7000 year old human bones found in Spain

[L.D.Brousse]

http://www.huffingtonpost.com/2012/...id10|htmlws-main-bb|dl1|sec1_lnk1&pLid=174863

 

[Christiaan]

http://www.huffingtonpost.com/2012/...id10|htmlws-main-bb|dl1|sec1_lnk1&pLid=174863

http://dienekes.blogspot.nl/2012/06/mesolithic-iberians-la-brana-arintero.html

 

[Taranis]

Excellent news!

(I have moved this thread into the ancient DNA section, just so that everybody knows )

 

[ElHorsto]

http://dienekes.blogspot.nl/2012/06/mesolithic-iberians-la-brana-arintero.html

I can not draw clear conclusions yet. Some remarkable results from the Paper and Dienekes regarding the ancient Brana DNA are:

- "...euro7 calculator, the results are: 89.6% Northwestern"
- "Using the K12b, the results are: 45% Atlantic_Med, 41.6% North_European"

That means there is again the same splitting of "Northwestern" component into into Atlantic_Med + North_European, and this time for an ancient human. As in the Scandinavian Gok4 and Ajv samples there is no Gedrosia component as well.

More:

- "...fall closer to Northern European populations such as CEU and Great Britons..."
- "not related to modern Basques" (This claim makes no sense to me, because Brana allegedly clusters with Brits, and the partial Basque-British overlap is actually famous)
- "nomads"
- "Eastern African component" (There was some very exotic element in the Ajv sample as well. Perhaps those exotic nomadic tribes became just extinct until now?)
- "asian shift" (perhaps a northern finnic-saami connection)

A K3 analysis would be interesting. Maybe these nomadic hunter-gatherers might have had a very open lifestyle, allowing locals to assimilate. These particular ones might have mixed with Basques (most likely women). But again, I don't get how the Brana guy is "unrelated to Basques".

 

[spongetaro]

It looks like the Southern admixture is extra european.

 

[Knovas]

Even if extra European (Middle East or North Africa), it is as West Eurasian as the Atlantic_Baltic is according to the Fst distances. It's the most distant cluster from East Eurasia, and almost the same far from inner Africa compared to Atlantic_Baltic. So in my honest opinion, it's not clear where this element originated...surely somewhere in the Mediteranean, and not necessarily outside of Europe (Southern Europe).

Also, Atlantic_Baltic includes extra European influences. All this clusters emerge using modern genotypes, hence, they're all admixed.

 

[spongetaro]

What if I2a1 (from the Balkans?) had brought the southern admixture?
Though I wonder which Y haplogroups did these Iberians Meolithic folks carried: I*, I2*, I2a1 already?

 

[Knovas]

Now with the new nomenclature I'm still a bit confused. At least, I2a1a (I2a1 before) most likely originated in the Pyrenees. The previous clade could have originated in the Balkans, and it could partly give an explanation for territories which nowadays are high in Southern and have sizeable amounts of I2a1 variants (Southern France, Iberia & Sardinia specially). Worth to mention that even nowadays there are small pockets of I2a1 variants in the UK and even in Ireland, where it's quite unexpected. Note that the Southern element is very significant there too.

Recently was posted in another thread the possibility that Haplogroup E is not a native African Haplogroup, but Paleo-Eurasian (likely from the Arabian Peninsula). Dienekes' wrote about this and the Out of Arabia hipothesis regarding modern human origins. The main point was that most Sub-Saharan populations (except Pygmies and a few San from Namibia I think), seem to have an afiliation with West Eurasians, and Haplogroup E could be the reason according to him. So, in other words, E as whole could be linked to the Southern element as well, although I must clarify that the mentioned afiliation it's only noticeable in genetic maps at the moment. For instance, in those maps we can see groups like Yoruba deviating towards West Eurasians, being the Pygmies and probably a few San the most extreme African pole, but when you run an admixture experiment, the same Yoruba samples come out 100% African or something like this...maybe it's due to a limitation of the Admixture program or there should be better represented populations in the dataset. Let's wait.

