Distribution map of Y-DNA and mtDNA haplogroup in and around Europe circa 8000 BCE

[Maciamo]

It's been a while since I haven't made any new maps. Here is an attempt to show what Europe, the Near East and North Africa looked like in terms of Y-DNA and mtDNA haplogroups some 10,000 years ago. I delimited the (very) approximate borders of the first cereal/legume farmers in the Fertile Crescent, the first cattle herders and the first goat herders.

Of course these Neolithic people eventually expanded. G2a farmers moved west across Anatolia to Europe and east to Iran, where J2 hunter-gatherers eventually became Neolithised. G2b moved south to Egypt and Arabia. J1+T1a goat herders eventually expanded over all the Fertile Crescent, then colonised the mountainous regions of the Mediterranean, well suited for goats.

I believe that R1b cattle herders ended up squeezed between all these other groups, which forced them to move south to Africa (R1b-V88) and north to the Pontic-Caspian Steppe (R1b-M269) between 6000 and 5000 BCE.




Go to this page for a larger version (click on the map there).

[Tomenable]

Very interesting map, thanks!

Do you exclude the possibility of some subclades of R1a among First Goat or Cattle Herders ???

IIRC, according to Underhill et al. 2014, R1a originally spread from somewhere around Iran.

And what do you think about N1c - where was it hiding 10,000 years ago?

I think N1c could also be present in Mesolithic East Europe, but maybe in its northern part.

=========================

I find that OP helpful but I got a message: "You may not vote on any more threads today."

[Shetop]

My comments are about G2a, J2, and E-V13.

Firstly, both J2a and J2b had already existed in the Mesolithic, and I believe they should be viewed separately. Among two of them, it only makes sense to place J2a in Anatolia.
J2b is still very hard to understand.

Further, regarding G2a, I'm quite convinced it entered Southeast Europe before J2a. And whereas for J2a it is almost certain it entered SE Europe from Anatolia, for G2a I wouldn't be so sure about that.

Finally I believe E-V13 could also be already in Europe 8000 BCE . Though it would be impossible to say where exactly (I would bet against Balkans).

[Maciamo]

Very interesting map, thanks!

Do you exclude the possibility of some subclades of R1a among First Goat or Cattle Herders ???

IIRC, according to Underhill et al. 2014, R1a originally spread from somewhere around Iran.

I doubt that R1a was among the first cattle and goat herders for several reasons:

1) R1a wasn't part of the cattle-herding Yamna culture.

2) R1a was already all over European Russia in the Mesolithic (also R1b for that matter, but not yet R1b-V88 and R1b-L23).

3) The R1a phylogeny started to explode downstream of M417 from 3500 BCE (TMRCA for M417 according to Yfull), especially with the formation of the Z283 and Z93 subclades (formed c. 3000 BCE), which is far too recent for a Neolithic expansion.

4) If R1a were goat herders from the Middle East, there was be old R1a subclades alongside J1 and T1a in mountainous regions like southern Italy, Sardinia, North Africa, and even Sudan and Ethiopia, when in fact there is very little to no R1a at all in these regions.

And what do you think about N1c - where was it hiding 10,000 years ago?

I think N1c could also be present in Mesolithic East Europe, but maybe in its northern part.

As I explain in the N1c page, N1c1 appeared in Siberia c. 15000 ybp and probably reached the Volga-Ural region between 10,000 and 7,000 years ago. I once thought that the Kunda culture was the first N1c culture in Northeast Europe, but I have revised my estimate to a later date, with the Comb Ceramic culture (4200-2000 BCE). Therefore, around 9000 to 7000 BCE, N1c tribes were making their way across the Ural mountains, and may or may not yet have reached the Upper Volga, but I seriously doubt that they were much further west than that. After the samples from Mesolithic Karelia and Samara lacked N1c and didn't have much Siberian admixture.

[Maciamo]

Firstly, both J2a and J2b had already existed in the Mesolithic, and I believe they should be viewed separately. Among two of them, it only makes sense to place J2a in Anatolia.
J2b is still very hard to understand.

Of course they existed separately, but at present so much remains uncertain about the origins of J2b that I preferred to lumped both subclades under J2. Note that I didn't write J2*. Likewise, when I mention mtDNA H with the subclade, I mean that there are several subclades, either too many or too uncertain to specify.

Further, regarding G2a, I'm quite convinced it entered Southeast Europe before J2a. And whereas for J2a it is almost certain it entered SE Europe from Anatolia, for G2a I wouldn't be so sure about that.

