search this blog

Thursday, October 27, 2016

Strong mitogenomic continuity on the Armenian Plateau since the early Neolithic

Below is another abstract from the upcoming 6th DNA Polymorphisms in Human Populations conference in Paris. It was added to the abstract book after I first blogged about the conference here.

The authors are probably first testing mtDNA to check which of the samples have enough DNA for full genome sequencing or enrichment capture of genome-wide SNPs. So the fact that they managed to sequence so many mitogenomes means that we might soon see genome-wide data for most of the same samples.

The results make sense considering the genetic structure of present-day Armenians (see here). Indeed, I suspect that early Neolithic farmers from the southern Caucasus will come out looking very similar overall to present-day Armenians.

If so, this will probably be hailed by many as evidence supporting the Armenian Plateau Indo-European homeland theory. However, as I've already shown, it's very likely that the Armenian Plateau was affected by population movements from the Eastern European steppe during the Bronze Age which may have introduced Indo-European languages to the region (see here).

The origin of the Armenian people is heavily debated among historians and archaeologists. Despite a long history and vast archaeological records in Armenia, it has proven very challenging to infer the demographic events that led to the formation of Armenians as a distinct ethno-cultural group. To obtain a detailed understanding of the demographic events in Armenia across millennia, we study complete mitochondrial genomes from 49 ancient individuals covering 7800 years and compare them with that of modern Armenians (n=206) and seven neighboring populations (n=482). In this context, the lowest genetic distance was observed between the modern and ancient Armenians and this was also reflected in network analyses and discriminant analysis of principal components (dapc) showing genetic proximity between the ancient individuals and modern Armenians. We used Approximate Bayesian Computation (ABC) to test five different demographic scenarios of the Armenian population, and the simulations favored a model where both ancient and modern Armenians derive from the same source population. We conclude that there is a strong signal of continuity in the maternal Armenian gene pool during the last 7800 years.

Margaryan et al., 7800 years of Mitochondrial genetic continuity in Armenia, 6th DNA Polymorphisms in Human Populations, Talk Workhsop Genomic Demography, Musee de l’Homme, Paris, 7-10 December, 2016

Update 29/06/2017: The paper has just been published. It only has a few samples from before the Middle Bronze Age (MBA) and these don't match closely the mtDNA lineages from the Eneolithic/Bronze Age steppe, which is interesting and potentially important for the Indo-European homeland debate. See here.


Aram said...

Oh nice this are the Shulaveri period. I hope the Y dnas and autosomes will not be late.
My bet Shulaveri farmers are somewhere between Tepecik samples and Iran_Chl off course without EHG.

Rob said...

Agreed, but Late Neolithic Armenia will have 5-10% EHG

Davidski said...

Agreed, but Late Neolithic Armenia will have 5-10% EHG.

Why's that?

Rob said...

? Exogamy with the north
By the Copper age, it already had some

"Anatolia_Neolithic" 49.95
"Iran_Neolithic" 19
"Samara_Eneolithic" 16.55
"Satsurblia" 11.2
"AG3-MA1" 3.3
"Villabruna" 0
"Israel_Natufian" 0
"Loschbour" 0
"Levant_Neolithic" 0
"LBK_EN" 0
"Hungary_HG" 0
"Iberia_Chalcolithic" 0

MtDNA U3 & U6

Aram said...

Areni cave was a strictly sedentary farmer context. They were practising viticulture.
The fact that they have U4 women could mean that in the neighbourhood in an unsampled region there were some EHG reach populations. Who maybe were crossing Caucasus. Areni cave is contemporary with Leyla Tepe and it is only 300 km away from them. Leyla Tepe is in Kura valley which means they were more open to Steppe. Maybe Leyla Tepe people were EHG reach pops? Also they were contemporary to Maykop period.
But in any case I admit that this EHG situation is very intriguing. Impatient to see how Shulaveri people were looking.

Aram said...

Btw Areni cave's L1a-M27* is practically dead end. Only few persons in whole Near East & Caucasus are probably related to them. But even that there is no definitive proof because there is no Big Ys. There is also a very small cluster of L1a-M27 in Germany. But again the exact relationship is unknown.

Aram said...

I checked again. There are only U4 and U3 in ancient Armenia. No U6.
U6 is African non?

Davidski said...

And U3 is West Asian.

EHG markers are U4, U5a, and crazy stuff like C1.

Rob said...

My mistake U3, not U6.
Anyhow, U4 is there, and that fit works with pure EHG too. But it could have arrived mid copper age rather than Neolithic. Let's see..

Olympus Mons said...

if they used Aratashen or Arkanshen then is Shulaveri... if not... not really.
Shulaveri was mainly in Georgia (Kotias land).

A plausible Shulaveri origin read here:

Olympus Mons said...

"EHG markers are U4, U5a, and crazy stuff like C1"

Early chalcolithic (~3000BC) in Perdigoes, Portugal (most western part of Iberia) in (pit 7) showed U4, U5 and the obiquous H.

Davidski said...


Early chalcolithic (~3000BC) in Perdigoes, Portugal (most western part of Iberia) in (pit 7) showed U4, U5 and the obiquous H.