 

[ElHorsto]

What if I2a1 (from the Balkans?) had brought the southern admixture?
Though I wonder which Y haplogroups did these Iberians Meolithic folks carried: I*, I2*, I2a1 already?

I'm actually quite sure that Y-HG I, or at least I2* belongs to Atlantic_Med or South_Euro components, e.g. Basques and Sardinians. For the following reasons:

- Sardinians are one of the very few europeans with 0% North_European component. At the same time they are a major I2 hotspot in europe. Of course, there is the possibility of genetic drift, but if I2 was from North_European component, then Sardinians should have at least a tiny bit more than 0% of North_european. This basically makes a northern origin of HG I2a1 very unlikely in my opinion.

- The Basques also still have some significant Y-HG I, but it got mostly replaced by R1b, which can be probably explained by the detected gedrosian admixture, similar to other west-europeans. The Sardinians have the Caucasus component instead, which possibly could explain their Y-HG G lineages.

- Regarding possible Balkan origin of Y-HG I2a1, I have no opinion yet.

- Even if Y-HG I is a Cro-Magnon heritage, as some others propose, it still is consistent with Atlantic_Med component being its carrier, because I find it unlikely that all these southern glacial refuge areas suddenly became entirely depopulated, with the Atlantic_Med component being a completely new settlement of south-europe in the Mesolithic or Neolitic. It could also be that Atlantic_Med component eventually turns out to be a combined Cro-Magnon + something else component. But the North_European component really seems to be related more to the Finnic or Indo-European peoples, not Cro-Magnons, unless HG-Y I is not from Cro-Magnons.

- North and North-West Europe might have been very sparsely populated after glaciers receded. And we know how the Saami dwelled along a huge area and are anthropologically part-asian. Probably the european hunter-gatherers can be compared to american indians, which intermarried with foreign tribes, and which were rather sparse peoples and got almost exterminated by indo-europeans and/or farmers. The Braga sample might belong to nomadic visitors that carried exotic admixtures and came later and died out.

 

[sparkey]

What if I2a1 (from the Balkans?) had brought the southern admixture?
Though I wonder which Y haplogroups did these Iberians Meolithic folks carried: I*, I2*, I2a1 already?

I2a-M26 is the only good candidate for an I subclade 7000 years ago in that region that is known to still exist today. That's I2a1a in current ISOGG nomenclature. The M26 mutation occurred between 10,000 and 18,000 years ago or so, so it's a good candidate to have been in a 7000 year old population. A better question is whether or not L160 was present (I2a1a1).

Now with the new nomenclature I'm still a bit confused. At least, I2a1a (I2a1 before) most likely originated in the Pyrenees. The previous clade could have originated in the Balkans, and it could partly give an explanation for territories which nowadays are high in Southern and have sizeable amounts of I2a1 variants (Southern France, Iberia & Sardinia specially). Worth to mention that even nowadays there are small pockets of I2a1 variants in the UK and even in Ireland, where it's quite unexpected. Note that the Southern element is very significant there too.

The diversity of I2a1a IMHO strongly indicates a broadly Western European distribution prior to the Neolithic... from the Rhine to Iberia. That's consistent with finding outlier clades along the Rhine and Northern France/British Isles. I also don't suspect that any of its recent ancestor clades were from the Balkans. Maybe I* before it split into I1 and I2, but that's extrapolating back too far, I think.

Recently was posted in another thread the possibility that Haplogroup E is not a native African Haplogroup, but Paleo-Eurasian (likely from the Arabian Peninsula). Dienekes' wrote about this and the Out of Arabia hipothesis regarding modern human origins. The main point was that most Sub-Saharan populations (except Pygmies and a few San from Namibia I think), seem to have an afiliation with West Eurasians, and Haplogroup E could be the reason according to him. So, in other words, E as whole could be linked to the Southern element as well, although I must clarify that the mentioned afiliation it's only noticeable in genetic maps at the moment. For instance, in those maps we can see groups like Yoruba deviating towards West Eurasians, being the Pygmies and probably a few San the most extreme African pole, but when you run an admixture experiment, the same Yoruba samples come out 100% African or something like this...maybe it's due to a limitation of the Admixture program or there should be better represented populations in the dataset. Let's wait.