I agree, but this map stops around 7000 BCE, before G2a reached northwest Anatolia or Europe. European Neolithic farmers were overwhelmingly G2a because that was the true Y-DNA lineage of Fertile Crescent farmers. Many other haplogroups pop up here and there because they were just a minority of people assimilated along the way. That includes J2a, E-M78, H2, I1, I2a, and so on. Even the T1a from the LBK culture is surely just descended from a goat herding family that joined the G2a farmers in Anatolia before moving to Europe. It is interesting to note that the big expansion of agriculturalists outside the Fertile Crescent happened when pottery was invented (from 7000 BCE in the Near East). I am not quite sure why that was an essential prerequisite (maybe food storage when advancing to new lands), but there is surely a reason why it was the G2a farmers with pottery who colonised Europe first, and not the J1 and T1a goat herders (who presumably lacked pottery at the time).

Finally I believe E-V13 could also be already in Europe 8000 BCE . Though it would be impossible to say where exactly (I would bet against Balkans).

That's what I have said for several years. I mentioned E-M78 among Mesolithic Southeast Europeans and Southwest Europeans, and that includes E-V13. It isn't only E-V13 though. Europeans also have many indigenous subclades of E-V12 and E-V22 - hence the umbrella term M78.

[Shetop]

I agree, but this map stops around 7000 BCE, before G2a reached northwest Anatolia or Europe. European Neolithic farmers were overwhelmingly G2a because that was the true Y-DNA lineage of Fertile Crescent farmers. Many other haplogroups pop up here and there because they were just a minority of people assimilated along the way. That includes J2a, E-M78, H2, I1, I2a, and so on. Even the T1a from the LBK culture is surely just descended from a goat herding family that joined the G2a farmers in Anatolia before moving to Europe. It is interesting to note that the big expansion of agriculturalists outside the Fertile Crescent happened when pottery was invented (from 7000 BCE in the Near East). I am not quite sure why that was an essential prerequisite (maybe food storage when advancing to new lands), but there is surely a reason why it was the G2a farmers with pottery who colonised Europe first, and not the J1 and T1a goat herders (who presumably lacked pottery at the time).

It is not directly related to your map but this may be a good place to write something I think many would disagree with. It is an opinion I've had for some time about the spread of agriculture - considering a lot of information we have collected about haplogroups, both from modern populations and aDNA, it looks to me as if agriculture was not spread by human migrations, but it was more like a spread of a cultural phenomenon.
This is partly why my view is a bit different when G2a is in question.

[Fluffy]

We spread agriculture.

[sparkey]

This is a tricky time period to reduce to subclades, considering how few actual samples we have for the time period. I'd be interested to see if your predictions for SE Europe are accurate in particular, since we have to extrapolate there without much to extrapolate from. I guess I don't have a better proposal for the region at the moment.

I suppose we're mainly using the Motala samples to guess what Scandinavia would have been like. If so, I think it should include I2c, since that was found there (Motala2). It's a bit misleading to have I2c in Anatolia but nowhere in Europe when it's not clear yet which location is the origin point (and I vote Europe FWIW).

[Sile]

Of course they existed separately, but at present so much remains uncertain about the origins of J2b that I preferred to lumped both subclades under J2. Note that I didn't write J2*. Likewise, when I mention mtDNA H with the subclade, I mean that there are several subclades, either too many or too uncertain to specify.



I agree, but this map stops around 7000 BCE, before G2a reached northwest Anatolia or Europe. European Neolithic farmers were overwhelmingly G2a because that was the true Y-DNA lineage of Fertile Crescent farmers. Many other haplogroups pop up here and there because they were just a minority of people assimilated along the way. That includes J2a, E-M78, H2, I1, I2a, and so on. Even the T1a from the LBK culture is surely just descended from a goat herding family that joined the G2a farmers in Anatolia before moving to Europe. It is interesting to note that the big expansion of agriculturalists outside the Fertile Crescent happened when pottery was invented (from 7000 BCE in the Near East). I am not quite sure why that was an essential prerequisite (maybe food storage when advancing to new lands), but there is surely a reason why it was the G2a farmers with pottery who colonised Europe first, and not the J1 and T1a goat herders (who presumably lacked pottery at the time).



That's what I have said for several years. I mentioned E-M78 among Mesolithic Southeast Europeans and Southwest Europeans, and that includes E-V13. It isn't only E-V13 though. Europeans also have many indigenous subclades of E-V12 and E-V22 - hence the umbrella term M78.

