October 28, 2011

Sardinian continuity against a backdrop of European discontinuity

Ancient DNA research from Europe has tended to paint a picture of substantial population discontinuity. In a Neolithic sample from Derenburg, neither of the two major Y-chromosome haplogroups prevalent in Central-Northern Europe, R1 and I were found. The Tyrolean Iceman belongs to haplogroup G2a4, a minor lineage in modern-day Europeans. In a Neolithic site from the French south, Treilles, haplogroups G2 and I2 were found, the major R1 lineage again being absent. Megalithic mtDNA from France complements that from the Linearbandkeramik in suggesting a picture of discontinuity, or, at least, substantial change, in the occupation of Europe since the Neolithic. An upcoming study suggest that mtDNA haplogroup X2 did not arrive in Central Europe with the early LBK Neolithic, but with the later Bell Beaker folk. The picture of discontinuity, at least for the Mesolithic-Neolithic transition is supported from Scandinavia, as well as Luxembourg. To cap it all off, a pair of unexpected results from Hungary and Ukraine stretch the contact zone of West/East Eurasian populations well to the west of what has already been determined by ancient DNA work in the Tarim basin and Siberia.


There is, however, one population that stands as an outlier against the backdrop of discontinuity: Sardinians. Ghirotto et al. inferred population continuity in Sardinian mtDNA at least until the Bronze Age. Stories about the Tyrolean Iceman, confirmed in the NOVA TV documentary suggest that a 5,000-year old denizen of Central Europe was genetically closest to Sardinians. A study of craniofacial variation confirms population continuity in Sardinia since the Neolithic, with the population of peninsular Italy gradually diverging, consistent with the Iceman findings. Substantial frequencies of Y-haplogroups G and I are found in modern Sardinian newborns, the same two haplogroups detected in Neolithic southern France at Treilles. And, a study of European genetic isolates finds Sardinians to be lacking in some European polymorphic sites, suggesting a degree of genetic isolation compared to mainland Europeans; this is consistent with my own finding that Sardinians are maximally "western" along the east-west Eurasian axis.

Two cryptic bits of revealed information suggest that a major event may have happened in Europe. In an ICHG 2011 talk, David Reich revealed that his lab has unpublished work of ancient admixture in Europe and that "Europeans are anciently mixed just like South Asians." A blog post by Ewen Callaway suggests that Zink's Tyrolean Iceman investigators are asking "whether he and his kin died out and were replaced by migrants from elsewhere, such as the Middle East."

In the absence of concrete data, it is difficult to interpret such hints of things to come. But the idea of an episode of admixture in Europe from the East that would leave a South Asian-like cline of diminishing West Asian ancestry is supported by admixture studies of West Eurasians, showing that the Near-East to Atlantic-Baltic vector captures the principal aspect of variation in this region.


Interestingly, a "West Asian" ancestral component centered on the South Caucasus region has an Fst distance of 0.028 with the main European "Atlantic-Baltic" component, and of 0.058 with a "Southern" component present at non-trivial amount in southern populations from both sides of the Mediterranean and the Near East.

The possibility of a major east-west population movement into Europe that left Sardinians least affected is intriguing. At K=7 Sardinians have almost none of the "West Asian" ancestral component and most of the "Southern" one in Europe. This appears to be consistent with a population that was least affected by population movements from the northern parts of West Asia.

If I had to guess, I would propose that most extant Europeans will be discovered to be a 2-way West Asian/Ancestral European mix, just as most South Asians are a simple West Asian/Ancestral South Indian mix. In both cases, the indigenous component is no longer in existence and the South Asian/Atlantic_Baltic components that emerge in ADMIXTURE analyses represent a composite of the aboriginal component with the introduced West Asian one. And, like in India, some populations will be discovered to be "off-cline" by admixture with different elements: in Europe these will be Paleo-Mediterraneans like the Iceman, an element maximally preserved in modern Sardinians, as well as the East Eurasian-influenced populations at the North-Eastern side of the continent.


In a roundabout manner, the Caucasus and its environs may soon reclaim their position as the fons gentium that Blumenbach, more than two centuries ago, ascribed to them. More than one century ago, Italian anthropologist Giuseppe Sergi, proposed a model of the double origin of Europeans, supposing that the Mediterraneans of southern Europe (a branch of the "Eurafricans") were invaded by peoples from the east, the "Eurasiatics." Naturally, the ideas of Blumenbach and Sergi cannot be adopted today in their entirety, but they do well to remind us that inklings of truth can be found in the most unexpected of places.


And, when one reads this quote from Carleton Coon's classic 1939 synthesis The Races of Europe (section "The western Mediterranean Islands")... "Sardinia and Corsica were peopled at the beginning of the Neolithic by a race of short-statured, dolichocephalic, low-vaulted, brunet Mediterraneans, coming probably from several quarters, including the adjacent European coasts, North Africa, and the eastern Mediterranean. Subsequent immigrations of other Mediterranean peoples have affected the racial composition of these islands but little." ... one is inclined to say "Dr. Coon, you were right!"

