search this blog

Monday, May 2, 2016

The genetic history of Ice Age Europe (Qiaomei Fu et al. 2016)


Just in at Nature:

Abstract: Modern humans arrived in Europe ~45,000 years ago, but little is known about their genetic composition before the start of farming ~8,500 years ago. Here we analyse genome-wide data from 51 Eurasians from ~45,000–7,000 years ago. Over this time, the proportion of Neanderthal DNA decreased from 3–6% to around 2%, consistent with natural selection against Neanderthal variants in modern humans. Whereas there is no evidence of the earliest modern humans in Europe contributing to the genetic composition of present-day Europeans, all individuals between ~37,000 and ~14,000 years ago descended from a single founder population which forms part of the ancestry of present-day Europeans. An ~35,000-year-old individual from northwest Europe represents an early branch of this founder population which was then displaced across a broad region, before reappearing in southwest Europe at the height of the last Ice Age ~19,000 years ago. During the major warming period after ~14,000 years ago, a genetic component related to present-day Near Easterners became widespread in Europe. These results document how population turnover and migration have been recurring themes of European prehistory.

Fu et al., The genetic history of Ice Age Europe, Nature 02/05/2016; DOI:10.1038/nature17993

Here are three summaries of the findings from the comments:

Krefter: Europeans dating 36,000-30,000 years old from Russia, Belgium, Czech, and Italy form a clade with 24,000 year old Siberian MA1. But they're more related to each other than to MA1. Also, none of them have an especially close relationship to each other, unless they come from the same archaeological site/region.

DNA from a lady who lived in Spain 19,000 years ago, has a close relationship to people who lived in Belgium 30,000 years ago. She also has a close relationship to "WHGs". So, she looks like a mixture of WHG and those 30,000 year old people from Belgium.

After 14,000 years ago, Europeans from Northern Italy, Germany, France, Luxembourg, Spain, and Hungary all form a tight cluster. They are WHG. They also appear to have minor ancestry related to the 30,000 year old people who lived in Belgium.

Stone age Middle Easterners CHG(Caucasus) and especially EEF(Turkey) have WHG-related ancestry. They're more related to post-14,000 years ago Europeans than to earlier Europeans. There's an unknown link between the Middle East and WHG.

...

Matt: Trying to interpret this (and probably getting something wrong) the genetic history and cultural of Europe from this so far looks like this

- 42,000-37,580 BP - Oase1 type people (and possibly others) living in Europe (Romania)

- 35,160-34,430 BP - GoyetQ116 type people living in Europe (Belgium) (GoyetQ116 is classified to the Aurignacian culture wave).

- 34,580-31,210 BP - Vestonice cluster types seem to replace the GoyetQ116 type

(Except the GoyetQ116 type, Aurignancians(?) are likely to still be living in Iberia, as they're ancestral to the El Miron cluster, from later in history)

Vestonice types seem to all be classified as Gravettian.

- Glaciation then wipes out the Vestonice / Gravettian types around 24,000 BP.

- 18,830-18,610 BP - El Miron is an admixture of a GoyetQ116 and the Villabruna branch (the latter thought to be Near Eastern in origin due to greater sharing with present day Near Eastern people)

All samples from then on in Europe, are identified with the Magdalenian culture and with the same cluster membership as El Miron, as these guys expand across Europe after the retreat of the ice.

- 14,180-13,780 BP - Villabruna branch ancestry in samples increases a lot, and material culture transitions to other industries - Epigravettian, Azilian, Epipaleolithic and Mesolithic. Looks like another great migration to Europe (out of the East via Italy?).

...

Colin: There was no Basal Eurasian admixture in pre Neolithic Europe. The reasoning was that Ust Ishim (45kya Eurasian) does not contain Basal Eurasian and Ust Ishim is equally related to East Asians and pre Neolithic Europeans.

There was no ANE, Mal'ta cluster, admixture in pre Neolithic Europe (presumably they are talking about the portion of Europe that is West of Russia and Ukraine). The reasoning here is that the mal'ta cluster is equally related to all pre Neolithic "Europeans" starting with Kostenki (37kya).

-Consequently, ANE only entered "Europe" during the Neolithic and Bronze Age, probably from the Eurasian Steppe.

So far, GoyetQ (35kya Belgium) and Kosteni (37kya Russia) are the first ancient Europeans to exhibit closer relations to modern Europeans over East Asians. The older Eurasian samples of Ust Ishim and Oaste do not show higher affinity to modern Europeans over East Asians.

European samples from about 37kya to 14kya tend to be equally unrelated to non-European Eurasians. However, there are 14kyo samples, representing the Villabruna cluster, that show a higher affinity to the modern Near Easterners. The 13kyo to 10kyo samples from the Caucasus, representing the Satsurblia cluster, show more affinity to the Villabruna cluster than to the earlier Europeans. The Villabruna cluster can not be the result of admixture with the Satsurblia cluster since the former does not contain Basal Eurasian while the later does. GoyetQ (35kya Belgium) type ancestry shows a resurgence in the Miron Cluster of 19kya centered in Iberia. About half of the ancestry in the Miron Cluster derives from GoyetQ types.

Because of the large number of individuals in the Villabruna cluster, the researchers were able to identify structure within the cluster. One interesting tid bit is that certain individuals in the Villabruna Cluster showed extra affinity towards East Asians. The oldest of which is the 13kyo sample from Switzerland. Two other Villabruna individuals, Loschbour and La Brana, for example, show extra affinity to Han Chinese over Kostenki. Because all three individuals are equally related to Ust Ishim, hypothetical differences in levels of Basal Eurasian cannot explain the greater affinity Loschbour and La Brana have to the Han.

See also...

On the modern genetic affinities of Ice Age Europeans

For a typical West Eurasian PCA, the Epipaleolithic is the final frontier

79 comments:

Nirjhar007 said...

Cool! :D

Chad Rohlfsen said...

Supplementary info

http://www.nature.com/nature/journal/vaop/ncurrent/extref/nature17993-s1.pdf

Nirjhar007 said...

I have it. Italy has R1b from 14ooo ybp!!!!

Karl_K said...

When analysing the mitochondrial genomes we noted the presence of haplogroup M in a ~27,000-year-old individual from southern Italy (Ostuni1) in agreement with the observation that this haplogroup, which today occurs in Asia and is absent in Europe, was present in pre-Last Glacial Maximum Europe

Davidski said...

The figures and tables are freely available here.

http://www.nature.com/nature/journal/vaop/ncurrent/fig_tab/nature17993_ft.html

Nirjhar007 said...

Karelia again show extinct R1a1 , only more older!

Matt said...