Its highly doubtful that T1a where goat herders along with J1 because because there is zero evidence of any T1a in the arabian peninsula or eastern Africa of any T1a older than 2000years.
T-L446 appears to show greater variation in Europe than it does in the middle east. The T-Y7381 branch found in Saudi Arabia (and heavily tested) is relatively young (1400 ybp) so could be the result of a recent migration from further north.
The paper the Levant versus the Horn of Africa also states this.If T1a and J1 where together as goat herders , they would also be together in the Arabian peninsula , which they are not. 44% J1 and 3.5% T1a ..........J1 are the nomadic group. I see T1a as per Haak , that is 95% EEF ( farmers)

The 2 x T1a in early Neolithic Central germany are surrounded by 8 x G2a and 1 x H2 ..............the only conclusion is that T1a was around with G2a in the caucasus .............IMO the ancient T in northern Europe are from the Azeri lands today.
Were there is G2a in the Alps of Tyrol, you find 5% of T1a, where you find G2a in the mountains of central france you find 4% of T1a. Where you find G2a in the mountains of central Italy you find nearly 9% of T1a.

The other factor is that all T men have the marker TL-P326, this union and eventual split still sees T and L in places together in the present and in the ancient times.....Dagestan, Lezkins, Caucasus, levant, Anatolia, Tyrol Alps, Estonia, Bulgaria etc etc...............so where ever T you should also find L nearly
The TL formation first rose in the sind valley of South Asia and the split between T and L somewhere near by.

After being together with L and G2a group , the next marker it is with is J2 ( phoenician main marker )............be it 14000 years ago ( T1-Pages21 ) in the northern Levant or 9000 years ago in northern egypt it also trvelled with the phoenician J2.
Btw there is a lot of L marker in northern Levant.

[Maciamo]

This is a tricky time period to reduce to subclades, considering how few actual samples we have for the time period. I'd be interested to see if your predictions for SE Europe are accurate in particular, since we have to extrapolate there without much to extrapolate from. I guess I don't have a better proposal for the region at the moment.

I suppose we're mainly using the Motala samples to guess what Scandinavia would have been like. If so, I think it should include I2c, since that was found there (Motala2). It's a bit misleading to have I2c in Anatolia but nowhere in Europe when it's not clear yet which location is the origin point (and I vote Europe FWIW).

I2c could have had a wide distribution from Anatolia to Scandinavia via Central Europe. After all I2a and I2a1 had an evene wider distribution all over Europe. Anyway I have added to Scandinavia considering the evidence.

[Tomenable]

1) R1a wasn't part of the cattle-herding Yamna culture.

So far it seems so, however, R1a was part of the cattle-herding Khvalynsk culture.

Such a comparison of R1a vs. R1b samples from the Eurasian steppe known to date:

The first sample of R1a in the steppe appears in Khvalynsk culture, alongside R1b:

Abbreviations used:

EHG = Eastern Hunter-Gatherers
EBA = Early Bronze Age
LBA = Late Bronze Age

Steppe culture:	R1a samples:	R1b samples:	Dates of samples:	Approximate location:
Samara EHG (other EHG)	0 (2)	1	5650-5555 BC	Samara region
Khvalynsk	1	1	4700-3800 BC	Samara region, Khvalynsk II
Yamnaya	0	11	3340-2620 BC	Samara region, Buribay, Elista
Poltavka	1	4	2925-2200 BC	Samara region
Stalingrad EBA	0	1	2857-2497 BC	Stalingrad Quarry
Xiaohe Tomb complex	11	0	2558-1940 BC	Tarim Basin
Potapovka	2	0	2469-1900 BC	Samara region
Sintashta	2	0	2298-1896 BC	Orenburg, Chelyabinsk
Srubnaya	6	0	1850-1200 BC	Samara region
Andronovo	3	0	1800-1298 BC	Barnaul, Uzhur, Abakan
Mezhovskaya	1	1	1598-700 BC	Kapova Cave
Karasuk	2	0	1416-1261 BC	Altai Krai
Altai Scythians	4	0	1371-1011 BC	Mongolian Altai
Tanais Scythians	1	0	older than 1000 BC	Maeotia, Azov steppe
Afontova Gora LBA	1	0	926-815 BC	Krasnoyarsk region
Tagar	6	0	800 BC - 100 AD	Khakassia, Krasnoyarsk
Pazyryk	1	0	450 BC	Sebystei Valley
Sabinka II Iron Age	1	0	396-209 BC	Altai Krai
Volga Scythians	1	0	380-200 BC	Balakovo region
Tashtyk	1	0	100-400 AD	Khakassia
Caucasus Alans	1	0	400-600 AD	Krasnyy Kurgan region
Saltovo-Mayaki	1	0	800-900 AD	eastern part of Belgorod Oblast
TOTAL:	47 (2)	19	5650 BC - 900 AD	Eurasian steppes

We can also add 2 Bronze Age R1b samples from the Armenian Plateau to this list (1906-855 BC).

