August 22, 2013

Y-chromosome haplogroup Q and Native Americans

PLoS ONE 8(8): e71390. doi:10.1371/journal.pone.0071390

The First Peopling of South America: New Evidence from Y-Chromosome Haplogroup Q

Vincenza Battaglia et al.

Recent progress in the phylogenetic resolution of the Y-chromosome phylogeny permits the male demographic dynamics and migratory events that occurred in Central and Southern America after the initial human spread into the Americas to be investigated at the regional level. To delve further into this issue, we examined more than 400 Native American Y chromosomes (collected in the region ranging from Mexico to South America) belonging to haplogroup Q – virtually the only branch of the Y phylogeny observed in modern-day Amerindians of Central and South America – together with 27 from Mongolia and Kamchatka. Two main founding lineages, Q1a3a1a-M3 and Q1a3a1-L54(xM3), were detected along with novel sub-clades of younger age and more restricted geographic distributions. The first was also observed in Far East Asia while no Q1a3a1-L54(xM3) Y chromosome was found in Asia except the southern Siberian-specific sub-clade Q1a3a1c-L330. Our data not only confirm a southern Siberian origin of ancestral populations that gave rise to Paleo-Indians and the differentiation of both Native American Q founding lineages in Beringia, but support their concomitant arrival in Mesoamerica, where Mexico acted as recipient for the first wave of migration, followed by a rapid southward migration, along the Pacific coast, into the Andean region. Although Q1a3a1a-M3 and Q1a3a1-L54(xM3) display overlapping general distributions, they show different patterns of evolution in the Mexican plateau and the Andean area, which can be explained by local differentiations due to demographic events triggered by the introduction of agriculture and associated with the flourishing of the Great Empires.



Annie Mouse said...

Whereas I have no problem with Siberia as a source, it is what I assumed ultimately. This conclusion is seriously NOT supported by the data. IMO.

Q1b is the closest to the root in this data set, and it is found deep down in Panama/Andes but not in the Mongolian/Siberian samples. This is a widespread group. Assuming no recent contamination this suggests that Q1b folk could have arrived in the Americas any time after the Q1a/Q1b split, with the location so far south suggesting earlier rather than later. This is the stickiest point.

Q1a3a is exclusively Northern Altai populations(figure 3). This seems to mean the Koriak somehow.

Q1a3a1 is exclusively American. The core is Mexico/Andes. Looks like a migration from the "Northern Altai" Koriak.

I can buy the Koriak as proxies for Beringia (location in far, far eastern Russia) or the Gravettian (reindeer folk). Yes it is Siberia, barely, but Northern Altai?

Q1a3a1c is exclusively Mongolian/Southern Altai. Looks a back migration from the Americas.

Q1a3a1a looks fairly evenly distributed across the Americas (figure 3). You have to go deeper into the phylogeny to get regional information. Q1a31a1a4 and a5 are exclusively Andean for example.

Fanty said...

"Yes it is Siberia, barely, but Northern Altai?"


Northern Altai is technically in "Sibiria", as "Sibiria" is the whole Asian part of Russia. In other words: Almost all of Russia, except for that small European part west of the Urals is "Sibiria".

Aaron said...

@Annie Mouse

I agree with you that no where in this paper do they provide evidence that Haplogroup Q originated in Siberia. I wonder why they try to make this claim if the data is not there to support it.

mregdna said...

The Q1b have a unusual DYS392=16.
But this is also found among some jewish Q1B-L245.
So I suggest rather a jewish origin for the rare amerindian Q1B maybe related with the inquisition that reached South America after Portugal.

Clay said...

This seems like a very important paper to me. I am surprised at the few comments here. Annie, what do you think about the idea of Mexico as a staging area for the ultimate expansion of the Amerind population over the Americas? Has that hypothesis ever been advanced before?

Annie Mouse said...



Not seen the idea of a central staging area before. I was expecting expansion from North American staging areas down to the south.

But I suppose over a prolonged period of near isolation expansion out of the main civilizations (Inca/Maya) makes sense.

Annie Mouse said...

Hmm looked at this site:

And there was only 1 DYS392 16

Where do you get the Jewish reference?

The q1b in this study represents 2 different lineages, both with DYS392 16, but multiple other differences.

terryt said...

"Q1b is the closest to the root in this data set, and it is found deep down in Panama/Andes but not in the Mongolian/Siberian samples"

I notice the paper doesn't use the current ISOGG nomenclature. Q1b-M378 is now Q1b1, basically a European, South Asian and West Asian group. So mregdna is probably correct with 'So I suggest rather a jewish origin for the rare amerindian Q1B maybe related with the inquisition that reached South America after Portugal'.