U4 is an European-specific lineage, and not exclusively EHG. But it shows high frequencies in populations with high levels of EHG admixture, like those on the EBA steppe.

So when it turns up in an ancient sample from Armenia with obvious genome-wide EHG admixture, then there's not much to debate.

By the way, those Portuguese results are old, and they were done with PCR, which isn't exactly contamination safe. Let's see what happens when they're re-analyzed using modern methods.

Davidski said...


Shulaveri was mainly in Georgia (Kotias land).

So no one kicked them out. They're still there, because Georgians are more similar to Kotias than anyone else is.

Nirjhar007 said...

So R1b from Eneolithic too?.

Mher said...


Nirjhar007 said...

We have to see the dimensions of continuity . Y-dna would have been nice... but they are also coming.

Olympus Mons said...

Shulaveri will be a mix of hungary hg/ balkan farmer/ barcin/ kotias.

Matt said...

The Early-Late Neolithic Caucausus and Armenia is a black hole at the moment. Some post-Upper Paleolithic (9700 BP) Early Neolithic information from Caucasus and Armenia that's pre Armenia Chl (6000 BP) is required. To me it seems pretty possible they could have an increase in EHG type ancestry after the age of CHG but pretty early on before the Steppe Eneolithic has accumulated much in the way of CHG. (and could be some differentiated local ancestry).

Davidski said...

Not exactly a black hole IMO.

Iran_Chalcolithic formed when a population from near the Caucasus and eastern Anatolia moved into western Iran. It probably came from near the historic borders of Armenia.

So the people of the southern Caucasus Neolithic should be a lot like Iran_Chalcolithic. And indeed modern day Armenians share a lot of drift with Iran_Chalcolithic.

Iran_Chalcolithic f3
Georgian 0.165908
Armenian 0.165234
Avar 0.165039
Abkhasian 0.164335
Chechen 0.164174
Tabasaran 0.164111
Kabardin 0.163471
Italian_Bergamo 0.163435
Italian_Tuscan 0.163233
Greek 0.163195

Shaikorth said...

Is their IBS-ranking similar? Armenians might even top it.

Davidski said...

Similar, but with no European pops in the top ten...

Iran_Chalcolithic IBS
Abkhasian 0.718832
Armenian 0.718637
Georgian 0.717631
Lezgin 0.71751
Adygei 0.717308
Druze 0.71705
Avar 0.717021
Chechen 0.716728
Iranian 0.716545
Tabasaran 0.716439

Aram said...


What You think about dolmens? Present in North Caucasus and in BB. Are they part of Your theory also?

Olympus Mons said...

Oh no. Dont really know anything about those dolmens (ashame) in BB and N caucasus. But Dolmens seem to have a even more complex story...

To me the key point is architechure. that is core human behaviour.
They were gone to somewhere else. Don't really know about them being R1b... But there are only two options for Chalcolithic Iberia Architecture. Either the shulaveri or a derived, never spotted, population that existed millennia after Khirokitia in Cyprus.

Whoever got to Iberia by 3500bc made 5 meter round houses with wattle and daub with 1-5/2 meter round silos, with a central yard with an earth and buried their dead under the house floors (etc,cetc,etc,). I know the Shulaveri (4900 bc) did, I know a 200 year period in Tell tsaf (4800-4600bc) Merimde and El Omari in Egypt Nile (4800-4000bc) that got the same package (elaborate mix of animal husbandry and agriculture) and in South Iberia (3500BC).

Can anyone find that architectural trait in a different route? - Shoot!
Or Spelt? Or flat bone spoons? Or Carenated pottery?
Or even Tanged Arrows, or lever pressure blade making... or, or, or or...

If anyone can make a story more plausible than they sailed all day and night for 20 days, or ride trough the most roughed and dificult terrain for 3 months (and a lot of them) to get to Iberia, feel free to do so.
And especially show me the origin and the mid points where it is found. Again space and time. Space-and-time.

Grey said...

"but the autosome component associated with an early African R1b-V88 would be very enlightening."

I think that will turn out to be one of the most fascinating journeys.

Olympus Mons said...

"...around 3400 BC, and there is no reason to suppose that the people in the Nile Valley at 4800 BC didn't already speak...."

What are you talking about? what people in the Nile valley? Delta or Upper? Because one had nothing to do with the other.

Olympus Mons said...

... and I am talking about concrete and real objects... you are talking about hypothetical languages timelines?

Olympus Mons said...

What support do you have for the language of a Delta nile 4800bc-4000bc population that desapeared from the region and was replaced by population coming from the South?

Karl_K said...
This comment has been removed by the author.
Aram said...

It looks like megalithic structures like dolmens were made by EEF-like farmers.///

Makes sense. In Caucasus it is associated with NWC people who have the same G2a2 as EEF people.

Olympus Mons said...

Do you know the story of pre dinastic Egypt, right?

If Merimda and El-omari were caucasian speakers of Indo european language (I am not sure. dont know enough about it) then you can not, absolutly not, use any language reference to infer anything. They lost. the south badarian and Amratians coming to the north "won" and the language that has reference were theirs, not at all merindian.

And yes, I still think that el-omari were the first bearers of R1b L51.