2) R1a was already all over European Russia in the Mesolithic (also R1b for that matter, but not yet R1b-V88 and R1b-L23).

That R1a in European Russia was also not yet M198, but some more archaic clades, such as M420*, M459* and YP1272.

Well, maybe M198 was also present among some groups of EHGs - but it has not been found yet, AFAIK.

3) The R1a phylogeny started to explode downstream of M417 from 3500 BCE (TMRCA for M417 according to Yfull), especially with the formation of the Z283 and Z93 subclades (formed c. 3000 BCE), which is far too recent for a Neolithic expansion.

According to Genetiker, many R1a samples in German Corded Ware were xZ645 - perhaps CTS4385, L664, M417*, M198*.

It seems that some clades of R1a which are rather rare today, could be more numerous in the past. Maybe even M198*.

I was surprised by such results. It now seems that R1a-L664 in Western Europe all came from Western Corded Ware.

[elghund]

Maciamo, thanks for all the hard work you put into making the latest in population genetics accessible to us laymen. The maps and summaries are enjoyable and informative.

[Tomenable]

I think that R1a from Xiaohe Tomb complex is going to surprise us!

I can't wait until they finally publish more precise data on subclades.

=========================

The 2nd oldest (after Khvalynsk) R1a sample from the steppe - Poltavka, dated 2925-2536 BC - is R1a1a1b2a Z94.

According to YFull Z94 formed 4800 ybp (ca. 2800 BC) and its TMRCA was also 4800 ybp (ca. 2800 BC):

http://www.yfull.com/tree/R1a/

[Maciamo]

So far it seems so, however, R1a was part of the cattle-herding Khvalynsk culture.

Good point, but the Pontic-Caspian Steppes were inhabited by R1a HG tribes when R1b cattle herders crossed over the Caucasus. It wasn't virgin land. Therefore it must have taken a few centuries for R1b herders to occupy the large territory that would become Yamna and drive out the indigenous R1a people.

At first it is to be expected that R1b and R1a lived side by side for a while. In fact, the way I conceived the Indo-European migrations when I started my migration maps in 2009 was that a minority (about 5-10%) of R1a people joined the R1b cattle herders before the Yamna culture, by the time (4200 BCE) the first horse riders started raiding the copper-rich towns of the Balkans. In other words, this was during the Khvalynsk culture (in the Volga) and Dnieper-Donets culture (in Ukraine). The reason that I found it necessary to included an R1a minority among R1b Steppe people is that the horse was domesticated c. 4600 BCE in the Middle Volga region, and that was probably too early for R1b to have been there. So there must have been a merger between cattle-herding R1b with their wagons and copper tools on the one hand, and the R1a horse-riders on the other. Riding horses were the perfect way to manage herds of cows on the open steppe. That's what I wrote in my R1b history.

The reason that I thought R1a people were just a minority is that the first wave of IE migrations to SE Europe, the one that eventually went up the Danube and invaded Central and Western Europe (Megalithic Bell Beakers) had to be predominantly R1b, as R1b is the dominant haplogroup in Western Europe. However R1b countries settled between 2500 and 2000 BCE (Austria, Czech Rep., Germany, Switzerland, Benelux, France, British Isles) all possess a minority of R1a-L664. That's why I wrote that L664 was the R1a branch that got integrated early into R1b tribes and accompanied the R1b migration to Central and Western Europe.

According to Genetiker, many R1a samples in German Corded Ware were xZ645 - perhaps CTS4385, L664, M417*, M198*.

It seems that some clades of R1a which are rather rare today, could be more numerous in the past. Maybe even M198*.

I was surprised by such results. It now seems that R1a-L664 in Western Europe all came from Western Corded Ware.

Another possibility was indeed that L664 wasn't integrated in the Steppe before R1b-L51 went west, but only got absorbed when R1b overran the Corded Ware around Germany c. 2500 BCE.

However since Z283 is so common in Germany and the Scandinavian branch, descended from Corded Ware, is undeniably Z283>Z284, I still stand by my theory that Corded Ware must have possessed Z283. It's also possible that other side lineages were present, including L664 and extinct branches.