"Two main founding lineages, Q1a3a1a-M3 and Q1a3a1-L54(xM3), were detected"

Current nomenclature has these as Q1a2a1a1 and Q1a2a1 respectively.

"I agree with you that no where in this paper do they provide evidence that Haplogroup Q originated in Siberia."

Even the extract Dienekes published says, 'The first was also observed in Far East Asia while no Q1a3a1-L54(xM3) Y chromosome was found in Asia except the southern Siberian-specific sub-clade Q1a3a1c-L330'. Q-L330 is now called Q1a2a1c and is basically a Ket/Selkup haplogroup. That fits the comment regarding a South Siberian origin. But even that doesn't actually cover Q's 'origin'. Q-M242(xQ1a2a) is more notably a West Eurasian haplogroup, so probably originated somewhere on the steppe west of the Hindu Kush/Altai line. Just the derived Q1a1a1-M120and Q1a2a-L53 are found east of that line and even within the last haplogroup Q1a2b-L940 is primarily Scandinavian and Q1a2c-M323 seems primarily Jewish.

Annie Mouse said...

The paper quotes 33 STRs with 10 being different for the two Q1b lineages (both are DYS492=16).
I got curious so I (painfully) checked out the Q1b STRs in YHRD. I entered 9 core STRs from these two lineages and got only one match (Columbia, similar to lineage 1 in the paper).

There are 5 additional lineages that differ at DYS439 (8 of the core STRs). 4 South Americans, 1 Lianing,China and 1 Lenkoran, Azerbaijan.

There are 12 additional lineages that differ at DYS492 (8 of the core STRs). 10 South Americans, 1 Hispanic North American, 1 Punjabi and 1 Cukurova, Turkey.

IMO this Q1b is not Western European, unlikely to be Jewish and looks very old in South America (tonnes of closely related lineages).

mregdna said...

Maybe I am wrong but I never heard of so many Q1b in south America.
In the database like FTDNA or Genebase etc, the only south America Q1bs I found are close to sephardic jews. And some match with morocco jews.
Some Q1a haplotypes are not so far from Q1bs, are you sure that those Q you found are not particular Q1a3?
If not, it will be an intriguing case because Q1b has not yet been found in Siberia.

Kristiina said...

I went to see the FamilyTree page Annie mentioned and found several people with DYS392 16:
1. Asmatullah Khan, b. 1810, Afghanistan
2. Tal Abitbol, Morocco
3. Suzman, Lithuania
4. Saadia, Iraq
5. Helman Michelson, Suvianiskis, Lithuania
6. Moshe Elfasi. b. 1860, Morocco

Q1b in Native Americans may not be prehistoric, but it does not look native Spanish or Portugese either. Am I right that you suggest that the ancestor of this lineage was a jew living in Iberia taking refuge in America when in 1492 and 1501 Jews and Muslims were forced to convert or leave.

Is someone able to tell us if the STR results of these Native American males match the Moroccan STR results or are at least closer to them compared to other STR results?

Annie Mouse said...

Yep you are right. I had a closer look later. But I did not dare send in another back to back correcting post. Dienekes would have had coniptions.

I did spend time looking at the exact sequence and the Tal Abitbol Morocco sequence is the most similar to the South American lineage but still differed by a couple of STRs.

Turns out a bunch of Moroccan Sephardi migrated out to the Amazon to farm rubber.

But the diversity in South America still looks deeper than anything that appears to exist anywhere else. So it is not clear to me which way the genetics travelled.

Perhaps this is a database imbalance. But Jewish folk are the most heavily tested ethnic group. I would expect South Americans to be less tested.

Did a very large number of Morrocans travel to South America giving all this diversity?

Or did a few travel back to Morocco carrying SA haplotypes giving rise the close Moroccan sequence?

Could all of this diversity (lineage 1 and 2 differ by 10 out of 33 STRs) have really arisen since Pizarro arrived in the 1400s? And there was even more SA variation in the database.

Annie Mouse said...

"Some Q1a haplotypes are not so far from Q1bs, are you sure that those Q you found are not particular Q1a3?" is an STR haplotype database. But it seems unlikely that such long strings of such similar STRs would be a different haplogroup.

Its free but now has a limit of 20 searchs a day. 2 without registering. I don't think it works well with explorer, I used Firefox.

Kristiina said...