I also want to explore the Haplogroup E = Southern element hypothesis more, especially now that we've found some ancient Haplogroup E in Europe indicating that it's Neolithic at the latest, and after my own analyses of Haplogroup I have found little diversity of it across Southern Europe (the closest exception being northern Iberia having a lot of I2a1a diversity).

 

[sparkey]

I'm actually quite sure that Y-HG I, or at least I2* belongs to Atlantic_Med or South_Euro components, e.g. Basques and Sardinians.

By "I2*" you mean I2 as a whole? (That's a misuse of "*" by the way... no I2* is currently known, unless you're going by the FTDNA standard which doesn't recognize L460, L416, L596, etc., and mean what ISOGG calls I2b and I2c, which I don't think you do.) Including I2-M223, which is, as a whole, very Northern/Central European, and an important component of Germanic populations? You could instead say I2-P37 as a whole, which removes the tricky P217 subclades, but even with P37, you have to account for its diversity not having a particularly Southern spread (look at the red dots here). I suppose it's true that the greatest expansions of I2-P37 have been southward... into Sardinia and the Balkans in particular... and that we can say that those expansions weren't just Y-line expansions. But watch out, that's a testable hypothesis, which will have us expecting significant Southern_Euro components in ancient samples from Northern France to the Netherlands, and little Southern_Euro components from ancient Balkans samples.

- Sardinians are one of the very few europeans with 0% North_European component. At the same time they are a major I2 hotspot in europe. Of course, there is the possibility of genetic drift, but if I2 was from North_European component, then Sardinians should have at least a tiny bit more than 0% of North_european. This basically makes a northern origin of HG I2a1 very unlikely in my opinion.

IMHO we'll see typical haplogroup "arrays" for different autosomal clusters, rather than obvious haplogroup-cluster mappings. That is, Sardinians indicate that the North_European component does not tend to have an array that has a lot of I2-M26, G2a, and J2... but that doesn't exclude certain autosomal North_European populations carrying the occasional I2-M26 as part of a different array, including, say, I2-M223 and I1.

- Even if Y-HG I is a Cro-Magnon heritage, as some others propose, it still is consistent with Atlantic_Med component being its carrier, because I find it unlikely that all these southern glacial refuge areas suddenly became entirely depopulated, with the Atlantic_Med component being a completely new settlement of south-europe in the Mesolithic or Neolitic. It could also be that Atlantic_Med component eventually turns out to be a combined Cro-Magnon + something else component. But the North_European component really seems to be related more to the Finnic or Indo-European peoples, not Cro-Magnons, unless HG-Y I is not from Cro-Magnons.

Why not map the North_European cluster to mainly Cro-Magnon heritage, including a haplogroup array that includes the more northern I clades? Don't forget that Haplogroup I is just as diverse as Haplogroup R, and it seems like we're establishing that two R subclades are important contributors to the haplogroup arrays that map to both the Gedrosian and Caucasian components. So clearly, different I subclades can give different contributions to the haplogroup arrays that correspond to autosomal clusters.

 

[Knovas]

We should also keep in mind that Ötzi was G2a and the autosomal data showed very high Southern. The number of SNP's available was low compared to modern samples, but the major components seem to reflect fairly well his ancestry.

I'm telling this because equating one component to one haplogroup is absolutely wrong. Atlantic_Baltic is not 100% I, and Southern, for sure, includes I even if in a lesser degree.

My thoughts:

Atlantic_Baltic = I (mostly I1), N and some ancient R1a. U maternal line would be significant in this cluster.