Actually the fact that Khvalynsk had some R1a integrated with cattle-herding R1b society reinforces my suspicions that R1a-L644 was integrated into R1b tribes before Yamna. But L644 could have been present both in the Pontic-Caspian Steppe and in the northern forested Steppe that would give rise to the Corded Ware. It's the same region so it's only natural that the same lineage should be present in both areas. Additionally Corded Ware could only develop thanks to an influx of R1b-L23 people who brought cattles and copper technology. I always said that Bronze Age R1b tribes possessed a 10% minority of R1a-L644, and R1a tribes a 10-20% minority of R1b-L23. That's also why about 10% of R1b-L23 is found in all Slavic countries today (not including later Celtic R1b-S116 and Germanic R1b-S21).

[Danelaw]

The part about North Africa is complete speculation. Ancient Guance Berber remains had quite a lot of R1b and I. Those were probably the original NW African y-dna haplogroups. E-M81 came with proto-afro asiatics from eastern Africa, alongside mtdna haplotype M1. Some of those guys mixed with R1b-V88 dudes in Marocco and later migrated in Mali as Chadic speakers. One of the neolitich farmer in Spain was V88 but he could have got it from the Levantine farmers (???).

[Twilight]

Very interesting, thank you for posting. Although I got to wonder what the archeologists are going to call the Mesolithic Caucasian culture not to mention Anatolia. I guess time will only tell ^_^

The part about North Africa is complete speculation. Ancient Guance Berber remains had quite a lot of R1b and I. Those were probably the original NW African y-dna haplogroups. E-M81 came with proto-afro asiatics from eastern Africa, alongside mtdna haplotype M1. Some of those guys mixed with R1b-V88 dudes in Marocco and later migrated in Mali as Chadic speakers. One of the neolitich farmer in Spain was V88 but he could have got it from the Levantine farmers (???).

Do you have a source claiming that R1b spread out into Africa as early as Mesolithic times?

[Maciamo]

The part about North Africa is complete speculation. Ancient Guance Berber remains had quite a lot of R1b and I. Those were probably the original NW African y-dna haplogroups. E-M81 came with proto-afro asiatics from eastern Africa, alongside mtdna haplotype M1. Some of those guys mixed with R1b-V88 dudes in Marocco and later migrated in Mali as Chadic speakers. One of the neolitich farmer in Spain was V88 but he could have got it from the Levantine farmers (???).

You are one to talk about speculation by placing R1b in Mesolithic Northwest Africa ! It's possible that I was present (considering the H1, U5 and V maternal lineages), but certainly not R1b.

As Mesolithic Egyptians almost certainly possessed high percentages of E-M78, it is very likely that it was found all over North Africa. Besides, E-V13 was found in Early Neolithic Spain and as far as we know it wasn't brought by Neolithic farmers, as it hasn't shown up anywhere else. So E-M78 must have been present in Southern Europe and North Africa during the Mesolithic.

[Tomenable]

Good point, but the Pontic-Caspian Steppes were inhabited by R1a HG tribes when R1b cattle herders crossed over the Caucasus. It wasn't virgin land. Therefore it must have taken a few centuries for R1b herders to occupy the large territory that would become Yamna and drive out the indigenous R1a people.

Although this theory is plausible, one of problems is the fact that so far there is no proof that HG tribes in the Pontic-Caspian tribes were R1a. Quite the opposite, the only known HG sample from the steppe - Samara HG - was R1b. Two samples of R1a HGs were found far away from the steppe, but they were xM198, so they are about as irrelevant to the Indo-European question as Chadic-speaking R1b-V88 in Neolithic Iberia. The oldest R1a from the steppe appears in Khvalynsk culture and it autosomally distinct from Mesolithic HGs, it is also autosomally the same as R1b sample from Khvalynsk culture (I'm not sure about autosomal DNA of that 3rd sample which was Q1a and was buried without any grave goods, he was also killed by 4 strikes against his head - perhaps he was an intruder from some hostile tribe, and not part of the community, unlike the other two samples).

However since Z283 is so common in Germany and the Scandinavian branch, descended from Corded Ware, is undeniably Z283>Z284, I still stand by my theory that Corded Ware must have possessed Z283. It's also possible that other side lineages were present, including L664 and extinct branches.

Territory occupied by Corded Ware culture was huge, extending from the Rhine to the Volga. Z283 was probably more prevalent in Central and Eastern CW, while xZ645 was probably more numerous in Western CW. Later there took place "in situ" migrations within the former Corded Ware zone, such as Baltic migrations (according to Gimbutas up to the Oder, IIRC) and Slavic migrations (beyond the Elbe). A common Balto-Slavic subclade R1a1a1b1a2 Z280 appears in Germany for the first time in I0099 / HAL36 Urnfield sample from Halberstadt, dated to 1113-1021 BC.