The recent study from Jilin University confirmed the existence of Q1a, Q1b and Q in Xinjiang 2000 ybp. Instead, c. 500 years earlier in Pengyang only East-Asian M120 was found.
Heigouliang, Xinjiang, 2000 YBP, 6 Q1a*, 4 Q1b, 2 Q
Pengyang, Ningxia, 2500 YBP, all Q1a1-M120

It would be interesting to know which subtypes the Heigouliang haplotypes are and how they are related to Native American and Siberian subtypes. I think that it is also interesting that Q1a and Q1b are found together as if they belonged to the same tribe. This is something unexpected if we postulate that Q1a arrived earlier from Central Asia through Altai and Q1b later on from Afghanistan/Pakistan.

The study is commented at least on Eurogenes Blog at

mregdna said...

There was also a migration of morroco jews in Amazonia (mostly males):

Some were assimilated in amerindian villages.
Maybe some jewish Q1b lineages disappeared from Europe?
Or there was Q1b amerindian lineages in south America and they alomost disappeared?

DocG said...

Not precisely on topic, but I'd like to call attention to a remarkable early paper dealing with classical blood type evidence pertaining to the population genetics of S. America, specifically the "Diego gene," as analyzed by Layrisse and Wilbert back in 1961:

The authors "suggest that Diego-negative populations were the first to arrive and to extend throughout South America, while the Diego-positive tribes came later."

Dienekes, I'm wondering whether you are aware of this very early research and if you know of any followup. The paper isn't cited at all in Cavalli-Sforza's History and Geography, and I wonder whether anyone has ever looked more closely at this evidence.

Donald Albury said...

Wikipedia has an article on the Diego antigen system.

Cuah123 said...

Amazing that "they must be Jewish" is the default ethnic mapping that occurs even amongst Native Meso American studies.

In Latin speaking countries being "indigenous" was seen as a lower caste. Being anything other than Native would be a blessing. I truly believe that alot of those Q "Sephardics" are actually Native Americans whose families have tried to erase their indigenous heritage or have forgotten their heritage so they go with what they know- Bible Education.

Has it not occurred to anyone that the hegemonic nature of the "they must be Jewish" across ethnicities that might seem to be arcana to popular knowledge doesn't serve either community properly.

It is so incredibly unlikely that a roaming religious group fill in good with any Meso American indigenous group, it goes against all that is known about both groups. What did these roaming groups also take up human sacrafice, cannibilism a

terryt said...

I see that ISOGG has recently updated their Q phylogeny. THey have not added the newly discovered Q-M557, Q-PV2, Q-PV3 and Q-PV4 to their chart though. But interestingly I notice they place Q-NWT01 as basal to Q-M120 (Q1a1a and Q1a1a1 respectively)rather than being on separate lines. And ISOGG still has Q-M120 and Q-M25 as a single branch within Q1a2-M346 rather than separating them as this paper does.

"Amazing that 'they must be Jewish' is the default ethnic mapping that occurs even amongst Native Meso American studies".

Yes. Even the first Q1a2c-M323 (called Q1a3b in this paper) was found in 'Yemenite Jews'. Of course this doesn't mean the haplogroup was 'jewish' or even originally from anywhere outside Yemen.

"Maybe I am wrong but I never heard of so many Q1b in south America"

That is the real suprise of the paper. If correct it may show that Q had spread from the Iranian Plateau to Beringia very soon after it first coalesced.

mregdna said...

I still believe that the Q1Bs found here are of sephardic origin because it's the first time that Q1b is found among amerindians.
And also because the migration of sephardic jews, marranos etc is well known and the migration of morroco jews in Amazonia is well documented.
And also some morroco jews settled in amazonian Peru (in Iquitos and around) on the border of Colombia not far from the amazonian ecuador.
And Q1b is clearly found among ashkenazim and sephardic jews (with more diversity!) including Morocco and Algeria jews and it seems absurd to believe to an hidden ameridian origin of the south american Q1bs that match with sephardic jews.

AKO said...

Great discussion guys. I really appreciate as my direct paternal line is probable Q1b. My brother tested at the 12 marker level so far and his close matches are the Q1b haplotype. In YHRD database he has four exact matches and they are from Slovakia, Kuwait, and the last two are from Brazil. I do become concerned that this branch of haplogroup Q is sometimes being pushed for an agenda. I wish those two Q1bs were not dismissed so easily in the study. The databases are top heavy with people of Jewish heritage testing. There is so much to learn out there.
The marker of 392 showing 16 instead of the usual 15 is rare, and should be investigated more thoroughly. I have also been told that the Jewish cluster is on marker 395S1. So it would be interesting to see what these samples had shown there. My other issue is that not enough have tested the SNP 245 to show positive or negative for sure. I am guilty here too, but as soon as funds permit will be testing the SNP and more STRs.