37 comments:

Charles Nydorf said...

Is it possible to characterize the Southern, West Asian and Atlantic-Baltic components via cranial morphology?

Maju said...

"If I had to guess, I would propose that most extant Europeans will be discovered to be a 2-way West Asian/Ancestral European mix, just as most South Asians are a simple West Asian/Ancestral South Indian mix. In both cases, the indigenous component is no longer in existence and the South Asian/Atlantic_Baltic components that emerge in ADMIXTURE analyses represent a composite of the aboriginal component with the introduced West Asian one".

I do not think that is possible. If in South Asia (a perfectly comparable case, as you say well) the ancestral component (ASI) is still clearly apparent, as well as the ASI-ANI duality, why don't we see anything like that in Europe?

The reason is that there has NOT been any significant West Asian immigration with Neolithic. And therefore the components we see (not always exactly the same but with tendency to repeat across studies) are actually pre-Neolithic stuff, with the West Asian component still apparent as such West Asian (and Red Sea) components.

These still make up (using your K=11 analysis) 44% in South Italy, 40% among Greeks, 36% in Central Italy, 30% among Tuscans, 24% among Romanians, 16% in North Italy, etc.

Alternatively, you seem to be able to discern that the North European component is closes to that of West Asia. Much closer than the Mediterranean or even the Red Sea component. This offers another interesting possibility: that North Europe was the part of Europe which was actually colonized by (blondish?) people from Turkey and such.

I'd rather think this only indicates maybe a second Paleolithic layer but that is what your own analysis has to offer - and not what you are saying.

Sardinians would then be 100% of another component, which may or not have arrived from the East (it's rare in West Asia, excepting Cyprus) and that we could describe as Greco-Italian for the two regions that have it above 30% (although you use the term "Mediterranean" instead).

But I do not dare to say more because I really do not like too much your analysis strategy, discarding important European components and retaining once and again exotic pointless components from Africa and East Asia.

pconroy said...

Yes, I agree it would certainly seem so!

That means that my prediction over 3 years ago about the possible source of R1b in Western Europe may be right on the money:

http://dienekes.blogspot.com/2008/07/haplogroup-sizes-and-observation.html


pconroy said...
I wonder if R1b in Western Europe is associated with the Cardium Pottery peoples from the Middle East?
Wednesday, July 30, 2008 9:00:00 PM



pconroy said...
Well if not Cardium Pottery people, whose origin was in Lebanon/Syria, then how about the Megalithic people.

It would seem that they might have originated in Southern Armenia, in the Syunik area, as there is a huge ancient megalithic structure there, called Karahunj (aka Zorats Karer, Angelakot) - http://www.nationmaster.com/encyclopedia/Karahunj - which may date to 8,000 yo, it seems to have also served as an observatory - much like Newgrange in Ireland or Stonehenge in England.

It would also seem that the population of this Southern Armenia area is 40-45% R1b:
http://www.ucl.ac.uk/tcga/tcgapdf/Weale-HG-01-Armenia.pdf
Wednesday, July 30, 2008 10:16:00 PM


pconroy said...
ebizur,

Thanks for the data, especially:
56/140 = 0.4000 Syunik
92/215 = 0.4279 Karabakh

So it seems that Southern Armenia was heavily R1b and had some of the earliest megalithic structures.

Maju,

I disagree that the megalithic structures were religious in nature, they were instead like great stone locks, which were used by agricultural people to measure the passing of the seasons, and determine when planting and harvesting should take place.

Can you point out any "stone circles" that are not associated with some plain, where agriculture took place?
Friday, August 01, 2008 11:36:00 PM

Fanty said...

"Is it possible to characterize the Southern, West Asian and Atlantic-Baltic components via cranial morphology?"

Not really.

For example, Scandinavia centered components spread out like Nordid (long skulled, long faced), Faelid (Long skulled, short faced) and Borreby (short skulled, long faced) skull shapes.

There is virtually no component with a spreading pattern comparable to "Alpinoid" (short skulled, short faced)

I am sure, those admixture components are not limited to alleles of a single skull shape.

But 23andMe has a survey running that involves facial (and Nasal) shapes. Possibly some alleles show up that might connect people with certain shapes of face or noses. But its limited to what rough shape the face shows from the front. No lengh of the skull etc.

But its actually possible to ask 23andMe to do surveys of your likening.

About admixture components linked to skull shapes...

the only thong I could imagine is, doing supervised runs with groups of individuals that are of totaly different heritage, but identical skull shape.

the software would be forced to produce components that contain allele patterns that these people share.

But maybe this doesnt work at all and my thinking is off.

*shrugs*

mooreisbetter said...

On the dead scientists: The theories of those scientists you cited, from days long past, have a remarkable way of staying relevant. The leaps that come from sheer intuition can often dwarf the incremental progress made by scrupulous testing.