That's also why about 10% of R1b-L23 is found in all Slavic countries today

I think this amount is closer to 5%, at least if we believe N. Myres et al. 2010:

R1b-M269 in Poland according to N. Myres 2010 (n=202):

L51 (= 11,91% of Y-DNA and 64,9% of R1b):

U106(xU198) ------------- 0,0594 (= 5,94%)
U152 ---------------------- 0,0347 (= 3,47%)
S116*(xM529xU152) ---- 0,0101 (= 1,01%)
M529(xM222) ------------ 0,0099 (= 0,99%)
L11*(xU106xS116) ------ 0,005 (= 0,5%)

xL51 (= 6,44% of Y-DNA and 35,1% of R1b):

L23(xM412) -------------- 0,0544 (= 5,44%)
M269(xL23) -------------- 0,005 (= 0,5%)
M412(xL11) -------------- 0,005 (= 0,5%)

M269 all ------------------ 0,1835 (= 18,35%)

And in Germany also according to Myres 2010 (n=321):

L51 (= 42,05% of Y-DNA and 95,8% of R1b):

U106(xU198) ------------- 0,19 (= 19%)
U198 ---------------------- 0,0187 (= 1,87%)
U152 ---------------------- 0,1028 (= 10,28%)
S116*(xM529xU152) ---- 0,0685 (= 6,85%)
M529(xM222) ------------ 0,0187 (= 1,87%)
M222 --------------------- 0,0031 (= 0,31%)
L11*(xU106xS116) ------ 0,0187 (= 1,87%)

xL51 (= 1,83% of Y-DNA and 4,2% of R1b):

L23(xM412) -------------- 0,0062 (= 0,62%)
M269(xL23) -------------- 0,009 (= 0,9%)
M412(xL11) -------------- 0,0031 (= 0,31%)

M269 all ------------------ 0,4388 (= 43,88%)

More data in Supplementary Table S4:

http://www.nature.com/ejhg/journal/v...2010146s1.html

Here a graph (made by user Hereward from The Apricity):

http://s21.postimg.org/bb07dljdz/Myres_Graph.png

[Danelaw]

You are one to talk about speculation by placing R1b in Mesolithic Northwest Africa ! It's possible that I was present (considering the H1, U5 and V maternal lineages), but certainly not R1b.As Mesolithic Egyptians almost certainly possessed high percentages of E-M78, it is very likely that it was found all over North Africa. Besides, E-V13 was found in Early Neolithic Spain and as far as we know it wasn't brought by Neolithic farmers, as it hasn't shown up anywhere else. So E-M78 must have been present in Southern Europe and North Africa during the Mesolithic.

Ok so V88 could have arrived with G2 neolitich farmers from the Levant. It makes sense. But I do believe that it arrived earlier than E-M81, which was brought by proto Afro Asiatic speakers from Eastern Africa about 5000 years ago. V88 probably joined these duded in their Southern migration to become proto-Chadic speakers in Mali and Niger.

[Tomenable]

As for Yamnaya culture - so far we have 11x R1b and 1x I2 from this culture, but all of this R1b appears to be "Eastern" ht35 (Z2103), right? In any case, we do not have any "Western" ht15 (L51) so far, just like we do not have any R1a so far. I think that one of problems is that we are getting only samples of chieftains buried in elite kurgans. Maybe all of them belonged to the same "ruling dynasty", descended from a common ancestor, and that's why all of them had Z2103. And later in cultures such as Srubnaya, we find only R1a, no longer any Z2103 - maybe the "ruling dynasty" changed?

Imagine describing British Y-DNA diversity and using only samples from Royalty, or perhaps Royalty + Nobility at best.

We probably need more samples from burials of "simple commoners", to get a more representative picture of Y-DNA diversity.

Another option is that L51 or pre-L51 emigrated westward from the steppe already before the emergence of Yamna culture?

PS: Yamnaya I2 sample was I2a2a1b1b2 - is this subclade common today, where is it most frequent ???

[Maciamo]

Just a quick note. I have replaced the Y-DNA F in the map by H2 as what I have always referred to as the Paleolithic European F-P96 has recently been renamed H2.

C1a2 and H2 are very old haplogroups associated with Cro-Magnon. C1 was found in the 37,000-year-old Kostenki 14 in Russia. C1a2 was found in Mesolithic Iberia and Early Neolithic Anatolia. H2 was present in Early Neolithic Anatolia and Hungary (Starčevo) as well as in Megalithic Spain. In my opinion these were all descended from Paleolithic Europeans. The fact that both haplogroups were found among the first farmers of north-west Anatolia shows that there was a genetic continuity between northwest Anatolia and Europe at the time.