Reading phys.org's coverage first: http://phys.org/news/2016-05-genetic-
analysis-ice-age-europeans.html

"The genetic data show that, beginning 37,000 years ago, all Europeans come from a single founding population that persisted through the Ice Age, said Reich. The founding population has some deep branches in different parts of Europe, one of which is represented by a specimen from Belgium. This branch seems to have been displaced in most parts of Europe 33,000 years ago, but around 19,000 years ago, a population related to it re-expanded across Europe, Reich explained. Based on the earliest sample in which this ancestry is observed, it is plausible that this population expanded from the southwest, present-day Spain, after the Ice Age peaked.

Two big changes in prehistoric human populations that are closely linked to the end of the last Ice Age around 19,000 years ago. As the ice sheet retreated, Europe was repopulated by prehistoric humans from southwest Europe (e.g., Spain). Then, in a second event about 14,000 years ago, populations from the southeast (e.g., Turkey, Greece) spread into Europe, displacing the first group of humans."

So WHG basically = somewhat more ancient Anatolians than EEF?

This would make it more mysterious as to how WHG can lack pigmentation SNPs found in both Early Anatolians and in Upper Paleolithic CHG....

Davidski said...

You guys need to stop quoting so much of the text.

Karl_K said...

But... several of the authors themselves shared the full text with me... why can't I share it with my friends?

Nirjhar007 said...

As you wish :)

Nirjhar007 said...

Because its a pay walled article. We don't want to make Publishers angry and frustrated :D.

Nirjhar007 said...

BTW It appears there was No ANE un till Neolithic in Europe.

Roy King said...

So Villabruna (Northern Italy) circa 14K bp has Near Eastern affinities and is R1b1. Wouldn't this suggest that R1b1 might have originated in the Near East?

Nirjhar007 said...

Karl,

Just read , whats the hurry?. You have it. Its a priced article , if you release everything it becomes meaningless. You can argue against and maybe you have good reasons , but since it is a priced research and not OA, we should show some respect to the respected Journal!.

Matt said...

@ Karl, I think there was suggestive stuff, but not sure it was that definitive. That's something to give pause in light of similar work earlier this year that there were large divergences in selected SNPs between Loschbour and Stuttgart, If they came from the same region, evolving with pretty much a similar climate? (or did they?).

Seems like it also makes EHG easier to understand as well? I.e. descended from in part the same population hypothesized to be from Anatolia / Greece, and also from the Mal'ta cluster on the other side?

Also: Section 8, testing whether Kostenki14 is "basal Eurasian" admixed, or later Europeans have East Asian admixture:

"Replication of the statistics in Seguin-Orlando et al. in our data "

"D(Kostenki14, North Eurasian Hunter Gatherer; East Asians, Mbuti) << 0
D(Kostenki14, Early European Farmers; East Asians, Mbuti) ~ 0 "

"Kostenki14 patterns are general in Upper Palaeolithic Europeans"

"This analysis reveals that the observed signals interpreted as Basal Eurasian ancestry are not unique to Kostenki14, and are also seen in later samples until around 14,000 years ago. Only after around 14,000 years ago (from Villabruna onwards) do samples with genetic affinities like those in the Loschbour cluster appear. "

"These results show that whatever history is driving the signals identified in ref. 1 is not unique to Kostenki14. Two scenarios can potentially explain these observations: "

1. "There is Basal Eurasian ancestry in Kostenki14 (consistent with ref . 1 ) as well as in all Europeans prior to some Villabruna Cluster samples."

2. "There is no Basal Eurasian ancestry in Kostenki14, and instead, gene flow occurred between the ancestors of East Asians and Europeans after 14,000 years ago. (Independently, there would need to be gene flow between the ancestors of East Asians and Malta1 to explain its affinities.)

"The strongest evidence of Basal Eurasian ancestry comes from UstIshim and Oase1, and does not support Basal Eurasian ancestry in Kostenki14", therefore

"Gene flow among eastern non-Africans, the Mal’ta Cluster & pre-Neolithic Europeans"


Gene flow between WHG Europeans and East Asians is what I thought was very much back on the table would be likely as soon as the stats came out showing Kostenki14 as related to Ust Ishim as WHG and ENA. Cool to see subsequent samples other than Kostenki14 have substantiated this relationship, and to have this written up to point at.

Davidski said...

BTW It appears there was No ANE until Neolithic in Europe.

No, because Karelia HG is Mesolithic, and has been given an older date than before.

It's the same sample as in the earlier studies, by the way.

Nirjhar007 said...

Yeah Roy, looks like a possibility but we have many things to see first. And above all some samples from E Anatolia and Kurdistan..

Krefter said...

List of the new samples.

https://docs.google.com/spreadsheets/d/1lMeojSM7Lrep4GYi8qV31QbDPIw5w-bEygm8esoSHF4/edit#gid=0

capra internetensis said...

lmao

Gioiello's prediction was spot on. There it is, R1b in Ice Age Italy.

Otherwise we have I, pre-I, C1a2, which is pretty much as expected.

Nirjhar007 said...

yeah, big eyes on the German ones...

Krefter said...

The Villabruna had no Basal Eurasian or Near Eastern(unless it lacked Basal Eurasian) ancestry. There's a section about how no pre-Neolithic Europeans had Basal Eurasian ancestry. Starting with Villabruna, about 14,000 years ago, Europeans had *slightly* more affinity to Near Easterners than older Europeans did, but this has nothing to do with Basal Eurasian.

Nirjhar007 said...

Capra,

Yes I am happy for him. He had professional knowledge and passion also. Now He seems to be proven right. Who knows perhaps 20000 yo old samples from Italy will also show R1b? ;).

Davidski said...

But they didn't assign this to a cluster as it was obviously admixed.

Yeah, Karelia HG is a mix of WHG and ANE. We knew that long ago. So ANE was lurking around Eastern Europe at least when Karelia HG was alive, but probably further back than that, because it'd be an amazing coincidence if this was the first person in Europe with ANE admixture.

Btw, the cluster assignments can be viewed here...

http://www.nature.com/nature/journal/vaop/ncurrent/fig_tab/nature17993_ST1.html

Krefter said...

There were distinct populations in Ice age Europe(besides WHG and EHG). Here are the conclusions by the authors based on treemix and admixturegraph.

Conclusion
>Paleo Euros as clade opposed to MA1
>Vestonice is mostly Kostinki with minor Goyet.
>El Miron is mostly Goyet with minor Villabruna.’
>Loschbour is mostly Villabruna with minor Goyet.

Vestonice: 30,000 years old Czech Republic(Central Europe).
Kostinki: 36,000 years old Russia(Western egde).
Goyet: 30,000 years old Beligium.
El Miron: 20,000 years old Spain.
Villabruna: 14,000 years old Northern Italy.
Loschbour: 8,000 years old Luxembourg.