I wouldn't quite say that Coon was right though. I do not see much North African component in Sardinians... Lots of Western European and Ancient European.

On the Sardinians: You may want to check out Cavalli-Sforza's two seminal books, in which he covers the Principal Components of European Genetics. In one book, he writes that were he to remain true to the data, he'd have give Sardinians their own pole. (His poles were Anatolia, Finland, Greece/Ireland (axis), Ukraine, and Basque Country).

He went on to say that Sardinians were so unique, they should be characterized as their own type of European really. He did not give them their own pole because it would complicate the data too much (as they were waaay off the charts in terms of uniqueness).

On other ancient Europeans: I believe there is a significant similarlity between Southern Italians from the mountains of Basilicata and North Calabria, with Sardinians. There is a lot of G2 and I2 in those remote areas, which, if you think about it, make sense as a reliquary. Add a Greek and Roman layer by the coasts, and you have the modern Lucano/Cosentino.

Dienekes said...

I wouldn't quite say that Coon was right though. I do not see much North African component in Sardinians... Lots of Western European and Ancient European.

It is true that Sardinians do not have much _recent_ North African ancestry, i.e., the component centered on Mozabites and labeled "Northwest African" in several iterations of my analyses.

But, at the appropriate level of resolution, Sardinians and Mozabites actually are the two populations with the maximum of a component that is also frequent in North Africans, and which also occurs in Basques and along the Atlantic facade and dissipates toward the eastern Mediterranean.

Here are the top populations for this component at K=8, using the same dataset as 'eurasia7', slightly modified for pruning relatives.

[150,] "Spaniards" "38.1"
[151,] "Portuguese_D" "38.3"
[152,] "Spanish_D" "38.8"
[153,] "IBS" "39.4"
[154,] "North_African_Jews_D" "39.5"
[155,] "Sephardic_Jews" "39.7"
[156,] "Basque_D" "41.2"
[157,] "Morocco_Jews" "41.5"
[158,] "French_Basque" "42.4"
[159,] "Tunisian_D" "43.9"
[160,] "Algerian_D" "46.9"
[161,] "Moroccans" "48.3"
[162,] "Moroccan_D" "52.7"
[163,] "Mozabite" "57.7"
[164,] "Sardinian" "63"

Dienekes said...

At K=9, this persists, a little altered

[150,] "TSI" "32.9"
[151,] "Morocco_Jews" "34"
[152,] "North_Italian" "35"
[153,] "Portuguese_D" "37.6"
[154,] "Spaniards" "37.7"
[155,] "Spanish_D" "38.2"
[156,] "Tunisian_D" "38.7"
[157,] "IBS" "39"
[158,] "Algerian_D" "43.2"
[159,] "Basque_D" "43.2"
[160,] "Moroccans" "43.8"
[161,] "French_Basque" "44.2"
[162,] "Moroccan_D" "49.2"
[163,] "Sardinian" "59.7"
[164,] "Mozabite" "62.8"

At K=10, a specific Mozabite-centered component emerges:

[150,] "Palestinian_D" "4.6"
[151,] "Sephardic_Jews" "5"
[152,] "Ethiopian_Jews" "5.3"
[153,] "East_African_D" "6"
[154,] "Ethiopians" "6.1"
[155,] "Portuguese_D" "6.9"
[156,] "North_African_Jews_D" "7"
[157,] "Egyptian_D" "7.5"
[158,] "Egyptans" "8.1"
[159,] "Morocco_Jews" "8.3"
[160,] "Tunisian_D" "26.8"
[161,] "Algerian_D" "33.4"
[162,] "Moroccans" "39.1"
[163,] "Moroccan_D" "40.9"
[164,] "Mozabite" "83.2"

as well as a Sardinian-Basque one:

[150,] "O_Italian_D" "39.5"
[151,] "C_Italian_D" "40.4"
[152,] "French_D" "41.4"
[153,] "French" "41.5"
[154,] "N_Italian_D" "43.1"
[155,] "Tuscan" "43.5"
[156,] "TSI" "43.7"
[157,] "Portuguese_D" "46.2"
[158,] "North_Italian" "47.2"
[159,] "Spanish_D" "51.1"
[160,] "Spaniards" "51.2"
[161,] "IBS" "51.8"
[162,] "Basque_D" "68"
[163,] "French_Basque" "69.3"
[164,] "Sardinian" "77"

What seems to have happened is that there is a common population stratum in the western Mediterranean, on both continents. In the north side, this was influenced by pre-Neolithic Europeans (whom I called "Ancestral Europeans") who form part of the so-called Atlantic_Baltic component. In the south side, it was influenced by pre-Neolithic Africans as well as Sub-Saharans. A few thousand years of different admixtures and separate evolution has been enough to create a separate blend north and south of the Mediterranean at its western end.

eurologist said...

"Interestingly, a "West Asian" ancestral component centered on the South Caucasus region has an Fst distance of 0.028 with the main European "Atlantic-Baltic" component, and of 0.058 with a "Southern" component present at non-trivial amount in southern populations from both sides of the Mediterranean and the Near East."