Southern = I (mostly I2), E, and G2a. H maternal line, which was surely present prior to the Neolithic, must have something to do here. Oh, and don't forget K, since its presence in Europe seems to be much older than expected according to the latest data (check mtDNA K thread: http://www.eupedia.com/forum/showthread.php?26612-mtDNA-K)

This is just a summary, I'm possibly missing something. Maybe Y-DNA T is also a minor part of the Southern element for example, although it's not clear due to its wide distribution (found even among Indian populations at not trivial amounts I think). Not easy.

 

[spongetaro]

I2a-M26 is the only good candidate for an I subclade 7000 years ago in that region that is known to still exist today. That's I2a1a in current ISOGG nomenclature. The M26 mutation occurred between 10,000 and 18,000 years ago or so, so it's a good candidate to have been in a 7000 year old population. A better question is whether or not L160 was present (I2a1a1).

I2a-M26 is indeed a good candidate to have been in a 7000 year old population but where?
M26 is linked to Neolithic culture (Treilles, Eure et Loire...) so far but was it already in western Europe BEFORE the Neolithic? I hope they will try to test the Y haplogroup of these two Iberian Mesolithics.

 

[ElHorsto]

By "I2*" you mean I2 as a whole? (That's a misuse of "*" by the way... no I2* is currently known, unless you're going by the FTDNA standard which doesn't recognize L460, L416, L596, etc., and mean what ISOGG calls I2b and I2c, which I don't think you do.) Including I2-M223, which is, as a whole, very Northern/Central European, and an important component of Germanic populations? You could instead say I2-P37 as a whole, which removes the tricky P217 subclades, but even with P37, you have to account for its diversity not having a particularly Southern spread (look at the red dots here).

I indeed meant I2 as a whole and I misused the "*", sorry. Actually I meant even complete I, including I1.

I suppose it's true that the greatest expansions of I2-P37 have been southward... into Sardinia and the Balkans in particular... and that we can say that those expansions weren't just Y-line expansions. But watch out, that's a testable hypothesis, which will have us expecting significant Southern_Euro components in ancient samples from Northern France to the Netherlands, and little Southern_Euro components from ancient Balkans samples.

Well, if I'm not mistaken, Gok4 from Scandinavia is already one such case.
Regarding Balkans, I did not say that Y-HG I = Atlantic_Med, but only that Y-HG I long time ago has become a part of today Atlantic_Med component. Probably Atlantic_Med itself is a combined component or a continuum.

IMHO we'll see typical haplogroup "arrays" for different autosomal clusters, rather than obvious haplogroup-cluster mappings. That is, Sardinians indicate that the North_European component does not tend to have an array that has a lot of I2-M26, G2a, and J2... but that doesn't exclude certain autosomal North_European populations carrying the occasional I2-M26 as part of a different array, including, say, I2-M223 and I1.

Of course not.

Why not map the North_European cluster to mainly Cro-Magnon heritage, including a haplogroup array that includes the more northern I clades?

Simply because all northern european populations I'm aware of still have at least 12% Atlantic_Med component (e.g. Finns). Even Mordovians still have 11%. On the other hand the Sardinians have 0% North_euro admixture. To me it really looks like the entire HG I is correlated with Atlantic_Med, slowly fading out towards the north-east, while North_euro component seems anti-correllated. That's just my amateurish observation.

Don't forget that Haplogroup I is just as diverse as Haplogroup R, and it seems like we're establishing that two R subclades are important contributors to the haplogroup arrays that map to both the Gedrosian and Caucasian components. So clearly, different I subclades can give different contributions to the haplogroup arrays that correspond to autosomal clusters.

Sure they can, but I have not found any evidence yet. For that I would like to find a North_euro rich population without any Atlantic_Med admixture, but at the same time still with any Y-HG I.

Note that I don't exclude that Y-HG I can stem from the Cro-Magnons. As I said, in that case there even might be an individual Cro-Magnon component which we just don't see yet, but so far it seems to be hidden inside the Atlantic_med component, but not North_euro component. By North_euro I mean the dodecad K12b component, not north-european populations of today per se, because the latter actually carry already considerable Atlantic_med (32.9% in Sweden!).