Matt said...

Yeah, lmao R1b1 in 14,000 YBP Upper Paleolithic Italy! What a surprise.

While the particular R1b1 clade common in Europe today seems like it is going to be due to Yamnaya (I'm still pretty sure) and ultimately their EHG side of their ancestry (less sure but still give it the higher probability), does this change our likelihoods at all on where R1 originated and diverged to the downstream R1a and R1b groups?

We now have R1b1 in an Upper Paleolithic European* (14,000 YBP), then R1a1 and R1b1 in Karelia (8000 YBP) and Samara. It looks like it's plausible that the Villabruna cluster effectively is WHG and that the Villabruna cluster contributed to EHG, along with relatives of AfontovaGora3... Mal'ta boy carries R, non-specific (not R1 or R2), at around 20,000 YBP. Seems like whether R1 diverged in Western Asia / Anatolia among ancestors of the Villabruna cluster, or Siberia may be back on the table? May need some more thought.

*and yes, later in a Neolithic male in Spain.

Nirjhar007 said...

IMHO R1a and R1b diverged in Western C Asia, R1 and R2 around SC Asia. I favor a non steppe origin of R1b-m269. Only getting samples from Europe will never solve the origin debates...

epoch2013 said...

So, El Miron doesn't have mtDNA H but U5b.

Nirjhar007 said...

news take :
Using 51 samples, the team was able to build up a far more extensive view of the changes over time. Findings, published in Nature, showed there were two major population changes in Europe between 45,000 and 7,000 years ago. The upheavals reveal how the Ice Age impacted the migration of populations and modern Europeans came from one founding population that lived in north west Europe roughly 35,000 years ago.

This founding group was displaced, however. It was found to have been living in south west Europe 19,000 years ago. As the ice started retreating, this same group spread northwards and repopulated Europe.

Another massive upheaval took place 14,000 years ago, when populations from the south east, including Turkey and Greece, spread into Europe. Their arrival introduced a new genetic component into modern Europeans. "These results document how population turnover and migration have been recurring themes of European pre-history," the study said.

Reich added: "What we see is a population history that is no less complicated than that in the last 7,000 years, with multiple episodes of population replacement and immigration on a vast and dramatic scale, at a time when the climate was changing dramatically."


http://www.ibtimes.co.uk/ancient-dna-reveals-ancestry-migration-history-modern-europeans-1557485

Chad Rohlfsen said...

The R1bs are new comers from the Balkans or West Asia, so it seems.

Nirjhar007 said...

Without aDNA from those key areas , we can argue all we want :) .

Matt said...

Quiet here (so I assume the dialogue is going on elsewhere).

Trying to interpret this (and probably getting something wrong) the genetic history and cultural of Europe from this so far looks like this

- 42,000-37,580 BP - Oase1 type people (and possibly others) living in Europe (Romania)

- 35,160-34,430 BP - GoyetQ116 type people living in Europe (Belgium)
(GoyetQ116 is classified to the Aurignacian culture wave).

- 34,580-31,210 BP - Vestonice cluster types seem to replace the GoyetQ116 type

(Except the GoyetQ116 type, Aurignancians(?) are likely to still be living in Iberia, as they're ancestral to the El Miron cluster, from later in history)

Vestonice types seem to all be classified as Gravettian.

- Glaciation then wipes out the Vestonice / Gravettian types around 24,000 BP.

- 18,830-18,610 BP - El Miron is an admixture of a GoyetQ116 and the Villabruna branch (the latter thought to be Near Eastern in origin due to greater sharing with present day Near Eastern people)

All samples from then on in Europe, are identified with the Magdalenian culture and with the same cluster membership as El Miron, as these guys expand across Europe after the retreat of the ice.

- 14,180-13,780 BP - Villabruna branch ancestry in samples increases a lot, and material culture transitions to other industries - Epigravettian, Azilian, Epipaleolithic and Mesolithic. Looks like another great migration to Europe (out of the East via Italy?).

------

Seems actually seems quite clear. These material cultures which differed - Gravettian, Magdalenian and the post Magdalenian cultures - were three different genetic clusters. Presuming the assignations are correct.
Anything I'm missing or simplifying here? (Maju if you're listening in, for instance).

One problem with this is seems: no Solutrean samples. That would help answer whether the GoyetQ116 type population who seem to be ancestral to Magdalenian / El Miron were also ancestral to the Solutreans, who were then ancestral to Magdalenians. Presumably there just are no remains.

Also of note: "Malta1 is consistent with being an outgroup to Upper Palaeolithic Europeans" and "No ancestral connection between Malta1 and samples from the Vĕstonice Cluster - Shared ancestry between Malta1 and the Vĕstonice Cluster would be predicted if the assignment of both to the Gravettian culture was due to movements of people (the Mal’ta site yielded Venus figurines stylistically similar to Gravettian sites such as Dolní Vĕstonice in Europe, thousands of kilometers to the west). Thus, if the cultural similarity that led to the assignment of Malta1 to the Gravettian culture is not a coincidence, it is likely to reflect communication of ideas rather than movements of people."

Arch Hades said...

What is the R1 carrying Italian autosomally like? Is that info in the supplementary material?

Also, is the 12,000 BC population essentially "Anatolian farmer"? So is this now confirmation that migrated to Europe before the Neolithic revolution in Europe?

Matt said...

@ Arch Hades, the 12,000 BC population is basically what we've previously just been calling "WHG". Here they're called the Villabruna cluster, after a type defining sample in their set, and are associated to various post Magdalenian material cultures in Europe. They're not "Anatolian Farmer"; they're "WHG". WHG / Villabruna replaced people who came before them, who were also West Eurasian, but with less affinity to both the Near East and East Asians.

As they (WHG / Villabruna) share more drift with the Near East than the people who came before them in Europe (the other clusters), which is why they are described (at least in press) as being likely having an origin in Greece or Turkey. See Section 11 of the supplement and particularly Table S11.1. They also seem to have more affinity to East Asians, which is another likely clue to an eastern origin.

Like all WHG, they are different from EEF in that they don't seem to have Basal Eurasian ancestry, which has been essentially redefined in this paper as a type of ancestry that reduces relatedness to Ust Ishim, basically.

Matt said...

Also, the R1b1 Italian sample, is Villabruna, the type sample for the WHG / Villabruna cluster.

Krefter said...

This is an amazing. Same techqnue as post-6000 BC ancient DNA papers but on Paleo genomes.

This is my summary of it.
Europeans dating 36,000-30,000 years old from Russia, Beligum, Czech, and Italy form a clade with 24,000 year old Siberian MA1. But they're more related to each other than to MA1. Also, none of them have an especially close relationship to each other, unless they come from the same archaeological site/region.