[Maciamo]

As for Yamnaya culture - so far we have 11x R1b and 1x I2 from this culture, but all of this R1b appears to be "Eastern" ht35 (Z2103), right? In any case, we do not have any "Western" ht15 (L51) so far, just like we do not have any R1a so far. I think that one of problems is that we are getting only samples of chieftains buried in elite kurgans. Maybe all of them belonged to the same "ruling dynasty", descended from a common ancestor, and that's why all of them had Z2103. And later in cultures such as Srubnaya, we find only R1a, no longer any Z2103 - maybe the "ruling dynasty" changed?

It is a very important problem in historical population genetics, especially when looking at the Y-DNA of patriarchal and elitists societies like Proto-Indo-Europeans. Elite Kurgan burials may not be representative of the common folk from the culture in question. If, as you say, there was one ruling dynasty that expanded its territory over time but always placed royal princes as local rulers (like the Mongols did much later), then obviously we get a very skewed view of the Y-DNA in the overall society. That may simply be the reason why R1b-L51 hasn't shown up in Yamna yet. But it also means that there could have been plenty of Mesolithic (R1a, I2a) and Neolithic (G2a, T1a) lineages that were part of Yamna, but that are invisible to us now. The same would also apply to Corded Ware, Sintashta and any other Bronze Age Indo-European culture. If the ruling dynasty lasts long enough in one region, over time it will become the dominant male lineage in that region, even if it starts with in single individual. I think that would explain why R1b got replaced by R1a in Central Asia, and how the overwhelming majority of Indo-European Y-DNA that made it to the Indian subcontinent were R1a and not R1b, even though the European component of Indian genomes is about half Yamna R1b and half EHG R1a.

PS: Yamnaya I2 sample was I2a2a1b1b2 - is this subclade common today, where is it most frequent ???

I2a2a1b1b2 (S12195) is also known as Cont3b. It has a very wide distribution all over Europe, and even places like Georgia, but is especially common in Central Europe.

[Tomenable]

Maciamo I'm still not so convinced by your association of all of R1a with EHG. About 99% of R1a alive today is under M198 branch, which emerged ca. 14300 years ago (per YFull), while that EHG Karelian sample of R1a, who lived 7000-7500 years ago, was negative for this clade (xM198). Ancestors of EHG R1a split from ancestors of M198 R1a over 14 thousand years ago, and then over 7 thousand years ago we discover M459* in Karelia. M459 itself emerged 17700 years ago, and the most basal form of R1a - M420 - emerged 22000 years ago, and TMRCA of both M420 and M459 was 14300 years ago. Therefore M420*, M459* as well as YP1272 (formed 14300 years ago as well) had a lot of time to migrate all over Eurasia as hunters.

But TMRCA of M198 is only 8000 years ago, when also its "son" M417 formed. And we don't know where they came from. In any case M198 split from the rest of R1a over 14 thousand ybp, and both EHG samples (Karelia and that one from Chekunova) found so far are xM198. So it is possible that all EHGs were xM198 and M198 came from somewhere else. It is equally possible that some EHGs were M198. We can't be sure for now.

11 samples from Xiaohe Tomb complex in the Tarim Basin were positive for M198, and negative for Z93. They are mysterious:

https://openi.nlm.nih.gov/detailedre...7-8-15-1&req=4

(...) The Xiaohe cemetery (40°20'11" N, 88°40'20.3" E) is located in the Taklamakan Desert of northwest China, about 60 km south of the Peacock River and 175 km west of the ancient city of Kroraina (now Loulan; Figure 1). It was first explored in 1934 by Folke Bergman, a Swedish archaeologist, but the cemetery was lost sight of until the Xinjiang Archaeological Institute rediscovered it in 2000. The burial site comprises a total of 167 graves. Many enigmatic features of these graves, such as the pervasive use of sexual symbolism represented by tremendous numbers of huge phallus-posts and vulvae-posts, exaggerated wooden sculptures of human figures and masks, well-preserved boat coffins and mummies, a large number of textiles, ornaments and other artifacts, show that the civilization revealed at Xiaohe is different from any other archaeological site of the same period anywhere in the world [3]. (...)