As I posted before, one has to be very, very careful about this. Judging European relations based on the extremes of Sub-Saharan Africans and South-East Asians, East Asians, Siberians, and in the end Amerindians is not fruitful at all.

It looks like the first 3 or 4 or 5 dimensions are defined with these extremes, which means that Fst differences are projected along these extremes, and really have nothing to do with European differences.

We should evaluate such differences based on European bases and components, alone - including Caucasia and perhaps "Red Sea".

Anonymous said...

G in Italy is not evenly distributed, it peaks in Alto Adige (South Tyrol)8% and other mountainous areas, it seems a relic of the past.

Dianekes, instead of using the Caterina Murino's pic , try this.
http://img201.imageshack.us/img201/3898/img1544au9.jpg

"Fuit enim gens antiquissima Italiae – Primo Italiam habuisse quondam qui Aborigenes appellabantur – Italiae cultores primi Aborigenes fuere"

There's a truth in the ancient oral tradition. It has been reworked and rewritten to policy needs in historical times, but it contains a nut of truth.

Onur Dincer said...

Without knowing what David Reich meant by "Europeans are anciently mixed just like South Asians" I am afraid we can not go beyond speculation on that issue. We do not even know what he meant by European. Maybe he meant all Caucasoids irrespective of continent and region, maybe some of them, just we do not know.

Also we do not know the autosomal genetic analysis results of Ötzi and those of the ancient skeletons from central Europe investigated with him.

In conclusion, we do not what we normally ought to know now just because of the commercial interests of a few, and unfortunately few scientists have the courage and apparently even also the desire to demolish or at least bypass the academic establishment.

pconroy said...

There's also this old map by Biasutti:
http://img26.imageshack.us/img26/4035/biasuttimap01.png

Which labels Southern Sardinia as, "Forme Europoidi arcaiche", which I think means, "Archaic European Form"

Dienekes said...

Without knowing what David Reich meant by "Europeans are anciently mixed just like South Asians" I am afraid we can not go beyond speculation on that issue.

That is why I am attempting to guess what he meant and correlate it with other pieces of information. We'll see what he meant when his research is published. I see a good reason to highlight what was just a passing comment, because Reich and collaborators have an unusually high level of statistical sophistication, and if he decided to say _anything_ about this in public, it's probably something more or less secure.

In conclusion, we do not what we normally ought to know now just because of the commercial interests of a few, and unfortunately few scientists have the courage and apparently even also the desire to demolish or at least bypass the academic establishment.

You are too harsh on scientists. They have to work in a particular system that requires certain tropes to be followed. I don't expect any scientist to turn kamikaze and sacrifice his career and publication record to make a point. Project leaders also have to worry about their postdocs and grad students, and only a fool would say to them all: "You won't get a Nature paper because I decided to make a point about open science".

Dienekes said...

Biasutti had identified, within the Mediterranean race, Paleosardinian, Berber, and Atlantic subraces. It may be a coincidence, but the Sardinian-Mozabite component mentions above fractures first into Sardinian-Basque/Mozabite and then into Sardinian/Basque/Mozabite components. Given the material they had to work with, I'd say that the old-time physical anthropologists had much that was interesting to say.

Onur Dincer said...

You are too harsh on scientists. They have to work in a particular system that requires certain tropes to be followed. I don't expect any scientist to turn kamikaze and sacrifice his career and publication record to make a point. Project leaders also have to worry about their postdocs and grad students, and only a fool would say to them all: "You won't get a Nature paper because I decided to make a point about open science".

Is not open science what all of us wish? They can at least make an effort for that. I know that some scientists are really doing that, but they are apparently in the minority.

dok101 said...

@ Dienekes

"[T]he old-time physical anthropologists had much that was interesting to say."

Took a look at Henry Field's "Contributions to the Anthropology of Iran." Specifically, the listing of cephalic indices for mostly West Asian groups, but also a few others. The most brachycephalic populations had the greatest frequencies of R-M269, T, and J1* (and possibly other HGs), among the populations listed. Though, it should be noted, Roewer et al. found significant frequencies of R-M269 in some of the northern Iranian minority groups (e.g. Talysh), not listed below. At least not separately.

1 Bakhtiari 88.38 T=13%, Roewer et al.
2 Assyrian avg 88.07 T/hg26~14%, hg1(perhaps 30-35% R-M269)=41.5%, J1*=10+% (Yepiskoposian et al., Mendez et al., Chiaroni et al.)
3 Lezgin 87.77 R-M269=29.6%, J1*=44.4% (Balanovsky et al.)
4 Druze 87.26 T=8%, R-M269 frequencies vary. Most recent=49%. Rough average of past studies, perhaps in the neighborhood of ~20% R-M269? (Mendez et al., Al-Zahery et al.)
5 Armenian avg 86.34 R-M269=26%, T=6%, J1* (mostly xP58) = 12%. Armenian FTDNA Project.
6 Laz 85.6
7 Alawi avg 85.25 R-M269 estimated, based on haplotypes in Dönbak et al, to be at least 30%-35%.
8 Syrian 85.11
9 Lebanese 84.88
9 Turkish avg 84.88
11 Georgian 83.2
12 Ossetes avg 82.17
13 Circassian 82.05
14 Iranian 80.06
15 Kurd avg 79.84
16 Persian avg 78.55
17 Azerbaijani 76
18 Lurs 74.25

mooreisbetter said...