DNA from a lady who lived in Spain 19,000 years ago, has a close relationship to people who lived in Belgium 30,000 years ago. She also has a close relationship to "WHGs". So, she looks like a mixture of WHG and those 30,000 year old people from Belgium.

After 14,000 years ago, Europeans from Northern Italy, Germany, France, Luxembourg, Spain, and Hungary all form a tight cluster. They are WHG. They also appear to have minor ancestry related to the 30,000 year old people who lived in Beligium.

Stone age Middle Easterners CHG(Caucasus) and especially EEF(Turkey) have WHG-related ancestry. They're more related to post-14,000 years ago Europeans than to earlier Europeans. There's an unknown link between the Middle East and WHG.

Krefter said...

FOrgot to mention. Genomes dating about 15,000 years old from Germany appear more like El Miron than WHGs. But samples dating 8,000-9,000 years old from Germany are WHG. Furthermore a genome from Southern Italy dating 11,000 years old is WHG.

SO, it looks like WHG replaced El Miron-types in Central Europe and Spain(La Brana-1). It's possible WHG originated in Italy, but that's a premature theory at this point.

rozenfag said...

Well, we need Paleolithic genomes from the Middle East. Given the climate, it's complicated, but there is hope.

Alberto said...

A few questions come to mind:

- The Villabruna types appear ~14,000 BP, but we have that "Red Lady of El Mirón" (18,830-18,310 cal BP) that already looks admixed with these types. So did these Villabruna types enter Europe just after the LGM and spread fast all the way to Iberia? Or where they already in Europe before the LGM?

- Now it seems clearer that these Villabruna types (WHG, as we know them) are indeed the ones that mixed with Basal Eurasians in the Near East. What we don't know yet is how that happened. Did these types enter Europe from the Near East, and later the Near East received an influx from Basal Eurasians? Or were Basal Eurasians in the Near East and these Villabruna types moved to the Near East from Europe?

- R1b1 in this Villabruna sample: interesting indeed, but difficult to know what it means exactly. A loner? Or something more than that?

- Afontova Gora 3: indirectly dated to 16,930-16,490 cal BP). "The mandible demonstrates a gracile structure: its longitudinal and transverse measurements fall into the category of “small” or “very small”. The gracile mandible is quite different from the children of similar age recovered of Sungir-2 and Sungir-3, which might indicate different affinities." Autosomally similar to MA1, but a modern version with higher affinity to probably everyone. Will be interesting to see D-stats with her. Her mtDNA was R1b (where is that from?).

Matt said...

@ Krefter, pretty good way to explain it (although of course naturally I like my summary and what I included better :) ).

Note on CHG, though for when you say "Stone age Middle Easterners CHG(Caucasus) and especially EEF(Turkey) have WHG-related ancestry", though "Section 12 - Population affinities of the Satsurblia Cluster" is important:

- "Satsurblia Cluster samples have Basal Eurasian ancestry" (as defined by Ust Ishim and Oase1, as they have abandoned East Asian affinity as a signal of Basal Eurasian, due to findings related to pre-WHG West Eurasian populations having a low affinity to East Asians. I find this somewhat questionable, as Ust Ishim and Oase1 relatedness seems plausible to occur through admixture without any need for ghost Basal Eurasians, but there we go.)

- "Satsurblia Cluster samples have West Eurasian as well as Basal Eurasian ancestry"

but

- "The Satsurblia and Villabruna" (WHG) "Clusters are not particularly closely related".

"In all fitting models, Satsurblia is inferred to harbor ~32% ancestry from a Basal Eurasian lineage that branched before UstIshim (Extended Data Figure 4). These results are consistent with the finding that Satsurblia Cluster samples have Basal Eurasian ancestry, while also rejecting the previous model in which Satsurblia is of entirely Basal Eurasian ancestry"

"All three of the best fitting models in Extended Data Figure 4 specify that the majority ancestry component in Satsurblia branched very deeply in the tree of West Eurasian populations, forming a clade with Malta1"
.

Satsurblia's 68% non-Basal Eurasian ancestry appears to be a very deeply divergent branch of the Malta1 / ANE cluster, if anything, not WHG. (Which may be of interest to Alberto and Sein, regarding their recent attempts to model Basal Eurasian in the CHG)

@ Alberto, yeah, we have no way of knowing if the R1b1 in Villabruna was a "loner". It seems just about possible to me that it could have been widespread in a Southeast European or Anatolian or West Asian population that seems to have given rise to WHG, but was not in WHG due to founder effects and/or mixture with the previous Magdalenian (El Miron) wave. Then it could have been replaced along with others by y-dna lineages from "Basal Eurasians" (if they existed), like G or J2.

But this seems not particularly parsimonious and really we have no way of knowing how widespread R1b was.

Colin Welling said...

Wow, super interesting article. Ill summarize the results.

There was no Basal Eurasian admixture in pre Neolithic Europe. The reasoning was that Ust Ishim (45kya Eurasian) does not contain Basal Eurasian and Ust Ishim is equally related to East Asians and pre Neolithic Europeans.

There was no ANE, Mal'ta cluster, admixture in pre Neolithic Europe (presumably they are talking about the portion of Europe that is West of Russia and Ukraine). The reasoning here is that the mal'ta cluster is equally related to all pre Neolithic "Europeans" starting with Kostenki (37kya).

-Consequently, ANE only entered "Europe" during the Neolithic and Bronze Age, probably from the Eurasian Steppe.

So far, GoyetQ (35kya Belgium) and Kosteni (37kya Russia) are the first ancient Europeans to exhibit closer relations to modern Europeans over East Asians. The older Eurasian samples of Ust Ishim and Oaste do not show higher affinity to modern Europeans over East Asians.

European samples from about 37kya to 14kya tend to be equally unrelated to non-European Eurasians. However, there are 14kyo samples, representing the Villabruna cluster, that show a higher affinity to the modern Near Easterners. The 13kyo to 10kyo samples from the Caucasus, representing the Satsurblia cluster, show more affinity to the Villabruna cluster than to the earlier Europeans. The Villabruna cluster can not be the result of admixture with the Satsurblia cluster since the former does not contain Basal Eurasian while the later does. GoyetQ (35kya Belgium) type ancestry shows a resurgence in the Miron Cluster of 19kya centered in Iberia. About half of the ancestry in the Miron Cluster derives from GoyetQ types.

Because of the large number of individuals in the Villabruna cluster, the researchers were able to identify structure within the cluster. One interesting tid bit is that certain individuals in the Villabruna Cluster showed extra affinity towards East Asians. The oldest of which is the 13kyo sample from Switzerland. Two other Villabruna individuals, Loschbour and La Brana, for example, show extra affinity to Han Chinese over Kostenki. Because all three individuals are equally related to Ust Ishim, hypothetical differences in levels of Basal Eurasian cannot explain the greater affinity Loschbour and La Brana have to the Han.

