencies before the purported influence from the South Urals and the language shift. And they're not a heavily drifted population at all.

mikej2 said...

According the Wikipedia there is only 843,350 (2002) Mordvins. This is quite little compared to ethnic Russians, 115,889,107. The ratio is 137. I suppose that after the Russian colonization their number was very small and started to grow later.

I could guess that there was pretty much gene flow from Russians to Mordvins. For comparison, Saamis have still their own language, although the yDna has shifted to Scandinvian. The ratio between Scandinavians and Saamis is quite similar.

As to the aDna, we dont know much about it 500 years ago.

Davidski said...

I can't see any evidence of massive gene flow from Russians to Erzya and Moksha.

At the finest scale, the Mordovian samples I have are certainly distinguishable from all but one of my Russians. But they're not distinguishable from each other, even though some are Moksha and others Erzya.

This small but clear difference from Russians, in a very similar and even way, points to an ancient affinity to Russians, rather than large scale recent mixing. Recent contacts would be picked up as much more uneven results, with some of the Erzya being more similar to Russians, and others to other Erzya. In other words, they would be a lot more choppy.

Onur said...

"This didn't happen across the border in what is now Chuvashia, where language replacement from Volga Bolgar to Turkic seems to have been accompanied by significant gene flow from Siberia."

"They were Volgar Bolgars within historic times, and now they're Turkic Chuvashs. Voila, that easy."

Were not Volga Bulgars already Turkic-speaking and spoke a Lir Turkic language similar to the Chuvash language? In fact, there is a good chance that the Chuvash language is a direct continuation of the Bulgar language (through its Volga branch).

Davidski said...

^ Fair enough, I made the correction.

Onur said...

I think whatever Turkic genetic impact that happened in what is now Chuvash lands happened essentially during the Bulgar colonization of that region starting from the 7th century CE. I think Chuvash people are a direct continuation of Volga Bulgars (Volga Bulgars themselves being a result of the admixture of the Turkic-speaking Bulgar migrants from the Pontic-Caspian steppe with the Uralic-speaking Volga natives, especially Mari people, they assimilated).

Onur said...

Volga Bulgars themselves being a result of the admixture of the Turkic-speaking Bulgar migrants from the Pontic-Caspian steppe with the Uralic-speaking Volga natives, especially Mari people, they assimilated

So I think you should compare Chuvash with Mari instead of Mordva.

Abbat said...

Erzya-Moksha DNA Project



Abbat said...

Photo examples moksha and erzya

Abbat said...

162897 Sergachev Z280+
199944 Kaftaev Z280+ Z92-
210272 Zhivaev ?
219253 Samarkin Probably Z280+
219252 Inzhuvatov Probably Z280+ Z92+
162898 Mochalov Z280+ Z92+ (In your project it Erzya1)
209011 Erin I1-T2 Z63+

Abbat said...

199944 Kaftaev in your project-RU9

Goga said...

I am really interested to learn more about this project. My mother is Erzya from Atlay Siberia. Please advise sources. Thanks

sidoroffs said...