One more thought:

Doesn't France hold the key to understanding a lot of this? I mean, I think sometimes even the most balanced and perhaps, um, Grecocentric scientists (I'm ribbing you, friendly, Dienekes) are nevertheless a bit Western-European centric.

Let us not forget that Europe can rather simply be cut into three quadrants: West, Central, and East. In central Europe, I and G are not really "minor" lineages. There is still significant I and/or G in Fennoscandia, Germany, Austria, and Italy.

It is Ireland, Spain, England, etc. where these are not so prevalent.

France has very bizarre laws as to DNA testing, and that means we don't really have good studies there. I would venture that Southern France bears a significant G and I presence to this day. That Bretagne looks like Western Europe, etc. I would even venture a strong, detectable Greek component near Marseilles (Massilia).

If you assume Central Europe had a healthy population of I and G (non R1 peoples) -- the traditional model, of R1 being present on the West-East poles of Europe, then expanding -- remains plausible to this day.

This debate is far from over.

bmdriver said...

Could it be that Europeans are closer to north indians? rather than north indians closer to europeans? If the aryan theory is false which it is, then it proves a migration out of india, and not into india, as abrahamic people think.

It seems that Abrahamic people, of whatever faith, have one unity that is to protect the origins of their ancestoral land, middle east, for jews, christians and muslims who believe that the sons of noah civilised middle east first and the sons of ham, the darkies, where civilised by them. It seems that you are very quick to establish that North Indians are closer to europeans but the truth is that Europeans are closer to north indians. That r1a is Indian origin, older in india, and that north indian population in general are alot older than in europe or central asia.

dok101 said...

Two interesting bits, in light of the details provided in my previous post, from, "The Urartian Substratum in Armenian." John A.C. Greppin, 2008.

"[I]t was quite likely...that both the Hurrian and Urartian languages were directly related to the languages of the North Caucasus."

And, now, the more intriguing of the two:

"[I]t also became clear that Armenian seems to have its best fit with the Lezgian group of the Daghestani languages."

On a separate, but possibly related note, Aramaic dialects, and specifically Northeastern Neo-Aramaic (NENA) dialects, are the only instances of Semitic, that, based on the most recent scholarship, appear to display some form/degree of ergativity. "The Debate on Ergativity in Neo-Aramaic." Edit Doron and Geoffrey Khan, 2010.

Ergativity is also found in (source:Wiki) the following languages. Not intended to be exhaustive:
Kartvelian
Hurrian
Urartian
Nakh–Daghestanian
Sumerian
Northeast and Northwest Caucasian
Several Southwestern Iranian languages, including Davani

Assyrian-Aramaic ("Sûret") is one of the NENA dialects.

Maju said...

@Mooreisbetter: What you say might make some sense about I, whose center may well be in Ukraine... but G is actually centered in Southernmost Europe, from Crete to Portugal

Brief list of >7% G samples in non-Caucasus Europe:
·Tirol: 7% (but not elsewhere in Austria)
·Korcula (Croatia): 10%
·Osijek (Croatia): 13%
·Corsica: 12%
·Central Crete: 8%
·Mainland Italy: 3-15% (mostly in the high range, depending on sample)
·Some locations in Sicily, but the island overall is lower than mainland Italy (c. 5%)
·Sardinia: 12-14%
·Malta: 8%
·Gagauzes (Turks) of Moldova: 13%
·Portugal: 7-12%
·Orlovaskaja Oblast (Belarus): 9% (but rare elsewhere in non-Caucasus Eastern Europe)
·Spain: 5-8% (not among Basques however)

As you can easily see, almost without exception these populations with notable Y-DNA G are from the Southernmost zone of Mediterranean climate. The main area of partial exceptions is near NE Italy (Austrian Tirol, Osijek). And then in very small pockets of Eastern Europe.

Actually within the Mediterranean there is a Western trend with this lineage, which is much more concentrated in Iberia and Italy than anywhere else of comparable size. Even Greece falls clearly behind.

Excepting the Caucausus only Turkey and Palestine compare in frequency of Y-DNA G. However diversity is clearly higher in and around the Caucasus (at all phylogenetic levels??? Is this another R1b-like "diversity" fiasco? Worth considering).

Unknown said...

I think we have to be careful here. The genetic similarities between Sardinia and other mediterannean countries are entirely consistent with the southern coastal migration which has been proposed for as far back as I can recall. Nothing new here, its just that the islandic Sardinians were more able to defend their turf from later population flows. They have a reputation for fierceness to this day.