Rob said...

Great paper & good summaries by others

@ Matt

"Seems actually seems quite clear. These material cultures which differed - Gravettian, Magdalenian and the post Magdalenian cultures - were three different genetic clusters. Presuming the assignations are correct.
Anything I'm missing or simplifying here? (."

Indeed. Industries are based on stool tools, hunting techniques etc. These were the lifeline of hunters, and thus we can expect a complex chaine opertoire, ie 'a socially conditioned genetic inheritance' (quoting myself)

* "One problem with this is seems: no Solutrean samples. That would help answer whether the GoyetQ116 type population who seem to be ancestral to Magdalenian / El Miron were also ancestral to the Solutreans"

I would have thought there should be Soultrean skeletons, can't recall on top of head. Shall dig around , but I think magdalenians might be a decent proxy for Solutreans, but cant be too sure. Don't' want to make simplistic conclusions.

" - 14,180-13,780 BP - Villabruna branch ancestry in samples increases a lot, and material culture transitions to other industries - Epigravettian, Azilian, Epipaleolithic and Mesolithic. Looks like another great migration to Europe (out of the East via Italy?)."

Yes, these post-/ Late Glacial industries seem to be native to Europe, but from southern Europe (Italy & Balkans). So the shift toward East and/or Near East is likely just low-level diffusions (procurement networks, occasional mate exchange) amongst mobile HG networks. Archaeology, nor the Y lineage data, doesn't seem to support wholescale arrival of new groups from Near East, as the otherwise awesome paper clumsiliy suggests. I'd like to see what you boys make of some number crunching; and further define how R1b* fits into this with more data from Russia, Balkans & *northern* Near east.

Colin Welling said...

"The R1bs are new comers from the Balkans or West Asia, so it seems."

This paper still reaffirms the idea that ANE and R1b M269 come from the steppe.

If you are instead talking about the origin of r1b, the common thread among ancient r1b peoples is having heritage from Europe to Siberia. Its not really about having heritage from the ice age Near East.

Davidski said...

This quote is worth repeating:

"However, a plausible alternative is population structure, whereby Upper Palaeolithic Europe harboured multiple groups that differed in their relationship to the Near East, with the balance shifting among groups as a result of demographic changes after the Glacial Maximum."

Chad Rohlfsen said...

The R1b is pre P297, and does look more like a loner. L23 is still the key. These groups look like they spread from SE Europe or Anatolia about 20k years ago, would be my guess.

Key points too are that there is no gene flow between Gravettian and Malta, and more surprisingly, no gene flow from Gravettian to Epi Gravettian. It looks like only minor Gravettian in LaBrana and Loschbour, about 20 and 15% respectively.

Chad Rohlfsen said...

Excuse me. 15-20% Magdelenian. I'll look that over again.

a said...

@ Colin Welling said...

"The R1bs are new comers from the Balkans or West Asia, so it seems."

This paper still reaffirms the idea that ANE and R1b M269 come from the steppe.

"If you are instead talking about the origin of r1b, the common thread among ancient r1b peoples is having heritage from Europe to Siberia. Its not really about having heritage from the ice age Near East."

Ancient samples,
How many R1b samples from Anatolia? - 0%
How many R1b samples from Italy?- 1, 14,000+/- YBP. Maybe V88+related
How many R1b samples from Spain?- 1, 7,500+/- YBP.
How many R1b samples from Samara Steppe region?-1, 7,500+/YBP. M73+

Now let's see how all R1b samples cluster/plot in Eurogenes; with all the ancient samples combined?

Rob said...

@ Chad

"no gene flow from Gravettian to Epi Gravettian."

That's because we don't have Balkan Gravettian samples.....yet

Seinundzeit said...

Amazing, a lot of new information/analyses to look at (I'm pretty late, finally got on my laptop after a whole day, and was surprised/glad to see that the aDNA drought has ended)!

Just as Matt noted, it seems that we were right about the connection between CHG and ANE, as Satsurblia seems to be predominantly MA1-related + some Basal Eurasian admixture (although I think CHG and South Central Asians/West Asians also have some ancestry from an unsampled population which makes Villabruna/WHG closer to contemporary Near Easterners, in comparison to more ancient Europeans with regard to contemporary Near Easterners).

And I suppose we can now safely say that EHG/SHG are a mix of something related to MA1 and Villabruna/WHG, in varying proportions.

Although, I'm surprised at how the MA1/Afontova Gora group is quite distinct/diverged from the whole rhizome of ancient European populations (from Kostenki14 and GoyetQ116 to Villabruna/Loschbour).

For whatever it's worth, they seem to now recognize WHG-ENA gene flow (also ANE-ENA gene flow, and EEF-ENA gene flow), and with that they've discarded the notion of Basal Eurasian ancestry in Kostenki14. Yet they don't dispense with the Basal Eurasian concept in totality, which I think is good (I now think that the Lazaridis et al. Basal Eurasian construct is mapping unto an actual cluster of ancient populations, although it makes more sense to refer to Ust-Ishim/Oase1 as "Basal Eurasian").

Krefter said...

None of you noticed the 33,000 year old Y DNA I from Southern Italy? Pretty big deal.

Nirjhar007 said...

Of course we noticed , but it isn't so surprising is it?.

imam-din said...

I think these ancient samples lysosomal clusters rule out the diversification of R1 lienages anywhere in europe or even near that region. R1 just like its brother clad R2 is most probably central eurasian in its origin due to strong correlation of ANE with the high incidence of R1 and R2 lineages in areas inhabited by isloated tribes like Kalash and Burusho.

Chad Rohlfsen said...

They missed important stats. I'll cover it tomorrow.

Krefter said...

The French WHGs dating 13,000-7,000yo are significantly closer to LuxembourgWHG Loschbour than to Italian, Spanish, and Hungarian WHG.

Matt said...

@Sein "And I suppose we can now safely say that EHG/SHG are a mix of something related to MA1 and Villabruna/WHG, in varying proportions."

As I expect you've already noticed, this is another topic on which this paper seems to, almost incidentally, offer clarification. But for anyone else who didn't yet:

EHG was postulated to be unadmixed because EHG showed as strong a relationship to Native Americans as Malta1.

However, from this paper's SI on p44 (see Table S5.8), we know AfontovaGora3 (16,930-16,490 cal BP) to actually be a better representative of the population that contributed to the Americas, and to Karelia / EHG, so the problem of how EHG could be admixed yet as related to Native Americans is resolved.