 

[sparkey]

I2a-M26 is indeed a good candidate to have been in a 7000 year old population but where?
M26 is linked to Neolithic culture (Treilles, Eure et Loire...) so far but was it already in western Europe BEFORE the Neolithic?

If I were to guess based on our limited ancient DNA and the modern diversity patterns, I would guess that it spanned most of Western Europe, with the Rhine being its eastern edge and northern Iberia the furthest it went to the southwest. But it's also possible that drift over time is throwing me off.

One important clue, I think, is that the calculated TMRCA of I2a-M26 is actually older than 7000 years. It's more like 8500 according to Nordtvedt and Cullen.

I hope they will try to test the Y haplogroup of these two Iberian Mesolithics.

Yes, hopefully.

 

[sparkey]

Well, if I'm not mistaken, Gok4 from Scandinavia is already one such case.

Gok4 was the farmer, though... I was thinking that we would find more Southern components in northern hunter-gatherers if your placement of I entirely within Southern/Atlantic_Med components was right.

Regarding Balkans, I did not say that Y-HG I = Atlantic_Med, but only that Y-HG I long time ago has become a part of today Atlantic_Med component. Probably Atlantic_Med itself is a combined component or a continuum.

I have trouble imagining this sort of continuum (which I also think probably exists) picking up Haplogroup I as a whole, considering the wideness of Haplogroup I's dispersal.

Simply because all northern european populations I'm aware of still have at least 12% Atlantic_Med component (e.g. Finns). Even Mordovians still have 11%. On the other hand the Sardinians have 0% North_euro admixture. To me it really looks like the entire HG I is correlated with Atlantic_Med, slowly fading out towards the north-east, while North_euro component seems anti-correllated. That's just my amateurish observation.

But Sardinians also have basically 0% I1, 0% I2-M223, and 0% any Haplogroup I except I2-M26, so that doesn't say a lot about anything but I2-M26. Isn't there some correlation between I1+I2-M223 and North_European? At least the western extent of North_European, because I also observe that the North_European cluster extends beyond the extent of Haplogroup I (as does Atlantic_Med).

Note that I don't exclude that Y-HG I can stem from the Cro-Magnons. As I said, in that case there even might be an individual Cro-Magnon component which we just don't see yet, but so far it seems to be hidden inside the Atlantic_med component, but not North_euro component. By North_euro I mean the dodecad K12b component, not north-european populations of today per se, because the latter actually carry already considerable Atlantic_med (32.9% in Sweden!).

I think we'll need to wait for more northern ancient samples (directly testing Cro-Magnons would be ideal), because I don't know of a modern population that represents anything close to direct descent from ancient Cro-Magnons.

 

[Keegah]

If this is a very obviously foolish question, my apologies. I am still learning about the field of genetics, and at the moment understand little other than the basics.

Is there any chance the haplogroup was I2a-L161+?

 

[sparkey]

If this is a very obviously foolish question, my apologies. I am still learning about the field of genetics, and at the moment understand little other than the basics.

Is there any chance the haplogroup was I2a-L161+?

There isn't a lot of L161 diversity in northern Iberia nowadays, so I find it unlikely. But L161 probably existed somewhere in Europe 7,000 years ago, so who knows?

 

[ElHorsto]

Gok4 was the farmer, though... I was thinking that we would find more Southern components in northern hunter-gatherers if your placement of I entirely within Southern/Atlantic_Med components was right.

Actually my reasoning is opposite, assuming that Y-HG I has been brought to the north by farmers. One tiny evidence for this is exactly the almost complete lack of Atlantic_Med among the two scandinavian hunter-gatherer Ajv samples. But since Atlantic_Med is strong today in scandinavia (~30%), I assume that Gok4-like farmers are the original carriers.

Further, if guessing the Y-HG of those farmers, Y-HG I appears most likely, because there are relatively few haplogroups today in Scandinavia, while I1/I2 being part of them. In particular, the typical neolithic haplogroups E,J etc. are almost missing in Scandinavia today, such that I1/I2 would remain as likely candidates for neolithic farmers.
Of course that is still much speculative. Extinction of tribes is also something that I believe happened quite often in sparsely populated regions. So maybe the farmers could have died out, although I see no important reason to assume this so far.