Also as I recall Sardinia has known more recent influxes from Central Europe, from regions rich in G. I cant find the reference but I recall the Sardinian G was associated with (relatively) recent coastal settlements rather than the inland area (where you would expect anything truly ancient).

As I see it we have a southern coastal flow along the mediterranean mostly from the east, that started in the paleolithic (Aurignacian) and continued probably to a lesser extent until Roman times. I expect these were the first folk in Europe.

We have the flow along the central European rivers which would intially have been from the east but was probably two way after the initial flow. With a bias towards the direction of flow of the respective river. Aurignacian to middle ages.

We have the two way flow along the treeline stretching from East Asia to Western Europe and possibly on to pre-Clovis America. Gravettian mostly and a nice source of asian genes for Europe.

Then there is the Kalash/Chuvash component that may South Asian that travelled north, or North Asian travelling south (who knows).

Plus of course the South Asian from the Roma all the place.

And whoever the Finns were percolating across from the East.

And other layers upon layers.

IMO Sardinia evolved mostly from the paleolithic southern Europe. If we have to choose a modern population as a pole it should be Sardinia.

The Basques did also but had more of paleolithic northern Europe in the original mix.

I think G in France/Otzi is from a later coastal wave that only made it to France/Italy in significant numbers. But we shall see. Y haplogroups have proved to be extremely unstable.

eurologist said...

Could it be that Europeans are closer to north indians? rather than north indians closer to europeans?

I think hardly anyone doubts that Europeans ultimately mostly derive from Pakistan/India - the question is, when. Is it mostly 40,000 years ago, or that plus some West Asian in the neolithic, and additional migrations since the bronze age? And conversely, did West Asians and/or Eastern Europeans migrate back into northern India in substantial numbers since the neolithic/bronze age?

Given the many possible layers, it is not that easy to disentangle. To complicate matters even more, until after LGM the climate was extremely dry in much of the subcontinent, which would have naturally led to a split into two distinct northern and southern population.

Maju said...

@Bmdriver: I like the your questioning (on second read):

"Europeans are closer to north indians? rather than north indians closer to europeans?"

That has nothing to do with the Aryan invasion model, which is likely to stand (it has more to do with linguistics than genetics) but may have caused a much smaller genetic impact than though some years ago.

Those that propose a West Asian (not European) origin for the ANI component suggest a Neolithic time frame now, being in that case more likely related to Dravidian (Elamo-Dravidian?) languages than Indo-European, which had not yet began expanding (from the Urals probably).

But your idea of ANI being in fact the source, some 40-50,000 years ago of West Eurasian gene pool could make sense in principle. It would still demand a much older division of the ANI/ASI duality, implying that there would have been two clearly distinct populations in South Asia for maybe 80,000 years, almost never admixing before Neolithic.

This seems hard to argue for, really. But the possibility is there... if only archaeology could support this idea!

Otherwise, I fear we will have to settle for ANI being a distinct Neolithic component from West Asia, what does have some archaeological support indeed.

And this is what raises my eyebrows when it's claimed the same pattern for Europe but with no ANI-like single component that would justify this idea.

eurologist said...

diversity is clearly higher in and around the Caucasus (at all phylogenetic levels??? Is this another R1b-like "diversity" fiasco? Worth considering)

Maju,

Last I looked into this I could not find any substantiation that diversity is higher in the Caucasus. In fact, all subgroups present in the Caucasus are rather limited. As for many haplogroups and subgroups, the Caucasus each and every time looks like an ideal refugium - people came and claimed and fiercely defended a valley for millennia. Europe on the other hand seems to have some very old G2a* - pre-LGM (widely distributed there and in Asia) - and more recent G2a3b1*, G2a3b1a1* and G2a3b*, among others, likely from local Balkan sources at the beginning of LBK.

Pascvaks said...

As the Ice melts, follow the grass, follow the game, follow the hunter, follow the blood. But I have a strong belief that there were those in issolated areas who were well positioned to survive during the LGM in such areas as Sardinia and Sicily, and their blood would be of an older mix than others to the East.

Pascvaks said...

FYI - Not sure how accurate this is. Came across an article by a Robert Henvell, "Aspects of Sardinian Genetics", Aug 2011, when searching for more info. He appears to cover a pretty fair swath of history and prehistory in his articles. Link -

http://www.articlesbase.com/science-articles/aspects-of-sardinian-genetics-5111648.html

Andrew Oh-Willeke said...

Given the substantial discontinuity between Mesolithic and Neolithic in Europe, you need at least three layers - Mesolithic European, early Neolithic, and late Neolithic, although the "Ancestral European" admixture component could constitute a fixedly admixed combination of the first two.