Malta1 seems from more of a dead branch that is further from the real branches of the "ANE rhizome" that contributed to EHG / Karelia. Not unexpected, as we go back earlier in time, the more likely it is that a population's from a dead branch (Oase1 and Ust Ishim being the extreme cases of this).

@ Chad, in relation to the above, stats that I can see they have not included that would be:

D(Malta1, AfontovaGora3; Satsurblia, Mbuti) and D(Malta1, AfontovaGora3; Kotias, Mbuti)

which would be cool to see if they are zero or significant, as they identify CHG / Satsurblia cluster as likely having Malta cluster related ancestry, and others who do tend to be closer to AfontovaGora3. I think it's likely the CHG's Malta cluster related ancestry will be very deep divergent from AfontovaGora3.

I think it would also be cool to see something like the Table S11.1 D(European1, European2; Iraqi_Jew, Mbuti), which confirms the increase in relatedness to the Near East for the post-Magdalenian Villabruna group, rerun with both Anatolia_Neolithic or Han/Dai/Atayal or BedouinB in place of Iraqi_Jew

FrankN said...

If anybody intends to re-run some stats, here a few more suggestions:

1. The Near-Eastern realatedness may in fact be a North African one. Remember that sea levels were some 80-130 m lower during the glacial, so one could in fact almost walk from Sicily to Tunisia - alright, not to Tunisia, but at least to Malta and the Egade Islands:
http://mapmistress.blog.com/2012/04/14/the-maltese-islands-to-sicily-malta-ragusa-platform/

http://www.google.de/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&uact=8&ved=0ahUKEwia4Lu3sr3MAhXF1SwKHT_FBGEQFggnMAM&url=http%3A%2F%2Fwww.geoeco.ut.ee%2Forb.aw%2Fclass%3Dfile%2Faction%3Dpreview%2Fid%3D1168667%2FAlba_Mazza_poster.pdf&usg=AFQjCNH5vWrplJtQ765Z_dNA7mWrUeeAEQ&sig2=JK2AhdhEr4mrxWhTyqEbag

http://www.clim-past.net/7/161/2011/cp-7-161-2011.pdf

So, including Mozabite or other North African populations into the stats might be worthwhile. UP interaction across the Western Mediterranean (Strait of Gibraltar) is anyway archeologically evidenced.

2. There is a possibility of genetic relation between Mbuti and the Near East. The Mbuti traditionally keep Basenji dogs, a very ancient race of West Eurasian origin with Israeli Wolf admixture. Assuming that the Basenji didn't reach the Mbuti unaccompanied by humans, some levantine-Mbuti contact during the Late UP/ Mesolithic should be anticipated. Thus, it might be a good idea to rerun stats that involve NE populations whith Chimp instead of Mbuti as outgroup.

3. If anybody has the genomes at hand - it might be worthwhile to include Paleoindian aDNA (Anzick, Clovis, Sappaq) into the ANE/ENA-related analysis. Those samples are not only a better fit timewise, but should also remove effects of later SEA (Onge/Papuan) admix in Karitiana. Such admix has both been demonstrated by Raghavan 2015 and Skoglund/Reich 2015.

Olympus Mons said...

So, Villabruna (R1b) in Italy as people that cluster to people of Satsurblia in southern caucasus, near that Kotias (J2) and Satsurblia (J1b), Same R1b and J2 that are latter seen (as outliers) also in Yammna millennia later… It all seems to point to the fact that R1b has little to do with northern Caucasus or Ukraine steppes. That is territory of the R1A and nothing else. And since the planet is orthodromic distance … they didn’t really went that far, did they? Not to western Europe that is for sure.
R1b (and some J2) in northern caucaus are a remnant of the push from Ubaid/Uruk of southern Caucasus people over the mountains to northern caucausus. Just remnant people pushed between 4500 BC to 3500 BC .
Just because its colder up north (preserved DNA) does not mean increased frequency of anything. That is why we will find in the near future bucket loads of R1b in Southern Caucasus. Just need to look for it. Especially7/ 6th millennia BC. Go to Aratashen, Shulaveri and so forth. On the other hand it’s also the reason we keep forgetting that V88 still has an R1b before.

Alberto said...

@FrankN

There is a possibility of genetic relation between Mbuti and the Near East.

Going a bit off topic with this, but a recent paper re-evaluating the Kennewick Man genome had something of interest. In addition to using formal statistics and Admixture, they used that novel rare variant analysis. If I remember correctly from the paper about it, 2 things about this method were that it was more deterministic (vs. probabilistic) and more resilient to the pass of time than IBD. Now, in this other paper about the Kennewick Man apart from using it on that genome they also used it with the Stuttgart EEF genome (Page 14 from the linked paper), and one interesting detail is that Africans (Yoruba, Esan) share significantly more of these rare variants with Stuttgart than East Asians do. So this would kind of prove some gene flow between Africans and something related to Stuttgart, to the exclusion of East Asians. We don't know from this in which direction the gene flow happened, though. But some ancient Near Eastern to Africa looks quite plausible.

http://cdm16021.contentdm.oclc.org/cdm/ref/collection/p16021coll11/id/950

Would be interesting to see that rare variant analysis with other ancients too.

Davidski said...

Not sure what some of you guys are puffing on this week, but we have just seen R1b reported from Ice Age Europe. And we already know that Mesolithic Europeans carried both R1a and R1b.

Despite that, you're taking this news to mean that R1 is not native to Europe, but to pretty much everywhere else but Europe.

Sober up, and think about what these results really mean, not what they might suggest for places that are yet to be sampled (although do keep in mind that western Anatolia was devoid of R1b during the Neolithic, so there's that).

Angantyr said...

@Krefter "The French WHGs dating 13,000-7,000yo are significantly closer to LuxembourgWHG Loschbour than to Italian, Spanish, and Hungarian WHG."

Can this be connected to some other affinity (or lack thereof)? Like to any of the older HG clusters, or to ANE, or to East Asians?

Olympus Mons said...

@Davidski,

*If by Mesolithic Europe you mean the area near Asia part of volga/Urals, then yes, probably been there since Paleolithic.
*Finding R1b in Paleolithic Europe is not like if the R1b existed for 15,000 years under a rock in the Volga margins or de Dnieper basin just to pop up in western Europe. Sure they were around like others.

But, what it starts to look like is that the Yamna, Cwc, expansion was a short distance event spreading a merger 500-700km from base seat on top of R1a people… so, let’s start to look for another source of the Pie languages, shall we?
and let’s be attentive for other possibilities involving R1b expansion in Western Europe. Or any other expansion for that matter. Lets just drop the “from the steppe” uber important to current outlook of Europe. It would hurt.