Nomadic tribes then might have brought a little bit of Atlantic_med and HG-I further to northern russia. Very speculative of course.

I have trouble imagining this sort of continuum (which I also think probably exists) picking up Haplogroup I as a whole, considering the wideness of Haplogroup I's dispersal.

On the other hand, Atlantic_med is also very widespread. If you mean that HG-I frequencies do not match Atlantic_med concentration, I agree. HG-I is indeed not exactly "fading out" like Atlantic_med does. But I think we also concluded that south-west europe was originally a stronghold of HG-I2 before aggressive R1b indo-europeans replaced these lineages. The few places where considerable HG-I2 is still left today are Sardinia and - to a lesser extent - Basque country, both being modal in Atlantic_med component today, although admittedly with very particular I2-M26 HG only.
And regarding the balkans, it is not clear yet. If the slavic theory of central european origin of I2 is correct, then the high balkanic concentration could be the result of a big genetic drift.

But Sardinians also have basically 0% I1, 0% I2-M223, and 0% any Haplogroup I except I2-M26, so that doesn't say a lot about anything but I2-M26. Isn't there some correlation between I1+I2-M223 and North_European? At least the western extent of North_European, because I also observe that the North_European cluster extends beyond the extent of Haplogroup I (as does Atlantic_Med).

That's true.

I think we'll need to wait for more northern ancient samples (directly testing Cro-Magnons would be ideal), because I don't know of a modern population that represents anything close to direct descent from ancient Cro-Magnons.

[/QUOTE]

Absolutely.

 

[sparkey]

Actually my reasoning is opposite, assuming that Y-HG I has been brought to the north by farmers.

Actually, I don't discount this possibility. The TMRCA of I1 indicates that it didn't really start spreading until after the Neolithic had already reached its region. But there seems to be a lack of other possible hunter-gatherer haplogroups in the region, which makes me think it's more likely that I1 is a case of a hunter-gatherer that bottlenecked and expanded by chance within the population later. Either explanation is possible, though.

One tiny evidence for this is exactly the almost complete lack of Atlantic_Med among the two scandinavian hunter-gatherer Ajv samples. But since Atlantic_Med is strong today in scandinavia (~30%), I assume that Gok4-like farmers are the original carriers.

Isn't this circular reasoning? That Haplogroup I was probably carried by farmers, because because Atlantic_Med was weak in the Ajv samples, and that Atlantic_Med probably had a lot of Haplogroup I, because Haplogroup I was probably carried by farmers.

I think it will close your loop and establish your thoughts firmly if we start finding lots of Haplogroup I in ancient northern farmers with a lot of Atlantic_Med.

Further, if guessing the Y-HG of those farmers, Y-HG I appears most likely, because there are relatively few haplogroups today in Scandinavia, while I1/I2 being part of them. In particular, the typical neolithic haplogroups E,J etc. are almost missing in Scandinavia today, such that I1/I2 would remain as likely candidates for neolithic farmers.

Let's see... modern Scandinavians have about a 5:3 ratio of Northern_European to Atlantic_Med. Ajv had more Atlantic_Med than Gok4 had Northern_European. So in my judgment, unless the group that introduced the Gedrosian component to modern Scandinavians had a significant impact on that ratio (which probably happened, but which way?), modern Scandinavians are about twice as much descended from Ajv types as Gok4 types. Yet Scandinavians are at least 1/3 I1... more than we would expect from I1 corresponding to Atlantic_Med, unless we don't think that any other haplogroups made up Atlantic_Med.

I actually lean towards simply throwing out modern Y-haplogroup frequencies as indicators of which haplogroup arrays used to correspond to these components, in this case. It's likely that modern I1 is a more recent expansion than when some of these components were introduced to the Scandinavian population. So its frequency in modern Scandinavians is unlikely to tell us anything.