I suspect that a four major sequential component mix is more accurate: (1)Mesolithic (dominated by a few hgs of mtDNA U), (2) Epipaleolithic Southern Europe/Early Neolithic (Sardinian/Otzi), (3) later Eastern/Central European Dairy Farmer Later Neolithic (this was possibly also a source for the Basque population), and (4) Indo-European starting with Bell Beaker and continuing in multiple sub-waves through the Slavic migration.

I also think that there are something on the order of half a dozen minor waves sprinkled in with the major waves that are distinguishable: (1) Gypsy, (2) Jewish Diaspora, (3) Classical Era Pax Romana North African, (4) Moorish Iberian/Moorish Southern Balkan, (5) Northern migration from North Africa in the Epipaleolithic, (6) Uralic, and (7) Turkic in parts of Turkey in the Caucusas.

bmdriver said...

Pre-ANI-ASI split stage, the Ancestral Indians, located diffusely from Gujarat/ Sind to the South Indian tip. Yet one of its limb extended along Narmada, to the Sone river. Then along Sone to Ganga, then along Ganga to Brahma-Putra. Then Along BrahmaPutra river into Tibet. That is why we get some of the oldest genes like Y-hg D in Tibet.The other limb proceeded coastally to Kerala, then reached Andhra coast, then Orissa, then Bengal and then moved into the SEA. The Indian tribes (Austro-Asiatic Speakers) originated at North Andhra-Orissa-Chattisgarh-Jharkhand region. The northwest Indians evolved the Indo-European languages, and the South Indians developed the Dravidian languages.

ANI is thousands of years older than (contemporary) Europeans and Central Asians, and ANI is not any existing gene pool, but a re-constructed hypothetical gene pool, just like the Proto-Indo-European language. Now if we assign the term “Ancestral Indians” or AI to the first Indian settlers they were the first Eurasian people outside Africa.

Soon they expanded to fill the whole of India. Because of geo-climatic barriers within India, this population developed regional genetic diversions/mutations/ features, and these genetic changes can now be identified by means of DNA technology, because we can determine the age of the DNA mutations. For example mitochondrial DNA haplogroup M2 should be considered to have originated within ASI and M3 within ANI from Ancestral Indian M*.However, we get a mixture of both ANI and ASI genes or DNAs in modern Indian population.

The European, West Asian and Central Asian DNAs match with DNAs identifiable with ANI. This is because these populations emerged out from the population inhabiting North India before ANI and ASI admixture took place in India. These migrations took place between 55,000 years and 30,000 years back. With a more recent migration out of India into central asia, and europe, with the onset of agriculture.

Sahoo et al had actually written the following words:“The perennial concept of people, language, and agriculture arriving to India together through thenorthwest corridor does not hold up to close scrutiny.Recent claims for a linkage of haplogroups J2, L, R1a,and R2 with a contemporaneous origin for the majority of the Indian castes’ paternal lineages from outside the subcontinent are rejected, although our findings do support a local origin of haplogroups F* and H.” .They also rule out arrivals from Southwest Asia because West Asian haplogroups (like Y-Hg G) are not found in India.

Kivisild’s findings (2003) too had shown that humans could not have arrived from West Asia into Indiabecause of lack of West Asian Y-hgs E, G, I, J* and J2f. Kivisild
et al wrote,“When compared with European and Middle Eastern populations (Semino et al. 2000), Indians (i) share with themclades J2 and M173 derived sister groups R1b and R1a, the latter of which is particularly frequent in India; and (ii) lack or show amarginal frequency of clades E, G, I, J*, and J2f.”

Even terminology, such as Indo-European, would be better classified as North Indian language group. Because the western world had Military domination over Asia, it has written history in his image. So in other words, If India had conquered Europe, and Central Asia in modern times, today it would be known as a North Indian language group instead of a Indo-European classification.

Lets JUST accept for one moment that European culture is derived from North Indian population. That means a direct contradiction to the religous ideology of Jews, Christians and Muslims.

That changes everything,in history, in politics, in culture and most importantly it challenges the VERY basis of Abrahamic faiths, as being nothing more than a very old branch of indic culture.A connection diluted on purpose to give more support to a local middle eastern/ biblical home of jews, christians and muslims.

Thanks.

Unknown said...

"Given the substantial discontinuity between Mesolithic and Neolithic in Europe"

We cannot overlook the probable impact of air and cremation burials on Europe. These are likely (given the climate and legends) to be the dominant local practice. And they leave no DNA.

Maju said...

@Eurologist:

Caucasus and surroundings (I said and I meant): Turkey, Iran... Highland West Asia may be a better name.

The rarer high tier G subclades (G1, G2*/G2b, G2c) are all located in that area. And that is higher basal diversity at the G and G2 levels.

But how is it for the G2a level and downstream? After all I expect G to be some 40-50 Ka old (being an F basal derivative in West Asia), so G2a could well be half that age (for example).

However the largest subhaplogroup under G in Europe is G2a3a (all the rest seem to have West Asian/Caucasus focus again), and this one, still shows up in West Asia and the Caucasus. As do all relevant clusters I can find under this node.