Note: Davidski, if there is a good recent post that sums it up nicely your view, I wouldn’t mind reading it…. Thanks.

Davidski said...

You obviously don't realize this, but you're not making much sense.

For instance, it's impossible that the Corded Ware expansion was on top of R1a people for two important reasons.

- The Corded Ware people, in fact their elites, were mostly R1a, because most burials being tested are those of elites

- Most of the R1a in the world today is from a rapid founder effect dated to the Corded Ware period and Bronze Age.

Kurti said...

Roy King said

"So Villabruna (Northern Italy) circa 14K bp has Near Eastern affinities and is R1b1."

Is that a suprise? What did I say :). Didn't I say that R1b v88 in 5500 BC Iberia with EEF aDNA is unlikely from the Steppes and probably came from a region that was close to the homeland of EEF and they catched it up there? For instance Mesopotamia or the Iranian Plateau? :)

Olympus Mons said...

Sorry Davidski ,
what I meant was that Yamna/corded ware is a very specific cultural an demic event related to R1a and just R1a. But sort of irrelevant to most of other cultural/historical phenomenon that usually remain attached to those, we should in fact start to look for other origins/pathways. So , too much attention in the last decade to Kurgan hypothesis and Maria Gimbutas stance and if someone tries to say that PIE is a J2 expansion via Anatolia… hey let’s look into it (although I think you don’t agree as I don’t).
See I personally think that the key element was actually the Ubaid up until Uruk period that steered things up from the 6th Millenia BC onward (ending up as Mesopotamia and Sumer and ur), making (my opinion) for instance a specific package of R1B in Southern Caucasus run south and west and make the Bell beaker and R1b expansion out of Iberia. To me is all there, but the sampling of DNA. So, sampling the all southern region from 4500 bc up until iron age does not mean a thing if 7th and 6th millennia is not sampled.
Overall message is that there is too little sampling at this point in time to make much of an history or get upset with each other about just a story. It’s all pet theories. Well I have got mine:

http://blogs.sapo.pt/cloud/file/eb6b52b82097d41dfa0e5797a2fa7945/olympusmons/2016/From%20Shulaveri%20to%20Bell%20beaker.pdf

Kurti said...

As I said in the past by Neolithic era R1 lineages have been already widespred throughout Eurasia. Mark my words we will find R1b l23 that predates Bronze Age in West Asia.


And than according to this paper WHG is not Paleolithic European but a pre Neolithic Anatolian component that replaced Paleolithic Europeans roughly 14000 BC.

I hate to say it ( actually not :P) But didn't I say that the "WHG" in Neolithic farmers is not a back migration of mesolithic Europeans into Anatolia.
I said that this WHG is ancient in the region since it is also found in the Levant and even Arabia and probably represents a pre Neolithic Near Eastern component.

WHG = Mesolithic Near Eastern EF= Neolithic Near Eastern CHG = mesolithic_neolithic Iranian Plateau.

Davidski said...

If Villabruna is Near Eastern then I'm an Australian Aboriginal.

mickeydodds1 said...

'The more we find out, the less we know'.

With the unbelievably tortured and convoluted attempts to try and discern the affinities of ancient populations to the extant, all that seems to happen is that unearthing complexities just causes further complexities to be unearthed.

FrankN said...

@Kurti: The paper's SuppMat Section 13 states clearly that the Villabruna cluster wasn't a population replacement, but resulted from an admixture event that involved at least three distinct populations.

- The first population is anchored in Vilabruna (R1b) himself. While Vilabruna obviously incorporates novel elements of a putatively Near Eastern origin, he is nevertheless still essentially UP West Eurasian, i.e. from a population that branched after UI, Kostenki, MA1 etc., and contributed some 25% to Vestonice ancestry (SM Sect. 6, note also the TreeMix results in Fig. S6.6 that place El Miron and Vilabruna on the same branch). I interprete this as reflecting the LGM seperation between the Ibero-Aquitanian and Adriatic (Italo-Danubian) refugia, and Magdalenian and (Epi-)Gravetian cultures, respectively.

- A second admixture element was the El Miron cluster (Magdalenian), and ultimately Goyet Q116-1 (Aurignacien, 35 kya), with variable Goyet ancestry shares (20% La Brana, 16% Loschbaur, 0% Bichon - Tab. S7.3). Unfortunately, the study didn't attempt to determine El Miron (Magdalenian) vs. Vilabruna (Epi-Gravetian) ancestry shares in La Brana and Loschbaur. They are probably somewhat higher than the shares calculated for Goyet Q116-1.

- The NE admixture signal becomes already apparent with El Miron (which makes me think it may actually incorporate North African admixture). It is present across the whole Vilabruna cluster, but in widely varying extent, weakest in Falkenstein and Berry au Bac, and strongest with KO1 and Loschbaur. For KO1 and Loschbaur, the signal strength may already reflect initial contact with EEF. Rochedane, Berry au Bac and Falkenstein appear to mostly reflect NE-like ancestry incorporated via El Miron (Tab. S11.1). Still, the signal in Vilabruna, Bichon and, to a lesser extent, Ranchot88 is notable. I'd tentatively link it to the presence of domesticated dogs both in the Levante and in SW France.

- For some members of the Vilabruna cluster, namely Loschbaur, La Brana, Bichon and KO1 (but not yet Vilabruna himself), there is indication of East Eurasian admixture. Interestingly, this admixture seems neither related to MA1, nor mediated via the Caucasus.

With all these sources identified, the authors nevertheless state: "We did not succeed at convincingly modeling the ancient relationships among these sources."

So, for the time being, and in the absence of contemporary aDNA from SE Europe, as well as NE Europe (Cheddar Man, Ahrensburgian, Swiderian etc.), I would be careful with making too radical conclusions. What is becoming clear, however, is that the role of the Adriatic (Italo-Danubian) refugium in the post-glacial repopulation of Europe seems to have been underestimated.

epoch2013 said...

I understand the conclusion that Kostenki 14 does have Basal Eurasian admixture. However, the Seguin-Orlando had more than just D-stats. I understand the value of D-stat and the relative lower value of ADMIXTURE. However, their ADMIXTURE run as well as others clearly show a Middle-Eastern componentn in K14. Not just that, D(f3,Mbuti,X, K14) presents Sardinians as having a clear K14 signal.

epoch2013 said...

Also this bothers me:

"There is no Basal Eurasian ancestry in Kostenki 14, and instead, gene flow occurred between the ancestors of East Asians and Europeans after 14,000 years ago.
(Independently, there would need to be gene flow between the ancestors of East Asians and Malta1 to explain its affinities.)"