Also the oldest cheese and the oldest glue found anywhere in the world so far, are from Xiaohe Tomb complex:

Cheese: http://www.livescience.com/43782-mum...st-cheese.html

Glue: http://www.ancient-origins.net/news-...0-years-020120

I have seen two conflicting versions when it comes to dating of Xiaohe samples: either 2558-2472 BC or 2020-1940 BC:

http://www.biomedcentral.com/1471-2156/16/78

http://sino-platonic.org/complete/spp185_silk_road.pdf

http://students.cis.uab.edu/ggabbert/site/xiaohe.html

http://people.ucas.ac.cn/upload/User...0433180905.pdf

Jean Manco used to have Xiaohe mummies dated to 2020-1940 BC, but have recently changed it to 2558-2472 BC:

http://www.ancestraljourneys.org/silkroaddna.shtml



I started a thread about this some time ago, but I assumed that they were M417, while in fact it is not certain:

http://www.eupedia.com/forum/threads...aohe-belong-to

They were only tested for M198 (here positive) and for Z93 (negative), but other types of M417 weren't tested.

[Maciamo]

Maciamo I'm still not so convinced by your bold association of all of R1a with EHG. About 99% of R1a alive today is under M198 branch, which emerged ca. 14300 years ago (per YFull), while that EHG Karelian sample of R1a, who lived 7000-7500 years ago, was negative for this clade (xM198). Ancestors of EHG R1a split from ancestors of M198 R1a over 14 thousand years ago, and then over 7 thousand years ago we discover M459* in Karelia. M459 itself emerged 17700 years ago, and the most basal form of R1a - M420 - emerged 22000 years ago, and TMRCA of both M420 and M459 was 14300 years ago. Therefore M420*, M459* as well as YP1272 (formed 14300 years ago as well) had a lot of time to migrate all over Eurasia as hunters.

Nevertheless, Indians have a few percents of Motala-like admixture. In some samples(Gujarati and Punjabi) it is as much as the Yamna-like admixture. And I have seen Bengali samples with no Yamna-like admixture, but a few percents of Motala-like and EEF-like admxitures. In contrast, Pathan, Sindhi, Burusho, Balochi and Tajiks all have considerably more Yamna-like than either Motala-like and EEF-like admxitures. It doesn't really make sense that Indians are so different. It's as if they were not descended from the same Indo-Iranian people.

Maybe it is because Pakistani and Afghans have more Proto-Iranian blood, a tribe that came from what is now Turkmenistan (relatively high R1b, so more Yamna-like), while Indo-Aryans descend from a more eastern Andronovo settlement around modern Kyrgyzstan and Tajikistan, where the levels of R1a are much higher. It's not a secret that R1b is much higher in Iran and northern Afghanistan than in India.

11 samples from Xiaohe Tomb complex in the Tarim Basin were positive for M198, and negative for Z93. They are mysterious:

https://openi.nlm.nih.gov/detailedre...7-8-15-1&req=4

Moreover, the oldest cheese and the oldest glue found anywhere in the world so far, come from Xiaohe Tomb complex:

Cheese: http://www.livescience.com/43782-mum...st-cheese.html

Glue: http://www.ancient-origins.net/news-...0-years-020120

I have seen two conflicting versions when it comes to dating of Xiaohe samples: either 2558-2472 BC or 2020-1940 BC:

http://www.biomedcentral.com/1471-2156/16/78

http://sino-platonic.org/complete/spp185_silk_road.pdf

http://students.cis.uab.edu/ggabbert/site/xiaohe.html

http://www.ancestraljourneys.org/silkroaddna.shtml

Thanks for the links. The dating is crucial here, as Andronovo starts from 2300 BCE. So if the Xiaohe tombs date from 2558-2472 BCE, the are older than Andronovo, which would be odd for an R1a tribe. That's barely the start of the Sintashta period. On the other hand, if they date from 2020-1940 BCE, then they could be an Andronovo offshoot. I don't have a problem with the fact that there were negative for Z93. As we discussed above, the different subclade could simply represent a different ruling family.

[Tomenable]

The dating is crucial here, as Andronovo starts from 2300 BCE. So if the Xiaohe tombs date from 2558-2472 BCE, the are older than Andronovo, which would be odd for an R1a tribe. That's barely the start of the Sintashta period. On the other hand, if they date from 2020-1940 BCE, then they could be an Andronovo offshoot.

Indeed!

Jean Manco on her website "Ancestral Journeys" used to have them dated to 2020-1940 BCE, but recently she has changed that information, and now she has them dated to 2558-2472 BCE. I vaguely recall that she has explained why she did that in some thread on Anthrogenica about Tarim mummies, but I can't find that post anywhere now. It seems that some new dating has been carried out, and the mummies have turned out to be older than previously thought (?). Anyway, here is the link to her website with Xiaohe mummies, and they are now dated to 2515 +/- 43 BC (few months ago, they were listed there as 1980 +/-40 BC):

http://www.ancestraljourneys.org/silkroaddna.shtml

I have asked her about this, but she has not responded so far. Maybe you will be more lucky?

I'm sure there was some good reason for that change.