So in the end we are surely before a true Neolithic marker as far as I can discern because we can hardly take apart the Palestinian, Turkish, Caucasian or European lines within.

My question was if it could be a problem of low resolution, but I can find nothing justifying this, unlike within R1b. But in this late case it took many years, a much greater gene pool to study and also huge popular interest. So there is still a possibility that the G2a case is also misunderstood - that's what I meant.

But on second thought, after revisiting some of the known data, I'd say that the possibility is slim (tell me if I am wrong and why).

Maju said...

@Bmdriver: the Indoeuropean language family is extremely solid (if that fails all linguistics is nothing but junk because it's the best studied and most debated family ever) and within it Indo-Iranian is clearly a single branch (trees vary but this is consistent across all tress, as is Balto-Slavic and as is the larger Western branch including Balto-Slavic, Germanic, Italic and Celtic).

However, this, I insist, has nothing to do with ANI in principle. Languages can spread regardless of genes (we are both speaking English and none of us have English ancestors we can discern, for example - at least not I).

I do not think we can estimate the age of ANI but we can estimate pretty well the genetic distance of ANI and ASI among them and with third components such as the West Asian one.

However I was under the induced belief that ANI was closer to West Asia/Europe component(s) but, reviewing what has been written on the matter I can find not a single demonstration of that claim. The closest I found was this old GNXP post (and graph in it) but seems quite inconclusive (and in fact they seem to claim ANI as WEA-SA admixed and not a genuine distinct component).

So you may well be onto something when pointing out that ANI is a genuine South Asian specific component, just that distinct from the dominant one in the Southern/SE part of the subcontinent (ASI). And you may (at least tentatively) be right when suggesting that ANI is ancestral to West Eurasian genetics, at least in a sense (ASI would not).

That makes good sense and is consistent with haploid genetics but I do not know how to research/demonstrate it.

Andrew Oh-Willeke said...

"Lets JUST accept for one moment that European culture is derived from North Indian population."

It is more plausible to reason that European genetics are derived from India, but that the culture we think of as Indo-European and Afro-Asiatic (and non-Indo-European) at much more shallow time depth are distinct.

Doug1 said...

I'm curious as to whether you're a Greek American, or blogging from Greece - or another place.

I suppose this is just idle curiosity.

You may have said, but I've only recently starting reading more than an occasional blog post of yours usually linked from a Razib blog post.

Anonymous said...

Dienke, an interesting paper.
http://www.pnas.org/content/early/2011/10/24/1113061108.abstract?sid=77d0f58a-4121-4ea3-81cc-aa4231a6b6d1
gd tms.

formerjerseyboy said...

"Lets JUST accept for one moment that European culture is derived from North Indian population. That means a direct contradiction to the religous ideology of Jews, Christians and Muslims."

May I ask what does this have to do at all with population genetics?

The rant about "Abrahamic faiths" glosses over the fact that adherents of Islam/Christinity/Judaism have spent more time and energy fighting amongst themselves, and have never exhibited the degree of solidarity that would be necessary to agree on and pursue any political agenda. In the absence of a more persuasive argument, why would it make sense to replace a so-called "Abrahamic bias" with an India-centric one?

eurologist said...

My question was if it could be a problem of low resolution, but I can find nothing justifying this, unlike within R1b. But in this late case it took many years, a much greater gene pool to study and also huge popular interest. So there is still a possibility that the G2a case is also misunderstood - that's what I meant.

Maju, I believe there is some very dilute upper-level G present at extremely wide distribution, and that attests to its ancestry. Conversely, I agree most G2 in Europe is neolithic, although some of it may very well originate from the Balkans (i.e., was present there from pre-neolithic contact with the general Black Sea/ Anatolia region - this conforms to my model that LBK picked up a lot of local middle-Danubian haplogoups which then ballooned, but later got overwhelmed by the haplogroups at the fringes). In that sense, a finer sampling of G could be very useful to understand some details of Cardium and LBK. (Sorry, I did not get your "surroundings" qualifier.)

Your message to Bmdriver:
So you may well be onto something when pointing out that ANI is a genuine South Asian specific component, just that distinct from the dominant one in the Southern/SE part of the subcontinent (ASI). And you may (at least tentatively) be right when suggesting that ANI is ancestral to West Eurasian genetics, at least in a sense (ASI would not).

That pretty much agrees with my model of the subcontinent, based on the expected climate-induced split into two distinct and largely separated niches.

Maju said...

But it's a good criticism: Coon specially but also other Nordic anthropometrists were blinded by their own ethnic perceptions, being able to discern subtle differences between various kinds of "Nordoids" and then dumping all Mediterraneans in one huge blank category (same in other regions of the World).

Some people claim that "all Chinese look the same". This is obviously not the case but for a non-Chinese (or non-East Asian) it may be a tempting easy trap: it's a lot easier to discern within a context you are familiar with since childhood than in a context that is alien to your personal experience.

Pascvaks said...

@askatasunaren said...

People are people. The BIG question is 'Why'?