So we need a gene flow from a population on completely the other side of Eurasia, in a time when the two populations were thoroughly seperated, and we also need a gene flow into ANE to explain that *some*, but not *all* WHG show affinity to Han. Mind you, it may be only slightly above the noise, but the Aurignacian Goyet-116-1 also shows some Han affinity.

Karl_K said...

This could also be the result of even one single unsampled group admixing into each of these populations.

We know that the EDAR allele was present in SHG populations, and in Native Americans, and in East Asians. Some low level of gene flow must have occurred between these populations.

Chad Rohlfsen said...

I'm curious if it has something to do with them splitting from ENAs on F ydna later than C. It would be some extra shared drift with the NO group, to the exclusion of the Aurignacian and Gravettian.

German Dziebel said...

@epoch2013

"So we need a gene flow from a population on completely the other side of Eurasia, in a time when the two populations were thoroughly seperated, and we also need a gene flow into ANE to explain that *some*, but not *all* WHG show affinity to Han. Mind you, it may be only slightly above the noise, but the Aurignacian Goyet-116-1 also shows some Han affinity."

At least someone here is focusing on the right thing. The authors' interpretation is indeed confusing. They needed to include Amerindians next to Malta and Afontova Gora to understand the exact affiliations of those ancient European samples. Goyet-116-1 is likely closer to Amerindians than to Han (just like all Europeans, ancient and modern). Upon leaving America, an ancient Amerindian population branched into East Asian, ANE and West Eurasians. This would explain the data well.

epoch2013 said...

@German

"They needed to include Amerindians"

They did. See figure 4b of the study as well as extended figure 3.

"Goyet-116-1 is likely closer to Amerindians than to Han (just like all Europeans, ancient and modern)."

No. See figure 4b of the study.

epoch2013 said...

@Chad

When - if - these genomes come available someone should rerun extended data table 6 with other Europeans than French. Baltics and Bosnians. Swedes and Dutch. The way Goyet-116-1 would pop up among its peers of same age would be interesting. It would also be interesting to see if Gravettians such as Vestonice 16 would give a similar signal.

German Dziebel said...

Thanks. I can't read Ext. Fig. 3. The crosses are too close to each other on the relevant charts. The PCA plot Fig. 3b shows Karitiana as closer to Goyet than Han is. Also This is consistent with Oase (the original study) being closer to Amerindians than to East Asians

pconroy said...

My theory is that Y-DNA P evolved in or around Northern India, then diverged into R in Pakistan, and R1, R2, Q in Northern Pakistan. Then R1 split into R1a North of the Caspian and R1b South of the Caspian.
Then R1b developed pastoralism as part of Mehrgarh Culture, moved through Turkmenistan, Northern Iran and on to Armenia (Syunik etc), and helped form the Kura Araxes Culture, then split, some going North to form Kavalynsk in Southern Russia (bringing mtDNA T1a and other Middle Eastern mtDNA groupings), and from there to Western Ukraine/Romania, then up the Danube, across the Sava (near Belgrade), and on to the Po Valley of Northern Italy, then to Liguria (aka France/Italy border and SW Switzerland). While others brought Kura-Araxes culture to the SW to the coastal area of Syria/SE Turkey and the foothills of the Taurus Mountains. From there to Cyprus, and split again, some going to coastal Albania and on up the Adriatic to the Po Valley and on to Liguria, others to Sicily and Corsica and on to Liguria.

From Liguria, there was a pastoralist population explosion, with some going SW to Iberia, some NW towards the Atlantic and Brittainy, some North up the Rhone towards Eastern France and then the Rhine to Belgium/Holland and Germany.

tew said...

re: Villabruna, I think people are getting too carried away by the word "Italy". The site is in Belluno, close to the Slovenian and Austrian border, and basically near the northern end of the Balkans taken as a region. The refugium is question would have been the Balkans. Plus, wasn't Gioiello's prediction specific to Tuscany and environs? That's many hundreds of miles to the southwest and in a very different ecological setting.

Davidski said...

Yeah, the Villabruna cluster probably comes from the Ballkan refuge.

The reason it shows a close relationship to an Near Eastern component, is because the Balkans were joined with Anatolia until more recent times.

batman said...

Please note that the dating given for Villabruna ("14.000 BP") may be a bit early. Please also note that the Magdalenian and Gravettian populations - along with the mamoths and cave-bears - all disappeared during the Younger Dryas, when the coldest millennia on record swept across the northern hemisphere.

http://www.academia.edu/3676838/Bodies_bits_and_pieces_Burials_from_the_Magdalenian_and_the_Late_Palaeolithic

Except from a small number of people, who happened to find a sustainable refugia at the shores of the Atlantic facade. Today we know the area of survival - as the Hamburgian-Bromme culture actually are proven to have survived YD in the Danish straits, from where they started to spread both north, south, west and east - right after YD - as of 12.000 years ago. To date, this is the only proven refugia known to have made it through the extreme climate-fall that happened at 12.900 yrs BP, not to improve substantially before 11.900 BP.

In the oldest known myths from the northern hemisphere such an extreme event is actually described - as the "Fimbulwinter".

Please note that the re-occurance of the human populations across Eurasia - after the Younger Dryas - where all descendants from the groups of "Cro-Magnons" that survived the YD bottle-neck.

http://horizon-magazine.eu/article/ice-age-europeans-roamed-small-bands-fewer-30-brink-extinction_en.html
http://www.livescience.com/9578-common-ancestor-blue-eyes.html


The result, re-populating the arctic and semi-arctic parts of Eurasia where all Caucasians. With time they obviously mixed with the tropical populations that had survived YD in the semi-tropical south.

Today we may link makrogroup C-F to this northern Eurasians, from which the subgroups/haplogroups GHIJK/L - and thus K2/R1/R2.

This seems to define the surviving male-line, north of the Alps, as hg C-F.
This seems also to define a common ancestor to the dynasties that formed the populations (etnicities) that's based on hg's G, H, I, J and O - paralell to the cultures known as Minoan/Hellenistic, Sumerian, Indian and Chineese, as well as Gothic (I), Vendic (N). Later the agriculture spread with the main-lines R1a (Scando-Slavic) and R1b (Germano-Gallic).

As the connections between north and south, east and west would reach across the Eurasian continent the spread of the IE symbols and languages seem to have followed due, as the networks of boat-, sledge- and barrow-travels as well as trade and cultural inter-change started to cover the Eurasian continent. This started already during the mesolithic and became duely productive throghout the Neolithic. It seems to have flourished during Bronze Age, before warfare and social unrest started to change the old, dynastical structures.

A similar correspondance can be seen in the sperad of megalihic construction and architecture, as well as metallurgy and 'advanced commuication', such as heraldics and writing.

A similar exchange of genetic material - after the YD - seems to have been the origin of the Caucasian genomes found in the pre-colombian civiliations in America.